D3S2ACorrelationKokouNealTu

September 2016

CBOE RISK MANAGEMENT CONFERENCE EUROPE

TRADING CROSS-ASSET VOLATILITY & CORRELATION

Trung-Tu rung-Tu NGUYEN NGUYEN

Kokou AGBO-BLOUA

Neale JACKSON

Portfolio Manager Directional strategies

Managing Director Head of Flow Strategy & Solutions Financial Engineering

Portfolio Manager, Mana ger, 36 South Capital Advisors LLP

AGENDA � ) �� .

��

�.

��

�.

��

��/�� .

��) �� .

�� , �� ... ��

�.

�� ... ��

�.

��

��) �� .

��

�.

�� ... ��

�.

��

�� , �� .. ��

September 2015

P. 2

A few words on correlation

•

Time varying property of Cross-asset correlation of volatility trading

•

How to deal with high correlation?

•

Term structure of correlation: the difference between long-term correlation and short-term correlation

P. 3

Multi asset short variance swaps

P. 4

Volatility assets become more correlated after 2008

P. 5

A simple trick to deal with high correlation

 Highly

correlated assets:

•

Long Stock indices future

•

Short Equity variance swaps

•

Short Volatility Indices future

•

Long CDX (IG, HY)

•

2 zones: US/EU

P. 6

Equity related: equity neutral (per zone) correlation matrix

P. 7

Term structure of correlation

P. 8

Correlation arbitrage

Tail Protection for Long Investors: Trend Convexity at Work http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2777657 P. 9

A few words

•

We should be aware of the evolution of cross asset correlation,

•

Some main factors could be removed using simple concept of beta-hedging to gain back the diversification,

•

The term structure of correlation is far from being flat, which could be used as an alpha strategy by using correlation arbitrage

P. 10

CROSS-ASSET CORRELATION WITH HYBRIDS Investment case for hybrids : 

A melting pot of derivative parameters

(correlation, volatility, forwards) 

Isolate specific outcome in joint probability distribution



Leveraged, yet with limited risk



Exotics book supply & demand distortion

Correlation drives the discount



“Worst-of” options provide the

most discount. 

% discount vs vanilla assuming

cross-asset correlation of 0%: 

Worst-of option: 70%,



Contingent option: 50%



Basket option: 30%

Source: SG CIB Financial Engineering P. 11

TRADING SYNTHETIC COVARIANCE SWAP COVARIANCE = VOLATiLITY X CORRELATION 

A three leg trade where: 

Long 1 x VarSwap Compo into composite currency



Short

w qto x VarSwap quanto on the index



Short

w fx x Currency VarSwap

 Varcompo − K compo ²   Varqto − K qto ²   Varfx − K fx ²   − w qto *   − w fx *   2K compo 2K qto      2K fx 

P&L =  

Proportions are determined so that the structure is VAR neutral

w qto



=

 K qto    & w fx  K compo 

=

pure COVAR trade

 K fx     K compo 

FTSE/ GBPUSD Covariance swap Backtest 12

And your final P&L can be written as:

1x UKX Com po $ Var vs . (1x UKX qto $ +0.5 GPBUSD) Var

 (RealisedCovarindex, fx − ImpliedCovarindex, fx   ⇔ Covar swap K compo  

P&L = 

a

10

8

6

4

2

0 Nov 05

Nov 07

Nov 09

THE VALUE OF YOUR INVESTMENT MAY FLUCTUATE. THE FIGURES RELATING TO SIMULATED PAST PERFORMANCES REFER TO PAST PERIODS AND ARE NOT A RELIABLE INDICATOR OF FUTURE RESULTS. THIS ALSO APPLIES TO HISTORICAL MARKET DATA.

Nov 11

Nov 13

Nov 15

P. 12

Alexander Calder, entropy and correlation

0.5

Pan asset class rolling three year weekly correlation Proxies are: S&P500, USGG10, DXY, CCI

Alexander Calder Small Sphere and Heavy Sphere, 1932-1933

0.4

0.3 n o0.2 i t a l e r r o0.1 C

0.0

-0.1

-0.2

N M N M N M N M N M N M N M N M N M N M N M N M N M N M N M N M o a o a o a o a o a o a o a o a o a o a o a o a o a o a o a o a v y v y v y v y v y v y v y v y v y v y v y v y v y v y v y v y 6 7 7 7 7 7 7 7 8 8 8 8 8 8 8 9 9 9 9 9 9 0 0 0 0 0 0 0 1 1 1 1 8 0 1 3 4 6 7 9 0 2 3 5 6 8 9 1 2 4 5 7 8 0 1 3 4 6 7 9 0 2 3 5

