DOWNSIDE RISK › AFIR › Colloquia › Zurich › Engle_presentation.pdfDOWNSIDE RISK IN THE CAPM...
Transcript of DOWNSIDE RISK › AFIR › Colloquia › Zurich › Engle_presentation.pdfDOWNSIDE RISK IN THE CAPM...
-
DOWNSIDE RISK DOWNSIDE RISK ––ECONOMETRIC MODELS AND FINANCIAL ECONOMETRIC MODELS AND FINANCIAL
IMPLICATIONSIMPLICATIONS
ROBERT ENGLEROBERT ENGLEJOINT ASTIN/AFIR MEETINGJOINT ASTIN/AFIR MEETING
ETH ZURICHETH ZURICHSEPTEMBER 2005SEPTEMBER 2005
-
22
RISK AND RETURNRISK AND RETURN
THE TRADETHE TRADE--OFF BETWEEN RISK AND OFF BETWEEN RISK AND RETURN IS THE CENTRAL PARADIGM OF RETURN IS THE CENTRAL PARADIGM OF FINANCE.FINANCE.HOW MUCH RISK AM I TAKING?HOW MUCH RISK AM I TAKING?HOW SHOULD I RESPOND TO RISKS HOW SHOULD I RESPOND TO RISKS THAT VARY OVER TIME?THAT VARY OVER TIME?HOW SHOULD I RESPOND TO RISKS OF HOW SHOULD I RESPOND TO RISKS OF VARIOUS MATURITIES?VARIOUS MATURITIES?
-
33
DOWNSIDE RISKDOWNSIDE RISK
The risk of a portfolio is that its value will The risk of a portfolio is that its value will decline, hence DOWNSIDE RISK is a natural decline, hence DOWNSIDE RISK is a natural measure of risk.measure of risk.Many theories and models assume symmetry: Many theories and models assume symmetry: c.f. MARKOWITZ, TOBIN, SHARPE AND BLACK, c.f. MARKOWITZ, TOBIN, SHARPE AND BLACK, SCHOLES, MERTON and Volatility based risk SCHOLES, MERTON and Volatility based risk management systems.management systems.Do we miss anything important?Do we miss anything important?
-
44
MEASURING DOWNSIDE RISKMEASURING DOWNSIDE RISK
Many measures have been proposed. Let Many measures have been proposed. Let rr be be the one period continuously compounded return the one period continuously compounded return with distribution with distribution f(r)f(r) and mean zero. Let x be a and mean zero. Let x be a threshold.threshold.
( ) ( )3/ 23 2Skewness= /E r E r( )( )
Probability of loss = ,
Expected Shortfall =
P r x
E r x r x
<
− <
( )x is the Value at risk if P r xα α< − =
-
55
PREDICTIVE DISTRIBUTION OF PREDICTIVE DISTRIBUTION OF PORTFOLIO GAINSPORTFOLIO GAINS
1%$ GAINS ON PORTFOLIO
-
66
MULTIVARIATE DOWNSIDE RISKMULTIVARIATE DOWNSIDE RISK
WHAT IS THE LIKELIHOOD THAT A WHAT IS THE LIKELIHOOD THAT A COLLECTION OF ASSETS WILL ALL COLLECTION OF ASSETS WILL ALL DECLINE?DECLINE?THIS DEPENDS PARTLY ON THIS DEPENDS PARTLY ON CORRELATIONSCORRELATIONSFOR EXTREME MOVES, OTHER MEASURES FOR EXTREME MOVES, OTHER MEASURES ARE IMPORTANT TOO.ARE IMPORTANT TOO.
-
77
MULTIVARIATE DOWNSIDEMULTIVARIATE DOWNSIDE
““Where are my correlations when I need Where are my correlations when I need them?them?”” –– a portfolio managera portfolio manager’’s lament.s lament.When country equity markets decline When country equity markets decline together more than can be expected from together more than can be expected from the normal correlation pattern, it is called the normal correlation pattern, it is called CONTAGION.CONTAGION.Correlations and volatilities appear to Correlations and volatilities appear to move together.move together.
