FIN3102 Investment Analysis and Portfolio Management
Case 1- The Dynamis Fund: An Energy Hedge Fund
Section C4 Group 1C
Group Members Chen Zu Qing (U098258E) Kwan Kin Weng (U090381H) Low Siao Chi (U098260J) Sim Wan Lin (U098374Y) Yong Jun Kang Eric (U098357R) Yong Lin Lin (U098312Y)
0
1. Why would a regional brokerage offer such instruments Compared to individual portfolios, such funds woo investors by offering several advantages namely: professional asset/money management, liquidity and more diversification than most individuals can create or afford in a personal portfolio. The brokerage’s motivation in recommending energy investments can be explained by the high commission that could be earned. Hedge funds charge a fee for assets under management and incentive fees based on a certain percentage of the profits earned. With good stock picking, the brokerage would be able to earn profits in both up and down markets. A regional broker would want to offer hedge funds because they are only lightly regulated and thus the fund managers can use more advanced investment strategies such as a leveraged and derivatives positions. It is stated in their selling memorandum that their mission is to exploit investment opportunities in publicly traded companies in the energy sector. Hence, the fund seeks to generate above average returns relative to both S&P Energy composite and the broader market through a variety of investment instruments. Also, the fund's use of various strategies will be designed to minimize risk while maximizing potential return, again increasing the commission that could be paid to the hedge fund managers. This potentially high level of compensation helps the brokerage retain talented brokers and specialists, raising the reputation of the firm. Furthermore, investing in energy funds serves as a diversification tool. This is because from historical records, energy prices have had a high correlation with inflation. In times of rising inflation, energy funds have been found to perform better than the market. Thus, they are able to act as a source of risk diversification. This explains the presence of a market for such energy funds.
1
In addition, energy funds have been a very popular fund with investors. The high dependence on oil in all parts of the world has made energy stocks a hedge for emerging market portfolios. With a demand for such energy stocks, a regional brokerage will want to cash in on this opportunity and offer energy funds. In order to cater to a larger crowd of investors, the brokerage firm will offer energy hedge funds to sophisticated investors and energy mutual fund to general public who will like to invest in energy fund, but are unable to do so given their smaller amount of capital. Investing in energy funds is often complicated and risky, given the volatility of such commodities. Brokerage firms have a fiduciary responsibility to research on such funds before recommending them to their clients. They have to ascertain if the investments are suitable for the clients based on their age, investment experience and tolerance for risk. In view of this, investors prefer a firm that can provide them with personalized services suited to their needs and risk tolerances. To be able to get these services, most of these investors go to regional brokerage firms. Such regional brokerages can deliver the attention to their clients due to their small size. Thus, with such demand in energy funds, regional brokerage firms would be able to make a profit out of offering such instruments. It will allow them to better position themselves in the market.
2. Why did S&S start a hedge fund in addition to its energy portfolio The Energy portfolio is essentially a long equity fund for investors to buy stocks. It is stated that in their selling memorandum that the Energy Portfolio will seek to earn above average returns by investing in smaller and medium-size companies that are growing earnings and cash flow in a dramatic way. Therefore, it can only stand to gain when the market goes up. On the other hand, the introduction of the hedge fund will provide benefits to both its investors and the fund manager in the following ways:
2
For the investors, the hedge fund acts as a better investment for reaping returns in both bull and bear markets by having both long and short positions. Also, hedge funds are lightly regulated as compared to mutual funds and thus the fund managers can pursue more advanced and a wider range of strategies including leverages, derivatives, short sales, options and futures contracts. The flexibility in managing the hedge funds allows fund managers to exploit opportunities within the energy sector. The potentially higher returns attract investors with higher risk tolerance. Hedge funds cater to sophisticated investors who earn a minimum amount of money annually and have a certain amount of net worth, along with investment knowledge. It helps to cater to the needs of the sophisticated investors and target a substantially different market from the mutual funds. This is in line with S&S’s corporate strategy of providing the best possible service to its retail customers and continuing to grow the asset management business. As for the fund managers, a hedge fund provides a radically different incentive package than the typical mutual fund. The fees paid by investors are higher as compared to that for mutual funds, including additional fees that mutual funds do not charge. There are no restrictions on the fees a hedge fund manager can charge, as compared to mutual fund fees which are regulated and transparent. The energy hedge fund charges a 1% management fee, which is for the same service that the management fee covers in mutual funds. This fee alone may form a substantial part of the fund manager’s profit, thus making the
management
of
the
hedge
fund
attractive
to
the
fund
manager.