Sources: Bloomberg, Standard & Poors, US Federal Reserve, US Treasury, Thomson Reuters

Source: Calder Foundation, New York; Bequest of Mary Calder Rower, 2011

Professional investors only. Private and confidential. Not for public distribution. Please see the full disclaimer at the end of the document

1 3

Low volatility – the great educator The floor (flaw?) in volatility control by correlation

SPX vs 10 %Volatility Control Index 2500

300

250

2000

Probability of rapid volatility expansion tends to 1

Volatility moves to anticipate future low realised volatility

200 1500 150 1000 100 500

0

50

D D D D D D D D D D D D D D D D D D D D D D D e e e e e e e e e e e e e e e e e e e e e e e c c c c c c c c c c c c c c c c c c c c c c c 9 9 9 9 9 9 9 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5

SPX

0

10% Volatility Control Index Source: Bloomberg as at 9 Sep 2016

Low volatility leads to greater amount known about ZERO , anything unknowns % Change in

that moves causes volatility, and therefore approaching zero this scenario is approximated

Volatility is greater off a lower base (Vol of Vol) Base rate volatility is known unknowns, either we know little is unknown or accept low premium

Source: 36 South Capital Advisors LLP


P. 14

Volatility regimes: Correlation and volatility Drift and Brownian volatility = 0 Correlation = 1

SPX as Perpetuity 12000

30 25 20

10000

15

US 10 Year @ 1.55%

10 5

8000

0 Jan

Feb

Mar Apr Asset 1

May

Jun

Jul

Aug Sep Asset 2

Oct

Nov

e u l a V 6000 X P S

Drift volatility = 0 correlation =1 10

4000

8

PE ratio today (approx. 20.5) 4.9% earnings yield PE ratio in 2012 (approx13) 7.7% earnings yield

6

2000

4 2 0 Jan

Feb

Mar

Apr

Asset 1

May

Jun

Jul

Aug

Sep

Oct

Asset 2 Source: 36 South Capital Advisors LLP

Any change in volatility to this state, will by definition cause a reduction in correlation

SPX Level

0

Nov

% % % % % % % % % % % % % % % % % % % 0 . 6 . 2 . 8 . . 6 . 2 . 8 . 4 . 0 . 6 . 2 . 8 . 4 . 0 . 6 . 2 . 8 . 4 . 0 1 1 2 2 3 4 4 5 5 6 7 7 8 8 9 0 0 1 1 1 1 1 1

Effective Yield ( 1/ PE) Source: Goldman Sachs 360; 36 South Capital Advisors LLP


P. 15

Geometric mean, and variance no proxy for risk

Definition of Investment “Geometric Zeros” 2 1.8 1.6 e1.4 u l a V1.2 o i l o f t 1 r o P0.8 l a t o T0.6

Source: Poundstone,W. 2005. “Fortune’s Formula: The Untold Story of the Scientific Betting System that Beat the Casinos and Wall Street” Hill and Wang, New York

0.4 0.2 0

Time as multiple of investment horizon

Smooth Investment Return Geometric "Zero" Breached

Geometric Zero of Capital: achieved by sufficient nonperformance (erosion) of capital such that expected rate of return for duration of investment time horizon no longer reinstates capital

Source: 36 South Capital Advisors LLP


P. 16

Factors affecting the gradient of the zero line Where correlation is an unsustainable rescue Winter White Dwarf Hamster

Geometric Zero of Capital: achieved by sufficient nonperformance (erosion) of capital such that expected rate of return for duration of investment time horizon no longer reinstates capital

EURIBOR

Life expectancy: 2 years

6

12,000,000

5

10,000,000 ) n 8,000,000 B , R U 6,000,000 E ( t b e 4,000,000 D

4

e t a R 3 t s e r 2 e t n I

1

2,000,000

0 J -1

EUR Debt- Social overhang

a n 0 6

J a n 0 8

J a n 1 0

J a n 1 2

J a n 1 4

J a n 1 6

0

M a r 0 2

M a r 0 3

M a r 0 4

M a r 0 5

M a r 0 6

M a r 0 7

M a r 0 8

M a r 0 9

M a r 1 0

M a r 1 1

M a r 1 2

M a r 1 3

M a r 1 4

M M a a r r 1 1 5 6

Source of charts: Bloomberg as at 9 Sep 2016


P. 17


��

�.

��

�.

��

��/�� .

��) �� .

�� , �� ... ��

�.

�� ... ��

�.

��

��) �� .

��

�.

�� ... ��

�.

