Guide 2.01
���� � ������� � ���� ����
���� � ��������� ����� �� � ���� ���� �������
�����������
������� ���� ����� �����
���� �
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Welcome We’ve written this guide for people looking for their first or second job in Quantitative Finance. The idea is that by the end of it you will know the things that many people wish they’d known a few years earlier. P&D is a headhunting firm, so we have a keen interest in helping people do well when they try for jobs in banks and hedge funds. This is a collection of things we’ve learned about the process, the pitfalls and the skills you need and how to present them. However, quantitative quantitative finance is a broad field and almost nothing we can say applies to every situation. This guide will help you, but we’re always available online to talk about your individual case, either through the website www.PaulDominic.com or directly as
[email protected] or
[email protected]. A number of people only think to ask for this book just before their really important interview, so the first section is a crammers guide to make use of the little time you have. A shorter version of this Guide is available within Paul Wilmott’s book : Frequently Asked Questions in Quantitative Finance, available from both Amazon.com and the bookshop on Wilmott.com. What’s in Guide 2.0 ?
2.0 is a complete rewrite; we’ve had a lot of feedback from people who already have a quant job, and so we’ve expanded the content for more experienced people, and also for interview technique, negotiating tactics and dealing with headhunters. We cover how to get more money, and why sometimes money does not matter. The book list is upgraded and we deepen the coverage on spotting bad jobs, as well as jobhunting in a difficult climate.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 2
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Welcome We’ve written this guide for people looking for their first or second job in Quantitative Finance. The idea is that by the end of it you will know the things that many people wish they’d known a few years earlier. P&D is a headhunting firm, so we have a keen interest in helping people do well when they try for jobs in banks and hedge funds. This is a collection of things we’ve learned about the process, the pitfalls and the skills you need and how to present them. However, quantitative quantitative finance is a broad field and almost nothing we can say applies to every situation. This guide will help you, but we’re always available online to talk about your individual case, either through the website www.PaulDominic.com or directly as
[email protected] or
[email protected]. A number of people only think to ask for this book just before their really important interview, so the first section is a crammers guide to make use of the little time you have. A shorter version of this Guide is available within Paul Wilmott’s book : Frequently Asked Questions in Quantitative Finance, available from both Amazon.com and the bookshop on Wilmott.com. What’s in Guide 2.0 ?
2.0 is a complete rewrite; we’ve had a lot of feedback from people who already have a quant job, and so we’ve expanded the content for more experienced people, and also for interview technique, negotiating tactics and dealing with headhunters. We cover how to get more money, and why sometimes money does not matter. The book list is upgraded and we deepen the coverage on spotting bad jobs, as well as jobhunting in a difficult climate.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 2
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
My Interview is in two days time Don’t bother with the reading list.
Revise Your CV The single best predictor of interview questions is the set of things you have said you can do on your CV. If some of this might have been forgotten, then fixing that is a good way of spending a finite time budget. Don’t try and learn some wholly new field of maths, or a new programming language. Not only will you fail, but instead of answering “sorry don’t know”, you may find yourself blundering about with some vague ideas and leaving the impression that you know very little. If you have any time left once you have polished your core skills, skills, then you should start You have been studying for 15-20 years, so you aren’t going to make a quantum leap between work on the things that people will assume now and your interview. The trick is to avoid you know based upon the buzzwords on your screwing up. CV. Remember that the people who will read your CV at entry level are usually not experts in your precise domain. We find that a lot of academic work has the word “statistical” in its description, attracting the questions given to a stats whizz. Conversely many employers still harbour the increasingly vain hope that a CompSci grad knows algorithms.
Brainteasers About the only thing which has a useful marginal utility in learning from new is the puzzles that will be thrown at you. Work out the intuition of the Monty Hall problem, understand coin tossing problems, and and spend a leisurely period period doing some some brainteasers brainteasers from Wilmott.com Wilmott.com.. Some of these are the algorithms we talk of above, and it does not just apply to those with some CompSci background. Any formal scheme of problem solving is useful.
Go to bed no w An easy way of losing or gaining a few IQ points is sleep, and it is never worth trading rest for study just before an interview.
Don’t read a textbook Buy a copy of the Economist to read on the way as you may get asked about your take on events affecting the market, maybe backed with the FT or WSJ. Do not watch Fox News for at least a week before going for a quant interview, if you must watch it at all.
Phone interviews Please try and do these on a land line. This makes you easier to understand, which is always good, especially if English is not your first language.
Don’t be late Plan to arrive early. It reduces your stress leaving more mental energy for the fight that awaits you. Take your credit cards and some cash.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 3
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
WELCOME ...................................................................................................................................................2 MY INTERVIEW IS IN TWO DAYS TIME .............................................................................................3 INTRODUCTION .........................................................................................................................................5 WHY YOU SHOULDN’T GET A JOB IN QUANT FINANCE ...............................................................7 HOW TO WRITE A GOOD QUANT CV ................................................................................................ 10 INTERVIEWS .............................................................................................................................................32 APPEARANCE............................................................................................................................................42 HOW INTERVIEWS GO WRONG..........................................................................................................49 WHICH PHD SHOULD I DO .................................................................................................................... 51
GOOD THINGS TO DO IN A PHD ..................................................................................................................59 BAD THINGS TO DO IN A PHD .....................................................................................................................60 WHY WE HATE MONTE CARLO. .........................................................................................................62 WHAT PEOPLE GET WRONG ............................................................................................................... 68 WHO WILL INTERVIEW YOU...............................................................................................................74 HR INTERVIEW ........................................................................................................................................76 THE DIFFERENT TYPES OF MATHEMATICS SEEN IN FINANCE ............................................... 80 READING LIST ..........................................................................................................................................84 WHAT YOU MUST KNOW ......................................................................................................................88 WHAT YOU MUST HAVE CODED UP ..................................................................................................89
FINITE-DIFFERENCE METHODS ...................................................................................................................90 ROGRAM OF STUDY FOR FINITE - DIFFERENCE METHODS .............................................................................91 P MONTE CARLO METHODS ..........................................................................................................................92 SUMMARY OF NUMERICAL METHODS .......................................................................................................95 PROGRAMMING SKILLS .......................................................................................................................96 C++ ...............................................................................................................................................................98
DEALING WITH OFFERS ...........................................................................................................................101 SPOTTING BAD JOBS ............................................................................................................................105 BEING FOREIGN.....................................................................................................................................109 BEING FEMALE ......................................................................................................................................112 BEING OLDER .........................................................................................................................................113 DEALING WITH HEADHUNTERS ......................................................................................................116 HOW TO QUIT .........................................................................................................................................121 IRED ............................................................................................................................................125 B EING F
USEFUL LINKS ........................................................................................................................................126 CHOOSING WHICH MATHEMATICS TO LEARN NEXT ..............................................................131 TRANSITION FROM IT TO QUANT FINANCE ............................... ................................................. 134
WHAT TV GAMESHOWS CAN TELL US ABOUT MATHEMATICAL FINANCE .............................................137 HOW TO COMPARE MULTIPLE JOB OFFERS ............................................................................... 138
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 4
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Introduction Quant finance has grown quite enormously over the last few years, and there are now tens of thousands of people who do it, teach it, study it, write about it and of course headhunt. For many people the issue is not just about getting a job, but getting the right job, or in some cases getting out of what had looked like the right job, but isn’t. Although the market is still pretty good, the “right” job will still be competed for, and this competition can be quite tight. If the recruitment process at the bank is working well (and no, it doesn’t always) then the people interviewed will tend to be roughly the same quality (however they choose to define that) and from comparable backgrounds. The people you are competing with typically have similar qualifications to you and are usually quite smart. The first thing this should tell you is that you need to stand out in order to win. Being headhunters we speak to a lot of recruiting managers and we find that the difference between the one who got the job and the person who came second is often very small for the employer, but obviously rather more You must get into the habit of using the advanced important for you. maths and finance techniques you have learned to make better decisions about your career.
But before you see anyone, your CV needs to get through the various filters, and again the necessity to stand out is there, and so later we will talk more on CV design.
Even with that filtering, there is a still quite a wide distribution of the people seen by the hiring managers. A lot of the interview process is trying to extract signals from very noisy data and so one person may be much better than another but for various reasons, this fact may be missed. The first result of this is that simplistic objective facts often get more weight than they deserve, and you must use this to your advantage. P&D quant Recruitment ship this guide to everyone who has sent us their CV, or who is studying at an educational institution with which we have an agreement to help people get jobs. If you got it some other way, but want to get updates, as well as improve your chances of getting a better job, send your CV to
[email protected]
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 5
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Paul & Dominic Quant Recruitment P&D has become the best known name for the recruitment of quants by principles that work in good markets and bad ones.
What we do Engage with the quant community as peers rather than sales droids. Take the trouble to understand what a candidate can and can not do Find out what they really want, which includes money but is not the only story. Filter for those will add the most value from an extensive database of candidates. Understand the work itself, rather than match buzzwords. Search everywhere for the best people, representing candidates from every place you can think of, plus a few others. Understand each candidate and explain to you why they are of value, and indicate both strengths and weaknesses.
What we don’t do We will not carpet bomb you with CVs that are of no value. Place people, then ring them a few months later to get them to leave. “Improve” CVs to make them unrealistic. All CVs are in the form we get them from the candidate themselves.
If you are interested in getting better quants into your organisation with less hassle, then contact us: Dominic Connor
[email protected]
©2005-2007 Paul & Dominic Ltd
Paul Wilmott
[email protected]
www.PaulDominic.Com
Page 6
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Why you shouldn’t get a job in Quant Finance This may seem a strange way to start this sort of Guide, but you really need to get your motivations clear in your own head before you try and explain them to managers who might employ you. The odds are about 50:50 that at entry level you will be asked a question of the form “why do you want to enter this line of work ?”. It concerns us that some people manage to get this simple question wrong, which is quite an impressive trick, though not a good one. It is clear that some people have simply not thought about what they are doing with their lives, which does not inspire banks to employ them. Indeed one gets the impression that sometimes people only find out that they don’t actually want to take on this career in the interview, when they stumble through half baked answers. Let’s review the classical reasons for doing quant finance from good to bad and all the way to scary.
Money Quant finance pays better than most other jobs You have no worth as a person at all. You are worth what you can get, neither more nor less. and sometimes extremely well. But if, somehow you have picked up the idea that banks are charities whose job is to give money to science graduates, this is the time to use a bit of economics. A classical defect in the sort of economics you will get from the media is the idea of inherent worth. This is a popular concept, generally in national economies that are on the slide. You may be smart, you may know a truly impressive set of things, and passed so many exams that they have merged into a warm fuzzy blur, but in banking you are worth what you can get; not a penny more not a penny less. You will get paid (or not) on the basis of how much it is believed you will make the bank richer, and in particular the part of the bank run by the people interviewing you. Some people are paid to manage risk so that the bank avoids getting much poorer, but it is a brutal fact of this market that they earn less than those who help get the money in. We have seen candidates who sail through the maths and programming, demonstrating that they understand finance on the way, but somehow leave the interviewer very unsure about what they can actually do for them. Banks and hedge funds pay more than other types of employer because they expect more from you. That’s not just at entry level where they seek out smart people, and you will be working longer hours than in most graduate jobs, and harder. Part of this is self-imposed because banks and HFs have a bonus culture and are a rough approximation to a meritocracy. People often say that they like meritocratic rewards and promotions, without always realising that the other side of that coin is that when so much of your progress is based upon what you deliver, you can’t get away with just thinking clever thoughts, or keep referring back to the one masterpiece you did in the past. Since there is no fixed upper bound to what you can achieve, you find that the person pushing you hard is not your boss, but yourself. A large chunk of quant finance is taking existing standard work and modifying it to fit these local conditions, but this is neither the most interesting work, nor the best paid. It is also the work that is most likely to be outsourced or replaced by a really cool Excel macro. Sometimes you may find that your job is basically to act as nurse to the sick old Excel spreadsheet that is important to the business.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 7
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Seen in that context, money is a poor motivator when it is some ugly hour of the morning, and you face a day of brutally hard work, for which no one will show any appreciation, and turns out to be entirely pointless because someone else screwed up. Maybe the failure turns out to be yours, but that doesn’t really make it better. Since the nature of much of quant finance is trying to get to grips with current and volatile market conditions, you can expect that a sad % of your work will be thrown away without ever being used because the situation for which you have worked no longer applies. This should be one advantage of a scientific education, that you have conducted experiments with negative results, and not been too worn down by it. Although hiring managers like to see success, a bit of “character” is also a good thing, and in this context “character” means the mistakes you have learned from. Picking yourself up after a knock is a critical personal trait in finance, so if you have an example of this, then it is worth mentioning it in your interview.
Glamour You will be working with large amounts of money, and often the firm you work for is seen You need to ask yourself honestly if the work itself as impressive by people with more mundane attracts you. Money and all the other nice things jobs. It is also the case that in investment of banking do not get you out of bed at 5:00 AM for long. banking, you will probably get an imposing job title much earlier than in any career, since its very rare in most other global companies to get “director” anywhere in your job title a few years after leaving education. In some exceptional circumstances that happens straight away. If you are really lucky and work very hard, you may just get issued with a business card by a major investment bank with “Director of Global Power” on it, which is pretty damned cool. There are actually several such gentlemen, and they work in energy derivatives, a market that is currently doing rather well. It is perfectly fine to offer the view that you like working in an environment where your work actually matters, or that you have been aware of the bank for some time. But be prepared to answer a follow up question of what, precisely you heard. A common theme in this guide is that anything you say verbally or on your CV may itself be the subject of a question. Our all time favourite question on this topic was asked of one candidate being interviewed for her first job in the City of London. She was asked by the very junior bod interviewing her why she was interested in this rather self-important firm. The candidate offered the view that the work would be interesting, and possibly of some importance to the clients of the firm. The bod clearly did not like the answer, and interrupted her part way through, and said “no, you’re wrong, you will be near important people doing important things”. The candidate was somewhat less than impressed, and at that point she came to know that she did not ever want to work for this firm, since her idea was to be one of the important people doing important things one day, not merely to use the same lift to get to the office. We like this anecdote so much that it has its own section in the sequel to this guide, “How to interview people”, or in this particular case “How not to interview people”. Glamour does not survive repeated viewings, your job is your job.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 8
Guide 2.0
Paul & Dominic’s Guide to Quant Careers The grass is greener
A surprisingly large % of entry level quants are heading in this direction because they are unhappy with where they are. Academia can be frustrating and badly paid, but this can apply to almost any type of job. Wanting something better is of course a perfectly good reason to leave where you are, but it is not sufficient for choosing banking. It is a fine reason for looking around, so think through why you picked this direction, and be prepared to answer questions on it. This is an opportunity to say things about what you are good at and what you enjoy working on. Put like that it sounds small, but can be a critical part of showing them that you would make a good part of the team, as well as making sure they haven’t missed some of your finer points.
Problem solving Of the standard answers saying that you like solving problems is one of the least bad, but we are not sure that it makes it all the way to “good”. Most people with a scientific or engineering angle enjoy solving problems, so a career where they pay you to do this seems ideal. It may be, but Get into the habit of thinking about your career the problem solving is only a small proportion using the skills you learn i n finance. of the work, so you need to ensure that you can face the hard and dull work as well. Also if you push this position, you are inviting them to come at you with their hardest brainteasers, which is good if you get past them.
Failure modes It is of course also true that some people that get hired who are really not as good as they appear at first, or that great people miss jobs for which they are well suited. There is profound sampling error because an interview is a relatively small set of samples taken over a short period and we know from experience that even the best people do experience off-days. Often this is a preventable situation so later in the Guide we give some tips on how to reduce the chances of that happening to you. For almost all jobs, there is competition varying from three or four to a dozen, and so it follows that you are unlikely to get any given job. Since your chosen career path involves a lot of probability, your mathematical intuition should therefore not let the subset of your peers who get jobs on their first try lead you to believe your career is doomed from the start. Early versions of the Guide didn’t contain this reassurance, but we’ve encountered a surprising number of people who are otherwise smart, but don’t think to apply any of the many maths or finance techniques they’ve learned to their career decisions. A significant part of the online career counselling we do is using these ideas from finance and maths to make decisions and plot better schemes for improving your position. Although we speak as if being interviewed were a purely stochastic process, you should be learning, and a big reason we have put the effort into writing this guide is to gather together our experiences so that this curve is as steep (and short) as possible. There is a lot more data available on why people don’t get jobs than why they do. This is why we talk of failure modes in several sections, not because we are morbid pessimists (usually), but because we observe that in today’s market, if you are competent, you will get a job, the variable is which one. A good mental discipline to follow is that a failed interview has bought you time to learn more f or the next try. It is also the case that banks can take a long time to get back to you, or require a series of interviews over a few weeks, so do not waste this time.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 9
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
How to write a good Quant CV Your CV is not merely some passive list of how you have spent the last 20 years. Its job is to tell a recruiter why he should interview you and it also helps to set the agenda for the questions you will be asked when they meet you. Thus it is important to choose what you disclose as a balance between what you think they want and the areas in which you are confident answering questions.
Start with a big list Write down everything you’ve done in some detail, and everything you think you can do. Unless you have a perfect memory, the first pass will miss many items, so you should revisit it a few times over the course of a couple of weeks. If you are studying QF, then it is worth sharing ideas about what to put on your list with your classmates. Making it is a valuable exercise, even though most of it won’t make it to your CV. Typically you will be trying to get as much of this into two pages as possible, which brings us to a highly Make sure that every single thing you include on important discipline in CV design that each your CV earns the space it occupies. If it does not, item you include must contribute to the then delete and replace it with something useful. message, making it as important choosing what not to include as to what you choose to say. At P&D we like to see these lists, especially at entry level, since before you have worked in finance you cannot be confident as to what is useful, and we can offer suggestions. It is also the case that the most useful skills vary with particular jobs, so you need to think what you have to offer to each one, and so will end up with multiple CVs The list is also useful when you write a covering letter. Assuming you see some sort of job specification, then it is worth explicitly taking each point in turn, and saying why you believe you meet the filter. This is good for both direct and headhunted applications, since it makes life easier for those people involved later in the process, which is always a good idea. A particularly useful outcome of this approach is if you are one of the majority of entry level quants who have made some changes of direction in order to end up here, and so have studied subjects that are not immediately useful. If your headline subject does not seem relevant, we may have seen that including useful topics within it can make a dramatic difference to your chances of getting an interview, by making it easier for recruiters to work out that you are useful.
Writing style As we say later in the “Being Foreign” section, good CVs use a version of English that is not taught in schools, or that you may read in academic or technical books. Indeed it is often what you have been taught in school not to do. Part of this is driven �� ��� �� ��� ���� by the need to keep everything in two or three pages. One example is to use short lists, rather than trying to pack in the conjunctions (and, as well, plus) �������� ��������� ���� which use up space for no payload. “Payload” is in fact the term you should ��� ������ ������ consider when building your CV. If something does not convey useful reasons ��������� ����� to hire you, then it should be removed to allow you to put more important text into the document. ��������� ���� ���
������ ��� ������� �� �� ���� ��
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 10
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Choices
It is important to show yourself as someone who has done things using clear and direct language rather than give the impression you are blurring things. Read each phrase, and ask yourself if it pushes you forward. The “passive voice” is often preferred in academic writing because it is impersonal and dispassionate, which is exactly why it is not ideal for CVs. Choose: I cured cancer at Kindergarten. Not: Cancer was cured at my Kindergarten. A hiring manager is interested in what you have done, and what you can do, not random bits of group history. There are two ways that the hiring manager can read this, and both are bad for you. The first reading is that you didn’t do anything at all, but an impressive thing happened near you. The second is that you are trying to get credit for something you didn’t do, or had a small part in. There is even a third interpretation that you simply can’t express yourself well. Choose: I brought about world peace. Not: We ran out of people to kill. This is of course spin. Try to use language that shows what you did was a success, or at the very least that you completed it. “Complete” is what some call a “power verb”, though of course no two sources agree on which verbs have the right power. Good verbs to use give the reader the understanding that you, personally did something. Strong: “delivered”, “finished”, “built” Weak: “Tried” Pathetic: “a sort of”, “a simple”, “like a real” Choose: I built a death ray to defend the Earth Not: Using a NN-221, we tried a simple method for the reduction in bioviability of a certain range of a sort of cyanobacteria in compliance with an overall mandate of xenophobia, under GTYQ. You are addressing a highly technical audience, but they don’t know what GTYQ might be, and as connoisseurs of CVs we sometimes wonder if (like in this case) the acronyms are just made up.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 11
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Choose: I wrote a best selling novel about cheese Not: This cheese loving, best selling author will be an asset to any team.
Do not talk about yourself in the third person, you may have been taught that talking about yourself that way is a good way of selling yourself. We sell people and we are telling you that it is not; it can even sound as if you are suffering from slight mental instability. As a point of interest, the “boasting” section of CVs is by far the most likely place to find tragically bad errors in grammar and even spelling, which is pretty bad in these days of spell checkers.
CV design principles To sell you effectively, a CV should be answering questions for the potential employer and attracting their interest. The first question when writing CV is “ who is going to read this? ”, followed up by “what do I want them to know at the end?”, and supported by “how do I show this to be true?”.
Who is going to read my CV? The cute answer is no-one. CVs are rarely read word by word like a novel, but skimmed by someone looking for the highlights. It is quite possible that you may turn up at the interview, and have to wait a few minutes whilst the managers read your CV for the first time. You may do yourself a favour by bringing a copy with you, since it may not actually reach the person interviewing you, or have been “improved” by a headhunter. The most important people to read your CV have a deep understanding of this area, but it is worth remembering that many headhunters have zero or negative understanding, an issue we expand in “The Klingon CV”. The obvious audience is the hiring manager, and of course critical. Most managers of quants are themselves from a highly numerate scientific background. We observe that an artefact of the inflation in qualifications is that the average entry level quant has about two years more education than the person interviewing them. Many senior quants started in finance when an undergraduate degree was ample. Conversely, quant finance is very unusual in how early in your career you will get to interview potential new hires. If you were working in the retail operation of a bank that would be anywhere from two to five years, investment banks will often send you off to look at new people within months, and in one occasion a candidate of ours was asked to look at other candidates before they were hired themselves. However, your CV will be read by headhunters as well as various people in HR and other groups, who will have only the most schematic view of this line of work. This means you should focus on clarity, and make sure your covering letter helps them route it the right way.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 12
Guide 2.0
Paul & Dominic’s Guide to Quant Careers What do I want them to know?
Try to get any information on the job you are applying for and make sure that your CV supports the idea that you are a reasonable choice for it. A useful tip is to remember that this person knows nothing about you and isn’t going to spend very long trying to find out. Short is good, and write it so that they don’t have to think too hard to find out how good you are. Make sure they can work out that you are suitable for the job, and this is not just some random speculative shot at getting hired. You want them to know you are interested in this job, so any specifics you can include, the better.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 13
Guide 2.0
Paul & Dominic’s Guide to Quant Careers How do I show this to be true?
Although this is the 3rd most important criterion, it is the one that requires the most attention. A surprisingly large number of CVs we read say things like “exceptional knowledge of derivatives pricing”, “expert C++”, “superb mathematician” and other phrases that are supposed to make us think that you are smart, or in some way useful. This goes wrong if you make these bold claims and then completely fail to provide any reason in your experience or education that you have even seen the things you claim to be a world leader in. Each time you say “I can do X”, make sure that there is at least one clear reason in your CV why we should believe you, and the level you got to with it. That can be done well by examples, or if you have space, adapting your claim to expertise to say something like “expertise in dealing with noisy data sets gained by extracting a signal from our experimentation upon cows”. For the best jobs, a manager would prefer someone who has worked on real data on cows to heavily cleaned up series that were selected to support a theory in finance.
Almost no one will ever read your CV. It will be skimmed by many people, so make sure it’s easy to pick out your good points.
This is actually one signal we have seen for working out which jobs are better than others, because the closer you are to the money, the lower the quality of the data. Real time market data is terribly messy, but in many cases if you wait for it to be cleaned up by a vendor then the chance of making money from it is lost. Thus model validation and risk often use clean historical series, whilst the people actually trying to make money must deal with data points so horrible they read like accounts produced by Arthur Anderson. As there may be a lot more applicants than jobs, the early stages are optimised to filter out those who stand no chance of getting in. Thus you must take considerable care to make sure you don’t fail at an early stage because of trivial errors. It is hard to tell from a two- or three-page CV whether you have the right stuff to be an ace quant, but little defects in spelling, grammar, or even layout can do more harm than you might think. Although at P&D we’ve actually done this stuff, it is often the case that CVs are filtered by people with wildly variable skills in finance with little time to spend analyzing every application. Often they resort to looking for keywords. Thus you should not rely upon them working out that if you have done subject X, you must have skills Y and Z. If you believe those skills are critical for this job, then make sure this can easily be spotted. One result of the relatively light attention given to CVs is that they may not spot things that are implied from your experience. It’s pretty obvious that if you have done swaps that you will have fixed income knowledge, and anyone interviewing you, will know that. But the people doing the filtering may not. Given that many HHs and others who look at your CV have absolutely no quant skills of any kind, they do pick up on the “feel” of a CV, which includes quality of English. This is particularly important if your CV gives them grounds for thinking that English is not your first language. The legal issues are such that no one is likely to ask, “you’re foreign, can you speak English properly?”, which unfortunately means that they may silently and unfairly judge you.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 14
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Read the job specification carefully and if they specifically ask for a skill or experience, then include whatever you can to illustrate your strengths. If you believe a particular skill is critical then you should mention it in your covering letter as well (or if you believe the headhunter is particularly dim). We could name names on the “dim” list, but that would be cruel.
Make sure you can be contacted You absolutely must make sure your contact details are reliable and that you regularly monitor the relevant email account(s) and telephones. It is sad when someone’s CV shows great promise, but they don’t respond in time to be interviewed. You have no idea how frustrating it is for us, but it’s far worse for you. This does happen. If you are at university, be aware that your current email address may stop working soon after you complete your course, in some cases the very next day. GMail is to be preferred over Yahoo or AOL for personal email addresses. The Private Message system on www.Wilmott.com also gets to us quickly. Dominic has the handle DCFC, and Get your CV checked by someone else who is a Paul is of course Paul.
native speaker of English. Encourage them to spot silly mistakes in dates or failure to make sense.
Be aware that the site privacy policy means that recruitment is separated from user registration, so we only know who you are if you make that clear in your message.
Get your CV checked Have your CV and covering letter proof read by a native English speaker, not just someone who has been taught English. This is important because many people really do judge your ability by how you express yourself. The first part of your application process is that someone will be trying to read your CV and if it looks poor then that hurts you. Quant Finance is an international sport, with speakers of every language so the ability to communicate difficult ideas is important and if you can’t get the name of your university correct, it makes the interviewer wonder if you can explain your views on jump diffusion or copulas. Yes, people do get their university name wrong, which is very nearly impressive in a scary sort of way. The same applies to subjects studied. The important thing is to take the view that someone who has found something wrong with your CV has done you a big favour, and you should thank them and encourage this reviewer to find more.
�� ��� ��� ����� � �������, ��� �� � ����� �� ���� �� ���� ������� �� ��� ���� ������. �� ���� ��� �� � ���������� ����������� ����������, ��� �� ���� �� � ���������� ���.
At P&D we do read all CVs and where we spot a tragically bad error, then we ask the candidate to re-write that part. However we do not ever fix it ourselves, as we want to present you, in as honest a way as possible. It is very easy for someone to make an error whilst “fixing” your CV, and we don’t want to add noise to the signal, especially in a domain where there are a lot of specialist acronyms (PDE, CDO, ABS, MCMC, FDM, RTFM, DCFC, PDQ, etc.) which are pathetically easy to get mixed up if you are not directly involved. So we believe that the only person who can improve your CV properly is you. If you genuinely can’t when given direction, then this line of work is not for you.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 15
Guide 2.0
Paul & Dominic’s Guide to Quant Careers CV services
So far, we’ve not been impressed by what we have seen in consultancies who offer to help you write your CV. It is unlikely that they have any knowledge of your specialist area, and they act more as a barrier to you showing why you are unique. You are unique, right ? One piece of feedback we’ve had from employers, is that the job some University careers services do with CVs is too good. Most of them help you write a CV that conforms to a good general standard, and that is the problem: They are very similar to each other, and you need to highlight the reason to hire you.
Weird science Having spent years mastering a subject, it will be obvious to you that it is relevant to quant finance. It may not be obvious to anyone else. Some subjects have names that make them sound as if they have nothing to do with finance. This can hurt quite a lot. An early success for Make sure that it is easy for a casual but busy reader to work out how each part of your CV our approach was as an able candidate who makes you suitable for the job. had fine qualifications, but few interviews, and none of them for a good job. He’d done a “Pure” maths PhD, and with a few select changes (like deleting “Pure”) he had a decent position within the month. It is of course rarely that simple, but you should have enough material to demonstrate a link. If all else fails list the finance books you have read.