In addition to the management fee, incentive fees are charged based on 20% of the fund’s profits. It serves to reward the fund manager for the stock picking ability and for good performance. The potentially high returns that hedge fund offers from the wider range of investment strategies also increases the potential incentives that the fund manager gets, explaining the main reason for starting the Energy Hedge Fund. 3
3. Does the fee structure for the Dynamis hedge fund make sense? Comment on the differences in the compensation between the energy portfolio and the Dynamis fund. The fee structure of the Dynamis Fund includes management fees as well as performance fees. Management fee is equivalent to 1% of assets under management while performance fee is equivalent to 20% of profits earned by the hedge fund. The performance incentive will only be allocated unless any previous losses have been recouped, implying the existence of a high-water mark. Also, there exist a hurdle rate of 10% and the full 20% profit incentive to be allocated to the fund manager begins only at 12.5% return. The main difference between the fee structure of hedge funds and mutual funds would be the implementation of the performance fee. Hedge funds are more actively managed than mutual funds and this performance fees can be seen as additional compensation to the fund managers for the additional efforts involved. Hedge funds have higher flexibility than mutual funds and hedge fund managers are allowed the use of short selling, leverage as well as derivatives, from plain vanilla options to very exotic instruments.1 Also, the incentive fees act as an attraction to lure talented managers so as to deliver high returns to the clients.2 In a research paper3, it was also found that incentive fee provides managers with strong incentive schemes: the higher the incentive fee, the better the fund performance. In fact, a 1% increase in incentive fee will increase the average monthly return by 1.3%. It is necessary to determine if this fee structure can indeed help deliver superior profits that are demanded. In theory, the fee structure helps align investors’ goal to that of the fund manager since the fund managers’ income is tied to the profits of the fund. Also, the high-water mark ensures that investors
1
Past present and future, Rene M Stulz, 2007 Are hedge funds fees too high?, Francois-Serge Lhabitant, 2007 3 On the Performance of Hedge Fund, Bing Liang, 1998 2
4
do not lose out since any losses have to be recouped before the performance fees can be charged. It also ensures that managers engage in limited risky activities. Without the high-water mark, the managers gain from the upside but do not suffer from the downside and this does not lead to goal congruence. However, in the case of huge losses, goal alignment will be out since fund manager can simply abandon the fund and start another fund since the high-water mark will prevent the manager from earning the incentive fees in the coming periods. Furthermore, towards the end of an evaluation period, managers have incentives to increase the variance of the portfolio and take on more risks when the asset value of the fund is not far below the high water mark, in order to receive the performance fees 4. The cyclical nature and volatility in the energy sector also makes the usage of high water marks an unfair means of compensation to the investment managers of energy hedge fund. By capping the fund managers’ compensation to the profits will prevent managers from simply increasing the portfolio and assets under management so as to increase management fees. To determine if the fee structure has been useful to investors, we look at the returns of the hedge funds over a period of time to determine its performance. The Hennessee Hedge Fund Index actually outperformed other indexes such as the S&P500, NASDAQ and Dow Jones Industrial Average. A thousand dollars invested in the hedge fund index would have grown to USD19, 647 as compared to $5,299 if invested in the S&P500.5 Apart from its superior returns, the annualized standard deviation of the Hennessee Hedge Fund Index is lower than most of the indexes other than the Barclays Aggregate Bond Index. This would indicate that hedge funds are good value for money since investors are getting the most from their investments. Also, it is indicative that hedge funds are able to minimize the investment risk as well.