-
88
MEASURING JOINT DOWNSIDE MEASURING JOINT DOWNSIDE RISKRISK
Let Let yyii be the return on asset ibe the return on asset i
Tail dependenceTail dependence (lower tail dependence) is (lower tail dependence) is defined as the limit as this probability goes to defined as the limit as this probability goes to zero. What is the probability that one asset has zero. What is the probability that one asset has an extreme down move when another has an an extreme down move when another has an extreme down move?extreme down move?
( )Find x such that i iP y xα = <( ) ( )i i j j j j i iP y x y x P y x y xαλ = < < = <
-
99
DEFAULT CORRELATIONSDEFAULT CORRELATIONS
Define an indicator for default and measure the Define an indicator for default and measure the correlation between these indicatorscorrelation between these indicators
For extremes, the default correlation is the same For extremes, the default correlation is the same as the tail dependence.as the tail dependence.
{ } { }( ){ } { }( )
( )( ) ( )
,
2
,
1
/ 1
i i j j
i i j j
Di j y x y x
y x y x
Corr I I
P I I
α
ρ
α
α α
λ α α
< <
< <
=
−=
−
= − −
-
1010
P1,T
P2,T
Probability that the portfolio loses more than K
W1P1+W2P2=-K
-
1111
P1,T
P2,TK1
K2
Put Option on asset 1 Pays
Option on asset 2 Pays
Both options Payoff
-
1212
P1,T
P2,T
Symmetric Tail DependenceSymmetric Tail Dependence
-
1313
P1,T
P2,T
Lower Tail DependenceLower Tail Dependence
-
1414
P1,T
P2,TK1
K2
Put Option on asset 1 Pays
Option on asset 2 Pays
Both options Payoff
-
1515
CREDIT DERIVATIVESCREDIT DERIVATIVES
IT IS WELL DOCUMENTED THAT THE IT IS WELL DOCUMENTED THAT THE MULTIVARIATE NORMAL DENSITY MULTIVARIATE NORMAL DENSITY UNDERPRICES JOINT EXTREME EVENTS SUCH UNDERPRICES JOINT EXTREME EVENTS SUCH AS DEFAULTS. AS DEFAULTS. INDUSTRY HAS ADOPTED A TINDUSTRY HAS ADOPTED A T--COPULA TO COPULA TO PRICE CREDIT BASKETS and CDO PRICE CREDIT BASKETS and CDO tranchestranches..TAIL DEPENDENCE IS ESSENTIAL IN THESE TAIL DEPENDENCE IS ESSENTIAL IN THESE MODELS.MODELS.
-
THE PURPOSE OF MY TALK THE PURPOSE OF MY TALK TODAYTODAY
TIME SERIES ANALYSIS OF TIME SERIES ANALYSIS OF DOWNSIDE RISKDOWNSIDE RISK
-
1717
PURPOSE OF MY TALK TODAYPURPOSE OF MY TALK TODAY
TO SHOW HOW DOWNSIDE RISK CAN BE MODELED AS TO SHOW HOW DOWNSIDE RISK CAN BE MODELED AS A TIME SERIES PROCESS A TIME SERIES PROCESS USING SIMPLY TIME AGGREGATION OF STANDARD USING SIMPLY TIME AGGREGATION OF STANDARD TIME SERIES MODELSTIME SERIES MODELS
CONSEQUENTLYCONSEQUENTLY
DOWNSIDE RISK CAN BE PREDICTEDDOWNSIDE RISK CAN BE PREDICTEDDYNAMIC HEDGING AND DYNAMIC PORTFOLIO DYNAMIC HEDGING AND DYNAMIC PORTFOLIO STRATEGIES CAN BE IMPLEMENTED.STRATEGIES CAN BE IMPLEMENTED.