Technical issues Some of the CVs we get at P&D require significant technical resources to read, such that it is unlikely that managers at hiring firms can open the files in any useful way. Sometimes the issue is that the characters embedded to support non-Latin character sets cannot be rendered on a Western PC running Windows. This can result in strange squares ruining the look of the document, or a flat refusal to open the file at all. Be very clear on this; whatever software you use to write your CV, it will be read by Microsoft Word, or Adobe Acrobat, and nothing else. This is of course pathetically easy to deal with; simply mail your CV to someone running different s/w. At P&D we can cope with RTF and the mutant XML formats used by the latest versions of Microsoft office, and as you may have spotted, this guide started life in a beta version of Word 12. However be aware that this is far from always true in the industry, so please stick to PDF or MS Word format. We still get occasional CVs in plain old text format, which although simple to deal with usually looks pretty amateur.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 16
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Bad words
Recently we have spotted a pattern, that some entry level candidates with good backgrounds are failing to get interviews, or any interest at all from banks. We do not claim to have a comprehensive model of this yet, but we do see some correlations. This centres around “Bio” and “Pure”. Biology is the science that usually offers the least good background to aspiring quants, but there is substantial research funding for medical and biological applications of maths and physics. Thus you may have some parts of your academic life where you have done biophysics, or explored some mathematical aspects of the way our bodies work (or more commonly how they fail to work). We believe that the simplistic electronic and human filters used to screen CVs for online applications, are set up to reject you in the same way as someone whose research has been cutting up animals or working with test tubes. Much the same effect seems to occur with “pure mathematics”, even when the candidate has done this to an exceptional level, and capped it with good education specifically in quant finance. Whereas it is obviously true that number theory, computability, topology et al have yet to find much use in banking, the prejudice against them seems to happen as an automatic rejection, rather than a reasoned evaluation of their abilities. Thus, whenever you can it is better to refer to biophysics as physics, pure mathematics as mathematics and so on. We wish to make it clear that you should never say anything untrue on your CV; not only is it unethical, the odds of getting away with it are not good enough. The sort of person who interviews quants knows which sorts of mathematics they will find useful, or at least prove that you are smart, so you are not fooling them. They may still of course turn you down, but at least that will be a more reasoned refusal than a keyword robot can manage, so your goal is to get past the dim gatekeeper so that they can make this decision, and you simply don’t want to confuse them. If you have a straight maths or economics background, it is worth making some reference to the physics you must have studied at some point in your life, since we also believe that these buzzwordbots seem to like it.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 17
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Teenage qualifications Lose them now
Some educational systems, in particular the British, or those derived from it, give all sorts of certificates at age 16 and 18. When people write their first CV, these are a large % of everything they have done with their lives, and know that their academic history is pretty much all they have to sell themselves with. You may even have been told by a careers tutor that they are important. This is true for some firms, particularly consultancies and the big accountancy outfits who for no obvious reason care a lot about what you did at 17. But investment banks don’t really care that much about what you studied before university. They may use the quality of the university as a proxy for your value, but no one in banking has any interest in where you went to high school. This matters because it takes up space that you should be using for something far more valuable, like which subjects you did at university. We see quite a number of CVs where equal space is given to achieving competency in history or woodwork at age 16 as to fluid dynamics at university. A depressing % of the CVs we see have lots of this junk and simply say they did Most things you did as a teenager do not earn “Physics” at a given university, with no detail their space on your CV, except where they offer at all. The only time your undergraduate you the chance to show that your English skills degree should be compressed is when the are better than your peers. things you have done later are worth more space, or if something went terribly wrong, and you’d like them to focus upon later or earlier parts of your life which look better. The converse is sometimes true, that employer assume someone who gives little space to their undergraduate degree has something to hide.
Except… Since finance is such an international game, it may be quite hard to tell at a casual glance where your “home” country might be, and occasionally it cannot be determined at all, which may be good or bad. You can get hit by uncertainty on where you are legally allowed to work, and whether your language skills are good enough to work in your target location. Thus a short entry on early education can be useful to indicate that you do speak English well, and/or that visas will not be an issue. An important thing to remember about international issues is that a lot of assumptions are made directly from the CV without asking you.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 18
Guide 2.0
Paul & Dominic’s Guide to Quant Careers The Klingon CV
Sadly, it is usually the case that most head hunters have absolutely no idea what most of your CV means, and from their perspective it may as well be written in colloquial Klingon. Between you and the hiring managers it will pass through the hands of headhunters, human resources, secretaries and if you’re really lucky, an intern who has some vague idea that “Mean Reversion” is not a film starring Clint Eastwood. Thus they will judge your ability in mathematical finance based upon buzzword matching and your adequacy at English. That’s unfair isn’t it? Life is unfair, and if you think banking is better watch “Wall Street”. At P&D we look below the surface, but the harder you make it for us, the more you reduce your chances and even for us it is worrying if someone puts “good communication skills” and writes text that makes the Teletubbies sound like Shakespeare. You should also know that CVs use a particular style of English, which is subtly different from the one you learned in school. This must be done right. In your first year alone you will cost your bank more than a Porsche. A lot of the people who will see your CV won’t be Would you spend 100,000 on “high able to actually read it in any useful way. specification” goods where the advert for them was rubbish? Do you think Rolex adverts spell “chronograph” incorrectly? (No, we didn’t have to look it up, and don’t ask what Paul has on his wrist?) Your CV is an advert for a high-value product. You. On the Wilmott forums recently we read of one person who applied to Goldman Sachs, but who got his dates of employment wrong. It could mean he mistyped, it could mean that he was tired or distracted whilst typing, other things may be true or false. One thing that is 100% true is that he didn’t get hired. Looking at CVs, we see this is not as rare as it should be. Several % of people have dates that do not look right and that does worry us. Of course, if you’ve done a part-time Masters in Finance, it may well be that you appear to be doing two things at the same time, but it is worth putting in a short explanation. Short is the important word, as we find that long explanations never sound as convincing.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 19
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Fonts and layout
Some things on your CV are important and you may want to draw attention to them. Your eye was drawn to this part, but Do not do this excessively. It is really irritating !! Exclamation marks are bad too ! (yes they are)
The only time breaking this rule has worked was a when hardcore programmer who learned the Postscript language that Windows uses to talk directly to printers developed a program that printed his CV as concentric spirals of text in varying size. Viewed on screen it would slowly spin. Yes, Dominic hired him… If you’re not prepared to spend at least a month learning reverse Polish notation, use a standard template. (Stick to two main font families, a sanserif, such as Arial, for large headings and serif font, such as Times, for main body text.) On the other hand, we get a small % of CVs in monospace fonts like courier. It is readable, but ugly, and looks very amateur. We prefer Times New Roman and Arial.
Make sure your CV is easy to read by getting a friend to look it over, better still get someone who doesn’t like you to pick holes i n it.
A surprisingly large % of CVs we get have fonts that “drift” due to cut and paste. Thus we will see the font face or size change within a paragraph, or between two blocks of text. This gives your CV a slightly untidy look, and gives the impression that it was knocked off without much care to detail.
Blue CVs Don’t send us a blue CV, don’t send anyone a CV that is black on a dark blue background. Just don’t OK? The default in every known word processor is black on white even for people who use Apples rather than proper computers, and for very good reasons. This message has been in our Guide from it’s early form, and in those days we’d only been sent one of these monstrosities. Since we published this advice we have got two more. We don’t know why, but every reason ��� ���� ��, ��� we can think of is bad.
����� ���� ���� �� ����� ���� �� �����. ���� ������� ���� ���� ���� �� � ������ ��� ����� �������� �� ���������� ���� ������� ��� �� ����� ��� �����.
We see shading on some CVs, and although it can be useful for drawing attention to important parts of your experience, we often see that it can make these actually harder to read or spot the very things you are trying to make more obvious.
(���, �� ������� ��� �����...)
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 20
Guide 2.0
Paul & Dominic’s Guide to Quant Careers PDF
Make a PDF if possible. These have a more professional feel than Word documents, they do not have virus problems (yet) and they retain original fonts and layout. They will also usually prevent people messing with your CV, provided you set the security options. Whatever software you use, print it out to make sure that what you see is really what you get. Perhaps view on and print from another PC to double check. A free 30-day PDF maker is available from Adobe.com and there are numerous utilities available at sites like Download.com. This is also important when dealing with some headhunters. Some like to “improve” your CV and this can be good, especially if a professional makes it better. But some HHs have a rather impressionistic view of quant finance and also may be tempted to “add” skills, as well as “correct” spelling. This does not always end well; sometimes the “improver” gets a bit carried away and it does really happen that “PDE” can accidentally get “fixed” to so that you apparently know a lot about “PDF”. Funny, yes. Good for your job prospects? No. We of course know that PDF can be a probability density function, and for all we know it also has meaning in cosmetic dentistry, but the important point is that you retain control over how you are presented.
You can help set the agenda of your interview by what you say you can do on your CV. Make sure this works for you, not against.
Document name Give your document a name that will be meaningful to the recruiter. Call it YourNameHere.pdf but not CV.pdf or worse still “temp”. No, we do not know why people do this, forming part of the magical mystery of being headhunters. Making it easier not to be lost is in the spirit of making it easier for the recruiter. It’s not nice for the recruiter to have a large number of files with the same name and it’s actually quite easy to get your CV written over by the CV someone else who also called it “CV”. Yes, this happens as well and if your CV gets lost this way, then you don’t get the job with 100% certainty. At a conference recently, Dominic was told by one senior manager at a top tier bank that just one large headhunter alone sent him 4 CVs per day.
It is easy to get lost in that.
Dates Make sure your dates “join up” as much as possible. Some people in the recruitment process worry about gaps. Do make sure that they make sense, you are supposed to be good at numbers remember? Of course sometimes they need a little explanation and at P&D we really like to know about when you were studying and working in the same period, or as with one candidate working on a prestigious national TV science show. Managers often ask and not only do we not look good if we can’t come up with answer, you don’t either.
Be honest If you claim skills in some area, it’s a good bet that you will be asked questions about them, so your CV should be a fair and positive statement of what you have to offer. No one expects you to share your history of skin diseases, but you’re going to be expected to back the talk with action and as we say earlier it can help push them into asking questions on your best subjects. Even honesty is not quite enough. You may say you have a “basic” level of skill, but some managers just see the buzzword, and fire hard questions anyway. The only thing to do is make sure you don’t let it rattle you.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 21
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Books There is a growing trend towards “self certification” in quant finance. We see a lot of CVs that list books the candidate has read. In general, that is a good thing, and certainly as we say above, worth more space than your teenage qualifications. You do however need to do this right, and if you do, it can help you do well at interview. When you say you’ve read Shreve, you are saying in effect “go on, ask me any question about his take on stochastics”, same applies to Jaeckel on Monte Carlo, Sutter on C++, or Gatheral on volatility. There is a good chance that not only will your interviewer have read these books too, but that he has a copy. That means he has a set of interview questions already made up for him, so you should be prepared to answer any of the chapter exercises, and of course have read all the book. It looks great when you have mastered one of these books, and really awful when you say “but I didn’t get as far as that”. This is part of the theme in this guide of trying to set the agenda for your interview. You cannot hope to succeed 100%, but the more you can help them to choose questions that you might be able to answer, the less the chances of being asked a question that you have no hope at all of understanding, and the greater of them walking into a trap that you have carefully prepared.
Think through your career, and try to find examples of things you have seen through to completion without help. A couple will be useful for your CV, but also you may get to mention them in the interview.
Show that you can do things By this point in your life you’ve soaked up a lot of information and acquired skills, which is of course, necessary, but it is not sufficient. A big question in the inquisitor’s mind is whether you can translate this into real actions that are finished, complete and correct. One can pass most exams by getting answers wrong, but by showing good working and an understanding of the principles. However, banks aren’t really all that happy if you lose a pile of money by getting “nearly” the right answer, where you put a minus where it should have been a plus, the sort of mistake we’ve all made on a paper. They like to see projects where you’ve started, worked to a goal and completed without having to have your hand held. This is a big reason why many banks like PhDs since it’s a powerful argument that you can complete substantial projects. However, if you’re going for a PhD level job, you still have to beat others who have reached that level. Although it is their job, managers often don’t really want to manage people much if they can avoid it, and will select (and pay for) people who need to be told what to do less frequently. Of course they tend to like people who also do what they are told. Projects completed are good and you should be prepared to answer questions on them. The people who interview you will often have your sort of academic background, so these questions may be deep. You may have changed direction in your career and you should be prepared to explain why you made any given choice. It is important to be able to show that you didn’t just “give up” on an area when it got tough. This is an important message to get across at interview, so if you can gently drop into the conversation a couple of examples of things you have done, rather than just studied, so much the better.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 22
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Interests and hobbies
Several of the people you meet will want to understand what sort of personality you have, or perhaps even whether you actually have one. Don’t try to be someone you aren’t and certainly it’s very hard to guess what exactly they want anyway, just aim to be a good version of yourself. In finance you spend more of your waking hours with your colleagues than the person you marry, so it is good to present yourself as interesting as well as smart. They all want to feel you can work with others, so the cliché of “reading, walking and listening to music” doesn’t really do anything for you. Certainly you shouldn’t fake an interest in something, but do try to find something with which you can speak with a little passion. One candidate had acquired a formal qualification in stage combat, we liked this since it is unusual and interesting without being kooky. Although it’s relatively rare these days for quants to engage in duels with swords on the trading floor, it’s Hobbies are slightly useful, but do not be tempted the sort of thing that catches someone’s eyes to cite one merely to impress an employer, for and can make a crucial difference. It also gives fear of risking that he asks questions that expose the non-specialist people who you will meet that you’ve been over optimistic. something they can talk to you about. This may sound trivial and sometimes it makes no difference, but even a small difference can be important. If you think that it might look out of place, then again all you have to do is ask. People do get “labels” in the process at banks, assigned almost randomly based upon things spotted on CVs. These are used because employers cannot hope to remember all the names of candidates from CVs, and so they jog each others memories this way. Examples we have seen include “postman”, “death”, “farmer Giles” and a distinguished female candidate whose earlier work led to the label of “fish girl”.
Last job first In your CV your most recent/current employment should stand out and be relevant to the job for which you are applying. Since we emphasise that CVs are not always read thoroughly it is easy for someone with substantial experience to be mistaken for a much more junior candidate because they put their first job at the top. Someone reading your CV may never get beyond this one piece of information. Make sure your dates are correct. As part of the pre-employment screen at most banks, they check your past employment and people have had offers withdrawn because of mistakes on this. Make sure you read this from the point of view of someone who knows nothing about your life. One way this goes wrong is where people put their most recent employment during summer holidays, which is fine if an internship at bank, but much less so if working in a supermarket, which can lead to the recruiter not bothering to read down to your PhD.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 23
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Summer jobs
Most of us did summer part time jobs whilst studying and occasionally they are useful beyond the welcome input of cash. However if they don’t serve to indicate something nice about your abilities, they should not take up much space, if any at all, since those lines can instead be holding something useful.
Paul & Dominic When applying to P&D, we also like to see a simple list of the various skills you have acquired, together with some estimate of how good you are. If you’re new to QF then it won’t be obvious which are most important, that’s our job, so include as many as possible. They are for our own internal use only, so don’t worry about formatting.
Multiple CVs Do not put all your career eggs in one basket, and There is no reason why you should have only try to make each application at least slightly one CV. Presumably your entire life doesn’t fit specific to the role you are trying for. on two pages and having produced the long list above, it can be hard for you to work out what to include. Thus the optimum may be to produce a variety of CVs, each emphasizing different aspects of your experience and education. You may take this as an exercise to work out the optimal number of variants and you will quickly find out that your chances don’t hit a maximum at one. This is made more important by the fact that failed CVs get little if any feedback. This applies both to good CVs and bad ones. Think of applying for jobs as shooting in the dark. If you don’t hear a scream when you fire in one direction, you aim somewhere else. This is a useful tactic when you’re not sure whether some skill or experience is good or bad. For instance you may want to avoid jobs with too much programming, but it’s hard to get a job with no programming skills. Thus you can vary the amount of this you put on your CV.
If you are able to find out any details of what the job requires, and the sort of person they are looking for, make sure your CV covers these explicitly. We would emphasise again that CVs are often not read as you would a novel, but quickly scanned for relevant items. This process is far from foolproof, and you don’t want to lose out at such an early stage because of someone else failure to spot your relevance.
Finding banks In this document, we use the term “bank” for the firm you want to work for. It is course the case that quants work for many different types of outfit, including brokers, governments, hedge funds, insurers, thrifts, consultancies, regulators, ratings agencies, building societies and of course in the case of P&D, for a headhunting firm. The wilmott.com website mentions any number of firms and before you approach anyone it’s good to do a few searches so that you know the nature of the target. If you’re still linked with your college then it has many resources to help you. Most have a careers office with directories of banks and they will have some contacts with banks in that country. Your college library will have directories and of course there is Google and Yahoo for getting a list of targets. All large firms have entry level programmes of some form and you can relatively easily find a good number to apply for. At this stage playing the numbers is important, since the ratio of new entrants to the market to jobs is quite high.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 24
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Direct applications vs headhunters Of course, being headhunters we despise direct applications since they take away money from feeding our children. The optimum course is a mix of direct and brokered approaches to banks. There are a lot of CVs sent to a bank and typically each one does not receive that much attention, so it is useful having someone to sell you and prod the manager into making the effort to interview you. This is not trivial. On the day we wrote this, one of our candidates was being interviewed by a top 5 bank 7 months after applying, and after a certain amount of pushing from us. Apart from finding the right hiring manager (who has changed in this period), gently bugging him, their application has twice been thrown away by their email system. That happens more than you might think. Your Gmail account has a couple of gigabytes (please tell me you’re not on AOL), but the average corporate mailbox used by the people who will employ you has a limit of 100 megabytes and often this is policed by random deletions of mail. Yes, really. A simple personal web page which carries useful
Soft copy CVs
information about you and your work can be a useful supplement for a short CV.
Now that we are well into the 21st century, it is now very rare for a CV to be sent to an employer on a piece of paper. You can use web links to your advantage if you have work that you would like to show off. Obviously it is ideal if you have some innovative piece of financial software to show off, or papers you have published often which can help your cause. We have seen that good papers can make a critical difference, if they are indeed good. Once you fire off your CV, to a given target its contents are fixed, but a website allows you to keep updating the content. One candidate went as far as to create a website based upon one of the books in our reading list, complete with errata. He now has a pretty decent job at a very large investment bank. As well as allowing you to show work and experience that cannot fit in your CV, it helps with the theme of showing that you can do things. You should not have a vast array of links, since they won’t all be clicked on, so pick those with the most impact. Some headhunters will of course remove this from your CV, since they are control freaks, but at P&D we don’t meddle with your CV. Our view is that the more good things there are to say about you, the easier it will be to help you get a job. The corollary of this is that the links must work. Please test them before you send us your CV, and ensure that they stay live for as long as you are looking for a job.
Putting your CV on the web This has both good and bad points. It costs nothing, and it may be found by someone who is looking for someone just like you. However, very few hiring managers bother to trawl the web in this way, since they already receive quite a few applications. Certainly you can expect a lot of spam to the email address, both Paul and Dominic have “public” email addresses and so we get a lot of spam. Once you put it up on the web, you lose any control, and once someone has taken a copy of your CV, taking it off your site won’t make their copy go away. But with a bit of care you can use a “disposable” email address, and not give your physical home address or telephone number.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 25
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Being Googled
It is mildly common nowadays for employers to Google on your name, so it is worth considering some of the more “interesting stuff” you might post on bulletin boards. That’s one reason that the usernames on Wilmott.com are kept highly confidential and never linked to recruitment.
What will happen to your CV At P&D, we have a hardcore policy that we don’t change your CV. At all. Even one tiny bit; the manager gests your words with no “adjustments” by us. To be sure, we freely offer advice and may ask you to expand certain parts at the expense of others. However you should be aware that most headhunting firms will make at least some changes to your CV. Sometimes this is small stuff like removing your contact details or the name of your current employer, but you should be aware that sometimes they will adjust it, from minor “fixes” to retyping the whole thing for their “standard format”. At interview, it is better to shine at a smaller set of
The value of skills on your CV
thing than to disappoint in a wider range because you were too optimistic when writing your CV.
Some parts of the recruitment process are very buzzword driven, this may be by computer or by human, so make sure you include a variety that indicate what you can do for the bank. An important parameter to what you should choose is whether you can answer questions on it. In this guide we tell you this more than once because it is so important, the terms on your CV typically determine about 1/3 of what you will be asked at interview, and since you have chosen to mention them, you will be expected to show competence. Overall it is a balance between having the number buzzwords to get the interview, but the ability to deliver on these promises once they start firing questions. Elsewhere we cover things you should know, but here are a few that are less valuable than you might think:
SQL Databases For a straightforward quant some ability to manipulate data is of course good, and SQL is often a good way of doing this. But there is little point getting very good at it, since much of the data points you use in quant work comes from non-relational sources. Relational databases like MySQL, MS SQL or Oracle are typically quite poor on tick data, as is Sybase. Bizarrely, toy database systems like Access can sometimes be better for high frequency data points, but few people are impressed by it. A mention of SQL, maybe up to stored procedures, is good but not much space should be given over to it. It is more useful for a quant developer to know databases, and there is a good market for people who can build or tame the tick databases used for the huge amount of high frequency data used in quant finance.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 26
Guide 2.0
Paul & Dominic’s Guide to Quant Careers IT operations.
It is good to show that you can deal with computers in a practical way, but you should not give the impression of too much “hands on” work. Network management is a very different discipline at a large bank to any university, thus combining being useless with the possibility of opening a door to a job you really don’t want, babysitting the chaos caused by the latest demented attempt to outsource IT support to EDS. However, low level network programming (packet/socket level) is highly valued by some managers, and at the very least removes some questions about your ability to be do “real programming”. Delete mentions of “backups”, “cabling”, and “user admin”.
Microsoft Powerpoint We see this a lot on entry level CVs, presumably because it is seen as a “business” skill. Yes, well maybe, but not this business. To some managers it looks a bit desperate to put this or things like MS Word or Adobe Acrobat. Yes, HTML is a programming language, and occasionally a quant will write a little bit of it, but it is not that useful, and may give the wrong impression. This applies to quant developers as well, and you probably don’t want the sort of job where they are very interested in Powerpoint.
Too many hobbies It is mildly useful to add some background colour to your life, and if part of your interview process involves meeting non-specialist people, it gives them something to talk about. We see some CVs where too much space is wasted on this, and in a few cases one is left with the impression that their work was an inconvenient interval between sport. One thing that surprised us when researching the Guide was the number of interviewers who reported that candidates were dishonest about their hobbies and interests, such that when casually asked about them, quickly showed that they had either just made it up or exaggerated to an embarrassing degree. Bankers don’t really care about the level you’ve played sport or learnt languages or which books you choose to read, but they focus like a laser upon dishonesty. In such cases, it is invariably fatal to your application. The “books” thing was mentioned quite often which seems just bizarre. We really do hear of applicants who list “reading” as a hobby but look confused when asked what was the last book they read. Not only does it expose their integrity, but just how difficult is it to think of some book you have read at some time in your life ?
Foreign languages Aside from the obvious fact that finance is dominated by English, there are those that think that competence in French has value simply because so many quants are French speakers. Given that in our experience they almost always speak good English, we don’t see languages as an important part of the academic armoury of a quant. If you are from a non-English speaking nation it can be less valuable to put yet more languages on your CV, as it may worry managers that you can’t do English well. We note that well educated people from the Indian subcontinent often speak several languages to a business standard, buts sadly these have little use in quant finance. For those from the People’s Republic of China, it is worth putting a little more about any time you have spent immersed in an English environment, since experience is that you can benefit from experience in speaking English. This is distinct from reading and writing skills, and can only be acquired from extensive conversation with English speakers.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 27
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Politically incorrect organisations
There are bodies in various countries that are perfectly respectable locally, but do not always reflect upon you well when you go abroad. We have seen a couple of entry level people who quite happily announce junior membership of the local equivalent of the Gestapo, to the extent that even the most politically indifferent manager might wonder just what sort of person you are. This is another case where you need to get your CV checked by a native English speaker. Military service on the other hand is usually seen as neutral, and some managers actually favour it, especially if you do difficult things or get to show some form of leadership. However, it rarely justifies very much space on your CV due to lack of relevance.
Top 37 Colleges …or some other number, we don’t care. As we say elsewhere, there are many rankings of universities, and we aspire to be ignorant of every single one of them. This reflects the indifference and ignorance of our clients who have better things to do with their lives than read academic league tables. The fact is those who filter CVs have a variety of heuristics, experience, gut feel, and simple prejudice to judge the brand value of where you have studied, and you aren’t going to budge them easily. They have warm fuzzies towards Indian IITs, Oxbridge, Warwick, QMC, the French Engineering schools and the Ivy League, except when they don’t because they went to a different place. There is great diversity in what hiring managers and headhunters interpret as “top”. If you like, you can say you went to a school in some list that placed it 37 th or 9th or some other number and it will be soundly ignored, or even possibly seen as a sign that you realise the place you studied was rubbish. We have learned from the forums that there are accreditations for various Masters degrees. We learned this from the forums, rather than structured and diligent research because we don’t care about these either. We are especially proud that we can’t remember what the lists are called, much less which programmes are on them. About the only technique that has positive value is if you mention a lecturer who they have heard of. For this to work properly you have to be very familiar with their work, since of course that’s where you can expect some tough interview questions.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 28
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Really Quite Amazingly Bad Things to put in your CV
Aside from the infamous “Blue CVs” we occasionally get CVs which catastrophically undermine the applicant. At P&D we want to find you a job, it’s how we make money, but some candidates seem determined to make this as hard as possible. Stating that you are “pretty” is easily top of this list and, yes, we apparently have several “pretty” quants on our database. Stating that you hate computers is commendably honest, but in modern banking close to suicidal. Text has poor “emotional “emotional bandwidth”, and you must ensure that a desire to work in a particular part of the world does not come across as xenophobia. In general, it is a good idea to let your headhunter know about “red lines” in terms of location, business area, type of firm that you are not interested in, but this is best done verbally unless you are 100% confident that this is not going to come across as being “difficult” or “arrogant”. Although earlier we talk of a positive spin on the facts, you should keep this within reasonable bounds, and claiming that “leadership” “leadership” in a part time job in a fast food bar just looks silly. Make sure the personal characteristics you are Although it does occasionally happens that selling from yourself are both true and suitable teenagers find themselves in positions of great for working in a results driven environment. environment. personal responsibility, responsibility, we think it unlikely they will be in charge of 750 people working on their nation’s bid to host the Olympic Games. That example is of course fictitious, but some of the claims we see are not very far short of this, and if we were not obsessive about personal information could give examples of several extremely implausible claims. “Leadership” is another one of those points on a CV where people inflict damage upon themselves. Firstly, it is not a characteristic that is critical to an entry level quant, since you won’t be leading anyone. Teamwork is good, and we especially value those who show that they can work in difficult conditions without having their hand held.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 29
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Covering Letters
We have read quite literally thousands of cover letters, and in the main they are a wasted opportunity. The first and most obvious bug is being simply generic empty statements that you are a person worth employing. employing. Instead you should should take this opportunity to explain why you are good for this job in particular. If you are replying to an advert, then detail which parts of your experience and education are good for each point in the specification. It does not hurt at all if you mention a good thing both in your CV and your cover letter. The other common bug is to say you have mastered some desirable skill like stochastics, C++, signal processing, credit or risk management, but there to be no mention of this in your CV. In some cases we simply cannot work out how and when you might have got the skills you claim, which does not help your case at all. A surprisingly large number of cover letters have words to the effect that you have mastered derivatives pricing, aside from that not being likely in someone who has never worked in a bank, and we then look at a CV which has no mention of any A covering letter is great for pushing harder the finance at all, not even books you have read. Thus a good cover letter says in effect: “You want X, I did X at Bank Y for two years”
important part of your skill set, but makes sure that anything you declare is supported by the CV itself.
This is right at the centre of our theme of making life easier for those who may employ you. The less they have to treat your application as a puzzle, the better your chances. It’s also a good idea to mention where you saw the advert, since it shows you have some business sense, and reduces their idea that you have just applied to every position you can find.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 30
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 31
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Interviews The most important tool for getting an interview is your CV, which is why we have a whole section to getting this right, but it has to get in front of the right person, which means a multi levelled approach, and simply sending your CV to one target bank and expecting a job offer by return is not going to work so we first talk about…
Kissing frogs Like trying to find a prince by kissing frogs at random, you have to accept that it is rare for your first attempt to succeed, and so be prepared for the long haul and to pursue multiple options at the same time. This means applying to multiple banks and not being deterred by failure to get into a particular organisation. We talk to candidates who have fixated on a small set of the most glamorous banks, and their morale can be a little fragile when that doesn’t work out. This happens even to very good people. Also when you get offered a job in your target firm, it may not be what you want. Networking and personal contact is a great way if you are looking for your first job. Another way in which job hunting is different from fairy tales is that even if the frog you kiss does turn out to be a prince, it may take months for you to find out, so your tactics need to be flexible enough to cope with this lag.
Getting interviews Milk rounds and graduate recruitment fairs These are opportunities to find out useful things about your target banks, and get yourself known to people who may be useful to you. One thing that is worth knowing about investment banks in general is that they are profoundly unlike a supermarket chain or manufacturer. IBs are highly diverse with many component businesses, many of which are not only operationally separate, but often do not even know of the existence of each other. It is not unknown with the work we do with banks that people in similar business areas in the same bank only find out about each other during our recruitment events.
Thus the routing of your application is a complex process, which is of course why headhunters exist, since we have some model of how to get you to the right inbox. But P&D do not want you to think that you must only use HHs because at entry level that simply is not true, since a large % of jobs for newbies do not go near any headhunter, and any firm that tells you otherwise is not showing the level of integrity that you should require of someone trusted to advise on your career. It is worth engaging with HR at these events. Indeed, at some events they are the only type of bank staff there, and the nature of their work means that they know more a lot of hiring managers and what they are looking for. Do make sure you try to get their business card for future reference. HR people rarely have a deep understanding of quant work, but do have a pattern they are looking for, and a reasonable gambit is simply to ask them, and see how well it matches your skills. However, you must remember that the set of skills used by any bank is huge, so it is worth specifically asking which areas might use your skills.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 32
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
HR staff are professionally reluctant to give out personal contact details for managers in their firm, so your questions will work best if you talk about business areas that you would like to work for, rather than appearing to pump them for names, which can spoil the mood a little. Armed with this knowledge you can develop a version of your CV that more closely corresponds to what they are looking to send them as follow up. It is worth gently emphasizing that you are looking for quant opportunities, rather than the more common graduate stream, since the majority of people hired by the bank go into other lines of work, and you do not want to miss out on the right stream because they misrouted you into retail, IT maintenance, or wealth management. Alumni networks These are of highly variable quality, but are worth a try, and even a relatively small institution will have a usefully large number of former students who may be willing to help your CV You have to be prepared to make multiple get to the right place. applications, and not be thrown by the complete indifference shown to you by some recruiters.