4 5
High-Water Marks and Hedge funds management contracts, Gotetzmann, Ingersoll, Ross, 2003 http://www.hennesseegroup.com/indices/returns/dollargrowth.html
5
Evidently, in times of bearish market, empirical data has shown that while hedge funds did suffer negative returns, they fared far better than the S&P500.6
Since the performance fees compensate the managers of hedge funds for their efforts and form a substantial portion of their income, would the management fees be excessive as a result? For mutual funds, the management fees cover the advertising and administrative costs as well as the managers’ compensation and thus easily justified. However, reducing the management fees for hedge funds would be unfair for hedge fund managers since there will be advertising and administrative costs involved in the operations of the fund. In order for the management fees to be able to cover the operating costs, the hedge fund would need to have at least $500million worth of assets under management, according to a survey done by KPMG. 7 In view of the increasing operational expenses due to investors’ need for pertinent risk controls, efficient operational systems and studious compliance, we propose the following scheme of compensation: 1) If the assets under management have a total value of $500million and below, it would be a better and fairer strategy to charge a management fee of 2 % in order for the fund manager to completely cover its operating costs, including costs for facilities, infrastructure and information technology, business support, marketing and public relations fees. 2) For assets under management above $500million, management fees will remain at 1% of assets under management. 3) Lastly, the use of hedge fund stock options to compensate the managers helps provide fair rewards and proper incentives. It works by giving the hedge fund manager an option to purchase a fixed number of shares of common stock at a price that is equivalent to the fair value of the shares on the date that the share option was granted. When the fund manager chooses to exercise
6 7
http://www.hennesseegroup.com/indices/returns/downmarketanalysis.html http://www.securitiestechnologymonitor.com/issues/19_78/22897-1.html
6
the option, he will realise a gain that is equal to the spread of the difference in the stock price on the date of grant and on the exercise date 8.
4a. Which index would you recommend as a benchmark for the performance of the energy portfolio? What, if any shortcomings does the index you recommended have? Energy Portfolio v S&P 500
Diagram 4a-1 on the left is a scatter plot of
20.00% y = 0.7827x + 0.013 R² = 0.1082 15.00%
energy portfolio against the current benchmark, S&P500. From the diagram, it appears that the
10.00%
S&P500 is not a very good benchmark for
5.00% 0.00% -5.00% 0.00% -5.00%
-10.00%
Diagram 4a-1
5.00%
10.00%
Energy Portfolio as the points were scattered
-10.00%
randomly. Table 4a.1 shown below also shows
-15.00%
that the r-square value and correlation coefficient
-20.00%
are the lowest among all other indexes. In
addition, its tracking error, which is used observe how the portfolio has diverge from the benchmark, is the second highest among all other indexes. All these evidences imply that S&P500 is not an adequate benchmark to measure the performance of the Energy Portfolio fairly. S&P 500 RSQ
Russell
Oil Comp
Wells
Drillers
Prod
SPOILC
SPOILW
SPOILD
SPOILP
Crude
10.82%
27.88%
40.31%
59.43%
59.33%
58.64%
0.3289
0.528
0.6349
0.7709
0.7702
0.7657
0.4134
Tracking Error 7.0200%
6.2900%
5.8400%
4.7200%
6.1100%
4.8000%
7.5300%
Correl
17.09%
Table 4a-1
8
Compensating Hedge Fund Managers with Stock Options: A new path of alignment of interests with investors, Earle,2010
7
Also, from Table 4a-1, we observed that
Wells v S&S Energy Portfolio
SPOILW has the highest r-square value. Thus, if we were to plot the Energy Portfolio return
y = 0.9547x + 0.00920.00% R² = 0.5943 15.00%
against index return, we will able to get a more
10.00% 5.00%
clustered plot (diagram 4a-2), but it is still not
0.00% -10.00% 0.00% -5.00%
-20.00%
good enough as the r-square values of SPOILD
10.00%
20.00%
-10.00%
and SPOILP in respect to energy portfolio are
-15.00%
also close to the r-square value of SPOILW and
Diagram 4a-2
-20.00%
excluding the SPOILD and SPOILP indexes might result in the loss of important indicative factors.
