a game theoretic approach to robust option pricing peter demarzo, stanford ilan kremer, stanford...

36
A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Upload: trace-belew

Post on 15-Dec-2015

213 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

A Game Theoretic Approach to Robust Option Pricing

Peter DeMarzo, Stanford

Ilan Kremer, Stanford

Yishay Mansour, TAU

Page 2: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Finance: Efficient markets,

Option Pricing,Black-Scholes-Merton

Universal Portfolios:Cover et al.

Game theory-Calibration, regret matching

Approachability:Hannan-Blackwell

Page 3: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Example 1: ApproachabilityYou repeatedly predict the outcome of a coin toss. The coin need not be a fair

coin. What success rate can you guarantee in the long run?

Simple strategy: Predict Heads with probability 50% and obtain 50% success rate.

Learning Strategy: At each point in time guess heads if there were more heads in the past, otherwise guess tails. In the limit the success rate is max{p,1-p} where p is the fixed parameter of the coin.

Q: What if the coin can change arbitrarily from period to period?A: You can still get an equivalent performance!!

* *. .lim {realized success rate} max ,1 0n n n a sp p

Page 4: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Example 2: Competitive Analysis (Regret Minimization)

When you go gambling each minute you choose which slot machine to use. There are N different machines, some machines may be better. You see the payoff of machines even if you did not use them.

Goal: In the long run obtain an average payoff that is no worse than the best individual machine ex-post. (No Regret)

Q: What if payoffs are not stationary (so a machine which has a high payoff may deteriorate over time)?

Page 5: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Regret/Approachability

Introduced by Blackwell and Hannan in the late 50s, rediscovered and used in:

• Computer Science- online algorithms• Statistic and Information Theory• Game Theory- Calibration, dynamic foundation

for Nash/correlated equilibrium • With the exception of work on ‘Universal

Portfolios’ not incorporated into finance.

Page 6: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Regret Minimization

• Dynamic optimization under uncertainty without a prior.

• Worst case analysis but specified in relative rather than absolute terms (as in Gilboa and Schmeidler)

Page 7: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Insights/Results

• Minimizing Regret can be expressed as robust upper bound for option pricing. Describe trading strategies that are based on approachability and the

bounds/regret they imply for call option with different strike prices.

• The optimal robust upper bound can be expressed as a value of a zero sum game. Provide a numerical solution and conjecture about closed form

solutions. Bounds are not wide and resemble empirical patterns.

Page 8: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Options: Basics

• What is an option:A right to buy a stock at a given price

strike price = K

At a given date (European) duration = T

• Option payoff: Max{ 0 , ST -K }St = Stock price at time t

• This talk:For Simplicity:

zero interest rate no dividends

CT

STK

Page 9: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Option Pricing: Arbitrage Bounds

• Upper Bound [Merton 1973]Current stock price: S1

• Lower bounds [Merton 1973]Always positive

At worst, payoff is zero.

Stock versus strike price: S1 – K

• Claim these bounds are tight

• Proof: Assume a huge change in first period…

• Better pricing needs more assumptions!

Page 10: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Example 1- Binomial model

• Suppose the risky asset can take only two values

BondCall (K=1)

Stock

$1

$1.2$0.8

$1

$1$1

?$

$0.2$0

0.5(0.8 1.2) - 0.4 (1 1) = (0 0.2)

Option price is 0.5-0.4=0.1

Page 11: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Example II: Black & Scholes

• Extend the tree to many periodsThe limit is continuous time

• Black and Scholescontinuous prices and complete markets

A specific stochastic model: random walk + drift

Page 12: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Regret

Regret- There is a given strategy and a set of alternative strategies. Regret is defined as the difference/ratio between the performance of the given strategy and the ex-post optimal strategy among the alternatives.

Regret guarantee- A lower bound for regret that holds even in the worst case scenario. This guarantee may be conditional on some restricted set of possible scenarios.

For the purpose of this talk we ignore any behavioral aspects. We do not argue that people behave according to our measure of regret or that they should behave in this way!!! We consider a specific regret measure to allow us to derive pricing bounds and compare them to the existing literature.

Page 13: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Regret and Financial marketsI have $100 which equals the price of IBM .

Should I buy one share of IBM or get a risk free asset?

Max{IBM, risk free asset}

Ex-post, compare to:

Loss: ratio

Alternatives: IBM, risk free asset

Page 14: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Linking Regret to Options

Note that holding Treasuries plus at-the-money call option on IBM leads to no regret:

Payoff = Max{IBM, risk-free asset}

Thus, regret minimizing trading strategies have implications for option values.

Page 15: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Regret and option pricing- An Example

• Suppose we measure regret by looking at the ratio of our performance to the best asset ex-post. In addition, suppose that the current IBM share price is $100 and the risk free interest rate is zero. Your goal is to minimize regret as compared to the best asset ex-post.