Random applications More politely called “speculative”, there is nothing to stop you sending your CV to managers whose names you have found. The odds of a given shot succeeding are pathetically small, but if you read of a manager whose work looks remarkably like your own, then it is worth a go. Your chances go up quite dramatically if you take the trouble to find out about the nature of the work in your target area, and produce a covering letter that details how you can contribute. Avoid phrases like “your firm”, since they imply that this is some form of spam, rather than a genuine interest in the work of the person you want to employ you. Stalking However you apply, you are well advised to check in with them every so often, just to check on progress, and implicitly to check that you have not been forgotten or lost. You should ensure that you do not come across as obsessive, especially since your application is a lot more important to you than to them. Do your best to sound casual, and do not mention that they have chosen not to return your calls, which is really what you have to expect. Trying again Just because a bank says “no” once, doesn’t mean it despises you, indeed it is far from unknown for them not to check whether you applied before, so it can be worth applying to the same bank later. Some ask, and some do not. Hiring is not a simple matter of you being good enough, since managers are often looking for a “pattern”, not just the smartest person. It is mildly common for people to be rejected because the manager feels that their aspirations can’t be satisfied in his area, and decide they’d leave and/or be unhappy. However, if you are applying to two distinct parts of the same large bank, it is worth letting them know. They will not be pleased if you didn’t mention it to them, since it can be embarrassing and that’s not a positive feeling to cause in someone who might employ you.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 33
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Being there
At entry level, you are expected to do most of the running. If you are from far away that means coming to your target city and seeing as many banks as possible. This is not stress free, but you should allow for as much time as possible. Some banks will be flexible about when they see you, others won’t. Some will go for initial screens by phone, but many managers really hate these and you will do your application far more good by investing in a ticket and a hotel room. Make sure you are contactable and if you have good reliable friend in your target city to receive mail, then that can make life easier since it shows that you can be contact. All the things about being easy to contact are very true if you’re operating remotely. When you submit your CV, the a priori probability of getting that particular job averages from about 2 to 5%. If you’ve got to interview this has jumped all the way to 10-15%. That is if you haven’t ended up with a carpet bombing headhunter. They do exist, and it’s not too Some research on the people you know will hard to find out who they are. The term interview will help you a lot more than trying to “carpet bombing” was given to us over a nice learn a new mathematical technique in a few dinner in Paris by a top quant manager at a days. large firm who had received so many CVs from one firm, he’d needed help from his technical support people to sort out the mess caused by a flood of CVs that not only had he not asked for, but had demanded they stop sending.
Be prepared Before you go for the interview, find out the names of the people you are seeing and do a Google on their name, as well as the bank/business unit you are joining. Try to avoid the error made by one candidate who could not understand why the interviewer was so interested in one part of her thesis. The candidate had quoted papers by the interviewer, but somehow managed to not connect the interviewer’s name with the paper. They were quite shocked when at last the connection was made for them. This was funny at the time, but she didn’t get the job. This is less rare than you might think. Many quants have written papers, especially those who may interview you and your thesis will reference any number of other papers, pushing the probability up. This can work for you as well. If you find some intersection, then there is a rather higher probability that you will be asked on this aspect of your work. Make sure you’ve read up on that area and thought about it before interview.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 34
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Show some market insight
This doesn’t mean you have to know the ticker symbols of all SP500 stocks, but it does mean you should be able to comment on the reliability of common models, what are their main pitfalls and how the quant and the trader might communicate about this. If you can quantify some practical phenomenon that is rarely discussed in academic literature then you will impress. (Tip: Because banks are often delta hedging, everything usually boils down to gamma and/or volatility.) A standard form of interview question is to ask about the assumptions of a model and whether they are realistic. You should start with the set of assumptions/defects in the classical Black model and work your way up from there. It is also worth reading the Economist for at least a month before interview. Some interviewers are keen to see if you have awareness of the world in general. The Economist may disturb some people since it covers other countries and has no astrology column and relatively little Get in the habit of reading the business press, coverage of golf.. For instance, you should partly to get a feel for what is driving the markets, reach the stage where you understand the but also to develop a notion of how money is relationship between the current price of oil in made in real life. nominal terms, as opposed to its high in real terms. If you have a physics or maths background, or are generally lacking in basic knowledge of economics, you should also try to reach the stage where you can guess what happens to the supply of oil when the price goes up. A surprisingly large number of market effects are best explained not by stochastics but by bog standard supply and demand, for instance the non-intuitive way yield curves behave in long maturity government debt. If you are completely stumped in trying to explain something, a good straw to grasp is this very basic concept.
Brainteasers There are several different types of brainteasers you might get asked, all designed to test how your mind works under pressure and to try and gauge how smart you are, rather than how much you have learned. This is an area where you should spend a significant amount of your preparation time. After >20 years of education, your ability at mathematics is not going to change all that much in the next few weeks and to an extent he same applies to finance knowledge. At Paul & Dominic, we have been studying why people don’t get a given job, to see how we can help increase the chances of getting our people in and match them better to the roles on offer. One of the largest variable factors is brainteasers, puzzles, etc. Our estimate is that with diligent study and assuming that you are basically quite bright, you can reduce by around 50% the chance of losing out this way and overall there is no single factor that you can change so quickly that has such a big effect. Some people are effectively weeded out at the brainteaser stage. Thus for the job seeker, it’s quite likely that the most valuable resource on wilmott.com is the Brainteasers section of the forums and they fall into the following broad categories.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 35
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Straightforward calculation Example: How many trailing zeroes are there in 100 factorial?
Lateral thinking Example: Several co-workers would like to know their average salary. How can they calculate it, without disclosing their own salaries?
Open to discussion Example: What’s the probability that a quadratic function has two real roots?
Off the wall Examples: How many gas stations in the USA? How many manhole covers are there within the M25? Why are manhole covers round? Why are manhole covers being stolen?
General Probability and Combinatorics Examples: You toss a coin 20 times, what is the probability that you get exactly 10 heads? There are 36 people in a room, what is the probability that at least 2 of them share a birthday?
Conditional Probability Example: Monty Hall problem. These seem to be the most common type and a lot of quant finance is probability so demonstrating a grasp will do you a lot of good. Also tree-type problems where coins are tossed.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 36
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Numbers
It’s worth having a few numbers at your fingertips for the “manhole covers” type of problem. One manager recently told me in rather despairing tones of the stream of candidates who didn’t have even a rough approximation to the population of the country they were Possibly the single most valuable resource on born and educated in. Several put the Wilmott.com for job seekers is the pool of population of Britain between 3 and 5 Brainteasers. million (it’s around 60 million) Actually it’s a bit more, a good trick when “estimating” is to pick numbers with which it is easy to do mental arithmetic. Sure you can multiply by 63.21, but why expose yourself to silly arithmetic errors? It’s also worth revising some basic probability and set theory if your recent study hasn’t covered them. Converging series can make apparently very long-winded calculations deliver quick results. It’s actually quite rare for them to really want you to add up the numbers between one and a thousand in your head, but having the formula at your fingertips is good. Indeed that sort of inductive proof is good practice for brainteasers in general. Quants tend to have greater depth in applied maths than pure and that can be a weakness when good brain teasers seem to come more often from pure maths and probability. In many types of question, they want to hear your train of thought with much less interest in the actual answer. Thus you need to share your thoughts about how you get to each stage. You also should “sanity check” your answers at each step and make sure they’re aware you’re doing it. This is a soft skill that’s very important in financial markets where the money numbers you are manipulating are rather larger than your credit card bill. Given the free form nature of the questions, it is very easy to slip into a set of numbers where you end up with silly numbers for a given quantity. That’s perfectly acceptable, provided that you catch yourself, indeed you may impress more this way. Thus it’s worth trying to estimate the rough bounds for a numeric answer. For instance it’s not likely than more than 1% of the population work in gas stations or are engaged in stealing or replacing manhole covers. Work through the Brainteaser Forum on wilmott.com. You can practice for IQ tests and the more you do, the better your score. Brainteasers are no different. And you’d be surprised how often the same questions crop up. We have spoken to job seekers who have been asked the same question at more than one bank.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 37
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Be confident
Relax, the hardest step is already past. Most applications are rejected without the candidate being seen by anyone. Bayes law is your friend here. Merely by getting through the front door, your probability of getting this job has gone up by a factor of three or four. Almost no one at banks actually enjoys interviewing people, some even see it as a form of punishment. That means they only interview you if there’s a good chance they will want to hire you. Most people who are considered for any job never even get a first interview. You don’t need to be arrogant about this, just know that merely by getting an interview you’ve beaten a good number of your peers and that your probability of getting the job is on the way up.
Be punctual This shouldn’t need saying. If you can’t be on time for your interview how can they expect you to put in 12-hour days? If you are going to be late (and assuming it was unavoidable) telephone ahead with an accurate ETA. The best strategy is to schedule having a coffee before the interview, a little caffeine and sugar may well help and this is a useful time buffer. Probably the worst bit about being late is not what it does to the interviewer, but what it does to you. The idea is to present yourself as cool, smart and in control. If you’ve been stressed out dealing with transport problems you knock a few points off your performance. Dominic once went to an interview where the headhunter had given him the address of a building the bank had left two years previously. He ended up very late, but out of sheer bloody mindedness found where they had moved, and went anyway, 90 minutes late. Banks also often have several buildings some distance apart make sure you are going to the right one. For instance Barclays has a large tower in Canary Wharf, London with “Barclays” written in big letters on the top. Do not go there. The Capital Markets Division is right next to the main station, but cunningly has no indication that they are inside and if you leave by the wrong station exit, it can take 10 minutes to find. No one knows exactly how many offices UBS has in London, but fortunately they are all close to each other. Because it is a listed building, Goldman Sachs main building in London has “The Express” written on it from when it housed a tragically awful national newspaper. The entrance marked “in” is almost always closed by a big metal gate. Amongst other things, the main Canary Wharf tower is the headquarters of Torchwood, not as many people believe in Cardiff. New York is worse since many buildings often have absolutely no indication of what is going on inside them.
���� ���� ������ ���� ��� ���� ���� ���� ���. �� ���� ������� �� ��� ������ ����� ������� ��� �������� ����� ���� ���, ��� ��� ���� ���� ��� � ���� �� ����� ����� �� ��������� ��� ��� ����. ������ ������� ��� ��� ���� ���� ������ ���� ���� ������ ���� ��� ���� �������, ����� �� ����� � ��� % ������ �� ���� ��������� ��.
The point here is that you should make sure that you have time to spare; so that this sort of thing is a little puzzle, or the cause of a little extra travel, not something that brings you late and breathless to the interview.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 38
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Carry the phone number of the bank with you, both of the manager to apologise and reception who will be able to help you find them. Also getting there vastly too early annoys some people. Dominic was present at one very large firm where someone decided to show how keen he was by arriving 45 minutes early. This needled the very busy interviewing manager so much that as she left to collect him from reception she was heard to mutter “well, we won’t be hiring him…”. Some reception areas get very busy and also seem to double as switchboard operators, so allow 10 minutes for this. It doesn’t hurt to try to get hold of the names of other people you may be seeing.
Set traps Although some questions are set in advance, most interviewers like to also drill down based upon your answers. Thus you should try to mention things that you feel confident in answering hard questions about. This is best done subtly, by phrases like “this is quite like X, but the A large part of interview is to show that your answer is Y,” where X is a bastion of your academic work translates in the ability to deliver competence; or by saying thoughtfully “this something useful. isn’t like X at all,” if you feel you are being drawn into an area where you will sink. The best way if you can do it, is to make sure you sound interested by a new problem, not in despair that you don’t know what is going on.
Show you can do things We mention this in the CV section and here’s a chance to “casually” drop in things you’ve done that show you can dig in and finish the job. It’s OK to mention problems you overcame and the lessons you learned from initial difficulties. Good managers are sceptical of people who claim to glide effortlessly through life and don’t want to be there when such a person hits a rock. Back when Dominic had a proper job, a senior manager overruled him over an interviewee who claimed to have found “no problems” when dealing with a dreadfully difficult piece of work. Not only was the person hired, but three months later had to be removed from the building when he attacked his PC with a screwdriver and his feet. Screams were heard. It wasn’t pretty. That is an extreme case (no Dominic didn’t say “I told you so”(much)), but you do yourself more good by showing how you got through hard problems than by saying you found it all easy. However, it is generally not a good idea to blame all your problems on your previous manager, as we shall see later. If you’re new to finance then you may find it hard to find examples that are directly relevant. It is worth spending a little time trying to find good examples from your academic work especially projects that you have completed. Practical ability is therefore something that you will need t o demonstrate a little more than theory. You wouldn’t have reached this point if you didn’t have a respectable record for absorbing many different concepts. So the next step is to see if you can apply what you’ve learned. When asked your motivation for moving into finance, it’s worth asking yourself if this is true for you.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 39
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Questions for the interviewer
It is a good idea to have a question thought out in advance – it makes you look interested in the position. You have three objectives when they ask if you have questions for them.
Getting the message across A question can be a good way of bringing in things you want them to know, or to emphasise a point you want them to remember. You can ask the importance of your experience in MC, C++ or PDEs to the work you’d be doing. This gets the message across, either as a reminder or to bring it to their notice.
Find out more about the job Good questions are on the direction for the team over the next year and how your work would help them get where they want to be. It shows interest and may give a better insight into what Interviews are a two way process, and a key you really will be doing. Although they are objective is to find out whether this job really is interviewing you, it is also the case that they the one you should take. are selling the job to you, since they want you to accept if they offer. So it’s up to you to work out whether it’s a good job or not. Remember - do not ask things that you should already know, unless this is a ploy to reinforce some point you want them to remember. You should discuss the job and the bank as much as you can with your recruiting consultant ahead of the interview and consult websites and any recruitment brochures. You don’t want to give the interviewer the impression that you aren’t interested enough in their bank to find out about it before the interview. Interviewers often say that this is the thing that really irritates them most at interviews. The same applies to the business area. At entry level there’s at least 50/50 will be asked “why do you want to be a quant in electricity futures?” With that level of probability you should spend a little time on some sort of answer. A smart manager will know that you don’t have a deep understanding of the business he’s spent 10 years mastering, but you should try to make a good shot at it. Where possible it is good to preface a question with a statement about some achievement that the bank is proud of (i.e., talks about at length on their website or in recruitment materials) e.g., “I know your office won the International Finance Press Award for Being a Bank last year, but could you tell me…”
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 40
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Do you want this job?
The third thing to get from your questions to them is some view of the quality of the job itself. You will often hear and read various bits of gossip about firms, but the problem is that the quality of this information is low, not only wrong, but often imprecise. It may not even be wrong. So it may be entirely possible the fixed-income division at a bank to be doing very badly at a time when equity derivatives are making a big pile of cash, indeed there are known correlations based upon the markets that cause such effects. In many ways an investment bank can be modelled as franchise of many businesses sharing a common infrastructure. Thus you will see areas that are great and successful not far from those that are not doing well at all. Your interest needs to focus on the division and the people who you will be working with and who you’re working for. Of course the fact that they are recruiting is a useful positive signal that both the bank and area are in good shape, since many firms react to bad news by slowing down recruitment. If you can, try to find out whether this job is a replacement for an existing person, or is a new headcount, “new” being the better sign. Also it is worth gently probing to see what happened to the previous holder of the position. The press carries little if anything of use to the job seeker in terms of trying to work out the real nature of the bank you are seeing. Obviously if there are lay offs, then it’s a bad sign, but there is simply not the level of detail you need to make a good decision. There’s no magic formula for spotting a pothole in your career, although there are a set of things that you should think about. High staff turnover is a bad sign of course, but if you ask this directly it may offend them and you may not get an entirely useful answer, other than “pretty average.” One candidate we worked with last year failed to spot that they weren’t being interviewed for the job at the bank at all. The team was leaving to join another firm, and thought they would use the time before they left, getting their junior members selected. It is a good idea to ask how long a given person has been with the team. This makes for polite conversation and you will get some idea how people’s work turns out in practice, as opposed to the ideal that they tell you about. As a very rough guide, the politeness we mentioned earlier goes two ways. If the bank is stupidly inflexible over interview times, or you note a lack of buy-in from their attitude, these are bad signs. You do however need to be sure that this is the bank, not the headhunter. Of course if you’re at entry level, you may be stuck in a rigid process as they trawl through what they may feel is an endless stream of identical candidates, so you have to exercise a bit of judgment. Do not forget that interviews are a two way process; they want to hire good people just like you want to get a good job. This means they are trying to sell to you and they are often very sharp people, with strong motivations, so bear this in mind when they are being nice to you. We have noted a reasonable correlation between the quality of the job and being offered coffee or water at interview.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 41
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Appearance It is entirely possible that in your interview process that every person you meet is not wearing a suit, some may not have shaved. That doesn’t always make it wise for you to turn up in “smart casual.” How you look is not a big deal for quants, you’re being paid to think. However, some people do get remembered for the wrong reason and it can undermine your application a little. You should feel comfortable and if that means a bit of perfume or good cufflinks then that’s fine, but see below…
Neatness is good More important than colour of cloth or design of tie, is t he general impression of being in control of how you look. This means wearing it well and being ordered in your appearance. It is worth checking this before you go into the bank. Most banks have a toilet near reception you can use, since you’ve taken our advice and made sure you have time to spare, you may want to check yourself out in the mirror.
Colours Black is the new black. White is nice for shirts and for no other visible item of clothing. Shoes should be clean and preferably black for men and muted tones for women. A particular issue for women is the poor workmanship in most of their shoes. Do not attempt to walk long distances in new shoes that hurt your feet so badly they bleed (we know one person who stained the carpet with her blood). Make sure your clothes fit – badly fitting clothes do not look presentable and if your trousers are too tight you (and everyone else) will find this distracts from the matter at hand. There are some complexions that are generally complemented by certain colours and apparently in some circles “brown” is seen as a colour for your clothing. It is not; it merely says things about you that are never said to your face. Dark blue is good as well. Ties are best boring, novelty is bad. Another reason for white shirts is that they don’t show sweat, some colours do this terribly and it’s not the image you want to project. A good shirt doesn’t crease badly in wear. (Dominic wears Thomas Pink shirts, two-fold Egyptian Poplin, though Twill can look quite good). This is of course a fine example of how you should dress if you want to look like the best known headhunter in global financial markets ☺
Jewellery This will never help you get a job, no matter how expensive or fashionable. Thus if you have any doubt at all, don’t wear it. If you’re female and you have some brooch or bracelet, that’s fine, but there’s no upside for a man at all in bling. Cufflinks of course are fine, as long as they are not “novelty” – you have no idea as to the sense of humour your interviewer may have: he may not have one at all. Some banking people spend quite appalling amounts on their watches, so don’t even try to compete.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 42
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Perfume and aftershave
Feel free to smell nice, but make sure that it’s not too strong. Some meeting rooms have poor ventilation.
Traveling issues One piece of feedback we got from the first edition of the Guide was on smelly candidates. Public transport can be hot and sweaty and suits aren’t really ideal clothing for keeping cool. For a short journey this doesn’t really matter. However, if you’re travelling for several hours to get to your interview, on perhaps the EuroStar combined with the London tube, you can end up with a quite noticeable whiff. This has affected the interview of more than one candidate, so it’s worth doing what you can to keep cool. One suggestion we have if you are interviewing after a long journey, is to carry your shirt/blouse and wear a cheap t shirt to throw away.
Make-up Make sure you are comfortable in what you are
The following is for women. If you’re a male wearing, and aim to be slightly better dressed reader, you really should not be reading this than those interviewing you. paragraph and we are rather concerned that you are. Unless you really never wear makeup, a small amount is a good idea. Again, this gives the impression that you are making an effort and will possibly counter the deadening effect of all the monochrome clothing you are wearing. It should be discreet (i.e. no bright colours) and presentable rather than intending to make you look prettier. There are jobs that you can obtain by being attractive, but they are rarely fun and never intellectually rewarding. Any make-up should always be well applied – if you can’t get eyeliner on straight, don’t put it on and never wear nail polish if there is any chance it will chip before the interview.
Ignore this section But ultimately all this stuff about your appearance is froth not beer. Unless you apply some malign creativity to your appearance, it’s unlikely to seriously hurt your application. We include this section because we see that it stresses some people and unnecessary stress does possibly does more harm. Be neat, turn up on time and you’ve already beaten a few percent of your competitors. Things change when you go for your second job. Often your employer will have a “smart casual” dress code, or maybe no code at all. This means that if your interview takes place at lunchtime or directly after work, you will turn up in normal working clothes, rather than a suit. This doesn’t matter, even slightly, though you should make sure the HH knows this so he can ring ahead and set the right expectation.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 43
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Equal Opportunities
Banks really hate being sued for discrimination, it costs money, ties up lots of resources and looks bad. Thus all banks impose on their headhunters contractual terms that say they should not do this sort of thing. This means that some banks will send you an email directly asking you to disclose your sex, race, age etc. This is not something to be feared, it will not be seen by the hiring managers, or indeed your headhunter, and is there so that the bank has some sort of statistical view of the diversity of both applicants and those who get offered jobs. At one market meeting where a large bank explained that it was now adopting this approach, some HHs who dealt with candidates from outside Europe and N.America were concerned that their people might be quite spooked by this, since in some places this sort of information is gathered specifically for the purpose of discrimination. We are not so naïve to belive that big banks don’t ever discriminate, but they are smart enough not to set up a large publicly visible system to do it. P&D has two equal opportunities policies. One is a rambling, worthy and very dull document, lost in some subdirectory of Dominic’s laptop. The other is our simple view that we do pimping for the money. Even if it were legal, discrimination would cost us money since we’d not be giving each vacancy our best shot, and thus lose the competition to fill that slot against some other HH who is more objective. We see losing money for that reason as immoral.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 44
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Understand the process
Interviewing people is a major industry all by itself; multiply the number of applicants by the number of interviews they attend and you sometimes wonder how any useful work ever gets done. Certainly this thought occurs to interviewers on a regular basis. They want it to end. Although it is important to get the right people almost no one enjoys all of the process. This is made worse by the fact that more than 80% of the work is wasted on those you never hire. The record (as far as we know) is one bank who interviewed thirty seven people before they called us in and by the end they were getting more than a little bit tired of it all. Some firms like DE Shaw seem to interview quite literally hundreds of people for every one they hire, on a different extreme GS interviews fewer, but favours a lot more interviews for each. In a different direction, we “adopted” one candidate from the forums who was suffering from a rather longer than normal process from application until start. Seven months in total. He got there in the end, although at more than one point it did look bleak. That is the current right hand Try to find out from your headhunter how many side of the distribution, though we now have interview sessions you will be expected to attend. set up a couple of students with research sponsorship that will lead to a job nearly 18 months after we first talked to them. This means that although you may feel uncomfortable in the interview, it’s worth remembering that the interviewer might well want to be somewhere else as well. Anything you can do to make their life easier is helpful. Not only does that mean being as pleasant as possible, but also turning up at a good time and seeming like the sort of person they want to work with. Most firms are less extreme than GS or DE Shaw, but you must still expect to be interviewed by quite a range of people.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 45
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Telephone Screens
Mobile phone interviews We’re old people (>35) and thus sometimes use quaint ’phone technology which involves long copper wires physically connecting us to a huge ancient UNIX computer miles away (yes, we still use miles). A typical quant has done enough physics to know that you can actually talk down metal wires rather than a 1 mm thick cell phone that has more processing capacity than its owner. Sadly, the quality of cell phone speech is hideously degraded and on many systems you can’t both talk at the same time. This is occasionally awkward when both speakers have the same first language, but if both have English as second language neither comes out of the conversation impressed with the other. The worst case is where the interviewers are sharing a speaker phone to you on a mobile, which is better imagined than experienced. Although a large % of all the mobile phones are made in China, this technology is particularly unkind to native Chinese accents, so try to get it on a land line; this matters.
We have found that most telephone interviews Mobile phones not only reduce the quality of the are to screen you to see if it is worth bringing sound, but also the way the interviewer perceives you in and consuming the time of several your quality. Try to use a land line. managers. Thus they tend to optimise for “selecting out” rather than finding your good points. In our experience this means brainteasers are easily the most likely thing you will encounters since they eliminate a wide range of people quickly, and yes this can be unfair and error prone, and no, this is not the only time we see this in careers process. Some brainteasers can easily be given over the phone, and so you can be expected to be thrown this sort of logic/lateral thinking and math type problems. Some firms like Credit Suisse screen a large % of those who apply at entry level by sitting standard examinations.
Take Home Tests You may be sent an exam to do before you pass on to the next stage. It is certainly pathetically easy to cheat on this, but that’s not the point. If you can’t get through these questions then you will be chopped up by the formal interview anyway, so they have great value for the bank as self selecting filters that involve them in no effort.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 46
Guide 2.0
Paul & Dominic’s Guide to Quant Careers First Interview
The bank likes you enough to spend time and effort on you, and at this stage you will be usually given the harder maths, programming and finance questions.
What you need to prove •
•
•
•
•
You are smart. This will often take the form not only of maths questions, but also brain teasers. Your CV makes a lot of promises, so now you need to deliver on them. It follows that anything you put on your CV is a potential source of interview questions. But they will ask questions that your CV shows you can’t do. Life is not fair, nor is investment banking. Sometimes this is done on purpose to see how you will react. You can work with people . Smart is important, but you can expect to spend more time with the people you work with than who you sleep with, so they would prefer to hire someone who has decent social skills. Most of your academic work has of course been on your own, so this is an area where many entry level people can seem weak. You can get things done. In exams you can get good marks for nearly completing a problem, or having the right idea and making a small error. That doesn’t work so well in finance where “nearly right” can be really quite unpleasant. In writing your CV, it is worth sprinkling in a small number of things you finished that worked. If possible refer to this sort of thing in your interview, as examples of your ability. A common failure mode in entry level. You can manage yourself and your time. You’re not a factory worker assembling Barbie dolls. Your output cannot easily be measured and good managers don’t micro manage quants. Note that we do not say all managers are good, hence the “My Manager Wants Me Dead” series on Dominic’s Blog. Being able to point to examples of where you worked alone without guidance or invasive management will help your case greatly. You are committed to this line of work. This may sound like a strange thing to have to demonstrate. But more than once we’ve had feedback from managers that the candidate seemed to have drifted into applying for a quant job with little idea of what the work was like, or what they could offer an employer.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 47
Guide 2.0
Paul & Dominic’s Guide to Quant Careers How to prove it
Hopefully they’ve read your CV, but this is only a very short summary. So as well as answering questions, it is good to be able to show examples of where you’ve applied a skill, or where you’ve coped with a particular situation. This bit of preparation before you go for interview is well worth the time, and does far more for your chances than trying to cram in bits of maths and payoff diagrams the night before. Throughout the guide we emphasise the competitive nature of the market these days. When Paul Wilmott worked on starting a university mathematical finance course it was rare for a university to have such a thing. Now it is rare for them not to have at least one. As far as we can tell, the University of London alone has something like 20 separate Masters in Finance programmes. Emanuel Derman’s autobiography My Life As A Quant is compulsory reading for anyone wanting to move into this field. The market for quants is of course much larger than when Derman Examples from experience are a fine way to show started working as a quant, but the supply side you can do things, and most of those who have has increased at least as fast. Thus you must had any training at interviewing staff tend to not also show yourself to be smart, but also actively look for these examples. more useful than the others going for the same job. The market is currently still relatively good (Q4/07), but if you’re reading this as a student, it might not be as good when you hit the street. Even if it remains in your favour, the best jobs are always going to be harder to get. Transitions You need to be able to explain why you made various choices in life. Why this subject, why that university, and of course why quant finance. If you are moving away from your existing bank then you absolutely must know what you are going to say about your reasons for leaving your current firm. We handle some of the bad reasons people have for getting away from their boss.
2..Nth Interview Quants tend to have more interviews than most other types of staff because their work brings together more disciplines, but also because you may be working across more than one area controlled by more than one manager. So you next few interviews will be a mix of making sure you are good and that the right managers “buy into” choosing you to hire. It is far from unusual for them still to be considering whether to hire anyone at all. This reaches a maximum at Goldman Sachs who interview staff far more times than anyone else. We hear entertaining stories of headhunters who keep telling hapless candidates that they only need to see one more manager and all will be well, and that the process will be over soon. GS moves at the pace GS feels like moving, and any HH who tells you otherwise is not being honest with you. One side effect of this is that you will get offers elsewhere before GS makes up its mind about you. It is quite possible that through the headhunter they will send the message that you ought to turn down the other definite job offers so that you have even a slight chance of working for them, and rarely will they move faster to get you once they learn of competition. So on balance it is not a good idea to tell them of other offers in the first place.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 48
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
How interviews go wrong Zeroth law of holes When you find yourself in a hole, stop digging. You will be asked questions for which you can’t think of any answer at all. Some interviewers make the questions harder until that point is reached. The trick is to cut your losses. With any luck they will just move on, unless it’s a critical topic. Of course if it’s critical then it’s game over anyway. What you must avoid is wasting time wandering like the lost spirit of ignorance over a vast formless expanse of your own incompetence. A good response is to look them in the eye after a little thought, then simply say, “Don’t know, sorry.” The exception to this are the “all the tea in China” questions where you are asked to Almost all interviews go wrong at least a little bit, estimate some quantity like the number of bull some managers even cause this on purpose. testicles consumed by McDonald’s customers Almost no one ever blows a job due to one per year. You aren’t expected to know the mistake. Remember this when you make yours. answer to these, indeed knowing it would seem rather strange. They want to see how well you can estimate an unknown quantity and how you think. But the biggest hole catches people who get very nervous when things go wrong. This is about the most negative personality defect you might have in a bank. When you realise you’ve said something tragically dumb, stop. Then say something like “let me think about that for a second,” and correct yourself. Make the pause work for you. Think the answer through and show that you are capable of recovering. Remember that no one can talk about things at the edge of their competence for five hours without saying something silly. You don’t have to be defect free, but self knowledge and recovery from a slip will score you vital points.