Regression on Multiple Indexes By running a multiple regression of the energy portfolio’s returns on the SPOILW, SPOILD and SPOILP indexes, it is possible to get a relation to see how the three indexes explain the returns of the energy portfolio. Regressing the S&S Energy Portfolio returns on the three indexes would yield the following equation.
The right hand side of the equation can be used as the multi index benchmark to measure the performance of the portfolio manager. By substituting the various returns of the SPOILW, SPOILD and SPOILP indexes over the time period into the equation, we would get a list of returns for this new benchmark. Further regression of the original energy portfolio returns on this new list of returns for the new benchmark would yield the following result.
R, Correlation R square
Multi Index Benchmark 0.88513 0.78342
SPOILW 0.7709 0.5943
SPOILD 0.7702 0.5933
SPOILP 0.7657 0.5864
Table 4a-2
8
As shown in table 4a-2 above, this new multi index benchmark will serve as a better benchmark since it has a higher R and R square value. A larger portion of the energy portfolio returns has been explained by this new benchmark.
Problems of Multi-Index Benchmark The returns of the three indexes used above are correlated and hence the model might not be that accurate. In fact, given that the SPOILW, SPOILD and SPOILP are indexes that track the performance of oil-related companies, it is highly possible that returns are highly correlated. Numerically, the correlation between the SPOILW and the SPOILD index is 0.7761 that of the SPOILW and SPOILP is 0.5892. Given this high correlation, the multi index benchmark will not be an accurate predictor for future hedge fund results. To create a more accurate benchmark, the variables have to be non-correlated and independent. This leads to the next benchmark that will be proposed, the multi factor benchmark.
Multi-Factors Regression The multi-factor regression model we used are stated below,
The indexes selection process are as follows: 1. We examine the major qualitative similarity S&S Energy Portfolio Small-Mid Cap Services & Supplies Offshore Drilling Offshore Rig Construction
S&P 500
Russell
Large Cap
Small Cap
Oil Comp (SPOILC)
Wells (SPOILW)
Drillers (SPOILD)
Oil Services
Intergrated Service Co.
Prod (SPOILP)
Crude Oil Benchmark
Intergrated Service Co.
9
Capital Equipment
Oil Equipment
Exploration
Exploration
Exploration
Exploration
Production
Production
Production
Production
Table 4a-3
As observed from Table 4a-3, the Energy Portfolio shares some similar characteristics with SPOILC, SPOILW, SPOILD and SPOILP. 2. Then we compare their quantitative characteristic, S&P 500 RSQ Correl
Russell
Oil Comp
Wells
Drillers
Prod
SPOILC
SPOILW
SPOILD
SPOILP
Crude
10.82%
27.88%
40.31%
59.43%
59.33%
58.64%
17.09%
0.3289
0.528
0.6349
0.7709
0.7702
0.7657
0.4134
Table 4a-4
From Table 4a-4, although SPOILC, SPOILW, SPOILD and SPOILP share similar characteristics with the energy portfolio, only the R-squared values of SPOILW, SPOILP & SPOILD are high and close enough. Hence, after taking into consideration of both the quantitative and qualitative factors, it appears that these 3 indices mentioned above can be used as the major factors in the new benchmark. Thereafter, we process the data with procedures as follows: (1) The SPOILW is used as the main proxy for the hidden factor due to its superior R square value. It is important to note that there will be other hidden factors present in the other two indexes given their high correlation with the portfolio’s returns. (2) Find the factor that is present within the SPOILD index but not explained by the SPOILW. To do that, we regress SPOILD on SPOILW so as to get the residual values (ResDonW). (3) Regress SPOILP on SPOILD and SPOILW so as to find the residual values of SPOILP that cannot be explained by both the SPOILD and SPOILW (ResPonWD). (4) Regress the Energy Portfolio Return with SPOILW, ResDonW, ResPonWD and obtain the regression model as below:
The resulting graphs and statistics are shown in diagram 4a-3 and 4a-4.