-
1818
-
1919
AN ECONOMETRIC AN ECONOMETRIC FRAMEWORKFRAMEWORK
MODEL THE ONE PERIOD RETURN AND MODEL THE ONE PERIOD RETURN AND CALCULATE THE MULTICALCULATE THE MULTI--PERIOD PERIOD DISTRIBUTIONDISTRIBUTIONRETURN FROM t UNTIL t + T IS:RETURN FROM t UNTIL t + T IS:
1
T t
t T jj t
R r+
+= +
= ∑
-
2020
ALL MEASURES CAN BE DERIVED ALL MEASURES CAN BE DERIVED FROM THE ONE PERIOD DENSITYFROM THE ONE PERIOD DENSITY
EVALUATE ANY MEASURE BY REPEATEDLY EVALUATE ANY MEASURE BY REPEATEDLY SIMULATING FROM THE ONE PERIOD SIMULATING FROM THE ONE PERIOD CONDITIONAL DISTRIBUTION:CONDITIONAL DISTRIBUTION:
METHOD:METHOD:Draw rDraw rt+1t+1Update density and draw observation t+2Update density and draw observation t+2Continue until T returns are computed.Continue until T returns are computed.Compute measure of downside risk
( )1t tf r +
Compute measure of downside risk
-
2121
A MODELA MODEL
MEAN ZERO, TIME VARYING VOLATILITYMEAN ZERO, TIME VARYING VOLATILITY
ASYMMETRY ASYMMETRY FOLLOWS FROM ASYMMETRY IN SHOCKSFOLLOWS FROM ASYMMETRY IN SHOCKSHOWEVER FOR MULTIHOWEVER FOR MULTI--PERIOD RETURNS, THERE IS PERIOD RETURNS, THERE IS ANOTHER SOURCE ANOTHER SOURCE –– ASYMMETRIC VOLATILITY.
( ) ( )1 1, ~ . . .
0,
ε ε
− −
=
= =t t t t
t t t t t
r h i i d
E r h V r
ASYMMETRIC VOLATILITY.
-
2222
The ARCH ModelThe ARCH Model
The ARCH model of Engle(1982) is a family of The ARCH model of Engle(1982) is a family of specifications for the conditional variance.specifications for the conditional variance.The The qqthth order ARCH or ARCH(q) model is a order ARCH or ARCH(q) model is a weighted average of squared returns.weighted average of squared returns.
Notice that it is a simple generalization of both Notice that it is a simple generalization of both constant variance and rolling variance estimates.constant variance and rolling variance estimates.
2
1
ω α −=
= + ∑q
t j t jj
h r
-
2323
-
2424
-
2525
GARCHGARCH
The Generalized ARCH model of The Generalized ARCH model of Bollerslev(1986) is an ARMA version of Bollerslev(1986) is an ARMA version of this model. this model. The GARCH(1,1) is the workhorseThe GARCH(1,1) is the workhorse
21 1t t th r hω α β− −= + +
-
2626
GOTHIC ARCHGARCH
-
2727
Asymmetric VolatilityAsymmetric Volatility
Often negative shocks have a bigger effect Often negative shocks have a bigger effect on volatility than positive shockson volatility than positive shocksNelson(1987) introduced the EGARCH Nelson(1987) introduced the EGARCH model to incorporate this effect.model to incorporate this effect.I will use a Threshold GARCH or TARCH I will use a Threshold GARCH or TARCH which is like a GARCH but where negative which is like a GARCH but where negative returns get an extra boost. returns get an extra boost.
( )12 2
1 1 0 1tt t t trh r r hIω α γ β
−− − −<= + + +
-
2828
WHERE DOES ASYMMETRIC WHERE DOES ASYMMETRIC VOLATILITY COME FROM?VOLATILITY COME FROM?
LEVERAGE LEVERAGE -- As equity prices fall As equity prices fall the leverage of a firm increases so the leverage of a firm increases so that the next shock has a greater that the next shock has a greater effect on stock prices.effect on stock prices.This effect is usually too small to This effect is usually too small to explain what we see.explain what we see.
-
2929
WHERE DOES ASYMMETRIC WHERE DOES ASYMMETRIC VOLATILITY COME FROM?VOLATILITY COME FROM?
RISK AVERSIONRISK AVERSION–– News of a future News of a future volatility event will lead to stock sales volatility event will lead to stock sales and price declines now. Subsequently, and price declines now. Subsequently, the volatility event occurs. Since events the volatility event occurs. Since events are clustered, any news event will are clustered, any news event will predict higher volatility in the future.predict higher volatility in the future.This effect is especially relevant for This effect is especially relevant for broad market indices since these have broad market indices since these have systematic risk.systematic risk.