Sleep regularly, sleep often Probably the most common error we’ve seen is not getting enough sleep the night before. As we said earlier, the difference between you and your competitors is tiny and losing a small percentage of your thinking ability through being tired has a non linear effect on your probability of getting a job. Hours in a bank can be quite hard, so it’s really not a good idea to mention feeling tired as an excuse for a poor answer. Not only will they not be impressed, but if you get drawn into a conversation about how it degrades your performance it won’t end well.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 49
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Barbarians
The word barbarian comes from the ancient Greeks who took anyone who didn’t speak Greek as making “bar bub bar” noises, like a drunk Homer Simpson, not barbarian as in the icy commanding tones of Governor Schwarzenegger. Although Dr Simpson has enjoyed careers as an astronaut, rock star and nuclear engineer, few of us would hire him as a quant. It’s important to get the right balance between gushing at people so fast that they have trouble following you, or being too quiet. You should try to practise looking at the reaction of people talking to you and if the interviewer is clearly trying to move on, you usually should let them. If you think of the conversation style used when first meeting someone you find attractive, you won’t go far wrong. (Just remember it’s a first date) It is also the case that no one wants to discriminate against those who aren’t English speakers. This is good, but means that if you aren’t understood they may just skip over what you say, rather than pass comment on your accent. This is especially true when having a telephone interview where you will not get visual feedback and the sound quality is degraded.
Read your CV We lied when we said no one would read your CV, you will. Read it to work out which questions it might provoke them to ask, “why did you pick X?”, “I see you’ve done a lot of Y, here’s a hard question about it.” This applies even to things you have done years ago, and mostly forgotten about. We are often asked “what should I read for my very important interview on Monday”, and a good answer is to make sure that you quickly revisit things they know you have done, to make sure you can make a stab at answering questions. Upon-reading this guide we see that we’ve said this more than once, it survived because it is so important, and we wish all holes were so easy to avoid falling into.
Make eye contact You need to make sure you look at your interrogators, they can smell fear. No need to stare at them, just remind yourself to look at them when they or you are speaking.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 50
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 51
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Which PhD should you do ? We don’t know. Recently Dominic has become convinced that he is wise in all things. From high energy physics, through numerical analysis, post-Black Scholes pricing, neural networks, econometrics, compiler design, copulas, cutting edge string theory (how do you cut with string?), the formation of galaxies, physics from solid state to plasma, and the life and times of the electron. He is apparently not alone in this view of his good judgment. Almost daily, people contact him seeking advice on how they should proceed in their research in these fields, as well as many others. Recently Dominic has expressed important views on the future direction of molecular biology and climate change. It is now certain than that soon he will have advised more people on Choosing a PhD topic purely on the basis of how you think it will help get you a job is a tragically their PhDs than any person now living. bad mistake.
If you are one of those who have had the good fortune to meet with Dominic, you will not find it in any way strange that he has no problem providing this advice to any who seek it, but one is tempted to ask (quietly) why the advice of someone who has won relatively few Nobel prizes is so sought after ? It is of course because a PhD is now seen by many as a step to getting a good quant job, and so it is rational for them to ask for advice on which topic of research has the best prospects of helping their careers in a bank or hedge fund. Many of the most successful quants have PhDs, and a large % of job ads ask for one as well, so it is tempting to think of a PhD as part of your career plan. We are seeing that as an increasing trend as more people become aware of quant finance work earlier in their careers, and understand that the choice of PhD topic can seriously improve your chances of getting the best job, and the set of employment choices you can make when you have completed it. There are both real and imaginary conflicts between doing a PhD “well” and steering yourself towards being an easy sell to employers in banks and hedge funds. The nature of PhDs should be to do something new, and by definition that means we can’t say “fluid dynamics is good” or “labour market analysis is bad”, though to some extent we do give that opinion that to people when they ask. Creating software models of large complex physical systems is good, but the current market does not require that you model financial systems, indeed some managers have expressed the view that finance PhDs often are less rigorous than physical sciences, even if they are more focused on finance.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 52
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Market risk
The simplest form of risk in any market is where the value of an asset goes down without you having any hedge or diversification. Current conditions (Q4/07) are a useful example of this. The credit sector has been a major source of lucrative employment for quants over the last few years. As this sector grew, it was quite noticeable that relatively relatively few new PhDs were entering the market having studied the relevant topics, and these entry level candidates found themselves sought after. This information fed back into universities, and thus the number studying some form of credit has gone up substantially. Sadly they are hitting the job market just as this sector is having a bad time, indeed since this is the first time this market has experienced substantial turmoil the people alread working there had come to believe that this would never happen to them. Thus putting a PhD into finance terms, it can be a substantial undiversified, unhedged position, with a serious time lag between getting the information that drives the trade and the payoff. This means that great care in choosing a topic can still fail because the market moves in the It is worth doing some market research on how wrong direction.
Choosing a supervisor
your PhD topic and choice of university is i s likely to be perceived.
Of course to a large extent, you don’t choose a supervisor, they choose you, but you can make some effort to get one that will help you in your career. This requires a little care since if you are smart enough that they want you, they may well want you to follow them into academia. A “name brand” supervisor has value in addition to the expectation that he knows the subject rather well. There is a virtuous circle that he is more sought after to supervise, and thus if you score him you have demonstrated that someone wise in your field thinks you have merit beyond the sterile numbers of exam grades. He will more often have good networks of contacts useful for internships and first jobs. This makes him still more sought after. It is the case that at P&D we occasionally get directly approached by such illuminati which tends to commend more of our attention which is of course good for them. As headhunters we recognise that having a hook to sell you to the bank can make the difference between getting an interview i nterview,, or not. That is very useful, but once you are being interviewed it does not get you anything further, and of course you have set a high expectation. The possible downside is that some PhD supervisors are frequently travelling, sharing their wisdom on important committees, or engaged in writing yet another important book. This can mean that you do not necessarily get very much of the great one’s time. This is not restricted to the great and the good, and some supervisors do not engage as directly with their students as they might. Thus if you possibly can get to meet with his current students you will can research this with a little care. There may also be a trade off to be made between the supervisor you want, and the brand of the university, which leads us to…
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 53
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Location of universities
This is very much a second order issue. It is useful to be in or near London, Chicago or New York because you can network with people in finance, and attend seminars and other useful events. This applies even if you are not studying finance, indeed it is slightly more important for people making one of the standard transitions into finance than for those directly studying it. But to be of any value, you do have to put in the effort to go to these events and networking opportunities, since most won’t come and find you. We do meet people at London events who’ve come from Amsterdam or Paris, but most stay where they are put. There is more quant work than you might think on the West Coast of the USA, but one is not really part of the scene there. Given the choice between two equally attractive programmes, you should optimise with respect to distance from one of the big three centres, but do not pick an inferior option just to be near the right people. If you are serious about this line of work, and prepared to do a bit of travelling, In the years before you finish your studies, do try go where you will get the most value added. and make the effort to attend some of the many finance events to help you learn where you should be focusing, and to make contacts.
An advantage of big universities is that banks will do campus recruiting for people like you, and you can get exposure to fields that are useful even when not contained in your narrow field of study.
As headhunters we often see these job specifications that say “PhD from top university” but we all know that many smart people are not at the “top universities”. This is especially true at PhD level where a leading researcher may not be at a leading university, and so his students don’t get that label. You will see job ads that say that they want graduates of a “top 5” university. We have seen demands for the “top 7”, “top 8”, “top 10” and one day we are certain to receive “top 13”, it may even somehow be fractional. These are a sign that something has gone at least slightly wrong with that firm’s recruitment process, not that it has high standards. Does “Top 5” (or Top 17, or whatever) mean in the US, or in the World, or in the UK ? We are rarely told, and interpret it as emphasis rather than a formal specification. Do you feel able to compare universities in Argentina with those in Poland or Canada ? There are many rankings, and include the universities “achievements” in dead languages, feminist poetry, astrology etc, et c, so that most of these lists have no real utility at all, but it helps the more lazy people in the process do less work. But “branding” does have value, since on average a more prestigious place place is more more likely likely to have smarter smarter graduates. graduates. A PhD is a fine shiny badge to wear, and will help you get to many of the places you want to be, but it is not not a risk free option, and it does matter matter what you do.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 54
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Failure modes
You may be assuming that you do finish the PhD, which doesn’t always happen, which can make focusing on its career implications a dangerous way of thinking about your future. PhDs are quite unlike taught degrees. You will be working to a much larger extent on your own, and pursuing targets that are far less structured and well defined than the courses you have taken so far. If the topic does not really interest you, or is chosen on some ill conceived conceived notion of what will get you get a job, there is a good chance that you will hit a vicious circle of low motivation leading to slow progress which causes you to lose hope that you will ever finish; resulting in a nasty cocktail of burn out with little learned or discovered. At interview, this may express itself as an “attitude problem”, or simply simply not being very very smart. A depressing depressing % of PhD candidates candidates quit, and that that rarely looks good on the CV. However, some employers see your PhD as a form of filter, since to get into a good programme you have demonstrated some extra level of being smart and accomplished, and we sometimes hear that they don’t care all that much about whether you finish. But the odds are not in your favour, and having a recoverable crash that consumes two or three years of your life is not anyone’s idea of a Financial Markets are in constant change. change. A topic great plan. Thus by far the most important that appears exciting and lucrative today may thing in this section, and quite possibly in well be yesterday’s news in 4 or 5 years when you this whole book is to help you decide to do finish your PhD. something that you genuinely find interesting, and which will engage you both at an intellectual level and attract some passion, since you must at some emotional level genuinely care at least a little about your work. Seeing it as simply the longest homework assignment of your life is buying years of misery, and quite possibly a less satisfying career at the end of it. If all that does not convince you, then it’s pretty clear to us that you can expect to earn less money. Which as henchmen of the global finance community we are not in favour of, one little bit.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 55
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Physics vs finance
A common dilemma is whether to change track from physics or engineering to finance for the PhD. We have yet to see anyone go in the opposite direction, which does contain significant information, as does the fact that some hiring managers actively prefer physics over finance PhDs. But again there is the pattern that the headline topic is less important than how you do it. A common issue with finance PhDs is quite the opposite of what you might at first expect, in that they often seem to be further away from use in banking than physics or maths. Often there is a striking lack of any attempt to relate the theoretical model to observed movements in the market, and often we see people who have simulated markets to support their research. One reason is that financial data can be expensive, and often requires work to turn it into something you can use. But this work itself makes you more valuable since it is one of the tasks you must do as a quant in a bank, and a critical part of being a successful quant is bridging theory and reality. There is nothing wrong with simulation, and getting to the position where you understand any system Make sure your field of study has a broad base of well enough to simulate how it may behave is maths, not just those to support the thesis you valuable, but it cannot be a completely ware working on. adequate substitute for feeding real observed data into your model. Managers have told us they feel a mixture of sadness and frustration when an obviously smart newbie explains their work but has made no effort to work out whether the model is in any way right. When choosing a PhD topic you should do everything you can to ensure that real data is available to you, personally to validate and calibrate your model. It is not enough that data exists somewhere, you must be clear that it can get to your computer. Since such a large % of working quants have a background in the physical sciences, it should not surprise you that by now nearly every mathematical and algorithmic technique used anywhere in that field has been thrown at finance to see if it is useful. People with a good physics background thus tend to have an understanding of a larger subset, then people with pure finance. There are some large holes, typically in statistics where physicists are sometimes not only ignorant but quite proud of not even grasping the basics beyond calculating the mean, and seeing GARCH et al as for people who don’t understand what is going on. Not all types of physics are in any way useful, and without doing some serious investigation, it is far from obvious which will help you. Thus is if you are in this field, at the earliest stage you must start trying to think how to apply techniques in a more general way. That is not to say that if you feel a technique is not useful, you should bypass it since often it is only when you really understand a formalism that you can work out how to use it in any situation other than the carefully set up framework of a lecture and exam question.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 56
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Finance PhDs have the advantage that most things you will learn will be set up in the context of money, and that you will pick up the motivation behind the processes. It is critically important that you achieve the right level and direction of focus if you are to make the most of this. It is necessary and useful to get deeply into your specific topic, and balance this with the wider context and a general literacy. Some of the feedback we get from hiring managers is that the general knowledge of finance they see from someone who has 3-5 years of specialist study is far lower than their expectation. That means exploring digital options even if you are doing credit, knowing the growing importance of volatility as an asset class whilst studying market microstructure, and even when there is no apparent connection between them and your own topic, you should try to see if there is any way they can be used either to help you model, or to turn it into a viable trading strategy. This means looking at these extras for hedging, or putting limits on your exposure. This is why articles on ideas like using FX as a predictor of sovereign debt default, or linkages between credit and equity markets should be part of your normal reading if neither asset class are in your area. Applying ideas across markets is one of the things that separates a respected quant from an Excel jockey whose working life is looking after broken risk report spreadsheets.
©2005-2007 Paul & Dominic Ltd
There is a huge delay between choosing a PhD and looking for a job. Your career ought to last decades, so make sure that even if you are doing a cutting edge highly sought after field, that you hedge with skills that have longer term value.
www.PaulDominic.Com
Page 57
Guide 2.0
Paul & Dominic’s Guide to Quant Careers So what is the right question ?
Although it matters what PhD you do, a more important term is how you do it. There are elements that you may be able to include that make you more employable without making your PhD in any way less valid. Elsewhere in the guide we list the maths, finance and programming that you ought to learn, but you will note that none of them are tied to finance as a unique discipline. Nearly everything in QF has been stolen from other subjects, so it is entirely reasonable to get these skills in whatever form best suits you. Thus the right questions are at quite a low level of granularity.
Above all to your own self be true. What are you good at? We don’t mean the large collection of exams you’ve passed by now, but what do you delight in doing ? What do you outclass your peers in ? What can you do by brute force and hard work ? Those are the sort of components that a good academic PhD and The content of the PhD is more important than one that will help your career are made precisely which topic you choose. Use the right from. There is inevitably a trade off between methods and learn the right techniques and you your comfort zone and what will earn you a will beat people who do more obviously relevant living, but it does not have to be all one subjects. sided.
Make sure you have the full skill set In many universities you are free to sit in on courses without taking the exam, and this may be very useful for helping you adjust your PhD study to be more useful to your career. However this is not always the case, and we find that often the finance courses are a great money spinner, and so they will be off limits. This is not always fatal, since most of the maths can be got from other courses which aren’t so in thrall to accountants. In particular, the programming on most finance masters programmes varies from the mediocre to the rather funny, so you may do better by getting this elsewhere. That often does not include CompSci departments who usually have lost the plot and will try and foist Java upon you. It is worth checking what the policy is on auditing courses, as well as the attitude of your supervisor. Make sure that your full skill set makes it to your CV, especially if it is not obvious from the title of your PhD.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 58
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Good things to do in a PhD
We are not suggesting that you create a PhD topic around these following points, but instead when you get a choice to do something that will help you get a job in quant finance these are the classes of experience that you might choose.
Variety in techniques As a quant you will rarely have the luxury that the problem you are working comes with the name of the method you should use on the label. A good quant PhD works with lots of different mathematical techniques. Constructing PDEs to describe systems is of course good, as is use of finite difference and Monte Carlo to solve them. But any analytical technique that models things that happen is of at least some value in finance as are copulas. Thus econometrics is good as part of a time series analysis portfolio of skills, and signal processing methods are showing good demand.
Depth of technique Conversely, managers recruiting for the best jobs like to see some degree of excellence, the distance you get ahead of others may often b more important than its direct relevance to finance.
If you can show that you have both developed a mathematical model, and made a worthwhile effort to make it consistent with real experimental or market data then you will outshine those who just do lots of equations.
Real data Over the last couple of years we have seen an increase in the demand by employers for entry level people who have “got their hands dirty” with real data. Another word for “real” might be “raw” data that has not been pre-cooked to make your life easier. Real data has many defects, not just noise, so the more scars you can show from this, the more in demand you will be.
C++ If you are going to be doing coding, then it should be in C++ if at all possible. You should try more than ordinarily hard to make sure there is some programming in your work. First because it will help you get a job, and second because you need to find out whether you can actually cut code because some very smart people simply can’t. Given that quants do a substantial amount of development, you don’t want to find that out once you’ve burned the bridges for staying in academia.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 59
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Bad things to do in a PhD Master of the big machine
PhD students often become wise in the ways of ultra sophisticated technology used in bashing particles together, heating or cooling stuff to unreasonable temperatures, or the super powerful computers used to try and work out what happened when you did these things. The important separation here is whether you are an engineer or mechanic. The huge magnets at CERN require exquisite physics to function, involving lots of problem solving and difficult mathematics. Writing programs to interpret the data, removing “bad” data points, detecting structures and patterns in real experimental data, is good especially if you run your software on high end clusters. However, running these clusters efficiently is also often a task requiring smart people, and the same thing applies to the vast array of arcane equipment that supports modern science. It is not actually that surprising that the World Wide Web came out of CERN, Try to build generally useful skills, and avoid as given their requirements for high end data much as you can the intense detail of packages or management. systems that you happen to use in your studies. Mastery of these devices is heavily fact based, in any system there are things you just have to know, rather than work out, but beyond that particular project, they are of zero value. That problem applies to other parts of science, as much as the rather larger jump into banking. For instance Beowulf clusters are used in banks, but the maintenance of them is rarely a dream job. If you fall into technology maintenance the best path for you is that your job is outsourced to Bangladesh. If you’re very unlucky, you will be sold like a chattel to a firm like EDS who has successfully sued many of its former employees who had the temerity to leave for more money. Bonuses ? Oh how we laughed. Avoid maintenance, no matter how impressive or important the machine. If you have some emotional need to look after something, have kids or cats. Cats are better, they don’t die if you leave them on their own for the weekend. Thus during any PhD project, you should have a look at your portability, even if you think you will probably stay in science. Projects end, it is the way they are. We guess the average age of Guide readers is 26, implying that your career should be 40 years long. A big reason for short careers is tying yourself too closely to a project that dies.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 60
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Software packages
The biggest issue with most packages is that they work too well. This may come as a surprise if you’ve used some maliciously buggy tool without documentation that your supervisor has foisted upon you. But as we see later with Monte Carlo there is a trade off between getting results to publish and learning things that help you be a better quant. The second problem is related to Master of the Big Machine, that they contain much that does not teach you principles of deep things, or have any relevance outside your project. However, if the package comes with source code, like many in academia, then you have the opportunity to work on a large buggy program. This is better than it sounds. Really, honest. As long as this does not take up a huge % of your time, it will both make you more attractive to employers, and more useful once they have hired you. If you do this, be sure to include it on your CV, because “optimised and fixed bugs in X”, looks very much better than “used X”. It follows that if you get the choice between open source tools and proprietary, you will get more out of using one whose workings you can follow and learn from. It is important to make sure this message comes across as showing your ability to overcome problems, and not as complaining about the toughness of the work. But we would emphasise that this should be a useful component of your work, and the core. This approach of course has a price, since tools like Matlab have a rather better user interface than their open source equivalents, so you will spend more time fighting their strange ways. The optimum is to use some of both. It is worth looking at job ads to see which tools are used in finance, and if you can use them in your work. We notice Matlab and Root falling into this category, however we would emphasise that nothing beats learning C++.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 61
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Why we hate Monte Carlo. A lot… Pretty much any numerical problem can be solved by MC, even deterministic ones that on the surface seem to have no relationship to randomness at all. We know that some equations are extravagantly hard to solve, some have no analytical solution at all, and many just look so ugly that you reach for you random number generator as a reflex. You can guarantee a result that you have confidence in, which can be vital if it is standing between you and finishing a project. The problem that we have as headhunters with MC is with precisely that power, because like any other power it may corrupt you. Since you can use MC for any problem you may be tempted to use it for every problem. Some people we talk to seem to have used it as a way of avoiding understanding the thing they are investigating, and relentlessly bashed it with a vast horde of random numbers until it surrendered. As a tool for getting to a new position to look at something MC is fine, but when it becomes the focus of your work, it may not be clear that have really understood anything. Also of course, if you don’t have any idea what’s really going on at all, an MC simulation of the physical process might give you insight, or at least let you feel as if you are making progress. However it is a common component of the Value at Risk (VaR) of most large financial firms, which perhaps gives you some idea of how risk can sometimes be a very black art. MC is often quite slow, and whereas in academia, it is perfectly acceptable to have a result that takes a week, in a bank it is very rarely so.
Neural networks, genetic algorithms There are several non-analytical techniques which although distinct from Monte Carlo share the risk that if misused deliver a result, but without opening up your understanding of the thing you are studying. Their ability to hack through problems that previously defied solution is acknowledged, but at the price that their solutions, although optimal may be brittle, and since there may be no way of knowing “why” it picked that result, or even framing “why” as a sensible question, the results are less trustworthy. There is demand for these skills, but for these reasons they are less common than you might think. If you have already committed yourself to holding these skills, then you need to arm yourself with good techniques for validating the results. There seem to be two main failure modes for NN and GA practitioners. From earlier you will have seen that being a user of a package is nowhere neat as good as someone who contributes to it’s development. This applies strongest to NN and GA systems, particularly if you are from a CompSci background, since one of your selling points is supposed to be a superior programming ability. Skilfully driving a package someone else wrote, really does not count at all. Thus you should get yourself to a position where you can contribute to the development of these packages, even if they are not written in C++.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 62
Guide 2.0
Paul & Dominic’s Guide to Quant Careers The good side of Monte Carlo
It doesn’t have to be as bad as we indicated earlier. Monte Carlo can be a rigorous field, with a range of techniques that give better and faster results than a simple, big loop and averaging the result. In our reading list we have two good books on the subject. What makes us sad is talking to entry level people who’ve apparently used MC at considerable length in their education, yet to whom the terms “variance reduction”, “antithetic variables” or “Mersenne twister”, are unknown, or worse still are seen as advanced topics, far beyond their comprehension. This does not just apply to physics PhDs who take a MC shortcut to solve a gruesome numerical analysis, who at least have the excuse that it was a sideline but to finance PhDs who simulate markets rather than work with real data from a price feed. We see some people with specialist masters in finance degrees who have no idea at all about efficient MC, and that is frankly shameful. Monte Carlo is by far the easiest class of algorithm to spread over multiple CPUs, and so all the major banks have substantial MC farms. Understanding of this is useful, but again be careful not to get too sucked into operational rather than algorithmic or analytical work. It was Phelim Boyle, then an actuary, who first proposed using MC for finance problems. That was back in the 1970s. Professor Boyle has written a book on the history of derivatives, with his son as coauthor.
©2005-2007 Paul & Dominic Ltd
If you are going to put Monte Carlo on your CV, you should be able to answer questions on antithetic variables, importance sampling, variance reduction, and why the RNGs in most standard libraries are garbage.
www.PaulDominic.Com
Page 63
Guide 2.0
Paul & Dominic’s Guide to Quant Careers What you should know about Monte Carlo
When should you use Monte Carlo, and when should you choose some form of Finite Difference method? This is actually a mildly common interview question even though it should not be, since it is pretty close to trivial. Given how basic this is to the whole field it is scary and depressing how many people stare blankly at the idea. No, we’re not going to tell you how. There are a couple of relatively simple answers, and if you don’t know, then you need to hit the books again, and/or send a nasty email to the person who taught you numerical methods. In the main book list we suggest the two best technical works on Monte Carlo in Finance, one by Peter Jaeckel and the other by Paul Glasserman Here are a few key points: •
•
•
•
•
•
•
Monte Carlo uses random numbers to solve problems. Usually we do something lots of times with random, random numbers and look at the statistical properties of some end result, such as the mean of a function of all those random numbers
Do not fall into the trap of thinking that because a simple MC technique can solve most numeric problems, that you do not need to understand the depths of the technique.
The problem under examination can involve probabilities, in which case the role of random numbers is rather obvious, or may be deterministic. It’s not immediately obvious that there is a connection between deterministic problems and random numbers but there often is. We see this in finance where the Black-Scholes model for the value of an option when the underlying asset moves randomly can be interpreted as a a partial differential equation for which there is only one, deterministic, solution. Monte Carlo is usually the easiest to code up for any given problem. However it often is not the easiest to code well . If you are trying to find the statistical properties of some function (or functional, look it up) of the random numbers or asset paths then the value of statistical property in question will depend on how many random numbers or paths you generate. The more, the more accurate your estimate will be. The rule of thumb with simple random numbers is that the accuracy is of the order of one over the square root of the number of random numbers/paths. So if you want to halve the error to improve accuracy it will take you four times as long. MC can be a bit slow to converge for simple problems. For the simpler derivatives valuation problems you may be better off with the harder-to-code-but-faster-to-run finite difference methods. MC really comes into its own when you have a lot of complexity or multiple underlyings to simulate. Some financial problems can be recast as integrals that you have to evaluate. Again the rule of thumb above applies to basic MC. However…
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 64
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
•
•
There are some MC techniques that take advantage of special properties of numbers which look random but are not. Confused? The subject of quasi MC or low-discrepancy series is that of finding sequences of numbers that when looked at casually look random but are structured so as to have statistical properties that make any calculations using them converge much faster. Random numbers generated by classical MC means will result in clumping of numbers (too many are similar) or barren zones (some regions will not have any numbers), that’s just the nature of randomness. But the low-discrepancy numbers are designed so that they fill out the required space in a way to avoid such problems. The benefit is then seen in convergence to a solution, usually of the order of one over number of paths now, much faster than standard Monte Carlo. You should understand that quasi is not the same as pseudo, when working with sequences of generated numbers. The standard random numbers
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 65
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Choosing a Direction You need to work out how committed you are to a particular sector of the market. Some people find that they prefer fixed income to equity products, others feel that commodities offer the best opportunities for them. Some managers feel that you should decide at this early stage which product class should be your home. This is frankly a bit unrealistic. A newbie rarely has the information to make that sort of decision, and of course anyone who’s smart enough to predict the relative success of several business areas of the next few decades is going to end up ruling the world anyway. But just because it is not a reasonable thing to demand, doesn’t mean that some managers don’t think that way. You don’t have to make up your mind yet, but when you accept an offer you are making some other doors quite hard to open in future. It is possible to move between the main types of business area, but not always easy and harder still to do it from a position of strength in your later One of the industry standard lies in recruiting is career. Thus getting a job is a competitive that if you take this unattractive job now, the bank international sport, where it is not enough to will recognize your potential, and move you in the succeed. Others must fail. right direction in a year or so.
Some people do have an excessive preference for a given bank, often Goldman Sachs. This is of course a fine ambition, and they do make great big piles of money. But that does not mean they want you, and discovering this when you’ve spent a lot of time on getting a job there, but made no attempt to hedge that position can have mildly unpleasant effects on your career. So the best job for you may not be at your target bank. They may offer you a role considerably below your level. Of course it may be worth your while working in a lower quality area to get into the one you want, but be aware that moving areas in a bank is really quite hard. If you’re good your current manager does not want to lose you, and if your unhappiness with your current role impacts your work, you’re not that attractive to your target manager. Some managers are entirely selfish on this point, and will use the bonus mechanism, and anything else they have available to keep you in your current position. That can include killing your bonus for that year if they think you will leave anyway, and using the money to make your colleagues happier. Bottom line is that if it were easy and efficient to change roles within a bank without having to leave, there would be a lot of poverty stricken headhunters, and Dominic would not have been able to build a castle for his sons. Also, you don’t know how good you really are, most people have some rough measure, but given that you’ve not met 99% of your competitors, it’s hard to judge, therefore our advice is to apply to roles that you think may be just out of reach, the cost is low, and the upside is good.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 66
Guide 2.0
Paul & Dominic’s Guide to Quant Careers When to Apply
The time to start the process is as soon as you know roughly when you are going to be on the market. Sooner is almost always better than later. Our current record from receipt of a candidate’s CV to a job offer is just over a week, but we also set up one quant in employment more than a year before they expected to start at the bank. These are extremes, but they illustrate that the process is flexible, yet hard to timescale with any precision. An increasing number of firms have structured training programmes for quants. Because they have fixed dates, a late application means you either have to wait, or miss out on joining the programme. These should not be confused with the vastly more common graduate training given to other types of workers in the bank, and it is entirely possible that your application will end up there, and thus getting you an interview on a help desk, sales, or in one extreme case interviewed for manning a counter for retail customers.
Apply before finishing your PhD Masters programmes typically have a very fixed end date, but most PhDs do not. It is also the case that the final PhD write up is observed to be at least a little non-deterministic. It is best to finish your PhD before starting work and not even slightly a good idea to try working on your write up and do a hard new job at the same time. Apart from the stress of serving two masters, a critical thing to your career is the early impressions your employer makes of you. The quality and type of work you get given is a function of what they think you can do, so you want to be 100% committed to doing early work well. This conflicts with the “finishing touches” to your PhD and given that your first job may well leave you tired, the PhD will drag on for longer, all the time reducing the impact you’re making on your new boss. Employers understand the sequence of events around a PhD, since a good % of them have one themselves, and they’ve hired people like you before. Some of course won’t be interested until you are free, but there are enough who can manage their time to make it worth while.
When to apply if you’re doing a Masters Now.
Apply early, apply often. Some employers are quite happy to be flexible about people studying masters part time and although they are a minority, they are worth grabbing whilst you can. If you’re doing a full time then if you haven’t started sending out your CV, you should do it now. You may be also able to wangle an internship. If you’re reading this right at the start of the year, then a good % of employers of headhunters will either ignore you or tell you to apply later. But, at the risk of repeating ourselves, as long as the probabilities remain >=0 every time you try, you increase your chances. Because some masters programmes track undergraduate terms, your classmates (or competitors as we like to call them), will all be hitting the market at the same time. They see the same adverts, talk to the same headhunters, and use the same careers service, and will be going for the same set of jobs and unless you are easily the smartest person in your year, that’s a bad thing. You feeling lucky punk ?