10
S&S Energy Portfolio v S&P 20.00%
Diagram 4a-3
15.00%
10.00%
10.00%
5.00%
5.00%
0.00% -5.00% -5.00%0.00%
5.00%
10.00%
-20.00%
0.00% -10.00% -5.00%0.00%
-10.00%
-10.00%
-15.00%
-15.00%
-20.00%
Energy Portfolio v S&P 500 New Benchmark
Diagram 4a-4
RSQ 10.82% 78.34%
Correl 0.3289 0.8851
y = x - 2E-17 R² = 0.7834
20.00%
y = 0.7827x + 0.013 R² = 0.1082
15.00%
-10.00%
New Benchmark
Slope 0.7827 1.0000
10.00%
20.00%
-20.00%
Tracking M2 Jensen Error Measures Alpha 7.02% 0.02% 1.30% 3.44% -0.21% 0.90%
Table 4a-5
From diagram 4a-4 and table 4a-5, we can observe that the r-square value and correlation coefficient have significantly improved by 7 times and 2.7 times respectively, from the original measurement using the S&P 500 as the benchmark. The proposed benchmark is a superior way in measuring and predicting Energy Portfolio returns compared to standalone indexes.
Pros and Cons of Proposed Benchmark The strengths of the benchmark are (1) it is more indicative of the portfolio performance, thus allowing the investors to know how well is the fund manager performance. (2) It allows the company to have a better gauge of fund manager performance, thereby rewarding them in a fairer manner. However, there are also shortcomings of the model, which are (1) investors, the general public, might not fully understand the highly sophisticated model, thereby rendering the model valueless. (2) Fund managers might feel exploited as the company measures their performance with a different benchmark when their job is only to “beat the market”, which is the S&P500. In addition, when the fund was started, investors
11
might have been told that its purpose was to outperform the general market, symbolised by the S&P500. Changing to a different benchmark thereafter would not make sense. Looking at these factors, the S&P500 is recommended as the benchmark for public reporting purpose because it is simple to understand, especially since the ultimate motive of the public to invest in the Energy Portfolio is to get a better return than the market. However, for internal purposes, when the company measures the performance of the managers, they can still use the new benchmark to measure the performance because the risk adjusted return would be more indicative of the manager performance in term of their timing, asset allocation and stock picking skills.
4b. Which index would you recommend as a benchmark for the performance of the Dynamis hedge fund? What, if any shortcomings does the index you recommended have? R R2 Beta
S&P 0.402676 0.162148 0.916135
Russell 0.59564 0.354787 1.767116
OSX 0.91738 0.841587 0.867105
Crude 0.261659 0.068466 0.377185
SPOILD 0.852843 0.727342 0.748641
SPOILP 0.720669 0.519364 1.257605
SPOILW 0.832547 0.693135 0.913691
Bonds 0.198899 0.039561 -2.18323
Table 4b-1
In selecting a suitable benchmark to evaluate the performance of the fund manager, it is important to select a benchmark which has a high correlation to the returns of the fund. Regression of the returns of the hedge fund on the various indexes will yield the results as shown above. From table 4b-1, It can be seen that the OSX, SPOILD and SPOILW indexes are highly correlated with the hedge fund’s returns. The R-square values of 0.842, 0.727 and 0.693 of the OSX, SPOILD and SPOILW indexes indicate that 84%, 73% and 69% of the hedge fund returns can be explained by the indexes respectively. At this instance, if a single index benchmark has to be chosen among the various indexes, it would have to be the OSX index due to its higher correlation and R square value. However, given that indexes like
12
the SPOILD and SPOILW also have rather high correlation values with the hedge fund’s returns, it would be better if these indexes can be taken into account as the new benchmark is created.
