A Universal A Universal Framework Framework For Pricing For Pricing Financial Financial and Insurance Risks and Insurance Risks Presentation at the ASTIN Presentation at the ASTIN Colloquium July 2001, Colloquium July 2001, Washington DC Washington DC Shaun Wang, FCAS, Shaun Wang, FCAS, Ph.D. Ph.D. SCOR Reinsurance SCOR Reinsurance Shaun Wang, Shaun Wang, 2001 2001
33
Embed
A Universal Framework For Pricing Financial and Insurance Risks Presentation at the ASTIN Colloquium July 2001, Washington DC Shaun Wang, FCAS, Ph.D.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
A Universal A Universal Framework Framework For Pricing Financial For Pricing Financial
and Insurance Risksand Insurance Risks
Presentation at the ASTIN Colloquium Presentation at the ASTIN Colloquium July 2001, Washington DCJuly 2001, Washington DC
Present Present a new a new formulaformula to to connectconnect CAPM with Black-CAPM with Black-ScholesScholes
Piece together with Piece together with actuarial axiomsactuarial axioms
Empirical findingsEmpirical findings Capital AllocationsCapital Allocations
CAPMCAPM
Black-ScholesBlack-Scholes
Price DataPrice Data
??
Market Price of Market Price of RiskRisk
Asset return Asset return RR has normal distribution has normal distribution
rr --- the risk-free rate --- the risk-free rate
={ E[R] ={ E[R] r }/r }/[R][R]
isis “ “the market price of riskthe market price of risk” or excess ” or excess return per unit of volatility.return per unit of volatility.
Capital Asset Pricing Capital Asset Pricing ModelModel
Let Let RRii and and RRMM be the return for asset be the return for asset ii
and market portfolio and market portfolio MM. .
MMii RR ),(
The New TransformThe New Transform
extends theextends the “ “market price of riskmarket price of risk” ” in CAPM to risks with non-normal in CAPM to risks with non-normal distributionsdistributions
))(()(* 1 xFxF
is the standard normal cdf.is the standard normal cdf.
If If FFXX is normal( is normal(), ), FFXX** is another is another
normalnormal(( ))
E*[E*[XX] = ] =
If If FFXX is lognormal( is lognormal( ), ), FFXX** is is
another lognormalanother lognormal(( ) )
))(()(* 1 xFxF
Correlation MeasureCorrelation Measure
Risks X and Y can be transformed to Risks X and Y can be transformed to normal variables:normal variables:
)]([*
)],([*1
1
YFY
XFX
Y
X
Define New Correlation
*)*,(),(* YXYX
Why New Correlation ?Why New Correlation ?
Let X ~ lognormal(0,1)Let X ~ lognormal(0,1) Let Y=X^bLet Y=X^b ( (deterministicdeterministic)) For the traditional correlation:For the traditional correlation:
(X,Y) (X,Y) 0 as b 0 as b + +
For the new correlation:For the new correlation:
**((X,Y)=1 for all bX,Y)=1 for all b
Extending CAPMExtending CAPM The transformThe transform recoversrecovers CAPM for risksCAPM for risks
with normal distributionswith normal distributions extends extends the traditional meaning ofthe traditional meaning of { E[R] { E[R] r }/r }/[R][R]
New transformNew transform extendsextends CAPM to risks with CAPM to risks with non-normal distributions:non-normal distributions:
MMii RR ),(*
To reproduce stock’s current value:To reproduce stock’s current value:
a)a) ApplyApply transformtransform to to FFXX to get to get FFXX**, ,
then derive then derive FFYY** from from FFXX**
b)b) Derive Derive FFYY from from FFXX, then apply, then apply
transform transform to to FFYY to get to get FFYY**
ApplyApply transformtransform with same with same ii from from
underlying stock to price optionsunderlying stock to price options
Both Both ii and the expected return and the expected return ii drop drop
out from the risk-adjusted stock price out from the risk-adjusted stock price
distribution!! distribution!!
We’ve just reproduced the B-S price!!We’ve just reproduced the B-S price!!
Recover Black-Recover Black-ScholesScholes
Option Pricing ExampleOption Pricing Example
A stock’s current price = $1326.03. A stock’s current price = $1326.03. Projection of 3-month price: 20 outcomes:Projection of 3-month price: 20 outcomes:
The 3-month risk-free rate = 1.5%.The 3-month risk-free rate = 1.5%.How to price a 3-month European call How to price a 3-month European call option with a strike price of option with a strike price of $1375$1375 ? ?