��

�� , �� .. ��

September 2015

P. 18

EQUITY / RATES CORRELATION IS CONVEX ! 1yr Equity vs US 10yr rates Correlation on Weekly returns since 1962



Challenges for multi-asset portfolios as Efficient Frontier shifts due to correlation shifts (Harry Markovitz) High inflation 70s/80s

QE distortions Tapper Tantrum



Multi-asset portfolio are now more volatile and subject to correlation shocks 

Need for new assets with more stable correlation against risky assets = volatility or variance swaps (equity , fx, rates, credit, commodities).

Source: SG CIB Flow Strategy & Solutions P. 19

Pistols at dawn Nash equilibrium proves necessity for a convex response

s s e c c u s f o y t i l i b a b o r P , σ

d, % move towards target Source: Duel Between Burr And Hamilton, 1870s Engraving. (Photo by: Universal History Archive/UIG via Getty Images)

�� ′�� �  � −  � � Professional investors only. Private and confidential. Not for public distribution. Please see the full disclaimer at the end of the document

P. 20

Long volatility strategy in a portfolio Volatility paid vs expected outcomes Nash equilibrium probability of hit vs % move 1 0.9

Self liquidating hedges

0.8 0.7 y0.6 t i l i b a0.5 b o r P0.4

Efficient hedge dynamic correlation gap

0.3 0.2 0.1 0 100

90

80 70 60 50 40 30 20 % of time from peak to trough

Convex Response

10

0

SPX volatility Days from peak to trough 15/04/2005 13/06/2006 05/03/2007 15/08/2007 12/11/2007 06/02/2008 17/03/2008 14/07/2008 10/10/2008 17/02/2009 01/09/2009 30/10/2009 08/02/2010 20/05/2010 16/03/2011 16/06/2011

48 29 29 31 30 51 29 30 50 43 32 31 29 42 30 26

08/08/2011 01/06/2012 28/12/2012 25/02/2013 20/06/2013 03/02/2014 31/07/2014 15/10/2014 15/01/2015 29/06/2015 24/08/2015 28/09/2015 08/01/2016 11/02/2016 24/06/2016

31 30 27 31 31 31 33 29 30 34 30 6 51 28 30

Linear Response Source of charts: Bloomberg and 36 South Capital Advisors as at 9 Sep 2016


P. 21

Correlation and regimes

Normal

correlation vs Tail correlation

P. 22

Asymmetry in the tails matters



The risk of the long and short VIX position are equal



Volatility (standard deviation) does not pick this up!



Idea : risk premia = compensation for taking on a negatively skewed position

Risk Premia: Asymmetric Tail Risks and Excess Returns

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2502743

P. 23

Classical diversification

P. 24

High moment diversification

P. 25

Skew diversification problem

 Allocation

solution “Tail risk Parity”

•

Define independent sources of tail

•

Equally allocation on tail scenario

•

Combining tail-anti-correlated strategies as natural hedge: •

Trend following (positive skew)

•

Risk premia (negative skew)

A Primer on Alternative Risk Premia http://www.thierry-roncalli.com P. 26

Natural hedge with convexity of trend following

Start date

End date

RP (%)

TF (%)

20080801

20081101

-31.0

17.4

19871001

19871101

-13.1

2.6

19900701

19900901

-12.1

5.7

20140701

20150101

-9.6

17.9

19971001

19971101

-9.6

-0.7

20130301

20130601

-7.2

1.3

20100401

20100601

-5.9

3.0

20070501

20070901

-4.9

-0.2

P. 27


��

�.

��

�.

��

��/�� .

��) �� .

�� , �� ... ��

�.

�� ... ��

�.

��

��) �� .

��

�.

�� ... ��

�.

��

�� , �� .. ��

September 2015

P. 28

VOL OF VOL: STRANGLE ON VSTOXX / VIX SPREAD V2X – VIX term Structure

Var upward vs Vol downward slopping TS 5

12

V2X - VIX Spread today

10

V2X - VIX Spread 1w ago

8

0

-5

6

-10

4

V2X Term structure - VIX Term Structure (5m future - 1m future)

-15

2 Jul16

Aug16 Sep16 Oct16 Nov16 Dec16 Jan17 Feb17 Volatility of volatility term structure

-20 Jun 09

Average Jun 10

Jun 11

Jun 12

Jun 13

Jun 14

Jun 15

Jun 16

Spread of term structures consistently downward slopping

�� 1�/01/2017 1�/01/2017 1�/01/2017 1�/01/2017 15/02/2017 15/02/2017 15/02/2017 15/02/2017 15/03/2017 15/03/2017 15/03/2017 15/03/2017

�� 5.3� 5.3� 5.3� 5.3� 5.13 5.13 5.13 5.13 4.5� 4.5� 4.5� 4.5�

��

�� 3 4 6 � 3 4 6 � 3 4 6 �

�� 0.43 0.66 0.�6 0.3� 0.55 0.�2 0.�5 0.4 0.71 1.03 0.67 0.32

�� 1.5 1.�4 2.22 1.55 1.6� 2.06 2.23 1.5� 1.�0 2.33 2.02 1.46

Source: Bloomberg, SG CIB Flow Strategy & Solutions

Past performance is not a reliable indicator of f uture returns.