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 67
Guide 2.0
Paul & Dominic’s Guide to Quant Careers What People Get Wrong Why your boss is a jerk
If you’re leaving your job, it’s of course likely to be because you’ve stopped enjoying the work, or that your job doesn’t offer you the opportunities you want. Money may well be a factor and as HH’s we don’t have a problem hearing this. Money has many problems but at least it is relatively simple as a motivation. Sometimes however we hear from a candidate that it is because his current boss is a shambling moron whose personality is an unstable mix of dishonesty and ignorance barely held together by malicious greed. His management style draws upon both forms of Marxism, both Groucho and Karl. He can recite “The Art of War” from memory and he frequently quotes from it at meetings (in the original Chinese of course). You feel you have to leave now or you and he will Talking about your old job is a balance between settle your disputes with knives. The IT at showing a willingness to move on for a better your department looks like it’s run by EDS, offer, and looking like a refuges from a financial the management are in league with Al Qaeda, disaster. compliance has been infiltrated by Accenture and Jack Bauer has told you that the back office wants you dead. Today you found a live rat in your coffee.
Why he isn’t At P&D we like to hear the truth about how you see your current role and how it’s developing and an honest summary of your reasons for leaving. That means its fine to share with us your views on the personalities involved and how things could be better. This sort of information helps us fit you to a new job rather more accurately. However, when you go forward to a bank, part of your value is your experience at this firm, indeed the branding you get from working at your current firm may have been a major factor in landing the interview. Some managers like to see people who come from firms that they know and respect. That’s not always the case of course. Dominic once hired a network manager away from a major outsourcing firm on the grounds that anyone who could achieve anything there must be remarkable. At the very least they usually won’t hold your current firm against you. So if you knock your current firm too much, you are to an extent, knocking yourself. You may not even have a reason for leaving, and that is perfectly fine with us. What we care about is interest in the new job, or at least an open mind to other career options. That applies at interview, not just a chat over coffee with us. Someone who is happy where they are, but sounds positive about a new job presents themselves far more successfully.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 68
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Quantitative finance is like sex because…
Jobs are not unlike relationships with the opposite sex and it’s stupid to start a first date by enumerating at great length the defects and personal failings of your ex. They may indeed have slept with your best friend, or applied on your behalf for a job at EDS, but leave that baggage behind. We have experimental data to support this. It is much better to talk of the positive reasons for leaving and your enthusiasm for the opportunity that you are now looking at. To get a good job, you need to be able to show your good work in some environment. You should think through the positive things about your current job, even if that’s hard. This is always good idea, even if your current boss is not a terrorist, just so that you can think through in which ways (if any) that this new job matches what you really want. You don’t get to change jobs all that often, so you want to make the right decision. This obviously goes beyond money, though that doesn’t hurt. You should try to see what parts of the new job you would particularly enjoy doing. You may well be asked why you want this job Make sure that the reasons you give for leaving and although it’s a soft question, people often your old job are not those which scare off the manage to get it wrong. people who will give you the next.
Another reason for not sounding too negative about your current job is that if you seem too eager to jump ship, it undermines your negotiating position.
Apply for the right job You may feel you are unique individual and an obvious match for the job. Sadly, that turns out not to be the case. If you are applying for a job called “Henchman to Assistant Quant’s Minion P&D0701067,” then do try to include that in your application, prominently. If you don’t include this, then you are critically dependant upon whoever opens your application guessing what job is best for you. If any. It’s entirely reasonable to apply for a job where you don’t quite match exactly what they are asking for. Many job specifications are a wish list rather than a set of critical skills. It is possible to get hired because strength in one area overcomes weakness in another. But make sure that this is a decision, not sloppiness. It doesn’t come across well when candidates who are selling themselves on being smart, have failed to understand the one or two paragraphs of the advertisement.
Spam When a web site asks, “Shall I send this stuff to your friends?” do not say yes. Dominic gets a stream of emails via one candidate asking him to join some sort of SMS dating agency. The candidate name is on each one. This does not make you friends. However we have a presence on Facebook and LinkedIn to see if it adds value to the service we provide.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 69
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Focus
Forging a rapport with the interviewer is a good thing, but some interviews drift off topic as the people involved chat. However, there is a time budget for each interview and most managers have specific objectives in checking your ability. If they don’t get covered it can hurt your progress to the next stage. Although it is the interviewer’s responsibility to get things done, it’s your problem if he doesn’t. This is where the politeness we mention elsewhere is important. When you feel that time is moving against you, ask to make sure that everything they need to know is covered. This is a good thing to say in any case.
Asking questions Actually, there are stupid questions. Bad questions are ones which embarrass the interviewer, or force them into corners. That’s their job. Do not try to score points off the interviewer, either you fail and look silly, or worse still, you succeed. It’s a bad idea to bring up any screw-ups that the bank has been involved in, or where the manager has to admit that he hasn’t read your CV.
Buzzwords Your interrogator will often come from a similar background to you, but even within maths and physics there are many specialisations that are mutually incomprehensible. You’re just emerging from a discipline where you think in terms of these names and equations and it’s easy to emit a stream of noises that your interviewer can barely understand. It’s actually worse if they are from a similar background, since they may feel embarrassed to ask what you actually mean. Use of obscure technical terms may be good for impressing your buddies at college, it’s not a great way of getting a job. You can lose points here. But it is generally polite to enquire about the background of your audience when asked to explain some part of your work. This both shows consideration and prevents you making this error.
Learning Your first interview will not be as good as your last. As you will read in the interview section there are distinct “interview skills”. Thus your early interviews have less value than your later ones and if you suspect that your face to face interaction is not 100%, you might as well go for some interviews to gain confidence and make mistakes that matter less.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 70
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Be polite
Your mother told you manners would be important one day, and this is the day. “Please,” “Thank you,” and actually looking as if you are listening are good things. Fidgeting, playing with your tie, or looking like you’d rather be somewhere else is not polite. Standing when people come into the room is good, preferably holding out your hand to be shaken. Occasionally you will find it appropriate to disagree, this is good since it can show that you’re thinking about what they’re saying, but get in the habit of using phrases like “I’m not sure if that’s the case, perhaps it is…”. You can’t just wake up one day and be polite on a whim. (Hint: “Pretty Woman” is fiction, we know this for a fact.) Without practice, politeness may even come over as sarcasm. In some languages “Please” and “Thank you” are implied in the context of the sentence and that habit can spill over into English. Break that habit, break it now.
The Volatility Smile Version 1.x didn’t tell you to smile, that’s why You will spend more hours with the people you we have upgraded to version 2.0. Smiling is work with than who you sleep with. It is important good on several levels. It softens the arrogance in both cases that you get along. that sometimes is felt when interviewing smart people straight out of education. In conversation with hiring managers the word “arrogant” is surprisingly common. Yes, you are smart else you would not have got this far, but you are in the big boys game now, on a global scale. Merely being the smartest person you know, or getting awards from your national government for extreme cleverness in no way makes you unique here; a fact you may well not enjoy learning. Often when arguing over points in real subjects, the issue is that one person (or both) have failed to understand a given point, there is no real legitimate position that may be taken on either side. One is right, and the other doesn’t grasp what is going on. That is why most people in bars and dinner parties prefer arguing over soft subjects like politics, where almost any position can be validly held. If you can’t argue both sides you don’t understand the topic. Smiling without eye contact, is bad, and as far as we are aware this is true in all cultures. It is easily mistaken for arrogance, and looking away whilst smiling might be interpreted as some communication with yourself where you and yourself are sharing some private joke at the interviewer’s expense. That may be seen rude or mad, or probably both. So smile into people’s eyes. That may sound like an over reaction to a minor personal quirk, and maybe it is. But that appears to be a significant cause for this common criticism coming as feedback for candidates who did not get a job is that they exhibit this “arrogance”. Confidence is good, the trick is to find the right balance. Practise sounding positive about things. You can become better at presenting yourself as someone who’s easy to work with. This is so important because your team will spend more waking hours together than most married couples and senior people want to know you will “fit in.” Like much of this whole process, it’s a game. No one really cares if you have a deep respect for your fellow man, but if you can emulate it well under pressure it’s a difference that makes no difference.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 71
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Be true to yourself
You are selling yourself, so obviously you will be putting a positive spin on things. However, this is a career, not a job. If you feel the job may really not be what you want, then it’s important that you think that through. If in the interview you hear something that sounds bad, ask about it. This does not have to be confrontational; you can use phrases like “how does that work out in practice?” and “what sort of flexibility is there to choose the work?” when told you’re going to be counting buttons for the first six months.
Do not sound as if you work for Accenture Even if you do currently work for Accenture or Arthur Anderson, you don’t want to sound like you do. Avoid the sort of management consultant pseudospeak that resembles Dilbert cartoons. A common type of “soft skills” interview question is of the form: “You find that X has gone terribly wrong, what would you do about it.” An Accenture style answer is: “I would see it as a challenge that would allow me to work as a good team player, as part of the global strategy, being You are a science graduate talking to other careful to solicit input from business science graduates, even if their focus has moved stakeholders”; or perhaps you might respond: from expanding human knowledge to expanding their wallets. “I will grasp the opportunity to show excellent leadership in integrity” which is interview suicide. Or to quote one person: “I love pressure, that’s why I want to work for a bank” Sounds good, but at some level, also slightly worrying. This warning may sound quite silly, but there is a growing trend for some universities to have formal coaching in interview technique. In theory this should be very useful. In theory. In practice is comes across as rather scary. It frustrates interviewers a lot to be faced with an obviously bright candidate who parrots clichés that some consultant has fed into him. We say at the beginning that you need to stand out and given that the people you are competing with may well include people from your current institution, it does you very little good. By all means listen to these people, but take it with a pinch of salt. When you know little about the process, it’s easy to give too much weight to the few things you get told. We also ought to include something that nearly contradicts everything else we say on this subject. Don’t be too self conscious. Many of us have had recordings made of presentations on subjects we’re supposed to be experts in and given a professional lecture. Which we do, usually. Except, of course, no one can talk for hours without sentences that make no sense or lack verbs, or saying something that’s so flatly wrong, you’d swear you never said it. Don’t sweat about this.
Interview overlap It is tempting to schedule lots of interviews as close together as possible, because travel does eat into your budget. You should be very conservative about the amount of time you allow for each interview. It’s not easy to get a manager to speed up his process because you want to get across town to talk to one of his competitors. The worry about time, just like lateness, can reduce your effectiveness, so make sure this doesn’t come up.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 72
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Odd questions
You will be asked some rather strange questions, not just brainteasers. They include “Can you work in a team?”, “Do you want to work in a bank?”, “What motivates you?”, “What sort of team do you want to work in?” and variants on impossible questions like “As you don’t know what we do precisely, what difference is there between this kind of work and some other job you know nothing about?” You may think this is because the interviewer can’t think of anything intelligent or useful to say, and frankly this is sometimes true, but you cannot know that for certain, and you must not assume it. Treat them as serious questions and although some like “Do you want to work here?” have obvious answers, you do best by thinking about them, preferably in advance. You should be able to articulate something that makes you want to do this job for this business in this bank. You should have views on teams. For instance, now that you aren’t competing with me for jobs I can share with you that when asked about the ideal team size for me, I have said “For me, what is important is the quality of the team, rather than its size”. Which has the virtues of not cutting yourself off from whatever size of team you might join (by not actually meaning anything) and also sounding positive and upbeat. A bad answer is “I like small teams because you don’t really know what strangers are thinking”. Even the largest banks are typically a large number of smallish, 5-20 person groups, that’s basically because we evolved to work best that way. Vague questions are an opportunity for you to slip into the conversation plugs for skills or experience that have not yet been covered. Some managers ask them for just this purpose.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 73
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Who will interview you You will be interviewed by several people for any given job. They will have different levels of seniority and different areas of specialization. It comes as a surprise to some people that firms such as Goldman Sachs want you to meet anything from five to twelve managers and members of your new team. Other firms prefer to get it over with in one day.
Quant prejudices There are many types of quant and we don’t just mean that they work in different product areas. Very often there can be a great deal of technology transfer from one product area to another. For example, the quant skills needed for credit and for interest rates have a wide overlap. No, we mean that quants come from different mathematical backgrounds and it is this rather than product area that will dictate the nature of a technical interview. Quants come in two varieties: the pure mathematician and the applied mathematician.
The pure mathematician quant This quant will have studied finance as a branch of probability theory, often quite abstract probability theory. Their degree will probably be in either finance or mathematics from a probability theory department. In their interview they will talk about martingales and change of measure, Girsanov and numeraires. Ninety-nine times out of a hundred the punchline of the probabilistic approach is the simple concept that the value of a financial contract, such as an option, no matter how complex is the expected present value of all cashflows. The expectation comes in because cashflows will vary depending on how the underlying instrument behaves and the present value appears because cashflows will be in the future and we are valuing the option today. The catch in this is that the expectation must be with respect to the risk-neutral random walk of all underlying random variables, not the real random walk. Many problems in finance can be addressed by either the pure or the applied mathematician. However, because many interest rate models are constructed from a simulation point of view they are easier to deal with by the probabilistic approach. These models are known as the Heath, Jarrow & Morton (HJM) and Brace, Gatarek & Musiela (BGM) models. When it comes to number crunching and writing code, which is invariably what you will be doing, you will find yourself writing a lot of Monte Carlo simulation code. At a basic level, Monte Carlo code is easy to write. There are really just the two problems with it, making it fast (including finding the Greeks) and using it for contracts with early exercise.
Key areas to brush up on You ought to be able to understand and be able to discuss knowledgeably the following. Risk neutral versus real Change of numeraire Martingales Monte Carlo methods Techniques for speeding up Monte Carlo Monte Carlo methods for contracts with early exercise • • • • • •
The last two are popular areas of research, so have a few recent articles at your fingertips.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 74
Guide 2.0
Paul & Dominic’s Guide to Quant Careers The applied mathematician quant
The applied mathematician may have studied anything but finance. They will have degrees in physics, electrical engineering, chemistry and some field of applied mathematics such as fluid mechanics. They will talk about calculus and partial differential equations. To applied mathematicians the problem of pricing a contract is about finding a differential equation which must then be crunched numerically. The differential equation will almost certainly be of parabolic type, also known as the heat or diffusion equation. This equation also has the same interpretation as giving the expectation of the present values of cashflows but this is seldom mentioned. Applied mathematicians can’t easily write certain interest rate models in the partial differential equation framework. However, they have other counter-balancing advantages over the probabilist. Using the tools of calculus they can derive models with fewer assumptions than the probabilists, they can incorporate transaction costs, optimality, nonlinearity and other fun stuff into their models. Number crunching to the applied mathematician can mean either Monte The market can stay irrational, longer than you Carlo again, or finite difference methods. can remain solvent. Finite-difference methods are sophisticated versions of the binomial method. No, strike that, it would be better to say that the binomial method is a prehistoric version of a finitedifference method. Key areas to brush up on Diffusion equations generally Finite difference methods Finite difference methods in several dimensions The different types of finite-difference method: Explicit, implicit, etc. • • • •
You ought to have coded up some of these in C++ or at least VB or Matlab, otherwise you won’t sound at all convincing in the interview.
Final trick Often you will know the name of the person interviewing you. If you do a simple search on Google or wilmott.com may give you information about the style of the person, their interests, perhaps even weaknesses in their own knowledge.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 75
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
HR Interview Typically the HR interview is the last one before they offer you the job and after everyone else has signed off that they want you. Although it is not very common for candidates to blow it at this point, you really don’t want to fall down at the last fence. HR are typically the only people involved in the hiring process with any real formal training in staff selection. Thus although they are not in a position to evaluate your maths or programming skills they are still responsible for trying to make sure that the bank is not hiring a dud. From the forums we get the impression that many people see HR interviews as a complete waste of time, and yes, many are. They are filters and sanity checks and in theory they shouldn’t be necessary. Except when they catch a dud that everyone else has missed.
At that point people appreciate that HR interviews are not a complete waste of time. You should make very sure that all the date on HR are usually the people who understand the your CV make sense, and if they do not join up be internal bureaucracy best, so gather together able to explain why. all your issues about interview expenses, start dates and when precisely you will get your first pay cheque for this meeting. They are the only people who can tell you the truth about the pay and benefits. All sorts of people will in good faith tell you about these. But if you think that because a bank pays 600K a year to your new boss to understand complex financial issues that he might stand some chance of getting these things right, you are sadly being a little optimistic. Headhunters try, but we know from experience that this second hand information is not of the best quality. However it is best to leave your “give me money and time off” questions until the part where they ask you for your questions. In many cases, they will go through this without being asked because better HRs use this phase to convince you of the wisdom of choosing their firm.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 76
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Audit
Many HRs see their role in these interviews to be catching these duds and thus although the questions they ask may be extravagantly banal, they are in fact checking that you are what you say you are. Since they have so little objective data to work on, they seize upon any perceived defects and give them a good hard shake. Back when he had a real job bossing people around, Dominic once interviewed someone who not only could not remember why they had chosen particular options on their degree course; they actually struggled to remember half of their names. This would have been a mildly bad thing if it had been ten years after, but this candidate was between finishing their course, and formally graduating. They didn’t get the job. HRs often pick up on errors in your CVs like mismatched dates and are trained not to like gaps at all. If the dates do not join up, be prepared to explain why. Part of the background check that most banks perform involves checks on employment dates. Some people’s dates are rather confusing, for instance you may have studied at two places during the same period, and worked during vacations as well as internships. That’s fine, but worth telling P&D about it so that we can set the right expectation. The short version is that you can get away with a lot more that you state up front than if you are found out. HR will ask about how you made various choices in your life. Let’s be honest here, we all have “choices” that were in reality some combination of bad judgment, ignorance, showing off and improvements to the chances of meeting people of a compatible sex. They know this as well, but they are also checking that you are capable of expressing yourself, as well as trying to guess how you make important decisions. The better type of HR people see their role of supporting the business decision to take you on by a combination of checking you out and selling the bank to you. They actively want you to take the job, and you should make sure you get their contact details because sometimes they are in the best position to kick the process into moving when it jams.
Give examples A common theme in the way many people in HR approach the intangible issues is the notion of finding examples of the trait in question through your experience. Thus when you are asked whether you are happy working in teams, the wrong answer is “yes” (although it is obviously better than “no”). Most often they are looking for examples of when you worked well in a team and what you gained from it, or whether you’re not really that good at teamwork. There are a lot of questions on this theme and yes, they do include, “Why do you want to work for us?” People sometimes get that wrong. Yes, really. You should prepare yourself for the most obvious questions, since both HR and line management may ask them.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 77
Guide 2.0
Paul & Dominic’s Guide to Quant Careers And then we hid the drugs under our mother’s bed…
Even major governments find it hard to work out whether people they take on should be trusted; and that’s with quite intrusive checks on your life and people you are associated with. Thus you should feel some sympathy for the HR who has the job of working out whether you should be trusted with millions based upon a ten minute segment of an interview. But that sympathy should be tempered with your resolution not to dig a hole for yourself. As above they may well ask for some example where you were found yourself having to make a difficult ethical choice, or how you dealt with a situation where some wrongdoing happened around you. They may ask follow up questions just to check that you are a normal law abiding citizen. However, it may be a good idea to review your life and think of a good example. What you really do not want to happen is to find it hard to think of such a case, then remember something and start a story about the errant behaviour of your little brother and how you had to keep him out of trouble. That’s fine, but think the anecdote through, because it may just end with the way Everything you say about yourself in your CV at you solved the problem paints you as an interview should be provable from either incompetent drugs dealer. For obvious reasons education you have received or better from of personal confidentiality the “drugs” idea is experience in real life that supports your claim. radically changed from what someone actually said, but is loosely based upon what happened. The HR manager is handing you rope by letting you talk about your life. Do not tie a noose in it and stick your head through. Some HRs have been taught the technique of “aggressive silence”. Most people from Western cultures find long gaps in conversations embarrassing, and so will fill them in, often without thinking too hard about the content. If you start along an “interesting” track, and the HR goes quiet, smiles in a polite way, and keeps on being silent, you know that they are interested to see what will crawl out from your past. One candidate when asked about a test result that seemed well below their ability replied that they had been to a really good party the night before and had been in no fit state to take the exam. They didn’t get the job.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 78
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Background Checks
Nearly all banks do at least some sort of basic background check. This includes trying to make sense of your transcripts, references from previous employers and looking for unexplained gaps in your employment history. You may gain a little personal satisfaction from your progress in life, when your ex-boss at some grim fast food joint you worked one summer finds out that a really big bank is hiring you for such an important position that they contact him to check out your past. These investigations include a credit check. There are several reasons for this. The first is that the bank may simply not be allowed to hire people who are bankrupt, or have certain types of court judgment against them. It also serves as an extra confirmation that you are who you say you are and that your life bears some resemblance to the one on your CV. The final reason is that it’s cheap. For a tiny amount of money one can get more raw information quite legally about a person from the credit agencies than a £300 per day private detective It is very rare for entry level candidates to lose can get in a few weeks. At this point a certain number of you will be worrying that the rather stochastic view you took to paying your credit card is going to come back to bite you.
out on their desired job because of some minor debt whilst a student. If in doubt get both your headhunter and human resource specialist on your side before the credit check comes back.
Not yet. It turns out to be the case that HR people usually know a lot about student finances, most of them are graduates and many have a Master’s degree and thus have a relatively relaxed filter on low level student cashflow problems. So a few misplaced bills or arguments with landlords or phone companies rarely disturb them very much. That is unless the amounts are so large that they imply either substance abuse or an unacceptably flagrant contempt for money. Credit checks will of course flag up potential fraud and it should come as no shock to you that banks really don’t like dishonesty at all. AT P&D we’ve helped any number of people get past this, and we’ve yet to see a case where debt stopped someone getting the job they wanted. So again this is a case where telling the truth up front is usually the best way forward. By bringing it up you present yourself as honest, and you can put a good spin on it. This is better than the cold impersonal view that a credit report will have of your unpaid bills. For the avoidance of doubt, unless you come from a country that is currently at war, the odds are that the bank can get some sort of credit check on you. It costs a bit more, but the global economy has several large firms who exist to do this.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 79
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
The Different Types of Mathematics Seen in Finance The real-world subject of quantitative finance uses tools from many branches of mathematics. And financial modeling can be approached in a variety of different ways. For some strange reason the advocates of different branches of mathematics get quite emotional when discussing the merits and demerits of their methodologies and those of their ‘opponents.’ Is this a territorial thing, what are the pros and cons of martingales and differential equations, what is all this fuss and will it end in tears before bedtime?
Financial modeling Here’s a list of the various common approaches to modelling and a selection of useful tools. The distinction between a ‘modelling approach’ and a ‘tool’ will start to become clear.
Modeling approaches • • • •
Probabilistic Deterministic Discrete: difference equations Continuous: differential equations
Useful tools • • • • • •
Simulations Approximations Asymptotic analysis Series solutions Discretization methods Green’s functions
While these are not exactly arbitrary lists, they are certainly open to some criticism or addition. Let’s first take a look at the modelling approaches.
Probabilistic One of the main assumptions about the financial markets, at least as far as quantitative finance goes, is that asset prices are random. We tend to think of describing financial variables as following some random path, with parameters describing the growth of the asset and its degree of randomness. We effectively model the asset path via a specified rate of growth, on average and its deviation from that average. This approach to modelling has had the greatest impact over the last 30 years, leading to the explosive growth of the derivatives markets.
Deterministic The idea behind this approach is that our model will tell us everything about the future. Given enough data and a big enough brain, we can write down some equations or an algorithm for predicting the future. Interestingly, the subject of dynamical systems and chaos falls into this category. And, as you know, chaotic systems show such sensitivity to initial conditions that predictability is in practice impossible. This is the ‘butterfly effect,’ that a butterfly flapping its wings in Brazil will ‘cause’ rainfall over Manchester. A topic popular in the early 1990s, this has not lived up to its promises in the financial world.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 80
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Discrete/Continuous
Whether probabilistic or deterministic the eventual model you write down can be discrete or continuous. Discrete means that asset prices and/or time can only be incremented in finite chunks, whether a dollar or a cent, a year or a day. Continuous means that no such lower increment exists. For reasons that we’ve never understood, the mathematics of continuous processes is often easier than that of discrete ones. But then when it comes to number crunching you have to anyway turn a continuous model into a discrete one. In discrete models we end up with difference equations. An example of this is the binomial model for asset pricing. Time progresses in finite amounts, the time step. In continuous models we end up with differential equations. The equivalent of the binomial model in discrete space is the Black-Scholes model, which has continuous asset price and continuous time. Whether binomial or Black-Scholes both of these models come from the probabilistic assumptions about the financial world.
Simulations If the financial world is random then we can experiment with the future by running simulations. For example, an asset price may be represented by its average growth and its risk, so let’s simulate what could happen in the future to this random asset. If we were to take such an approach we would want to run many, many simulations. There’d be little point in running just the one; we’d like to see a range of possible future scenarios. Simulations can also be used for non-probabilistic problems. Just because of the similarities between mathematical equations a model derived in a deterministic framework may have a probabilistic interpretation.
Discretization methods The complement to simulation methods, there are many types of these. The best known of these are the finite-difference methods which are discretizations of continuous models such as BlackScholes. Depending on the problem you are solving and unless it’s very simple, you will probably go down the simulation or finite-difference routes for your number crunching.
Approximations In modeling we aim to come up with a solution representing something meaningful and useful, such as an option price. Unless the model is really simple, we may not be able to solve it easily. This is where approximations come in. A complicated model may have approximate solutions. And these approximate solutions might be good enough for our purposes.
Asymptotic analysis This is an incredibly useful technique, used in most branches of applicable mathematics, but almost unknown in finance. The idea is simple; find approximate solutions to a complicated problem by exploiting parameters or variables that are either large or small, or special in some way. For example, there are simple approximations for vanilla option values close to expiry.
Green’s functions This is a very special technique that only works in certain situations. The idea is that solutions to some difficult problems can be built up from solutions to special solutions of a similar problem.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 81
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Volatility Classically, volatility plays many roles in finance.
For the holder of an equity, for example, more volatility is bad since it is closely associated with risk. However, for the holder of an option, it can be good, the value of simple options increase as volatility increases. Unfortunately, volatility of a stock or an interest rate is impossible to measure at an instant in time. Yet we need to know its value, and forecast it over the future, in order to value options. This leads to people talking about (and muddling up) all sorts of different t ypes of volatility. If you can distinguish between these volatilities while all around you are confusing them then you’ll be a quant, my son. Actual volatility means the amount of noise in the underlying. You can’t measure it accurately. But you’d really like to know what it’s going to be so you can price derivatives. Realized or historical volatility is a backward-looking statistical measure of what volatility was in the past, but depends on the statistical technique you use. You might use historical, statistical measures of what happened in the past in order to forecast future volatility. If you are in the business of vol arb you will need such forecasts in order to make trading decisions in which case you may live or die according to the accuracy of such forecasts. Implied volatility is what quants talk about a lot, of some of them confuse it with actual volatility. Implied vol means the volatility put into a Black-Scholes formula in order to match a theoretical model price with a price seen in the market. It’s sort of what the market thinks is likely to be volatility in the future. (Although when speaking to a trader you should say that implied volatility is simply where supply meets demand. This isn’t ‘sort of’ correct, this is ‘exactly’ correct. Here we have an example of using different language depending on your audience.)