-
3030
NEW ARCH MODELSNEW ARCH MODELS
GJRGJR--GARCHGARCHTARCHTARCHSTARCHSTARCHAARCHAARCHNARCHNARCHMARCHMARCHSWARCHSWARCHSNPARCHSNPARCHAPARCHAPARCHTAYLORTAYLOR--SCHWERT
FIGARCHFIGARCHFIEGARCHFIEGARCHComponent Component Asymmetric ComponentAsymmetric ComponentSQGARCHSQGARCHCESGARCHCESGARCHStudent tStudent tGEDGEDSPARCHSPARCHAutoregressive Conditional DensityAutoregressive Conditional DensityAutoregressive Conditional Skewness
SCHWERTAutoregressive Conditional Skewness
-
3131
-
3232
-
3333
-
3434
-
3535
TWO PERIOD RETURNSTWO PERIOD RETURNS
Two period return is Two period return is the sum of two one the sum of two one period continuously period continuously compounded returnscompounded returnsLook at binomial tree Look at binomial tree versionversionAsymmetric Volatility Asymmetric Volatility gives negative gives negative skewness
High variance
Low variance
skewness
-
3636
ANALYTICALLY: TARCHANALYTICALLY: TARCHWITH SYMMETRIC INNOVATIONSWITH SYMMETRIC INNOVATIONS
( ) ( )( )
( )( )
( ) ( )
3 3 2 2 31 1 1 1
1
2 2( 0)
3( 0)
3 31 1 1 ( 0)
3 3
0 0 3 0
3
3 0
and the conditional third moment is3 0
t
t
t
t t t t t t t t
t t
t t t tr
t r
t t t t t r
E r r E r r r r r rE r h
E r r r I h
E r I
E r r E r I
ω α γ β
γ
γ
+ + + +
+
<
<
− + − <
+ = + + +
= + + +
⎡ ⎤= + + +⎣ ⎦
= <
+ =
-
STYLIZED FACTSSTYLIZED FACTS
-
3838
S&P 500 DAILY RETURNSS&P 500 DAILY RETURNS
-.20
-.16
-.12
-.08
-.04
.00
.04
.08
1960 1965 1970 1975 1980 1985 1990 1995 2000
-
3939
0
1000
2000
3000
4000
5000
6000
-0.20 -0.15 -0.10 -0.05 0.00 0.05 0.10
Series: RETUSSPSample 1/4/1955 TO 6/25/2004Observations 12455
Mean 0.000318Median 0.000375Maximum 0.090994Minimum -0.204669Std. Dev. 0.009179Skewness -0.926286Kurtosis 28.00273
Jarque-Bera 326200.8Probability 0.000000
HISTOGRAM OF S&P500 DAILY RETURNS
-
4040
TRIMMING .001 IN EACH TAIL TRIMMING .001 IN EACH TAIL (8 DAYS)(8 DAYS)
0
400
800
1200
1600
2000
2400
-0.025 0.000 0.025
Series: RETUSSPSample 1 12455 IF Y>LOWTRIM AND Y
-
4141
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0 25 50 75 100 125 150 175 200 225
SKEW_ALLSKEW_TRIMSKEW_PRESKEW_POST
SKEWNESS OF MULTIPERIOD RETURNS
-
4242
STANDARD ERRORSSTANDARD ERRORS
ARE THESE DIFFERENCES SIGNIFICANT?ARE THESE DIFFERENCES SIGNIFICANT?THE INFERENCE IS COMPLICATED BY THE THE INFERENCE IS COMPLICATED BY THE OVERLAPPING OBSERVATIONS AND BY THE OVERLAPPING OBSERVATIONS AND BY THE DEPENDENCE DUE TO ESTIMATING THE MEAN.DEPENDENCE DUE TO ESTIMATING THE MEAN.FROM SIMPLE ROBUST TESTS, SIZE FROM SIMPLE ROBUST TESTS, SIZE CORRECTED BY MONTE CARLO, THESE ARE CORRECTED BY MONTE CARLO, THESE ARE SIGNIFICANT.SIGNIFICANT.