P. 29

US10Y CMS VS SPX VARIANCE SWAPS

PnL Long USD 10CMS vs SPX VAR Swap

Source: Bloomberg, SG CIB Financial Engineering


P. 30

P

SPX VARIANCE SWAPS CONTINGENT ON USD 10Y < ATM + 50BPS

PnL Long USD 10CMS vs SPX VAR Swap

Source: Bloomberg, SG CIB Financial Engineering


P. 31

P

SPX VS SX5E CORRIDOR VARIANCE SWAPS

SX5E vs SPX LT Skew

Backtest Dec18 SX5E vs SPX Varspread Corridor vs vanilla

Source: SG CIB Flow Strategy & Solutions

Past performance is not a reliable indicator of f uture returns.

P. 32

P

Variance swap replication

P. 33

Simple variance swap

P. 34


P. 35


P. 36

The future of volatility Gradients in volatility- the fast collapses after an “event” 0.5 0 -0.5

t n -1 e i d a -1.5 r G

-2

5 years

5 years

-2.5

1.5 years

-3 Jan 04

May 05

Oct 06

Feb 08

Jul 09

Nov 10

Apr 12

Aug 13

Dec 14

May 16

Sep 17

THE FAST 30 – Days from peak to trough 15/04/2005 13/06/2006 05/03/2007 15/08/2007 12/11/2007 06/02/2008 17/03/2008 14/07/2008 10/10/2008 17/02/2009 01/09/2009 30/10/2009

48 29 29 31 30 51 29 30 50 43 32 31

08/02/2010 20/05/2010 16/03/2011 16/06/2011 08/08/2011 01/06/2012 28/12/2012 25/02/2013 20/06/2013 03/02/2014 31/07/2014 15/10/2014

29 42 30 26 31 30 27 31 31 31 33 29

15/01/2015 29/06/2015 24/08/2015 28/09/2015 08/01/2016 11/02/2016 24/06/2016

30 34 30 6 51 28 30

Source of charts: Bloomberg and 36 South Capital Advisors as at 9 Sep 2016


P. 37

Bull case for volatility Velocity of Money (M2)

Yield on SPX vs 2 year Gov

5

100

4.5 4 3.5 3 2.5

10

2

1986 1989 1992 1995 1998 2001 2004 2007 2010 2013 201

VIX 3 month futures - bull markets no longer 0.2calm

1

0 -0.2 0.1

6 8 0 2 4 6 8 0 2 4 6 8 0 2 4 6 8 0 2 4 6 7 7 8 8 8 8 8 9 9 9 9 9 0 0 0 0 0 1 1 1 1 9 9 9 9 9 9 9 9 9 9 9 9 0 0 0 0 0 0 0 0 0 1 / 1 / 1 / 1 / 1 / 1 / 1 / 1 / 1 / 1 / 1 / 1 / 2 / 2 / 2 / 2 / 2 / 2 / 2 / 2 / 2 / 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 5 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 / / / / / / / / / / / / / / / / / / / 3 / 1 / 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 8 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0

-0.4 -0.6 -0.8 2004

2006

2009

2012

2014

2017

Yield on SPX vs 2 year Gov Source of charts: Bloomberg and 36 South Capital Advisors as at 9 Sep 2016


3 8


��

�.

��

�.

��

��/�� .

��) �� .

�� , �� ... ��

�.

�� ... ��

�.

��

��) �� .

��

�.

�� ... ��

�.

��

�� , �� .. ��

P. 39

DVEGA/DSPOT (VANNA) AND DVEGA/DVOL (VOMA) DYNAMICS OF AUTOCALLS

Source: SG CIB Flow Strategy & Solutions

P. 40

DISTORTIONS FROM ASIAN STRUCTURED PRODUCTS



In Europe, issuances have also slowed significantly since Sep15.



Monthly vega from SX5E autocallable is now around $10mln (vs ~$20mln in 2014-2015)


P. 41

D3S2ACorrelationKokouNealTu

Recommend Documents