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 82
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Time Series Analysis
This is the evil twin of modelling, where instead of trying to apply a model to a system, based upon what you think is happening you are examining a very noisy data set to try and find patterns, or to extract some quantity like volatility. We find that as well as the econometrics skills of ARIMA, Fourier, de-lagging, spectrum analysis, principal component analysis, and the variants on GARCH that signal processing skills from electronic engineering are in increasing demand, so if you are pushing this aspect of your skill set, we would advise you to bone up on Kalman and particle filters. You should be wise in the ways of not only finding the various parameters of a mean reverting model, but how to detect regime change. There is even a healthy market for Quant C++ developers who understand how to work with very large, high frequency data sets, and no, this does not mean Oracle SQL. Often there is no notion of a “real physical” system at all, merely a result that says “x correlates with y”, or a rather heuristic algorithm for getting the right sort of value. Another aspect of the work is simply cleaning up the data that hits your system, and this varies from highly automated real time processes through quite literally looking at the graph and thinking about it, though of course that extreme isn’t usually given over to quants to work on. Market data is so noisy that classical models like Black Scholes can usefully treat it as completely random. The noise
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 83
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Reading List Background Derman, E 2004 My Life As A Quant . John Wiley & Sons http://books.globalinvestor.com/books/19770.htm?ginPtrCode=10202&identifier=18932b1e1d2bc6d5a1ae69160e8d a035 Poundstone, W 2005 Fortune’s Formula. Hill & Wang http://books.globalinvestor.com/books/22645.htm?ginPtrCode=10202&identifier=18932b1e1d2bc6d5a1ae69160e8d a035 Taleb, NN 2001 Fooled by Randomness. Random House http://books.globalinvestor.com/books/22645.htm?ginPtrCode=10202&identifier=18932b1e1d2bc6d5a1ae69160e8d a035 Brown A 2006 The Poker Face of Wall Street. John Wiley & Sons http://books.globalinvestor.com/books/23007.htm?ginPtrCode=10202&identifier=18932b1e1d2bc6d5a1ae69160e8d a035 Taleb, NN 2007 The Black Swan. Penguin http://books.global-investor.com/books/168555.htm?ginPtrCode=10202 To get a feel for the recent history of quantitative finance and the working environment for a quant see Emanuel Derman’s autobiography. This is a very personal account of his own career from physicist to Wall Street quant. However, it doesn’t really convey the role played by programming in a typical quant’s life, since Dr Derman is far from being a typical quant. Bill Poundstone writes about some simple money-management ideas well known to all gamblers yet under-appreciated by risk managers. To back up his case he describes the Long Term Capital Management fallout and the role played by leverage. This book will give you plenty of nuggets to drop casually into an interview to show that crucial market insight and how to relate markets and mathematics in a very pragmatic way. You will either love or loathe Nassim Taleb’s non-technical book on the part played by risk in our lives, not just finance. It is compulsory reading. But whether you love it or loathe it don’t let on to your interviewer until you know their feelings towards it. Again, a lot of useful interview material, but unfortunately nothing mathematical. Again, a lot of useful interview material, but unfortunately nothing mathematical. The same goes for the even more opinionated The Black Swan. This book was on the New York Times best-seller list for aeons. Your interviewer will have read it, and have strong opinions about it. Be prepared to agree or disagree (both vehemently), but make sure your reasoning is sound not just based on emotions. Don’t expect too much detail about either poker or Wall Street in Aaron Brown’s book. Nevertheless it is highly informative, you’ll get a lot out of this on the non-technical side of quant finance and, importantly, introduce you to ways of thinking that will help you to lose not too much money.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 84
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Introductory
Haug, EG 2007 The Complete Guide to Option Pricing Formulas 2 nd edition. McGraw-Hill http://books.globalinvestor.com/books/66956.htm?ginPtrCode=10202&identifier=80d939a1914d743ad6d46738510 85ba8 Hull, J 2005 Options, Futures and Other Derivative Securities . Prentice Hall http://books.globalinvestor.com/books/9233.htm?ginPtrCode=10202&identifier=18932b1e1d2bc6d5a1ae69160e8da 035 Paul Wilmott Introduces Quant Finance Second Edition http://books.globalinvestor.com/books/168530.htm?ginPtrCode=10202&identifier=a89ab6672523e235bd844e19210 a4358 Espen Haug’s book concentrates on presenting pragmatic results for the practitioner without going into all of the modeling details. This book finds a nice balance between theory and practice. Do you want the perfect model that is impractical because it takes too long to find an answer? No. Do you want a model that is quick but gives useless results? No. A chain is only as strong as its weakest link and this book focuses on models where all the links are similar in strength. Very practical. It is now in it’s second and greatly expanded edition. Paul Wilmott’s book is of course famous, partly for it’s more practical approach to quant finance and relatively relaxed writing style, presumably if you’ve reached this point you already have a copy, if you don’t then you should get one. According to the Financial Times, Paul is a “cult derivatives lecturer”. There is a now a considerably extended and revised second edition. Knowing Hull is compulsory. Everyone has read it. Unfortunately that means that to some extent that knowledge will have been discounted by interviewers. The book is great for definitions of things financial and mathematical. However, if you have only read Hull you will have a strange, fragmentary understanding of quantitative finance because of the pick’n’mix nature of the contents and because there is no common theme holding the book together. There is also too much on the binomial model. Although this is great for teaching it isn’t great for hardcore, practical quanting. If you think that you can get a job after merely reading Hull you are in for a big shock.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 85
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Advanced/research level
Jaeckel, P 2002 Monte Carlo Methods in Finance. John Wiley & Sons http://books.globalinvestor.com/books/14474.htm?ginPtrCode=10202&identifier=18932b1e1d2bc6d5a1ae69160e8d a035 Peter Jaeckel goes into the details of Monte Carlo methods. It can be quite heavy going at times, but for interview purposes you won’t need to have understood everything in this book. As long as you can show an appreciation of random number generation methods, how to speed up Monte Carlo and using Monte Carlo for pricing American options then you should be fine.
Glasserman, P 2003 Monte Carlo Methods in Financial Engineering . Springer-Verlag http://books.global-investor.com/books/16497.htm?ginPtrCode=10202 Paul Glasserman also writes in an accessible manner on the topic of Monte Carlo. He begins by covering the basic, then moves on to advanced techniques, e.g. for speeding up the technique, and then moves on to specific financial problems. Yes, the book is ‘yellow,’usually a bad sign of a simple subject made complicated, but don’t let this put you off.
Morton, KW & Mayers, DF 1994 Numerical Solution of Partial Differential Equations. Cambridge http://books.globalinvestor.com/books/21542.htm?ginPtrCode=10202&identifier=18932b1e1d2bc6d5a1ae69160e8d a035 The other main numerical methods you will need to understand are the finite-difference methods. There are books on finite differences in finance (such as Tavella & Randall) but for extra kudos you would be better off reading one of the classics such as that by Morton & Mayers. It has no reference to finance whatsoever, but explains beautifully the numerical solution of various partial differential equations. Neftci, S 2004 Principles of Financial Engineering . Academic Press http://books.globalinvestor.com/books/17107.htm?ginPtrCode=10202&identifier=9984462dc967c88392532199278 5d117 Salih Neftci’s book explains the probabilistic side of quantitative finance modeling. Although this subject can be a bit daunting because it appears so abstract, Neftci does a fantastic job of making the topic intuitive.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 86
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Schonbucher, PJ 2003 Credit Derivatives Pricing Models . John Wiley & Sons http://books.globalinvestor.com/books/15886.htm?ginPtrCode=10202&identifier=18932b1e1d2bc6d5a1ae69160e8d a035 Philipp Schonbucher is another challenging read. Entirely on the subject of credit risk modeling, if you are able to understand (and more importantly, show that you understand) just 25% of this book then you will go far, my son. Taleb, NN 1997 Dynamic Hedging . John Wiley & Sons http://books.globalinvestor.com/books/3768.htm?ginPtrCode=10202&identifier=18932b1e1d2bc6d5a1ae69160e8da 035 Nassim Taleb’s technical book (as opposed to his popular science books mentioned earlier) combines theory and practice very successfully. This is another book that will give you useful gems for your interview.
Wilmott, P 2006 Paul Wilmott On Quantitative Finance, second edition. John Wiley & Sons http://books.globalinvestor.com/books/23017.htm?ginPtrCode=10202&identifier=18932b1e1d2bc6d5a1ae69160e8d a035
Paul Wilmott’s three-volume opus is from the other side of the tracks from Salih Neftci, it shows how the applied mathematician thinks of quantitative finance. The book starts at the standard level of explaining Black-Scholes, moving on to discussing modern quant models as used in practice, ending up with some ideas from the cutting-edge of research. One important point about this book, which distinguishes it from others, is that it does criticize many of the models commonly used by practitioners. Again, you will find many modeling insights and bon mots in this book that will be useful during your interview ordeals.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 87
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
What you must know The following is not an exhaustive list and is only the general broad brush strokes, but it will ensure that you don’t make an idiot of yourself by having major gaps in your knowledge. Products Fixed income, bonds, swaps Equity, dividends, derivatives Currencies, role of foreign and domestic interest rates Commodities, convenience yield etc. Exotics, main types, Credit derivatives • • • • • •
Concepts Risk, return and efficient frontiers Delta hedging Risk neutrality The no-arbitrage argument Market price of risk for non-traded quantities Calibration Static hedging using exchange-traded vanillas • • • • • • •
Models • • • • • • • • • • • • •
Binomial model Lognormal Jump diffusion Stochastic volatility Interest rate model (single, multi factor, HJM, BGM) Credit models (hazard rate, structural) Transition matrices Numerical methods Monte Carlo simulation techniques Binomial scheme Finite-difference methods Numerical quadrature Which method to use for which type of contract
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 88
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
What you must have coded up Key areas You must be comfortable with the four main numerical methods used for pricing derivatives: Binomial/trees Finite-difference methods Monte Carlo Numerical integration • • • •
Although the first of these (binomial and trees) are rather inefficient they still have their proponents and you should show willingness to lower yourself to this level. We know of a world class numerical analyst working in computational fluid mechanics who left his academic job for the bright lights and ended up coding the binomial model. A certain amount of pride-swallowing was necessary in the short term, but his recent progress has certainly made it worthwhile. Aim to code up as many different types of exotic contract as possible, in as many markets as possible (essentially the lognormal markets of equity, FX, etc. and fixed income and credit). When pricing an exotic such as an equity Asian option you should use both Monte Carlo and finite difference. The two methods are completely different but the exercise will give you invaluable experience and plenty to talk about in your interview. (The next few sections are taken from Paul Wilmott On Quantitative Finance, second edition, 2006, John Wiley & Sons.)
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 89
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Finite-difference methods
Finite-difference methods are designed for finding numerical solutions of differential equations. Since we work with a mesh we find the contract value at all points in stock price-time space. In quantitative finance that differential equation is almost always of diffusion or parabolic type. The only real differences between the partial differential equations are the following: • • • •
Number of dimensions Functional form of coefficients Boundary/final conditions Decision features
Number of dimensions Is the contract an option on a single underlying or many? Is there any strong path dependence in the payoff? Answers to these questions will determine the number of dimensions in the problem. At the very least we will have two dimensions: stock price and time, for example. Finite difference methods cope extremely well with a smaller number of dimensions, up to four, say. Above that they get rather time consuming.
Functional form of coefficients The main difference between an equity option problem and a single-factor interest rate option problem is in the functional form of the drift rate and the volatility. These appear in the governing partial differential equations as coefficients. The standard model for equities is the lognormal model, but there are many more ‘standard’ models in fixed income. Does this matter? No, not if you are solving the equations numerically, only if you are trying to find a closed-form solution in which case the simpler the coefficients the more likely you are to find a closed-form solution.
Boundary/final conditions In a numerical scheme the difference between a call and a put is in the final condition. You tell the finite-difference scheme how to start. And in finite-difference schemes in finance we start, strangely, at expiration and work towards the present. Boundary conditions are where we tell the scheme about things like knock-out barriers. When we write our code we’d like it to be as general and reusable as possible. That means writing it so that it doesn’t have to be changed too much in going from one contract or model to another. So we might put things like final conditions in some external function, to be changed easily.
Decision features Early exercise, instalment premiums, chooser features, are all examples of embedded decisions seen in exotic contracts. Coping with these numerically is quite straightforward using finite difference methods, making these numerical techniques the natural ones for such contracts. The difference between a European and an American option is about three lines of code in a finite difference program and less than a minute’s work.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 90
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Program of study for finite-difference methods
If you are new to finite-difference methods and you really want to study them then you need a program of study.
Explicit method/European calls, puts and binaries To get started you should learn the explicit method as applied to the Black-Scholes equation for a European option. This is very easy to program and you won’t make many mistakes.
Explicit method/American calls, puts and binaries Not much harder is the application of the explicit method to American options.
Crank-Nicolson/European calls, puts and binaries Once you’ve got the explicit method under your belt you should learn the Crank-Nicolson implicit method. This is harder to program, but you will get a better accuracy.
Crank-Nicolson/American calls, puts and binaries There’s not much more effort involved in pricing American-style options than in the pricing of European-style options.
Explicit method/path-dependent options By now you'll be quite sophisticated and its time to price a path-dependent contract. Start with an Asian option with discrete sampling and then try a continuously-sampled Asian. Finally, try your hand at lookbacks.
Interest rate products Repeat the above program for non-path-dependent and then path-dependent interest rate products. First price caps and floors and then go on to the index amortizing rate swap.
Two-factor explicit To get started on two-factor problems price a convertible bond using an explicit method, with both the stock and the spot interest rate being stochastic.
Two-factor implicit The final stage is to implement the implicit two-factor method as applied to the convertible bond.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 91
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Monte Carlo methods Monte Carlo methods simulate the random behaviour underlying the financial models. So, in a sense they get right to the heart of the problem. Always remember, though, that when pricing you must simulate the risk-neutral random walk(s), the value of a contract is then the expected present value of all cashflows. When implementing a Monte Carlo method look out for the following: Number of dimensions Functional form of coefficients Boundary/final conditions Decision features again! • • • •
A common interview question is to ask the pros and cons of MC vs FD. This should be an easy point to score. Number of dimensions
For each random factor you will have to simulate a time series. It will obviously take longer to do this, but the time will only proportional to number of factors, which isn’t so bad. This makes Monte Carlo methods ideal for higher dimensions when the finite-difference methods start to crawl. Functional form of coefficients
As with the finite-difference methods it doesn’t matter too much what the drift and volatility functions are in practice, since you won’t be looking for closed-form solutions. Boundary/final conditions
These play a very similar role as in finite differences. The final condition is the payoff function and the boundary conditions are where we implement trigger levels etc. Decision features
When you have a contract with embedded decisions the Monte Carlo method becomes cumbersome. This is easily the main drawback for simulation methods. When we use the Monte Carlo method we are only finding the option value at today’s stock price and time. But to correctly price an American option, say, we need to know what the option value would be at every point in stock price-time space. We don’t typically find this as part of the Monte Carlo solution. Improvements to basic MC
It is worth revising antithetic sampling, variance reduction, importance sampling and Sobol & Quasi Monte Carlo methods. A good point to reach is where you can describe the effect on apparent Greeks by calculating prices by MC and what you can do to make this less bad. A key advantage of MC is that it is so simple to distribute the work across multiple CPUs that even Accenture consultants can do it (eventually). Thus even though some methods like finite difference methods can be more efficient, they are harder to make spread the load.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 92
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Program of study for the Monte Carlo path-simulation methods European calls, puts and binaries on a single equity
Simulate a single stock path, the payoff for an option, or even a portfolio of options, calculate the expected payoff and present value to price the contract. Path-dependent option on a single equity
Price a barrier, Asian, lookback etc. Options on many stocks
Price a multi-asset contract by simulating correlated random walks. You’ll see how time taken varies with number of dimensions. Interest rate derivatives, spot rate model
This is not that much harder than equities. Just remember to present value along each realized path of rates before taking the expectation across all paths. HJM model
Slightly more ambitious is the HJM interest rate model. Use a single factor, then two factors etc. BGM model
A discrete version of HJM.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 93
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Numerical integration
Occasionally one can write down the solution of an option-pricing problem in the form of a multiple integral. This is because you can interpret the option value as an expectation of a payoff and an expectation of the payoff is mathematically just the integral of the product of that payoff function and a probability density function. This is only possible in special cases. The option has to be European, the underlying stochastic differential equation must be explicitly integrable (so the lognormal random walk is perfect for this) and the payoff shouldn’t usually be path dependent. So if this is possible then pricing is easy… you have a formula. The only difficulty comes in turning this formula into a number. And that’s the subject of numerical integration or quadrature. Look out for the following. Can you write down the value of an option as an integral? •
That’s it in a nutshell.
Program of study for numerical quadrature Here is a program of study for the numerical quadrature methods. European calls, puts and binaries on a single equity using Normal numbers:
Very simple. You will be evaluating a single integral. European calls, puts and binaries on several underlying lognormal equities, using Normal numbers
Very simple again. You will be evaluating a multiple integral. Arbitrary European, non-path-dependent payoff, on several underlying lognormal equities, using Normal numbers:
You’ll only have to change a single function. Arbitrary European, non-path-dependent payoff, on several underlying lognormal equities, using low-discrepancy numbers:
Just change the source of the random numbers in the previous code.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 94
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Summary of Numerical Methods The following table summarizes when the various methods excel and where they fall flat. Subject
Finite difference
Monte Carlo
Quadrature
Low dimensions
Good
Inefficient
Good
High dimensions
Slow
Excellent
Good
Path dependent
Depends
Excellent
Not good
Greeks
Excellent
Not good
Excellent
Portfolio
Inefficient
Very good
Very good
Decisions
Excellent
Poor
V. poor
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 95
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Programming skills This is central to the notion of doing things. As a Quant, you will be writing code, it will be a significant part of your work. That is the way it is. Before you consider a career in quant finance, you need to be brutally honest with yourself on how you view dealing with computers. The standard figure across the industry is 60% of your working hours, and hopefully you understand statistics enough to work out that it follows that you will often be spending more than 60% of your time fighting computers, and some weeks will have no other work at all. We emphasise this because some new entrants to the market have somehow that programming is something they can avoid, or even worse, that it is somehow beneath them. Frankly that is rather like the view of physics you get in the media where no maths are involved. Some people really do seem to come to this line of work with a model that is like the one you might have of space travel you get from Star Trek. No zero g toilets for a start. Yes, people do think great deep abstract thoughts, but you can no more be a quant who doesn’t program than a physicist who doesn’t do calculus. Both It is as hard to be a successful quant who doesn’t animals exist of course, but they are rare, and do programming as a physicist who doesn’t do most attempts to pull this trick off lead to calculus. misery, unemployment or both. This is made worse by some of the various Masters in Finance programmes which either don’t teach programming at all, or fail to allocate enough time to it. Overwhelmingly the skill that employers tell us is lacking in newbies is programming. Sadly there is no quick fix. The Certificate in Quantitative Finance places a great emphasis on practical techniques, numerical methods and programming and has become something of a market standard for this reason, and now has a substantial C++ component. But no course, no matter how good will make you a programmer. Dominic may spend quite literally days explaining how to beat C++ into submission, but in the final analysis, Dominic teaching you about C++ is like Sharon Stone doing your sex education at school. It’s the practical that matters. However you come to learn about programming you have to write code with your own bare hands, everything else is like watching cheap 1970s porn on a tiny iPod screen. No one is going to pay you to do either of them based upon that level of expertise. It is very easy to spot people who haven’t written much code, and although often they get a job, they frequently miss out on the job they really want. The programming ability of quants varies enormously, but since a large percentage of Expect that 60% of your working life will be spent your time will be spent fighting computers, fighting computers. If you can’t take that, quit now. showing aptitude is frequently a deciding factor. Most quant jobs ask for C++, with much smaller demand for C# and Excel VBA and Java. Although Excel is the second most common skill, alas Excel VBA is regarded as “trivial” so few employers will be impressed by mastering it. This attitude is responsible for major efforts at the large banks to defuse the vast number of actively disruptive, yet critical spreadsheets that enjoy the reliability of a British train. Although VBA is very common in quant work, an employer who is particularly keen that you have it is slightly worrying. Some quants get sucked into roles that we call “Excel Jockeys”. Although some Excel work is cutting edge trading floor work that makes money every time you get the sheet to work properly, the majority is looking after risk reports, data ingest and sheets that even users don’t know why they have.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 96
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
MatLab is common in both academia and finance, and it does not harm to have used it, but again like Excel VBA the view amongst managers is that if you are smart enough to do real quant work you can pick up whatever MatLab you need though we do see a few jobs for extreme high end MatLab gurus. Fortran is still quite popular in academia, but has only a tiny market share in banks. If you have the choice for which language to use in a project whilst studying it is worth considering C++. However you should budget in extra time to master it properly. But if you’ve got good skills in numerical analysis in Fortran, then you’ve proven you can do the work. Some managers appreciate this, although others do not. Much “C++” code in banks is really much like C, and one does see “C++” code that is written using the C subset but trying to be like Fortran. Apart from understanding the subtleties of the language, you may need to demonstrate a general understanding of software engineering. Although efficiency is, of course, good, what matters more is that the thing actually works and in financial markets you have to show a pragmatic view of how you trade off these things. There are at least two pertinent programmers’ sayings: Fast, Reliable, Ready Soon: Pick any two Easier to make correct code fast, than to make fast code correct • •
Academic code is light on things like error handling and checks to ensure input data is of the right range. Get in the habit of thinking of “sanity checks” for all numerical values. “Working code” is not code that happens to produce the right result when you feed in carefully chosen values in a tight range. It means functions that have been tested across a wide range of values and whose output has been compared to known good values. A common symptom of a bug like uninitialized variables, or a stack error is for floating point numbers to contain absurd values like 6.12122*10 302 Deterministic routines should always return the same result and Monte Carlo should have known convergence characteristics. You should get in the habit of making your code tougher, assuming that people will actively try to break it, because that is rather close to the truth.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 97
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
C++ C++ is a topic in its own right and we’ll soon be releasing an upgraded version of the C++ for interviews notes. An outline of skills required for C++ is:
Variables You should know about scope, stacks, and static variables, that comparison of unsigned and signed integers can be “interesting”, that doubles are not reals, and that comparing them for equality is not safe.
Pointers You should know the difference between new and malloc, when to choose either, and that you must not ever mix them together. Returning a pointer to a stack variable, and allowing resource leakages are common interview bugs for you to find. Knowing ** as well as * seems to separate those who are seen at interview as having only a superficial skim from those who have understood the language. You should also understand the relationship between pointers and references.
Classes Obviously you should know private, protected and public; but you should know why the default is private, and use good programming practices like avoiding data members being public. Learn the difference between a class and a structure.
Passing by value and reference C++ passes by value, unlike languages like VB. This means if you change a vaue in a function, it doesn’t change as far as the calling function is concerned. Think about why it can be much more efficient to pass a reference rather than a value, and how const can make that safer.
Templates You should be able to create basic templates as well as use them. You must be very clear that the overhead in templates is extra code space which is normally trivial, but that that at runtime it is essentially zero. This is why STL code which uses templates is typically faster than the code you might craft by hand, which as well as not being obvious may be a misconception shared by your interviewer.
STL (strings, vectors and algorithms) Very dangerous to put C++ on your CV these days without at least basic competence in the use of the STL. Common, if silly, interview questions include being asked to write the declaration of a vector of linked lists of queues.
Inheritance up to the level of virtual functions You need to have a grip on polymorphism, and please do not confuse it with templates, expect to be asked “which function gets called” questions at interview.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 98
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Basic OO
OO is not just inheritance it is a way of thinking about things in a way that models behavior in a way that makes them robust and easy to use.
All of the above represents a lower bound, below which you should seriously consider whether to put C++ on your CV without a qualifier such as “basic knowledge.” C++ is a rich domain for questions about syntax that look like someone banged their head on the keyboard. It is better to have a good grasp of the core concepts of C++ than in a short time attempt to read up the whole language. Common interview questions are of the form “which function is called,” and “what’s wrong with this code.” Often you will be expected to describe things like virtual inheritance or template specialisation.
Boost If you want to be taken seriously as some sort of quant developer or for your claims of being an “expert” in C++ then be prepared to justify this in terms of keeping up with current technology by mastering at least some of this set of libraries.
Numerical algorithms It’s good to have at least a basic grasp of the basic methods for solving PDEs and other maths problems. Most QF uses a mix of Monte Carlo, trees and finite-difference methods. It’s worth getting to the stage where you can sensibly say when you’d choose one method over another and their limitations and advantages. There is also some demand for Matlab and Mathematica. These are mostly used for prototyping ideas, although they are also doing live work in trading. Similar issues apply to quiche languages such as Java and C#.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 99
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Reading list for C++
If you are going to be selling yourself partly upon your C++ skills, then you must read the latest edition of Stroustrup.
If you are making the assertion that you are strong in C++, then you should read one, or preferably both of: Exceptional C++ More Exceptional C++
By Herb Sutter, published by Addison Wesley.
Sutter takes the approach of taking specific aspects of C++ and drilling down so that you understand them at a deep level. Neither are good as your first C++ book, Stroustrup is still about the best for that, but a more gentle introduction, and one that is preferred by many is Bruce Eckel, Thinking in C++ http://books.globalinvestor.com/books/20552.htm?ginPtrCode=10202&identifier=889da2f4043309ec6be6ce8b0667 c79c It can also be downloaded free at: http://www.mindview.net/Books/TICPP/ThinkingInCPP2e.html
Two good books on C++ for quants are: Financial Instrument pricing using C++ By Daniel Duffy, published by Wiley. This combines financial methods with patterns and is best thought of as your second book in C++ after you have worked through Stroustrup. Includes interfacing with Excel, important because a large percentage of C++ is front ended by Excel. http://books.globalinvestor.com/books/17081.htm?ginPtrCode=10202&identifier=a49d71be25a99088bb7a3d4aac82 ff74 C++ Design Patterns and Derivatives Pricing By Mark Joshi, Cambridge University Press Mark manages to get books into both the mathematical finance and quant programming sections and this is a relatively gentle, yet rigorous first book of C++ quant development. http://books.globalinvestor.com/books/17705.htm?ginPtrCode=10202&identifier=a49d71be25a99088bb7a3d4aac82 ff74
If you want practice understanding some of the more gruesome things that people do with C++, then the International Obfuscated C++ contest will provide you with a different perspective of how you build code. http://www.de.ioccc.org/whowon2005.html
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 100
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Dealing With Offers The Tipping Point
At some point during the process, it will switch from you selling yourself to them, to the bank selling itself to you. After what may be a protracted series of interviews, tests and meetings they have decided they want you. You now represent the payoff of a serious amount of investment in time and effort and they will try quite hard to make sure it does pay off. It’s not uncommon for them to have interviewed anywhere between 10 and 20 people, and we hear tales of some firms sucking in 50-60 for every one they actually hire. It is also the case that your headhunter will be very close to actually getting paid, so he will want you to accept as well. A decent employer will not expect you to make a decision on the spot and will allow a week or two for you to choose the option that works best for you. Unless the other interviews are for very inferior jobs, then it’s not wise to cancel them, no matter how much a HH tells you they are a waste of time. The better type of HR people see their role of supporting the business decision to take you on by a combination of checking you out and selling the bank to you.
What is an offer? Currently the market for quants is pretty good, so a verbal offer of a job is perfectly acceptable. But times change and it is only a real offer when the paperwork arrives from HR. Do not turn down your 2nd or 3rd best offer unless and until you get a piece of paper from your new employer. This will usually take far longer than you might expect, because of bank bureaucracy. If the market goes into a decline, then you should be very conservative about shutting down any offers until you have a written offer in your hand. For the avoidance of doubt, that means continuing to go to interviews no matter what the headhunter has placed you says. Some get very pushy about ending your search when their client makes a verbal offer.
Delaying Job Offers This is not as impossible as many entry level candidates seem to fear. The trick is to put this into the process at the right point and in the right way. The two wrong times are at the start, and at the end of the interview cycle. Many headhunters flatly refuse to deal with candidates who are not available within the next couple of months. Partly this is because they personally may not be paid, and often it will make you harder to sell to the firm. Conversely, it can be a terrible mistake to wait until the offer is made before letting your hiring manager in on your plans. This is because it takes significant effort by the hiring manager to get official permission to hire you. Most large firms have excessively bureaucratic and political processes whose sole function seems to make hiring as hard as possible. The trick is to get the manager on your side, once you feel he is coming around to the idea of hiring. Your attitude should be to make sure that he knows what he is buying.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 101
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Process
It is sometimes the case that there is no formal approval for the position yet, so your new boss has to go through internal processes to get it approved. This especially affects quants since your work will often be paid for and used by multiple groups and each has to both approve of you and the headcount that you will fill. This may take place in a short meeting soon after they see you, or over a few weeks as they see others and someone tries to coordinate the process enough that a consensus arises. As these people will be in more than one business area and sometimes different countries, this is not always rapid and is quite badly affected by Christmas and the summer holidays. In a slow market, many firms make the recruitment process considerably more bureaucratic and slow, as a primitive Soviet style mechanism to push headcount down. You may not be the first choice, but especially at the entry level we have many people going for many jobs, so they may miss the one they want, although of course this is not usually communicated to you. They then tell HR to send out an offer. They will have processes of their own and it is not exactly unknown for the first thing for HR to know of your employment is when they are told to make you an offer. In some firms the generation of contracts has been outsourced. This is done for cost reasons, not efficiency in time or stress. We have heard HR managers use the same language about getting their outsourcer to actually send the offer letter that you might hear from someone lobbying their government to change policy, with the same air of resigned low expectation, of a good or quick result. All these steps may not sound like a robust or rapid process and we are sad to say that is quite often the case that things go wrong or take more time than it should. If you’ve got a headhunter helping you through this, we will (very gently) pester the various people to help move the process along and give feedback where possible. One person who applied personally without our help recently had to wait several months while the process ran its rather random course, the record is 5 months between verbal and written offer and we expect that to be beaten in the current market. If we’re involved we can make this a little less stressful, but in the final analysis the bank will do what it wants, when it wants to.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 102
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Turning down offers
You should be careful about turning down other offers until you are confident that it is a firm offer for the job you want most. Earlier versions of the Guide were considerably more relaxed about what exactly made an offer. We now strongly suggest that under no circumstances do you turn down any offer from anyone until a piece of paper arrives in your hands from the firm you want to work for. This will not make you popular with some headhunters, to which we respond “tough”, feel free to quote us on this. Legally, a firm can usually kill your offer, with no greater cost than a month or two or salary, so even a written contract is not a risk free instrument. As we say above in the description of the “headcount” process, there are now more approval barriers to you being hired, and if one gets in your way it can lead to severe delay, or the verbal offer never actually making it all the Things have changed since we wrote Guide 1.x way to a contract of employment.
Our advice now is not to even consider ending your search for the right job unless and until you have physical paper in your hands.