Multi-Factor Benchmark An accurate benchmark that can be used would be a multi factor benchmark. It is assumed that returns and patterns in the various indexes are caused by separate factors. The OSX, SPOILD and SPOILW indexes will be used as proxies for these hidden factors that can be used to explain the hedge fund’s returns. The OSX index is used as the main proxy for the hidden factor due to its superior R square value. It is assumed that there will be other factors present in the other two indexes given their high correlation with the hedge fund’s returns. Moving on to find the factor that is present within the SPOILD index but not explained by the OSX index, we regress SPOILD on OSX so as to get the residual values (ResSPOILDonOSX). Following that, we would need to regress SPOILW on SPOILD and OSX so as to find the residual values of SPOILW index that cannot be explained by both the SPOILD and OSX indexes (ResSPOILWonOSX, SPOILD). Given that the returns of the OSX index, the residuals of the regression of SPOILD on OSX as well as the residuals of the regression of SPOILW on OSX and SPOILD are now available, we can regress the returns of the S&S hedge fund on these 3 variables and we will get the following equation.
a = OSX = factor explained by OSX index, b = ResSPOILDonOSX = factor explained by SPOILD but not OSX index, c = ResSPOILWonOSX,SPOILD = factor explained by SPOILW but not the SPOILD and OSX indexes
13
Substituting the relevant values back into the equation would give the following graph.
S&S vs Multi Factor Benchmark
As seen in figure 4b-1, the new multi factor benchmark will have a R-square value of 0.8485,
10.00%
y = x + 5E-18 R² = 0.8485
5.00% -10.00%
0.00% -5.00% 0.00% -5.00%
5.00%
earlier. However, this multi factor benchmark is more accurate because the variables are not correlated and
-10.00% Figure 4b-1
10.00%
similar to that of the multi index benchmark used
-15.00%
as a result, the equation in the multi factor benchmark
better predicts returns. Thus, in performance evaluation, the multi factor benchmark should be used.
5a. How have Energy Portfolio performed? In measuring how has energy portfolio performed, we employed several tools, namely sharpe ratio, measures and Jensen
measures. The definition of those tools are defined as follows:
Sharpe Measure
S&P 500 36.31%
Benchmark Energy 33.39%
Energy Portfolio 29.56%
Table 5a-1
And from Table 5a-1, we can observe that the reward-to-risk ratio of energy portfolio is well below both benchmarks, meaning that for the same amount of risk borne by a particular investor, investing in S&P would be more worthwhile as it generate more return for per unit of risk borne.
14
M2 Measures Jensen Alpha Energy Portfolio v S&P 500 0.02% 1.30% Energy Portfolio v New Benchmark -0.21% 0.90% Table 5a-2
Further, from Table 5a-2, we are able to observe that the M2 measures and Jensen alpha appears to be better when the portfolio is measured against S&P 500. However, when we use the new benchmark to measure the performance of Energy Portfolio, the M2 measures and Jensen alpha is lower than what we obtained using the S&P500 as benchmark, meaning that the fund manager did not “beat the market”. The 0.90% of alpha might simply due to liquidity premium. Hence, we are able to conclude that, even though the energy portfolio did give a higher return to its investors, but, in fact, the fund manager neither not “beaten the market” as its M2 measure is negative, nor justified for the additional risk borne as the sharpe ratio is lower than the sharpe ratio of S&P 500.
5b. How have Dynamis Hedge Fund performed?
Sharpe ratio M2 Measure
OSX
SPOILD
SPOILW
S&P
Multi factor benchmark
2.80% 0.01%
0.32% 0.09%
2.73% 0.02%
8.32% -0.05%
3.57% -0.01%
Dynamis fund 3.28% -
Table 5a-3
As seen in table 5a-3, from the Sharpe ratios, it is clear that the hedge fund did not perform better than the S&P500 index. However, it outperformed the oil-related indexes such as the OSX and SPOILD indexes. In general, oil-related indexes did not outperform the S&P500 and noting that the hedge fund is invested in the oil sector, it might explain the fund’s poor performance. The Dynamis hedge fund has a M2 measure of 0.01% and 0.09% when adjusted to the same risk of the OSX and SPOILD index respectively. It further proves that the hedge fund outperform the oil indexes even though it underperforms the S&P500 index as seen by the negative M2 measure value.
15