-
4343
EVIDENCE FROM DERIVATIVES EVIDENCE FROM DERIVATIVES
THE HIGH PRICE OF OUTTHE HIGH PRICE OF OUT--OFOF--THETHE--MONEY MONEY EQUITY PUT OPTIONS IS WELL DOCUMENTEDEQUITY PUT OPTIONS IS WELL DOCUMENTEDTHIS IMPLIES SKEWNESS IN THE RISK THIS IMPLIES SKEWNESS IN THE RISK NEUTRAL DISTRIBUTIONNEUTRAL DISTRIBUTIONMUCH OF THIS IS PROBABLY DUE TO MUCH OF THIS IS PROBABLY DUE TO SKEWNESS IN THE EMPIRICAL DISTRIBUTION SKEWNESS IN THE EMPIRICAL DISTRIBUTION OF RETURNS.OF RETURNS.DATA MATCHES EVIDENCE THAT THE OPTION DATA MATCHES EVIDENCE THAT THE OPTION SKEW IS ONLY POST 1987.SKEW IS ONLY POST 1987.
-
4444
MATCHING THE STYLIZED MATCHING THE STYLIZED FACTSFACTS
ESTIMATE DAILY MODELESTIMATE DAILY MODELSIMULATE 250 CUMULATIVE RETURNS SIMULATE 250 CUMULATIVE RETURNS 10,000 TIMES WITH SEVERAL DATA 10,000 TIMES WITH SEVERAL DATA GENERATING PROCESSESGENERATING PROCESSESCALCULATE SKEWNESS AT EACH CALCULATE SKEWNESS AT EACH
HORIZONHORIZONANALYTICAL CALCULATIONANALYTICAL CALCULATION
-
4545
-1.4
-1.2
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
25 50 75 100 125 150 175 200 225 250
SKEW_EXSKEW_BOOT_EX
SKEW_EXSSKEW_BOOT_EXS
SKEWS FOR SYMMETRIC AND ASYMMETRIC MODELS
-
4646
Time Aggregation of TARCHTime Aggregation of TARCH
-
4747
IMPLICATIONSIMPLICATIONS
MultiMulti--period empirical returns are more period empirical returns are more skewed than one period returns (omitting skewed than one period returns (omitting 1987 crash)1987 crash)Asymmetric volatility is needed to explain Asymmetric volatility is needed to explain this.this.Skewness has increased since 1987, Skewness has increased since 1987, particularly for longer horizons.particularly for longer horizons.These findings match options markets.These findings match options markets.
-
MULTIVARIATE MODELSMULTIVARIATE MODELS
-
4949
DOWNSIDE RISK RESULTS FROM DOWNSIDE RISK RESULTS FROM TIME AGGREGATION WITH:TIME AGGREGATION WITH:
ASYMMETRIC CORRELATIONSASYMMETRIC CORRELATIONSCORRELATIONS RISE PARTICULARLY AFTER TWO CORRELATIONS RISE PARTICULARLY AFTER TWO ASSETS BOTH DECLINE. (Asymmetric DCC (ASSETS BOTH DECLINE. (Asymmetric DCC (CappielloCappiello, , Engle, Sheppard(2004))Engle, Sheppard(2004))
VOLATILITY SHOCKS ARE CORRELATEDVOLATILITY SHOCKS ARE CORRELATEDPURE VARIANCE COMMON FEATURES(Engle, PURE VARIANCE COMMON FEATURES(Engle, Marcucci(2005))Marcucci(2005))FACTOR MODELS (Engle Ng and Rothschild(1992))FACTOR MODELS (Engle Ng and Rothschild(1992))CREDIT RISK MODEL(Engle, CREDIT RISK MODEL(Engle, BerdBerd, Voronov(2005)), Voronov(2005))
-
5050
DOWNSIDE RISK IN THE CAPMDOWNSIDE RISK IN THE CAPM
The return on a stock can be decomposed into The return on a stock can be decomposed into systematic and idiosyncratic returns using the systematic and idiosyncratic returns using the beta of the stockbeta of the stock
If the market declines substantially, many stocks If the market declines substantially, many stocks will decline. There will be skewness in each will decline. There will be skewness in each stock and downside risk in the portfolio.
, , ,i t i m t i tr rβ ε= +
stock and downside risk in the portfolio.