They will typically set a deadline of 1-2 weeks for you to make up your mind and even if you have no other useful options, take some time to think it through. If you’re applying personally, then you have to make sure you tread a path of keeping in their minds and really irritating them. No one likes nagging emails, no matter how polite you are. Either they remind them of the fact that’s there’s more bureaucratic work to be done, or they’re waiting for someone else. The politeness we mentioned above is really valuable to you here. Once you have an acceptable offer, then this can form a useful hedge when trying to get other firms to speed up their processes (but not Goldman Sachs). Since you have a limit on your exposure, you can take a slightly more assertive line with the laggards. They will know that the offering bank will be putting pressure on you to accept and that you have limits on how long you can wait.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 103
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Getting More Money
Once you have a firm offer, you have hedged against the downsize so volatility has a strictly positive value, which means you can ask for more money, or some other concession from the firms that have yet to make a formal offer. A straight demand is almost certain to fail, but dropping it into the conversation is good, or if you are not dealing directly informing your headhunter he has a number to beat is the best, since he has a straight Heaviside function for placing you: no job no fee. It is important to understand the details of the utility function of any headhunter in thi context. You probably know that in 80% of cases, the HH gets a % of your starting salary from the bank or hedge fund, and thus the more you earn, the more they get to spend on drugs and alcohol, or in the case of Dominic, funding a remake of Captain Scarlet. But they also care about the probability of placing you, because of the binary nature of the option they are trying to sell. Once they know that the only way they can get any money is getting a better option, then they have a simple clear incentive, which is always a good way to manage people. Some managers and headhunters will try and get you to accept extremely quickly. If this is the job of your dreams, then no real harm is done by accepting quickly, indeed in a tight market, delay does have risks. But if you are being sold-to heavily then it’s worth making sre you are making this big decision with the clearest view you can get. It is of course flattering to have a major financial organisation offer you good pay and really want you to join them, but do not let It is as hard to be a successful quant who doesn’t that blind you to the situation. One very senior do programming as a physicist who doesn’t do manager at a global firm once described his work calculus. as “persuading smart people to join us”. These bosses can be very persuasive when they want to be. It’s fine to enjoy the moment, but as a new entrant to the market you can say that you are very keen but because you are new to this, you think it’s best for all concerned if you make sure that it’s the right thing for you. Sometimes a headhunter may say that you have to accept today, or they’ll walk away. This is not likely to be true, both the HH and the bank have invested quite a number of hours in finding you and they’re not going to let that go to waste, unless you provoke them. Do try to get the manager’s direct line, for just this sort of situation.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 104
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Spotting bad jobs Objective Information Once a bank has decided it is interested in you, then they will try quite hard to paint a good picture of the job, and its prospects and opportunities for you. They are selling to you, sometimes quite hard, and it’s important to try and get as much hard information as you can. The odds are that if one bank wants you, then so do others, and you may need this information to help you decide. The trick of this is to get useful data, without appearing arrogant or “difficult”. Your questions should be pitched as those of an interested person trying to understand a large complex outfit. There is no need to be confrontational, or to appear arrogant. We’re always happy to offer personal advice on how to evaluate offers via the website or email. Questions you should get answers to Most managers are not trying to deceive • •
• • • • • • •
you, but they have a lot to communicate Where will I be sitting ? in a short space of time, and as a What is the actual name of the area which I be manager they have a different point of joining ? view from the face workers. Who would be my boss on a day to day basis ? Whose group am I in ? Which bonus pool will I be in ? What sort of work is the group doing now ? How will the work of the team change over the next year ? What other groups do they interact with, and how much ? Will I be using Perl ?
Most of these questions are at least partly stealthed as positive interest in the work and your contribution, except maybe the Perl, and we’ll come to that… Not all of these questions sound useful, and we suspect are not very intuitive, especially the first: Where will I be sitting ? Interviewers like emphasize that you will be close to the money. Some use this exact phrase, or variants such as working closely with the traders. A slightly worrying version is when they say “working with the business”. That implies that you aren’t part of it, and is a common term used in IT roles. Physical location is a big thing in banks, and whereas the majority of quants aren’t on the trading floor, distance does matter.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 105
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
What is the name of the area which I will be in ? Which bonus pool will I be in ? These two questions have distinct answers, and you need toask both. Sadly we observe that headhunters are themselves sometimes lied to on to these important questions, or at the very least not given correct information.
The name of the area is also a real issue, and it is not the same as your bonus pool. It is pretty common for your work to be with Team A, but you to be in an entirely different bonus pool. It is not exactly unknown for you to be told you will be working for Credit Derivatives Trading, but for you to be in IT or middle office. This can make a huge difference in what you get paid, and your future career progression. The work will also be different, perhaps in subtle ways but careers look remarkably like Martingales, so if you start down one path, the expectation is that is your direction for a long time. And yes, it is not unknown for you to be told explicitly that you are in one group, but actually be put in another. Sometimes it is not an issue, but 80% of the time it means that you aren’t exactly where you wanted to be. Who would be my boss on a day to day basis ? Is of value, since you can focus upon the sort of work he does as a proxy for what you will be doing. It’s also useful from this part of the information gathering to try and form a model of the background of senior staff. It’s generally a good sign if senior people “look like” you. Of course “working for X” does not equal “in same bonus pool as X”. What sort of work is the group doing now ? How will the work of the team change over the next year ?
Ask yourself two questions about your bosses : 1) How much like me are they ?
This question has at least two payloads; not 2) How much do I want to be like them ? only does it give you firm information, it is a hook you may use to mention skills and experience you have that are relevant. Although you are being hired for years to come (hopefully), the idea that you may be immediately productive can be very attractive to the hiring manager. What other groups do they interact with, and how much ? As above quants often work in what is usually called “matrix management”, and thus you can end up with more than one boss. One candidate referred to this as a “mother boss” and a “day care boss”. The degree to which this happens controls the work you do, what you learn, and the money you get for it. A particularly bad case is where the boss that sets your bonus does not see your work. There is upside, since exposure to other areas helps you learn, and at entry level this should be a higher priority than money.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 106
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Which other groups will I interact with? Entry level roles involve a lot of learning, so it’s good for this set to be relatively large. A few entry level quants are tightly bound to one or two people. This is an amplification term rather than being simply good or bad. If you build up a rapport with experienced people in your chosen field, then they can teach you a vast amount. However, your diversification is much lower, and thus risk goes up as well, if it doesn’t work out.
As we say elsewhere in the Guide, you will spend more time with the people you work with than who you are sleeping with, and you need to try and evaluate if you can get along with them. In our experience the probability of this going badly wrong is actually quite small, but the tails are quite long…
Perl At P&D we had to think hard about whether we share this piece of information so publicly because it is so useful, but here goes. If you are a quant developer, then Perl is a good way of pumping and probing data with less effort than C++, even if it is a somewhat slower. However is you are a quant, then almost any mention of this tool should worry you.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 107
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
IT It’s normal for a good percentage of your interview to focus on programming. However it is not unknown for people who think they are being interviewed for entry level quant jobs to get sucked into IT roles instead. It’s worth asking what percentage of your time will be IT and what percentage analytical. Many people express to us the notion that they don’t want to program at all. This is a worrying view of the work of a quant. The consensus is that the average quant spends 60% of his time in some form of programming. This may be C++, Excel VBA or Matlab. More than one manager has described to us the first year of a quant’s career as learning to code. As Peter Jaeckel, a famous quant put it, coding is how we turn maths into money. You must be prepared to be self sufficient in coding. Many firms have specialist developers for the final implementation, but you will still have to prototype and manipulate your own data. Probably the greatest single concentration of difficult career decisions and pitfalls are those experienced by people involved with computers. As we say elsewhere programming is a critical component of being a quant, but it is very easy to be landed with a job that is pure IT with no quant aspect to it at all. Sadly it can start with the description of the job you get from a headhunter, which may “lose” important facts about the role. Though this is less of a problem than it might have been since most are tragically bad at hiding the truth. If you are in any doubt, ask for a copy of the original job specification, since these are usually more honest. However they may decline to do so, sometimes because they haven’t been given one themselves, and sometimes because they just want to get you in front of the hiring managers, just in case they get lucky and you take the job. There is no single simple test for what sort of job has the right balance of technology and other work for you, but we will give you some tips on avoiding the blackest holes. You may not be a star programmer, but you have to attain basic competence, if you want to make a success of this career. We’re always happy to offer personal advice on how to evaluate offers via the website or email.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 108
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Being Foreign We don’t care Quant finance is a business area where nationality or other superfluous characteristics are not of any real significance. Turn up to a Finance Focus or Random Walkers and you will see a very wide range of people, since the core issue is how well you do the work. Some managers at banks are quite smug at having succeeded in an environment they regard as a global meritocracy. Banks are sensitive to discrimination issues, not least because it has cost them a lot of money in the past when they have got this wrong. Thus you may be asked about your age, race and some guess made about your sex. We say “guess” because a scary percentage of people who’ve seen Dominic’s online details thinks he’s female. Whilst working at IBM’s research labs, the HR department became quite intense about “Dominic’s maternity leave” at one point. Dominic was not even an employee of IBM. Some banks now directly ask applicants to fill in “diversity” forms ; this information is mostly to make sure that they can honestly say that although only 4% of their quants are women this is because only a tiny percentage of the applicants were women and similarly amongst various other groups. Like many other HHs we don’t put irrelevant stuff on your CV, and our policy is that we don’t change your CV at all. If we spot a wretchedly bad error, then we ask you to fix it, but the important word is if …
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 109
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Visas
Most of the good jobs are in London, New York or Chicago, so a large percentage of people applying to be a quant, are not from the country that they want to work in. Although Paris had a viable finance centre, pay is typically around half that in the major centres. Banks will sometimes help with visa applications, but they generally don’t want to, partly for reasons of cost, but mostly because it’s more work. The more you can reduce the effort of hiring you, the better your chances. In the US market we see a general random tightening of the criteria for non-resident students wishing to pursue long term careers in the states. In most cases you will get a year, but that clock starts ticking rather quickly, and in that time you need to find a firm that will fight the good fight on your behalf. Employers are increasingly asking only to be sent details of candidates who can start work without these issues. The core of the problem is not merely the time and cost, which can easily be outsourced, but the apparent lack of any real way of knowing whether you might be allowed to stay or not. If applying to a bank in London then you should look at the Highly Skilled Migrant Programme, which is neutral about your country of origin and is optimized for people whose presence brings financial benefit. We are now finding that it actually is a useful minor qualification in its own right since it means that some official has actually checked your earnings and academic qualifications, and of course it mean they don’t have to worry about visa hassles, so please make sure that it is at the top of your CV. Most HHs and HRs understand it, which makes it a good “brand”. The system seems to be scrupulously fair, but alas it is run by a section of the British government whose own minister refers to as “dysfunctional” and “not fit for purpose”, occasionally they rearrange it’s structure on a random basis. Thus you must make sure you don’t expect them to think at all . You must include every bit of paper they ask for, and preferably support any fact by more than one written piece of evidence. You need to make sure all translations of qualifications have the right translation and that it is certified in the right way. You may wonder why we make jokes about Accenture and EDS in this guide. Guess who does Home Office systems and consultancy ? But at least it isn’t as bad as the US system… Fortunately, the large banks have people who can deal with this nonsense, so your goal should be to motivate them to press that button. What they do not want are surprises, so you need to research your current status properly so that you can give them a reasonable idea of chances and effort.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 110
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Highly Skilled Migrant Programme
The HSMP is a points based system for those intending to move permanently to Britain. You get points for education, age, and how much money you have earned. There are currently no quotas based upon country of origin. Although slightly bureaucratic it is scrupulously fair, but you should allow time for the process to grind away. To make the most of your HSMP, you need to make sure that you have every single piece of paper that they will ask for, and if possible more than that supports each issue. In one case a candidate we were helping, found that the British authorities would accept Indian pay statements, but not British ones as proof of earnings. No, we don’t know why either, we said “fair”, not “efficient”. This was solved by acquiring copies of the relevant tax records. If your documents are not in English you will need translations. The HSMP is recognised by nearly all banks, and with a bit of prodding hedge funds and consultancies can understand that this is the gold standard of visas. It is not dependant upon you having a specific job, and after a number of years you get a full British passport. The HSMP is actually treated in many cases as a form of academic qualification, since it means that your background has been officially checked and found to be good, so you should make sure that if you have passed, it is on your CV.
Education If you have spent any part of your education or work experience in an English speaking country, then it is worth including this in your CV, just to emphasise that your language skills are OK, even if otherwise it has little value.
Multiple Languages Again on the point of language skills, we sometimes see an impressive list of several tongues, but once you have demonstrated skills in English, it does not help your case to list many other languages since it may give the impression that English is not even your second language. Unless the role you are targeting will definitely use a given foreign language, it is probably best to leave it out.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 111
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Being Female This is actually one of the harder elements of the guide to get right, not least since “Paul & Dominic” aren’t female, but here goes anyway. As you’ve already spotted, the % of female quants is so low that we find it hard to accurately measure, but looking forward we see from our work with entry level candidates that the trend is sharply up, especially amongst people entering via the Wilmott CQF or a Masters programme. Although the banks are at pains to talk of their equal opportunity policies, informal conversations paint an entirely different picture. It is the case that the hiring managers we talk to are quite positively inclined to interview more female candidates, and so the prior probability of getting an interview is significantly higher. That sharply moves the game in your favour. For this to benefit you, the hiring manager must know that you are female. This is pretty easy if you have a name that he (and it almost certainly will be he), recognises as female, but it does not hurt to put “Miss” or “Mrs.” in front of your name. P&D has a variety of contracts with banks for the supply of quants, and almost without exception we are explicitly forbidden to add details about sex, age, race, etc. This applies to other firms as well. For some reason this is rare in hedge funds, but we assume that is because they have yet to experience large scale lawsuits. However, P&D do not meddle with your CV, so if you want to put this stuff in, that is your choice. This is not a sure fire 100% guarantee, since some managers and headhunters have an attitude towards women, but it is low cost since if they don’t want to employ you, they won’t, so there is little to lose, though we do advise against the use of “Ms” which is dying anyway. Most of the larger banks have adopted some sort of diversity strategy, and several have introduced networking groups, which can be useful, once you have a job. As with race, the banks track applications from women, but their systems are set up so that ticking the “F” box is separated from your application, and is used for statistics and reports. We have seen a variety of groups which purport to help women into finance. Firstly we genuinely do not believe that women actually need more help in quant finance, provided they have the right stuff, and second we have yet to find any female quant who has actually benefited from such help, but we are always happy to learn more.
Networking Related to this is meeting people in this industry, both to learn how things really are, tricks of the trade and to make useful contacts in terms of getting employment. It is easily noticeable that the % of women who turn up for Finance Focus, Random Walkers and other events where you will meet your peers and older hands is far lower than men. Maybe you have better things to do with your life on any given evening, but the cumulative effect is quite harmful to your career progression, made worse by the fact that the loss you incur by missing them may fall below a level that you notice.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 112
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Being Older In the USA people tend to emerge from PhD programmes in their very late 20s or early 30s, and in Britain slightly sooner, but in either case they mean that the average age of people entering quant finance is rather older than many people suppose. Further, people in banks have relatively short time horizons, so the fact that you have couple of fewer years left in your career simply does not matter to them, usually. Of course some managers will reject you for that reason, and the fact that it is against the law of is relatively little use to you. But in our experience the principal emotion in the mind of a manager when he finds an older candidate that can do the job well is relief that he can stop looking at get on with doing the job he is actually paid for.
How Older Quants Fail Some managers fear that you will not be able to cope with the hours and stress of working in a bank, and before you start down this path we would seriously urge you to think this through honestly for yourself. Banking hours vary quite a bit but 8 to 6 is pretty common, with 7 to 7 far from unknown, and this may include weekend work. Bankers like to characterise themselves as results driven but sadly a large number of managers like “face time” where you show commitment by being there not only when it’s not necessary but often working hours past your personal maximum so that you actually achieve less. If you feel your age is hurting your chances, then leave it off entirely. This works better than you might think since the first pass at CV reading by interviewers rarely is sophisticated enough even to include simple arithmetic from graduation and other dates. Again repeating a message given elsewhere, you should try to have more than one variant on your CV, and dropping age is an obvious and easy approach to adopt. Hours Typically the closer you are to the trading part of the business, the longer and less flexible the hours, but the more money you get per hour, though very few people are paid on an hourly basis. This means that there are roles in risk management, model validation and IT which allow a better work/life balance. There are also jobs in software houses, ratings agencies and regulators where although you will typically earn less, the hours will be rather better. The picture for consultancies is less clear, some work very close to “normal” office hours, and in some cases they seem to put in longer days than anyone else.
It is quite possible that age discrimination laws actually work against you here. Ideally a manager will be able to ask questions that allow him to work out that you can do the hours he expects, but of course HR will have told him that this is legally risky, so he will just make up his mind without your input. This is probably not going to help you, so you need to take the initiative, without drawing attention to age by saying things like “although I’m old, I can still hack it”. You should instead go for dropping into the conversation, a few examples of long hours and stressful situations in your recent history. Naturally, these should not seem to much like complaining how tough your life is, but as amusing anecdotes.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 113
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Things you need to show
The single most common failure mode for older candidates is where they do not show that they have done something with the extra time. Of course in an ideal world these are things that have direct relevance to finance, but it is not always necessary, if you can show that your non-finance career was a success, and that you are not bailing out into finance because it isn’t working for you. That means mentioning progression based upon success, projects completed and that you have continued to learn even after your formal education ended. You have the opportunity to demonstrate that you have done things, and if you do not, the silence will speak loudly. You also need to review your experience in some detail and think through how it may be applied to finance. This should be done in parallel with the reading we encourage you to do as you start this change in your career. This should not be as difficult as it might at first appear, since quant finance is a new field which has stolen most of its techniques from physical sciences, Make sure that you can demonstrate the extra engineering, standard mathematics, computer time in your career has been used to some good science and of course gambling. However, if it effect. does turn out to be very hard, then we would counsel you to seriously consider whether you are doing the right thing.
Being an Old Newbie You will have to be clear in your own mind that you are not going to be joining at anything that resembles a management position, no matter if you get called “vice president” or anything else. From above, the sort of person who makes this transition successfully has typically earned the respect of their peers, and carries weight in the organisation they work for. This ends now. When you turn up at a bank, you are a newbie, you will be taking orders from people much younger than you, and quite possibly less intelligent, and apparently for no good reason, you may even be right and them hopelessly and expensively wrong. Tough. Not only do you have to deal with this, it is good to express the fact that you can to the interviewer, and again if you can think of examples of the variety of people you have reported to, and how well this worked out. Please don’t use anecdotes of the form “we once had this kid as a manager, but we soon sorted him out”. Management experience has to be handled carefully, since in many organisations your use of analytical techniques and real programming drops off very quickly the more people you get to boss around, and although your current employer may value your skills at budget management, Powerpoint presentations, and the way you rapidly answer mail on your Blackberry, these do not help you in quant finance one little bit.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 114
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Being old at Interview
Some of the factors above are things that you cannot influence greatly, but Bayes law is also on your side here once you get an interview. If a manager is interviewing you then it means that your age is anywhere between neutral and positive, so there is no reason to be defensive. In any quant recruitment, you have to bear in mind that they want you to be useful. We’ve calculated that each candidate that walks into a bank or hedge fund for an interview process is going to cost them about $2,000 to 3,000, and at firms like Goldman Sachs with their more extensive process and rather expensive people it’s nearer $10,000 in time that could be spent making money. You should therefore assume that they won’t bother if you clearly don’t meet their specification.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 115
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Dealing with Headhunters Disclaimer: OK, we are headhunters so apply whatever discount factor you feel appropriate, but we’ve tried really quite hard to be objective.
We started a headhunting business because we saw that quants needed help from people who had hardcore experience of working in banks and as well as quant finance. This guide is part of our ambition to go beyond the standard of being a meat market to a resource that people turn to when going through a quite stressful and important part of their lives.
Warning On the Forums, P&D often r efer to ourselves self-mockingly as “pimps.” We’re fine about being called this in return. Most headhunters do not like this at all , so don’t alienate the person who might be getting you a job this way.
How to work well with a Headhunter When applying for a job realize that he has to try and sell you to the bank, so arm him with information to pitch you to the client. Explain how you fit, both verbally and in the cover letter, and ask him what else he knows about the job so that you can review your background and find more of what the client has told him you are looking for, and so build a CV for this particular target. Be aware that you are not the only person he is trying to place, so will not be devoting every single waking hour to your cause.
How to spot a good h eadhunter A good HH has good jobs and understands the work. At any point in time a HH may not have a role for you, simply because the market is so fragmented. A good HH will be honest about this and chat to you about the sort of direction to head in. The majority of HHs are sales people; they get paid for selling you, not for building up relationships beyond getting you placed. That focus can work to your benefit, since their objectives and yours are to get you into a job as soon as possible. He should be able to describe the job and should advance at least some idea why he thinks you are right for the job. A good HH will have read your CV and identified things from your experience that need to be expanded to make your application stand a bigger chance.
How to spot a bad headhunter Some HHs will spam your CV to every bank on the planet, regardless of which job you actually applied to. This is for their benefit rather than yours. They also try to push you into giving the names of the managers you’ve spoken to at other banks, as we see later that’s not good at all. They may ask you to make them your “exclusive” HH, which is very bad for you at entry level. Some HHs are more than a bit “optimistic” about a job’s quality. This ought to be hard to spot, but often it is surprisingly easy. This is because they lack the knowledge of banking to cover their tracks. Thus you may well find a job in housekeeping IT described as a quant job. Yes, really, that is a daily occurrence.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 116
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Multiple applications
HHs get paid for finding people like you. But banks really don’t want to get involved in dealing with different HHs who have sent your CV and both want money if you get hired. Some have a simple policy that if you apply twice, then your application is rejected for both. HHs hate that a lot and it makes them look bad as well as costing money. Some HHs won’t tell you where they’re sending your CV. This is bad news and can cause this sort of problem. Of course it may stop you applying to other banks because you don’t know where your CV is being sent. You should insist on knowing the name of the bank before they send in your CV, some will refuse. None of the motivations for doing this are good. Some HHs will send your CV out without asking you and this causes all sorts of problems, yes it is illegal but sadly the organisations tasked with enforcing it do not bother. It is deeply frustrating when a HH can’t get hold of a good person to get permission to send and banking clients are The way some headhunters work is by sending not known for their patience. Typically it is your CV as part of a stream so large that it some sort of race, with more than one HH activates the spam filters at some banks. chasing each job. In several countries, it is illegal to send in a CV without permission, but it does happen. We (P&D) don’t do this at all, hence the slightly frustrated tone when we tell you to make sure your contact details work reliably.
Carpet Bombing Some HHs take this to an extreme. We have reports of senior managers getting a couple of dozen CVs from each such spamhunter per week, one apparently sent so many CVs it triggered the bank Spam filter. It’s hard to imagine that the candidates as cannon fodder actually agreed to be part of this human wave, or that it serves their best interests. But it does work for the HHs that do this, even if it harms the people they claim to be representing. If you’ve been spammed into a bank, then it may be the case that only the spammer can represent you there. Since they work by carpet bombing, there is no notion of them trying to engineer the process, so it means you may well be stuck with them. You can see why they like this, but even then many banks have a policy that if two HHs submit the same candidate then you will not be considered at all. They do not want to be caught between two firms arguing over who “owns” your application. This sort of collateral damage may be acceptable to the spamhunters but can have drastic effects upon your career prospects. The flipside of this is that your contact details must work and be on your CV. We do get CVs without working contact details and it’s not hard to see that that this does not advance their position at all. You will change your employer or university at some point, and if you’ve given that as your email address it will mean that you stop getting future versions of the guide, and make it harder for us to get in contact with generally. It is possible to get yourself free of being “owned” by a HH who has spammed your CV to a bank, and blocks other applications. Obviously as HHs ourselves, we have a variety of techniques, but a surprisingly effective one is to simply ring the HH and ask them to let you go. The fact is that in some countries this is illegal, and in most cases violates the contract that the HH has with the bank, so they really don’t want a fuss.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 117
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Names
HHs need to keep their list of contacts at banks up to date in the face of continuous change. As above the HH needs to know if you’ve been put forward for the role he’s currently working on. Sadly some HHs get a bit carried away with this and put pr essure on newbies to give up the name of the managers they’ve been seeing. This can of course work against you, since he may have other people who he can use to compete against you for that job, so it is not a good idea to be too free with this. A HH only needs to know a little information such as the division and bank, “Fixed income, Credit Suisse” is easily precise enough to make sure there is not a clash. Big banks are in some ways like a franchise operation. Different divisions rarely share r ecruiting processes and in some firms they have separate HR people. Also if you are currently in a job, you should be very careful about giving out your current manager’s name. It does not make friends to have your boss pestered aggressively by a HH before you’ve got your next role. This is of course easy for us to say, but if you’re reading this guide, the odds are that you’re new to this sort of thing. Some HHs are good at making you feel as if they personally are the only way you’ll get a job and that you’ll starve in some isolated village unless they rescue you. Remember of course that the HH that’s pressuring you to give up names doesn’t want the names he “owns” spread around either. Thus a good line to use is that you were told this information in confidence and that you will treat the other HH with the respect you treat this one. But in the final analysis, there is no shortage of headhunters and if one is giving you a really bad time, it’s not that hard to find another. However, contact names are rather valuable information and if you feel a HH has done a good job helping you in a move, you may feel that you want to reward that with some useful work. Think of it as a tip for good service.
Multiple targets For your later jobs, you want a relatively high degree of precision and it can be entirely rational to deal only with one headhunter, or even none at all. However at entry level, it is a bad sign if a HH wants any sort of long-term exclusive on you. The market is too fragmented for any single source to offer the degree of exposure you need. If the HH is pushy on this, it is worth remembering that there are about the same number of HHs as banks.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 118
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Exclusives
Some Headhunters will try and push you into dealing only with them. At entry level we think that is a truly bad thing for you. It is far more beneficial to the headhunter than the candidate and can cause serious harm to your job prospects. The simple reality of the market is that any given headhunter only sees a relatively small percentage of jobs on offer and at entry level considerable recruitment is done directly, without any search firm being involved. Also at this point there is significant competition for each role, so your chances of any particular application succeeding is not high enough to rely upon one or two shots. At P&D we don’t demand such exclusives for these reasons and although it does mean we spend effort on people who get placed other ways we are not in the business of damaging people’s career prospects. A headhunter who wants exclusivity from a new graduate is not a good person. (We’d use stronger language, but we’d probably get zapped by spam filters.) Once you are established with a good amount of experience, then you will want to apply a more targeted approach and at the point it can be rational to deal with a small number of HHs, possibly only one.
Multiple offers You may get offered more than one of the jobs you apply for and sometimes this seems more stressful than actually trying to get a job. Some HHs will see their commission slipping away and try to put pressure on you. This is made worse by the fact that the bank has put a lot of effort into finding you as well and may blame the HH if you don’t accept. You’ve got to take the job that suits you best. A good HH will of course put a case, but will realise when the game is up and store goodwill for next time. However, it’s not always clear cut what you want and again P&D are quite happy to provide a sounding board for talking through the pros and cons. We do this quite a lot, even when we aren’t the firm placing you; our view is that you’ll remember good advice for next time. Typically the ideal job will be a mix of the offers you have, especially in your first couple of jobs. So there may still be a value to the relationship with the HH that “came second.” He will usually be keen to try and keep the game going, given that he’s close to the final line and never ever forget that HHs are paid for placing people and they don’t get paid if they don’t.
See it from their side You should be aware that recruitment isn’t just hassle for you, but that the people at the bank don’t usually like it that much either and have other high priority tasks as well. Having to interview 10 people is not uncommon and at some banks anywhere between five and 12 people are involved in the process, which makes for a lot of work that they didn’t want to do in the first place and won’t earn them any bonus. So they will be quite sad about your rejection, and try to make you accept. There are two basic scenarios, either you narrowly beat the next best person, or were the first one they found who was any good. If it is the second type, then they are going to try harder to suck you in and this can help the negotiations.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 119
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Auctions
Banks do not like being drawn into auctions. The obvious reason is pushing the price up, but it’s worth looking at how people make decisions on what they offer. Typically the hiring manager has to explain his boss that he believes that offering X will get you on board. Often he has had to make this pricing judgment before he’s interviewed anyone, sometimes before even talking to a headhunter. This exposes him to risk and yet more of the recruitment process he so keenly wants to escape. This is a bank, so he will value his reputation as someone who can make good judgments about money and a process where he has to keep asking for higher levels of pay or rank for a new hire may risk making him look weak or foolish. Since these increases will require justification, an unrealistic set of expectations can easily be generated. If on the other hand he loses a good candidate because he’ wasn’t allowed to pay for him, a politically astute manager may have an edge for a future discussion on what his people are paid. There is also the delicate balance between what you pay people to join your teams vs. the people already there. In a rising market it’s not impossible that last year’s intake may be being paid less than is necessary to get good people now. Many banks try to keep pay rises to a single period each year, and that can amplify this situation more. Thus hiring managers have strong incentives to avoid auctions.
All or nothing If two jobs are close, then it’s worth seeing if you can make the second-placed one better. Although auctions are likely to fail, you can structure it so the “losing” manager can have one last shot. If you have a good HH, you can work together to see if there is a change in your level or pay structure which means you will definitely take the second-placed job. For this to work it’s important that both the HH and second manager feel that this is a real deal. To get you a worthwhile change they will have to sell the idea and risk some of their reputational capital. If they think that even after getting the better offer you won’t accept, then they won’t even try.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 120
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
How to Quit Often you can simply tell your boss you are going, wait a month or so and never look back. However there are a few obstacles that may well get in your way...
Notice periods and Gardening Leave As banks have become more concerned about the intellectual property you gain doing a quant job is that notice periods seem to have gradually grown. This can slightly interfere with getting a new job, since often the new employer will want you to start sooner. In general most notice period clauses are enforceable, so you will need to check on these. However, it is rarely a big issue and of course if you’re so valuable your firm doesn’t want you to go, that looks good. However, it’s worth telling your HH the terms, so that he can set the right expectation with your new employer. Some firms have “no compete” clauses and these are a naked attempt to improve their negotiating position when you leave. It is of course quite hard to get these removed until you’ve reached a stronger negotiating position with the bank.
The elegant way to negotiate your exit is to agree with your boss certain fixed targets that can definitely be delivered in the time remaining, then you leave free and clear.
Notice periods are usually enforceable, as are exclusion clauses on who you may move to. However, in legal terms it is technically difficult for them to get this right, since there is a lot of employment law to deal with. One of us was once offered a contract of employment that effectively meant that he could not work for any financial institution, company, the British government, the American government, or anyone who might ever do business with any of these. That essentially meant that he could only work in a coffee shop in Libya and then only if he didn’t handle money. This is not enforceable, surprisingly enough, but most banks are not that crude, so please feel free to contact us if you get one of these, and if necessary we can put you in contact with one of our reassuringly expensive lawyers. It is worth remembering that notice periods are usually enforceable, but a rational manager does not want a seriously unwilling member of staff in his team. Although it is rational, your boss may need a little time with himself to get in the right frame of mind for rationality to kick in.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 121
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Negotiation
If you find yourself with an inconvenient notice period, then by far the best way forward is to talk things through with your boss. Some people take resignations very personally, but it’s in everyone’s interest to make this as painless as possible. Before you quit, look at your current work and try to work out which ones are important to be finished before you leave, which really require you personally and those that can be finished relatively quickly. Then make a hard, realistic estimate estimate of how long it will take. The best deal for both sides is that you finish these jobs and go. You have an incentive to finish and your boss has some confidence it will happen. One of these jobs is to help recruit your successor, since of course you know what skills are required. As we say earlier, your boss won’t be looking l ooking forward to the prospect of going through piles of CVs and interviewing people of rather variable quality. You can also write a good job spec and if you’re dealing with the smarter end of HH then they will want to talk to you about the right sort of person. This helps make your ex-boss remain less stressed and means you can fade out leaving a good impression. The trick is to offer these things as part of a package. You are offering to Try quite hard to leave a good impression when leave well and in return they won’t give you you leave. The future is a big place, and there is a hassle over inconvenient parts of the contract. good chance that paths may cross again. again. Most people in banks are quite honourable about these things, but they may be secondguessed by some more senior or generally awkward person in the bank. It is not exactly unknown for your manager to “remember” a different version of events. Thus you should confirm the agreement in a mail and make sure a copy is sent to your personal email. It is now rare for banks to “lose” inconvenient emails ever since a large bank got hit for a serious legal bill when it apparently genuinely lost some correspondence that would have helped its case. The courts apparently took the view that the bank could not have been as incompetent as it claimed. Clearly the judges had never dealt with an outsourced IT department.