ComputationComputation
Sample data: Sample data: =4.08%, =4.08%, =8.07%=8.07% UseUse =(=(r)/r)/ =0.320 as “starter” =0.320 as “starter” The transform yields a price =1328.14, The transform yields a price =1328.14,
differing from current price=1326.03differing from current price=1326.03 SolveSolve to match current price. We getto match current price. We get
=0.342=0.342 Use the trueUse the true to price optionsto price options
Using New TransformUsing New Transform ((=0.342)=0.342)
Sorted Orig. Trans. Option wtd wtdSample Prob. Prob. Payoff value value
– implied transform: Simplied transform: SXX(t) (t) S SXX*(t)*(t)
Venter 1991 ASTIN PaperVenter 1991 ASTIN Paper
Graphic IntuitionGraphic Intuition
Theoretical ChoiceTheoretical Choice
• extends classic CAPM and Black-Scholes,
• equilibrium price under more relaxed distributional assumptions than CAPM, and
• unified treatment of assets & losses
))(()(* 1 xFxF
Reality CheckReality Check
Evidence for 3-moment CAPM Evidence for 3-moment CAPM which accounts for skewness which accounts for skewness [[Kozik/Larson paperKozik/Larson paper]]
““Volatility smile” in option pricesVolatility smile” in option prices Empirical risk premiums for tail Empirical risk premiums for tail
events (events (CAT insurance and bond CAT insurance and bond defaultdefault) are higher than implied by ) are higher than implied by the transform.the transform.
2-Factor Model2-Factor Model
11/b /b is a multiple factor to the is a multiple factor to the normal volatilitynormal volatility
b<b<11,, depends on depends on FF((xx)), , with smaller with smaller values at tails (higher adjustment)values at tails (higher adjustment)
bb adjusts for skewness & adjusts for skewness & parameter uncertaintyparameter uncertainty
))(()(* 1 xFbxF
Calibrate the Calibrate the bb-function-function
1)1) Let Let QQ be a symmetric distribution be a symmetric distribution with fatter tails than Normal(0,1):with fatter tails than Normal(0,1):
Theoretical insights of Theoretical insights of bb--functionfunction
Relates closely to 3-moment CAPM.Relates closely to 3-moment CAPM. Explains better investor behavior: Explains better investor behavior:
distortion by distortion by greedgreed and and fearfear Explains “volatility smile” in option Explains “volatility smile” in option
pricesprices Quantifies increased cost-of-capital for Quantifies increased cost-of-capital for
gearing, non-liquidity markets, gearing, non-liquidity markets, “stochastic volatility”, information “stochastic volatility”, information asymmetry, and parameter uncertaintyasymmetry, and parameter uncertainty
Fit 2-factor model to 1999 Fit 2-factor model to 1999 transactionstransactions
Date Sources: Lane Financial LLC PublicationsDate Sources: Lane Financial LLC Publications
Yield Spread for Insurance-Linked Securities
0.00%
2.00%
4.00%
6.00%
8.00%
10.00%
12.00%
14.00%
16.00%
Transactions
Yie
ld S
prea
d
Model-Spread
Empirical-Spread
Use 1999 parameters to price 2000 Use 1999 parameters to price 2000 transactionstransactions
Fitted versus Empirical Spread
0.00%
1.00%
2.00%
3.00%
4.00%
5.00%
6.00%
7.00%
8.00%
Transactions
Yie
ld S
pre
ad
Model-Spread
Empirical-Spread
2-factor model for corporate bonds: 2-factor model for corporate bonds: same same lambdalambda but lower but lower gammagamma than than
CAT-bondCAT-bond
Bond Rating and Yield Spread
0
200
400
600
800
1,000
1,200
1,400
AAA AA A BBB BB B CCC
Bond Rating
Sp
rea
d (
ba
sis
po
ints
)
Model Fitted Spread
Actual Spread
Universal PricingUniversal Pricing
Cross Industry Cross Industry ComparisonComparison
and and by by industry: equity, industry: equity, credit, CAT-credit, CAT-bond, weather bond, weather and insuranceand insurance