-
5151
SKEWNESSSKEWNESS
Under the standard assumptions, the Under the standard assumptions, the skewness of return skewness of return ii is related to the is related to the skewnessskewness of the market by of the market by where where ρ ρ is the correlation between stock is the correlation between stock and market.and market.Notice that all stocks will then have Notice that all stocks will then have skewness but that it will be less than for skewness but that it will be less than for the market. the market.
3ρ=i ms s
-
5252
TAIL DEPENDENCETAIL DEPENDENCE
The probability that two stocks will both The probability that two stocks will both underperformunderperform some threshold can be some threshold can be calculated conditional on the market calculated conditional on the market return.return.When the market return is a fatWhen the market return is a fat--tailed tailed distribution, tail dependence rises.distribution, tail dependence rises.
-
5353
SUMMARYSUMMARY
ASYMMETRIC VOLATILITY IN THE ASYMMETRIC VOLATILITY IN THE MARKET FACTOR IMPLIESMARKET FACTOR IMPLIES
SKEWNESS IN MULTIPERIOD MARKET SKEWNESS IN MULTIPERIOD MARKET RETURNSRETURNSSKEWNESS IN MULTIPERIOD EQUITY SKEWNESS IN MULTIPERIOD EQUITY RETURNSRETURNSLOWER TAIL DEPENDENCE IN EQUITY LOWER TAIL DEPENDENCE IN EQUITY RETURNSRETURNS
-
IMPLICATIONS FOR IMPLICATIONS FOR FINANCIAL MANAGEMENTFINANCIAL MANAGEMENT
-
5555
Don’t ask….
-
5656
IMPLICATIONS FOR RISK IMPLICATIONS FOR RISK MANAGEMENTMANAGEMENT
MULTIMULTI--PERIOD RISKS MAY BE PERIOD RISKS MAY BE SUBSTANTIALLY DIFFERENT FROM ONE SUBSTANTIALLY DIFFERENT FROM ONE PERIOD RISKS.PERIOD RISKS.THE MULTITHE MULTI--PERIOD RISK CHANGES OVER PERIOD RISK CHANGES OVER TIME AND CAN BE FORECAST.TIME AND CAN BE FORECAST.BIG MARKET DECLINES ARE MORE BIG MARKET DECLINES ARE MORE LIKELY WHEN VOLATILITY IS HIGHLIKELY WHEN VOLATILITY IS HIGH
-
5757
IMPLICATIONS FOR IMPLICATIONS FOR DERIVATIVE HEDGINGDERIVATIVE HEDGING
AS EACH NEW PERIOD RETURN IS AS EACH NEW PERIOD RETURN IS OBSERVED, THE DERIVATIVE CAN BE OBSERVED, THE DERIVATIVE CAN BE REPRICED AND THE HEDGE UPDATED.REPRICED AND THE HEDGE UPDATED.GREEKS CAN BE CALCULATED FROM GREEKS CAN BE CALCULATED FROM SIMULATION PRICING TO SIMPLIFY THE SIMULATION PRICING TO SIMPLIFY THE UPDATINGUPDATING
-
5858
IMPLICATIONS FOR PORTFOLIO IMPLICATIONS FOR PORTFOLIO SELECTIONSELECTION
MEAN VARIANCE PORTFOLIO MEAN VARIANCE PORTFOLIO OPTIMIZATION WILL MISS THESE OPTIMIZATION WILL MISS THESE ASYMMETRIES.ASYMMETRIES.
HIGH FREQUENCY REBALANCING WILL HIGH FREQUENCY REBALANCING WILL GIVE GIVE EARLY WARNINGEARLY WARNING OF DOWNSIDE OF DOWNSIDE RISK.RISK.
-
5959
HOW TO DO THIS?HOW TO DO THIS?
SUBOPTIMAL METHOD 1SUBOPTIMAL METHOD 1MYOPIC ASSET ALLOCATION ON A HIGH MYOPIC ASSET ALLOCATION ON A HIGH FREQUENCY BASIS. FREQUENCY BASIS. AS VOLATILITIES RISE, YOU NATURALLY SHIFT OUT AS VOLATILITIES RISE, YOU NATURALLY SHIFT OUT OF RISKY ASSETS.OF RISKY ASSETS.