Buy backs It’s hard to work out what people are really worth and obviously part of your reason for leaving is that you think they have priced you wrong. If you are important to the team, then they will consider offering you more money. The same logic applies to the work you do and your conditions, or whatever else is driving you away. Some areas such as model validation are usually subject to quite high staff turnover because of a combination of all three, plus the feeling of many people who who work there, that if they don’t get out soon they never will. However it is generally accepted that about 60% of people who are bought back after quitting leave within a year anyway. Leaving your job cleanly is worth doing well. The people you leave behind may crop up again in your career and it is far from unknown for you to consider going back to the same same firm again again later. For For an example example of how not to show your unhappiness with your current firm : http://www.wilmott.com/blogs/dcfc/index.cfm/2006/5
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 122
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Money
This is the easiest to fix, which may surprise those who’ve been told quite strongly that there is no money available. This is often an industry standard lie, as you may find when you hand in your notice. Money has the advantage that it’s simple, to get more, if your boss can make the case with his superiors or allocate money from some other pot. It may be hard to do this, but is usually simpler than altering your work. Usually they will top the offer you’ve just been made, though with one person we know his boss only offered half the difference. No, we don’t know why they did this either, but as you can imagine it completed his resolve to leave. You have to ask yourself why you had to resort to this level of threat to merely catch up with the market. You can’t realistically go through this very often, so the clock is ticking on you leaving anyway. Investment banks are the nearest thing to pure meritocracies meritocracies on the planet and given that merit implies money, being underpaid means they don’t value you, which is not a basis for a long-term relationship.
Work Your boss may offer to change the work you do. This is typically very hard for him to do, since the “bad” work still has to be done. It is also hard to objectively define and such promises are often made in good faith, but not delivered. If there is a specific thing you want to work on, then it’s worth negotiating on. You may be offered the chance to move to a sexier group which is attractive, if true. This needs to be tied down well, a vague promise to try and get you more interesting work does not count for all that much.
References One obvious reason for being reasonable as you leave is that your boss may by asked to give you a reference. This is a murky legal area. Many banks are very hard line on giving references and require that all requests must be channelled through HR. They will issue a reference that says as little as possible to minimise possible problems. This is to try and avoid the possibility of being sued and even then is not foolproof. One large bank was sued after it appeared that they’d given a hostile reference for the wrong person. Thus many have a policy of minimal references, typically stating little more than when you started and finished. We have seen that some candidates feel quite offended by years of service being treated in a way that they perceive as abrupt and ungrateful. It is nothing personal, just the bank protecting themselves. However, HHs have the job of making sure you’re right for the job and that often includes includes trying to get a verbal reference and some managers will contact your previous boss for an informal chat. This is a dark art. Your boss will usually be very wary of saying anything that gets him or the bank sued, but most people are quite reluctant to actually be dishonest. Thus a code evolves of pregnant silences, silences, emphasis on words and euphemisms. euphemisms. This is far from perfect and quite a noisy signal, but is often all one has. The clarity of the reference is a function of how motivated your ex-boss is to say things and if he has strong views these come out.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 123
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Quitting under fire
We don’t recommend this, except in the most exceptional circumstances. It is a simple fact that it is considerably easier to get a job when you have one, than when you do not. Yes, it may of course be clear to everyone involved that your time is up, then of course you can apply the negotiating posture we discuss above. Quitting because you can’t work with you boss looks bad, especially to the person who is thinking about being your next boss. Thus you need to tough it out as long as you possibly can, whilst looking for another job.
Being a Blame Sink Some people worry quite a lot about how people will think about them when they leave, and however reasonable you are, in some places it is inevitable that blame for something will be attached to your name. Maybe it is because you left something in a less than perfect state, or that it is expedient for your boss to cite your departure as a reason for targets being missed. Some managers even go the other way, using the fact that they are losing staff as a concrete proof that they need more money for the people in their team.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 124
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Being Fired References Managers are mostly reluctant to say very bad things about you, even as we’ve seen in some cases where they fired someone because they could not bear to work with them any longer. This is a difficult area to give precise advice, but there are some general principles you need to try and use. Firstly in most firms it is quite hard to sack people, even if the country you are in makes that relatively easy. The bank does not want to be sued, as so often happens, and it’s messy for everyone involved. In particular, it is messy for your boss. He may by now be an enemy (we’ve seen the emails, it does happen “why your boss is a jerk” is entirely based upon experience), but you still both have interests in common. A manager can end up looking bad to his peers and seniors if he seems to be incompetently or even maliciously managing his staff. He cares about that (usually). You don’t want to be sacked, and your goal has to make sure that you influence the timing and conditions of your departure. Leaving with no job to go to has to be viewed as a last resort, since it is much easier to get a job when you already have one. If you are unemployed many will assume that was because you screwed up, whatever the underlying truth. Sadly, HR is rarely on your side, although we know of any number of cases where HR took an honourable stand against a defective manager. Quoting one HR director of a household name firm (with her permission), “our job is to protect the firm against the staff”. That is certainly how senior management would choose to define the job of people in HR. The big expensive legal cases you will have read about are often seen as being caused by a manager not getting proper legal and HR advice. These days managers are under instruction to get help early when it’s time to let someone go. Done properly, an end of employment can be relatively stress free. To quote one manager “our HR are pretty crap but they are very good at sacking people”. However this is not always the case, and one gets the impression that some journalists live off the pages they write about mismanagement by bank of this process. Thus it may be the right course to sit down with your boss, and talk it through. By this point, you are probably not going to be friends, no matter how much air is cleared. But you can reach an agreement where you will make sure that you hand over your work in a reliable way, and won’t make waves for him. In return you set a flexible timescale for your departure. Once there is a cap on the pain you are both enduring, work may progress better, and you should be able to buy yourself enough time to get another job. Whilst this process is going on, you will inevitably have some dark thoughts about your future. It is pretty rarely that bleak. Unless you are being sacked for doing something illegal, this is not the end, provided you try to negotiate where possible and are proactive in looking for another job. A much higher % of people with decent jobs in finance have been sacked than you might think.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 125
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Useful Links www.wilmott.com Lots of useful content, quant finance chat, articles, videos, job adverts. For the job seeker, the Brainteasers Topic is incredibly valuable. http://www.wilmott.com/blogs/dcfc/ Dominic’s Blog, on the job market. www.PaulDominic.com Register your CV with us. http://www.gotw.ca/gotw/index.htm Interesting puzzles and points in C++. http://www.ieor.columbia.edu/forms/JobsRiskColumn.pdf Derman’s view on the way things are now. http://www.wilmott.com/blogs/ The beta test for out new blogging environment, useful and interesting stuff to be found. http://www.fooledbyrandomness.com/ Interesting thoughts on probability, finance and well pretty much everything. http://math.ucsd.edu/~crypto/Monty/monty.html The Monty Hall Problem. http://www.mindview.net/Books/TICPP/ThinkingInCPP2e.html Good free C++ book. http://www.nr.com/ Numerical Recipes Home Site. This is a useful cookbook for number crunching. The coding style however is not one we recommend. At all. Even slightly. Don’t do it like that, OK? http://archives.math.utk.edu/topics/ Math Archives, lots in useful stuff. http://www.dartmouth.edu/~chance/teaching_aids/books_articles/probability_book/amsbook.mac. Pdf Basic Probability. http://www.dartmouth.edu/~chance/teaching_aids/books_articles/probability_book/amsbook.mac. Pdf Intro to Numerical Methods. http://www.7city.com/cqf The Certificate in Quantitative Finance.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 126
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
A few basic C++ questions What is Auto_Ptr for? Write a piece of code to reverse the order of characters in an STL string Why would you use a const ref?d Is it safe to test two doubles for equality? Why? What are lvalues? Declarations
We include the 2 general answers to the “what does this stupidly complex definition mean?“ In general it of course means that either the interviewer or the person who wrote the code is an idiot, but neither tend to make a good impression, so we present an overview. A simple one, what ‘s the difference between pd1 and pd2? double const * pd2; const double * pd1; Answer: None. int *a,b; typedef int* PINT; PINT a,b; * or & only bind to the token next to them. So in the first declaration, we have a pointer and an integer. In the second they are both PINTs, ie pointers to integer
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 127
Guide 2.0
Paul & Dominic’s Guide to Quant Careers double *(*(*Ex1)(double const*))[42]; Ex1 is obviously a pointer to a Function taking a pointer to a constant double and returning a pointer to an array of 42 pointers to doubles. But you knew that anyway didn’t you? An algorithm: double *(*(* Ex1) (double const *))[42]; 98631247 1: Find the variable name Ex1 Ex1 is
2. Look Right ) 3. look left * A pointer to
4. Look right ( A Function that takes Parse arguments (goto 2) That Returns
6. look left * A pointer to
7:Look right [42] an array of 42
8 Look left * Pointer to
9 Look left double double
int& (*fn)( const double * a , double& b ); Pointer to a function that takes a pointer to a constant double and a reference to a double and returns a reference to an integer.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 128
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
A Broken Class: Try to spot as many errors as you can
class TimeSeries { double * Price; double *TimeStamp; Tick * t; public: TimeSeries(const size_t count) { Price=new double (count); TimeStamp =new double (count); }; ~TimeSeries() { delete Price; delete TimeStamp; } protected: virtual double Volatility(void); public: TimeSeries &operator+(TimeSeries Operand); }; Feel free to mail Dominic@PaulDominic with answers.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 129
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Casting Off Finally fix the following snippets of code: unsigned int * pint = static_cast(&i);
This won’t even compile. Signed and unsigned integers are distinct types. Default in C++ is signed. You need reinterpret. double * d = reinterpret_cast (&i); Reinterpret will compile, but bad things will happen. You’ve cast the pointer, not the thing. Thus if you try to use the value pointed at, you will get some undefined value. If you assign to the pointer, it will be worse. char *pc = static_cast (& uc); This won’t compile, signed chars are distinct from unsigned chars and chars. Again you need reinterpret. It’s also worth looking at C style casts of the form Z = (double *) q; If presented with a piece of code and asked to talk about it you should home in on C style casts as sources of cunning and vicious bugs. Even if the error isn’t there, you will should earn a point for spotting them as bad coding technique.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 130
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Choosing Which Mathematics to Learn Next One question we get reasonably often is the choice between applied maths and statistics. There is a lot to be said for either, and our net position is roughly neutral. This is not because it is unimportant, but because the factors mostly cancel out, and the ones that do not are more to do with personal circumstances. It is the case that in most quant finance there is more applied maths than statistics, but conversely there are more applications from applied mathematicians and physicists. But, there is another more important point; what are you best at ? The career prospects when you are doing the type of maths in which you excel are much better than being mediocre at something more popular. Also of course, we would never advise anyone to stick to an entirely applied maths curriculum or pure stats if they want to succeed in quant finance. When you go for the best jobs, they will fire both types of question at you, and this reflects that what you are being paid to do is solve problems and think up new ideas. You are not being paid to be a statistician or an applied mathematician, or and we would emphasise you are not employed to be anything at all, you are employed to do, and the doing is the making of money. That means if the problem involves stats you do stats, if it's PDEs, you do MC or FDM, if it's programming you choose C++ or VBA. One of the brainteasers illustrates our position quire neatly. It is World War 2, and you are asked to work out where to put armour on the plane. You cannot put it everywhere because the plane would not be able to take off, and you have full access to the records of where bullets hit the bombers. We hear many interesting answers to this, based upon theories of which parts of the aircraft will be hit most due to things like the relative speeds of attacking aircraft and position of guns on the ground. Engines and fuel tanks are often mentioned. Like so many brainteasers, you must focus on the detailed wording of the question, and think why you were told about the records of bullet holes ? The answer is that for there to be a record, the plane must have made it home, after having have survived a hit in that place, so it probably doesn’t do much good to increase the armour there. The lesson here for designing your skillset is that at interview you will be asked questions about PDEs, even if you are a statistician who hasn’t done serious calculus in five or more years. Conversely we have seen candidates who can hack applied maths and programming but were dumbfounded when asked why there is an N-1 term in one of the types of standard deviation. In any reasonable data set the N-1 really ought not to be a critical item, but the sort of intellectual curiosity that makes for a good quant will lead you to find out why.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 131
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
A lot of what people call "creativity" is actually someone smart enough to steal ideas from different domains and glue them together. Look at how the Black Scholes equation was solved and derived, bits of economics nailed together with the maths of thermodynamics. You can get a Nobel prize for stuff like that. ...or at least a decent bonus.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 132
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Internships
Banks see most internships as an extended interview, allowing you and them to see whether you are good for each other. If they offer you an internship, it is because they think you’re probably worth hiring in the future, they even pay you for the privilege. This means that it you may find it harder to get an internship than a real job. The current market conditions (Q4 2007), don’t seem to have collapsed the number of positions, but the continuing increase in the number studying finance related subjects means that the competition is more intense than before, a trend we do not see changing any time soon. Most initial applications are handled as bulk processing which means application through bank websites and automated systems. If you pass the first simplistic filter then you may actually get to meet a human being. This is the area where applying as soon as possible is very much in your interest. It is not good enough to just get your application in by the end of the period because many banks work by going through applications until they are full, then stopping. They rarely admit this on their web site of course. Do however, use whatever facility is available to check your application made it into the system, as we hear a lot of stories about the reliability of these systems, none of it good, some of it funny (if you’re not personally involved). Thus the vast bulk of internships are not done through headhunters, and most HHs don’t handle them at all. P&D do a few, where a specific business unit wants a particular skill set, but the only reliable way to get an internship is to press every button available. It is of course possible that when you apply, they say that you’re too early. Fine, then you can apply again later. You should not assume that they will keep your CV because a combination of data protection laws, “entertaining” application management systems and the sheer volume of job seekers mean the odds are that your CV will not be sucked into the process when it officially starts. However, last year we became aware that one top tier bank was so unimpressed with it’s applicants that HR was asked to go through the database of last year’s set, and re-interview those who met certain criteria. We know this through talking to people who got jobs that way. And yes, before you ask, it lowered still further our opinion of the IT involved. Applying for a job does not have a high cost. The real cost is when you don’t apply early enough and miss it. Every single day, jobs you might have wanted are being consumed by people you are competing with. Obviously the odds are against you getting any given one, but you are interested in maximizing the area under the curve, as well as the chance of getting any specific job. This is of course a conflict, since the weight of applications is already oppressive to firms, but that is not your problem, google on “tragedy of the commons” if you don’t get this idea. There seems to be a clustering of PhD defences in the lead up to Christmas and it is tempting to apply after your PhD. That’s wrong on a couple of counts. Firstly the bonus round is going on so employers are distracted and Christmas causes a lot of processes simply to stop. Also you are competing with the others who also waited for PhD completion and that doesn’t help your chances. You’ve also missed the jobs that went out whilst you were waiting and they’re not coming back.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 133
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Transition from IT to Quant Finance Since such a large % of quant work is software development many IT people see this move as attractive, since it offers more money and greater choice of work. Quant finance often uses a different mix of technologies from housekeeping IT. Central to this is C++, we talk about this a lot, because lack of this skill is easily the most common barrier between an IT person and the lane change they are looking for. It is also one that is fixable, because although C++ is not a trivial language, the majority of its users are self taught. Although typically IT professionals have weaker maths, and of a different type to those used in finance the useful level of demand exists because of the importance of programming to this line of work. Note we are saying “programming”, and most other aspects of IT are of considerably less value, so operations, network management, and business analysis do not usually help you, and in some cases act as signals that you are not suited. For many IT people, this therefore means what they will see as a backward or downward step to “getting their hands dirty”. That’s just the way it is, and you will need to make sure that you can code well.
Learn C++. If you don’t then expect to die alone, poor and unloved. Probably your body will be eaten by your mongrel cat which by now is your only friend. Or it could be really bad.
IT Degrees As you may observe from the Wilmott.com forums, we have certain views on the quality of many Computer Science programmes. We hear tales of departments that teach operating system internals using Java because it is “easier”. The fact that there are no O/Ses written in Java doesn’t seem to impact those responsible for this imbecilic decision. At the time of writing we are unsure it is due to the low calibre of the lecturers, or the fact that their students are too dumb to learn C++, partly because although we have written (politely) to several lecturers at Kings College London, they are too cool to log in and use email. For any subject we are entirely unimpressed by anything that is billed as the “easy option”, unless the problem is itself impressively complex. Learning C++ doesn’t count as “impressively complex”. Java is a perfectly adequate first language for teaching data structures, patterns and algorithms, and even simple threading. There is a useful commercial demand for Java, but not in quant finance, where it barely shows up as noise. Even for those roles that require Perl, VBA, C# and occasionally Java, we often see employers ask for C++ though they expect only a small part of the work. This is because mastery C++ proves that nothing a computer does is likely to scare you. What seems to come as a shock to some IT graduates we meet is the expectation that they are supposed to understand computers. By understand, we mean the ability to think at multiple levels, as the task demands. That may be large scale grid architectures, processor architecture, patterns, functional languages or bog standard C++ loops. You should know that doubles are not reals. If this is news to you then don’t bother sending us your CV. That may sound harsh, but we meet people who claim to have studied CompSci who are ignorant of this but still think they are smart enough to do quant finance. They are wrong.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 134
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
It is of course quite possible that your lecturers took the easy option of teaching you some quiche language like Java, this is not quite your fault. However if you really understand programming then you can learn C++ all by yourself. More than 50% of C++ programmers are almost entirely self taught, so it is quite feasible. But that is not the same as easy. We include various good books on the reading list. The time to start learning them may already have passed, so you need to make up for lost time.
Quant Developer Skills to help you become a Quant Developer… Yes, we’re going to say C++ again. You need to have your numerical methods up to speed, understand O(N) notation, and be able to trade speed vs space. Some SQL does not hurt though be aware that if they really seem very keen on your database skills this job may be far more along the developer axis than the quant.
Excel One thing we have learned the hard way is that most specialist Excel developers are really quite amazingly incompetent. When we started P&D, we just had some vague worry about their quality, but one day a large bank said “We’re your largest client…”, the sort of phrase that never ends with something you want to hear. They had been interviewing Excel people, and by this point they truly hated them, using words like “chimps”, and harsher words that would stop this document getting through corporate firewalls. So we interviewed Excel people, and to be fair the majority of them managed to make it to the interview without getting lost, but that was pretty much the only skill that we could rely upon. Simple VBA confused them, and nearly half had to have “And” explained to them. Yes, really. DDE was seen as anything from black magic to a “new feature”. (It was around 1987) In the end we found a couple of good ones, but it was hard. That is not to say that there are no good Excel people in banks, but that most of them don’t do Excel as their main job, indeed the authorative text on Excel addins was written by someone now working as a trader at Tradition.
MatLab This is one of the most common IT skill we see on entry level CVs, which has an effect on its value in the market. To be effective on your CV as a QD, you should make sure you understand programming it to include GUIs and C++ addins.
��� �� ��� ���� ������ ������� �� ��� �������� ����� �� ��� ���� ����� ����� ���� ��� ������� ��� ��������� �� �++
������ ����
It is hard to get good experience with real market data sources before you get your first real job, but if you are doing a good specialist course in finance they should cover this, and yes we see this as a test for whether it really is a good course for those who expect to do a lot of development in their careers.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 135
Guide 2.0
Paul & Dominic’s Guide to Quant Careers ����� ����������
As the name suggests you will be looking at the models and systems developed elsewhere to see if there are any holes. You will be looking at the assumptions they make, and applying different data sets to see if their model holds up in different market conditions to the ones that the creators assume. The overall model will be made up of a collection of component sub-models, and so part of the work is to see what happens if you replace (for instance) the Hull-White Model for interest rates with the Vasicek. It is not so much whether this is “better” but if the proposed business model betrays very different behaviours then it throws open the question of whether it is robust enough to be trusted with real money. Most banks have a set of “standard models” for this sort of component and your work will include making sure that either this rule is obeyed or asking why. In theory, model validation should include verifying that the code replicates the mathematics and business logic. This is indeed what the banks say they do, and the regulators are dumb or The goal of Model Validation is to reduce model lazy enough to believe them. The simple facts risk, by looking at a range of market scenarios are that in only a small minority of cases do and identifying risk factors. MV departments have the resources in time and skilled staff to do this properly. To validate a complex program requires skills as least as sharp as the person that wrote it, but sadly MV does not pay anything like as well as front office work, and a depressing amount of software testing in banks is not done by the A team, but instead include a large number of people who couldn’t cut it as hardcore developers, or who are busy getting their skills up to speed so they can get a “real job.” MV does have several attractions as a job. Although few roles in a bank can be classed as easy, there is considerably less time pressure, and on average you can expect to work more reasonable hours than in a front office job. It is very easy for people who give careers advice to assume that money is the only motivation, but if you have a life, or want to acquire one then MV has a lot to offer. The staff turnover is many areas is high enough that you can become relatively senior simply by not quitting, and the nature of MV means that in bad times is hard to cut, and in good times there is little pressure to lose headcount. However to the higher parts of the MV food chain, it’s critical to get some direct experience of front office and other parts of the business. This is where the “look at your boss” model of predicting your career works rather well. Most heads of MV have pulled back from front office, and are typically former quants or in some cases traders. This means that a gentle drift up to the top is not very easy, unless you manage to get this experience. One reason for this is that you must have credibility with traders and other kids of risk management, the sort that comes from having done it yourself. Also you need to build judgement in what risks are worth taking, what risks you should walk away from, and know when to run. One solution to that is to get an internal placement into a related front office position to build t his experience. The smarter end of traders and quants recognise that a more informed MV group is good for them, especially if “informed” means seeing the world from the saMe point of view. The more cynical reader will of course see this as a possible route out of MV to a higher paying job, but of course their manager probably has a copy of this Guide as well…
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 136
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
What TV Gameshows Can Tell US about Mathematical Finance Monty Hall This is one of the oldest Brainteasers dating back to a 1960’s US TV gameshow. It goes like this: You are asked to pick one of three doors, A, B, C. Behind one door is a Ferrari, behind the two others there are goats. You pick a door, A, for example, and the host of the game show (Monty Hall), who knows what is behind each door, opens another door, B, to reveal a goat. He then asks “Do you want to change your choice of door, from A to C?” Is it to your advantage to switch? Or should you stick with the original A? This is a very subtle problem, the answer depending on the precise statement of the problem. Since there has been so much written about this problem we shall here only mention the various presentations of the problem and the corresponding results. Warning: We know of one person who was asked about the Monty Hall problem in an interview
and who knew what we are about to tell you inside out. Unfortunately the interviewer was not so well informed, he only knew the ‘classical’ result. He refused to believe what the interviewee was telling him, even though it was correct, an argument ensued and a job was not forthcoming. Presentation 1: We are told that Monty Hall will always open another door after we have made our choice, regardless of whether we have chosen the car or the goat. This is what is commonly meant by the Monty Hall problem. In this case one should always change door when (inevitably) asked. If you don’t change then the odds are 1 in 3 that you get the Ferrari, if you do change, the odds are 2 in 3. Presentation 2: If Monty Hall opens a door and offers the chance to switch only when you have first chosen the correct door (we are here assuming you want the car not a goat) then you should never change because then you are guaranteed to lose. In an interview situation the rules should be made clear right at the start. If they aren’t then either the interviewer is being careless, perhaps doesn’t know the problem in great depth, or is making a double bluff assuming that you sort of know the Monty Hall problem and are making assumptions you shouldn’t. If very brave you could turn the tables on a careless interviewer, but then you’ll probably end up like our friend above, so forget that piece of advice!
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 137
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
How to Compare Multiple Job Offers As you’ve probably picked up from the Wilmott.com forums, at P&D we frequently give one to one advice on choosing jobs. Each person is of course a unique case, and so before you contact us you should know that we almost never give a straight answer the form of “this job is clearly better than that one”, since if it was that clear and simple you would not be asking for our input. Instead we typically ask questions to help you make up your mind. We don’t do hard sell in any aspect of our business, it’s your life, and if you’re smart enough to be a successful quant you only need the help of our experience with fine details and awkward questions not a major personality upgrade. There are however some general principles for this important dilemma in your life. Central to our approach is to look further forward to the jobs you can do as a result of having done this one. This operates in more than one dimension. The obvious parts are what you will learn, and what you will be able to demonstrate that you can do.
History as a predictor of your future.
One of the industry standard lies is that if you take a job in model validation in a top investment bank, they will bring you into a sexy front office Stat Arb role. It’s not true, and not even a good lie. Most banks would rather have high staff turnover than develop their careers.
Looking at people who have been taken on by the firm is like looking out into deep space. The further up and away you look, the further back in time you are looking. You can’t watch a firm for 5 years to model the main sequence of career evolution, and more than even the most patient astronomer can observe a billion year interval for a star.
It is a good idea to ask to meet the people who are roughly your peers, or a year or two ahead to find out what people like you really spend their time doing. Of course this should be spun nicely, as something like “it would be great to meet the rest of the team”, not “I want to get the truth”. When looking at the levels above you, try to do some pattern matching. These people have done well in this environment both from a skill position and their ability to work in that culture and market. The first thing to ask is how much they resemble you. In some areas, the people running things have quite different backgrounds from those on the front line. For instance, to be a good manager in risk or model validation you ought to have experience on the trading floor, but if you join that sort of department, you may find that going directly up the food chain very difficult.
Martingale Careers Be aware that you are choosing the expectations people have of what you can do, since the default assumption has to be that your next job will be some variant of your first. Sometimes it can be quite difficult to get out of a track and into another. This means you should be quite conservative about accepting jobs that seem to lead away from where you want to be.
We’re not a charity. At this point, a cynical type might be asking why we at P&D do this, and it is remarkably simple. Not only are we nice people, but we recognise that if you can get multiple offers now, in a couple of years time when you are still more valuable, we’d like to be in touch. Short version is that we are quite prepared to take the long view.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 138
Guide 2.0
Paul & Dominic’s Guide to Quant Careers
Brand Name The prospect of working for a highly prestigious firm can lead people to take an inferior job, just because of the brand name of the employer. We know that may well be a rational trade off since having a big name on your CV is a good thing, but it needs to be a reasoned choice. When valuing a brand name, make sure that you are pricing it in the right market. This is simple when the firm is well known as a pure investment bank, but many enjoy a high reputation because they have a substantial retail operation which makes them better known, but this adds little value to the aspiring quant who needs to focus on the quality of the investment banking operation. Just because your target bank has a branch in your home town, doesn’t mean they are wise in the way of credit derivatives. Hedge funds are an important issue here, since DE Shaw, Citadel, Renaissance, et al are often entirely unknown to many students, but are important well known outfits in the eyes of those who you would want to employ you later in your career, this is It The is easier most important to upgrade term the in quality choosing of bank a job that is you net where you really need to go on the forums to present work for, value, than not to make present a cash radical flow. change of lane check out the level of respect your future from an area like IT to structuring or trading. employer commands in the market. If you are faced with a less attractive job at a more prestigious firm, you can use the job at the less well known outfit as a hedge to try and get the bigger name to be more flexible. Once you have decided that you can live without the job at the better bank, it is worth saying to them, that you want to work for there, but that you have a better offer at the less prestigious firm. You can try to get the bigger bank to see if they can find you a better job. We will be entirely honest with you that this strategy is most likely to fail. However, it can work sometimes, and in those cases is of huge value to your career. It is viable because you are walking away from the larger bank, so you have nothing to lose. You need to frame this as a business negotiation, not as some moral “I deserve a better job”, maybe you do, but that is not their problem. By saying that you would prefer to work for them, you are using a little flattery, and express your choice in terms of regret that it won’t work out this time. Try to put the ����� ����� ����� ���� ������ �� initiative with them, and ask if there is anything ���� ��� �����������, ����� ���� they can do. Perhaps there is a better role in the same area, or it is possible that they know a ����� ���������� ��������� manager in a more attractive business unit. �������� ��� ���� ����� �����
������. ���� ��� ��� ���� �������� �� ���� ���� ���� ��� ������ ���� � ������� ��� �� � ��� ���� �� ������ ������ ���� ��� �� � ���� ���� ����� ����. ���� ��� ��������� �� �� ��������� �����, ���� ��� ������� ����� ����������� ���� ��� ���������, ��� ��������� �� �� ������ �� ����. ��� ���� ��������� �� ����� ���� �������, ��� �� ������ ��� �� ���� ����������� ������ �� ��� ���� ��� ������.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 139
Guide 2.0
Paul & Dominic’s Guide to Quant Careers Or not…
Even if this does not work, this is some of the most valuable business experience you will get. As you progress up through your career, you stop being seen as someone who does a specific role that is defined by their superiors, and more by how you negotiate what you will be working on. This is an acquired skill, and here is a chance to start building it, especially as this is a situation where there is not so much downside if you don’t get it right. The time to mistakes is when it doesn’t cost you much, rather than when you are in a job negotiating with you boss.
©2005-2007 Paul & Dominic Ltd
www.PaulDominic.Com
Page 140