SUBOPTIMAL METHOD 2SUBOPTIMAL METHOD 2MULTIMULTI--PERIOD FORECAST OF RISK GIVES AN EXPERIOD FORECAST OF RISK GIVES AN EX--ANTE OPTIMAL PLAN.ANTE OPTIMAL PLAN.OVERINVEST WHEN VOLATILITY IS LOW AND OVERINVEST WHEN VOLATILITY IS LOW AND UNDERINVEST WHEN IT IS HIGHUNDERINVEST WHEN IT IS HIGH
-
6060
OPTIMAL METHODOPTIMAL METHOD
DYNAMIC PROGRAMMING:DYNAMIC PROGRAMMING:WHEN VOLATILITY IS LOW, UNDERINVEST, WHEN VOLATILITY IS LOW, UNDERINVEST, RECOGNIZING THAT THIS PLAN MAY RECOGNIZING THAT THIS PLAN MAY CHANGE WHEN THE SUBSEQUENT CHANGE WHEN THE SUBSEQUENT VOLATILITY IS OBSERVEDVOLATILITY IS OBSERVEDSEE COLACITO AND ENGLE(2004)SEE COLACITO AND ENGLE(2004)
-
6161
EXPECTED RETURNSEXPECTED RETURNS
EACH OF THESE METHODS REQUIRES EACH OF THESE METHODS REQUIRES EXPECTED RETURNSEXPECTED RETURNS--COORDINATION OF COORDINATION OF RISK MANAGEMENT AND ALPHA RISK MANAGEMENT AND ALPHA ESTIMATIONESTIMATIONTHE LISTED IMPLICATIONS ARE BASED THE LISTED IMPLICATIONS ARE BASED ON THE ASSUMPTION THAT EXPECTED ON THE ASSUMPTION THAT EXPECTED RETURNS ARE UNCHANGED.RETURNS ARE UNCHANGED.IS THIS REASONABLE?IS THIS REASONABLE?
-
6262
BUT IF EVERYBODY DID THIS?BUT IF EVERYBODY DID THIS?
IF ALL AGENTS FOLLOW THIS STRATEGY, THEN IF ALL AGENTS FOLLOW THIS STRATEGY, THEN EXPECTED RETURNS WOULD NECESSARILY EXPECTED RETURNS WOULD NECESSARILY ADJUST. RETURNS WOULD INSTANTANEOUSLY ADJUST. RETURNS WOULD INSTANTANEOUSLY MOVE ENOUGH TO RESTORE EQUILIBRIUM. MOVE ENOUGH TO RESTORE EQUILIBRIUM. CAMPBELL AND HENTSCHEL(1992)CAMPBELL AND HENTSCHEL(1992)IN A REPRESENTATIVE AGENT WORLD, THERE IN A REPRESENTATIVE AGENT WORLD, THERE WOULD NO LONGER BE A MOTIVE FOR WOULD NO LONGER BE A MOTIVE FOR ADJUSTING TO CHANGES IN RISK.ADJUSTING TO CHANGES IN RISK.
-
6363
IN GENERAL EQUILIBRIUMIN GENERAL EQUILIBRIUM
CHANGES IN RISK WOULD INSTANTLY LEAD TO CHANGES IN RISK WOULD INSTANTLY LEAD TO CAPITAL GAINS OR LOSSES. CAPITAL GAINS OR LOSSES. INVESTORS WOULD TAKE SMALLER POSITIONS INVESTORS WOULD TAKE SMALLER POSITIONS BECAUSE OF THE MULTIBECAUSE OF THE MULTI--PERIOD RISKS OR PERIOD RISKS OR WOULD REQUIRE HIGHER RETURNS. WOULD REQUIRE HIGHER RETURNS. WE SAY IN THIS CASE, WE SAY IN THIS CASE, ““DOWNSIDE RISK IS DOWNSIDE RISK IS PRICEDPRICED””..
-
6464
HOWEVER EVEN IN HOWEVER EVEN IN EQUILIBRIUMEQUILIBRIUM
THERE IS NO REASON TO BELIEVE DOWNSIDE RISK THERE IS NO REASON TO BELIEVE DOWNSIDE RISK WOULD DISAPPEAR OR COLLAPSE IN TIME.WOULD DISAPPEAR OR COLLAPSE IN TIME.WITH HETEROGENEITY, THERE WOULD STILL BE WITH HETEROGENEITY, THERE WOULD STILL BE REASONS TO REBALANCE.REASONS TO REBALANCE.DERIVATIVE REPLICATION STRATEGIES CONTINUE TO DERIVATIVE REPLICATION STRATEGIES CONTINUE TO BE USEFUL.BE USEFUL.DERIVATIVE PRICING FOR NONDERIVATIVE PRICING FOR NON--LINEAR PAYOFFS SUCH LINEAR PAYOFFS SUCH AS OPTIONS AND CREDIT DERIVATIVES WILL NEED TO AS OPTIONS AND CREDIT DERIVATIVES WILL NEED TO BE MODIFIED.BE MODIFIED.
-
6565
CONCLUSIONSCONCLUSIONS
ASYMMETRIC VOLATILITY AND CORRELATION MODELS ASYMMETRIC VOLATILITY AND CORRELATION MODELS ARE POWERFUL TOOLS FOR ANALYZING DOWNSIDE ARE POWERFUL TOOLS FOR ANALYZING DOWNSIDE RISKRISKONE PERIOD MODELS HAVE BIG IMPLICATIONS ABOUT ONE PERIOD MODELS HAVE BIG IMPLICATIONS ABOUT THE LONG HORIZON RETURNS THE LONG HORIZON RETURNS THE UPDATING OF VOLATILITY AND RISK MEASURES THE UPDATING OF VOLATILITY AND RISK MEASURES HAS A NATURAL APPLICATION TO DERIVATIVE HAS A NATURAL APPLICATION TO DERIVATIVE HEDGING, PRICING, AND POSSIBLY PORTFOLIO HEDGING, PRICING, AND POSSIBLY PORTFOLIO REBALANCING.REBALANCING.
-
6666
DOWNSIDE RISK – ECONOMETRIC MODELS AND FINANCIAL IMPLICATIONSRISK AND RETURNDOWNSIDE RISKMEASURING DOWNSIDE RISKPREDICTIVE DISTRIBUTION OF PORTFOLIO GAINSMULTIVARIATE DOWNSIDE RISKMULTIVARIATE DOWNSIDEMEASURING JOINT DOWNSIDE RISKDEFAULT CORRELATIONSSymmetric Tail DependenceLower Tail DependenceCREDIT DERIVATIVESTHE PURPOSE OF MY TALK TODAYPURPOSE OF MY TALK TODAYAN ECONOMETRIC FRAMEWORKALL MEASURES CAN BE DERIVED FROM THE ONE PERIOD DENSITYA MODELThe ARCH ModelGARCHAsymmetric VolatilityWHERE DOES ASYMMETRIC VOLATILITY COME FROM?WHERE DOES ASYMMETRIC VOLATILITY COME FROM?NEW ARCH MODELSTWO PERIOD RETURNSANALYTICALLY: TARCH WITH SYMMETRIC INNOVATIONSSTYLIZED FACTSS&P 500 DAILY RETURNSTRIMMING .001 IN EACH TAIL (8 DAYS)STANDARD ERRORSEVIDENCE FROM DERIVATIVESMATCHING THE STYLIZED FACTSTime Aggregation of TARCHIMPLICATIONSMULTIVARIATE MODELSDOWNSIDE RISK RESULTS FROM TIME AGGREGATION WITH:DOWNSIDE RISK IN THE CAPMSKEWNESSTAIL DEPENDENCESUMMARYIMPLICATIONS FOR FINANCIAL MANAGEMENTIMPLICATIONS FOR RISK MANAGEMENTIMPLICATIONS FOR DERIVATIVE HEDGINGIMPLICATIONS FOR PORTFOLIO SELECTIONHOW TO DO THIS?OPTIMAL METHODEXPECTED RETURNSBUT IF EVERYBODY DID THIS?IN GENERAL EQUILIBRIUMHOWEVER EVEN IN EQUILIBRIUMCONCLUSIONS