• Suppose we have a trading strategy such that if we start with $100 then at time T our payoff will always exceed max{80,0.8ST}, where ST denotes IBM share price at time T. Hence, the regret is guarantee is 20%.

• We later describe how one can construct such strategies and know only focus on the implications to option pricing.

Page 16: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

• By scaling we conclude that starting with $125 our strategy would have a payoff that exceeds max{100, ST}.

• max{100, ST} is like $100 plus a call option with strike $100

the value of the option is bounded by $25

Page 17: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Model

• Discrete-time finite-horizon model t=1..T

• A risky asset whose value at time t is St where St=(1+rt)*St-1, where rt≥-1.

• In addition agents can borrow and lend at zero interest rates.

• Restriction on price paths: (rt…rt) RT, is compact and 0

Example:2

1

ˆ{ | ( ) log(1 ) }n

jj

r q r r q

Page 18: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Model- cntd.

• A dynamic trading strategy has initial value G0=c. At time t invest a fraction xt in the risky asset and 1- xt in the risk free asset. Zero risk free rate implies that Gt+1=Gt (1+ xt

rt);

• Definition We say that c=C(K) is an upper bound if there exists a dynamic trading strategy that starts with $c and for all possible price path in its final payoff, GT, satisfies: GT≥max{0,ST-K} (super replication).

Page 19: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Blackwell- (recall Example #2)

• You repeatedly choose a single action among {1..I} possible alternatives; j,i denotes the payoff of alternative i at time j.

• Can use a randomized strategy which is described by a random variable j; j=i implies that you choose alternative i at time j; your time t payoff is given by ,j

Page 20: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Aggregate Regret so far

Regret vs. machine #1

Reg

ret

vs.

mac

hine

#2

Choose the two alternatives with probability proportional

to current regret

Period n+1 expected regret if

Machine #1 pays more

Play 2

Play 1

Suppose Machine #1 pays

more

Period n+1 expected regret if Machine #2 pays more

2 2 21 1 1,2 1,12

( | ) | | ( )n t n n nE A A

Page 21: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Finite horizon properties

Proposition Conditional on the set of realized payoffs :

Corollary Conditional on the set of realized payoffs :

2 21 22

1

( )n

n j jj

E A

2, , , 1 2

1 1 1

(max ) - max { } ( )n n n

i n i j i j i j jj j j

E A E

Page 22: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Asymptotic No-Regret

Theorem (Hannan & Blackwell) If payoff are uniformly bounded then there exists a randomized strategy so that :

, , .1 1

1 1lim [ - max { }] 0

j

n n

n j i j i a sj jn n

Page 23: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Arbitrary starting point-

Consider a variant of the previous strategy where instead of starting at (0,0) we start at an arbitrary point (-x,-y) for some non negative x,y

Corollary Conditional on the set of realized payoffs :

Useful in improving performance and in the

application for different strike prices

2 2 2 2

2( , ) ( )nE A x y q r x y

Page 24: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

A trading strategy

• Multiplicative model versus additive model: Let 0,t=0, 1,t=ln(1+rt).

• Remove randomness: Invest at time t a fraction of xt=E(t) in the risky asset

Proposition The payoff of a trading strategy based on the generalized strategy satisfies:

2 2 2

2 2 2

max{ ( ), ( ) } where

( ) exp( ( ) ), and

( ) exp( ( ) )

n nG r r S

r x q r x y

r y q r x y

Page 25: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Application- Upper bounds for at the money

options (K=1)

Using the same logic as the IBM example if

So we can choose x=y

max{ , * }

1then the bound is 1

n nG S

Page 26: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Application- Upper bounds for at the money

options (K=1)

Restricting price paths:

Using the expressions we derive before we can get an upper bound on the regret. Using the same logic as the example of IBM one gets:

Using the basic trading strategy:

Using a generalized strategy with optimal starting point (that depends on )

ˆexp[ ] 1q

ˆ{ | ( ) }r q r q

logq̂

ˆexp[ / 2] 1q

Page 27: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

K1

• Choose starting point where x=y+log(K)• That implies:

• It also implies a bound of 1/-k for the value of an option with a strike k

- Borrow $k and invest 1/ in the trading strategy.

max{ ( ), ( ) } where ( ) / ( )n nG r r S r r k

Page 28: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Optimal bound

• Let V(s,2,n) denote the optimal (lowest) upper bound for a call option with a strike k=1 when the current price is S. This is equivalent to having S=1 and arbitrary K. The restriction on the price paths is again:

• Let V(s,2) denote the limit as n goes to infinity.

{ | ( ) }r q r

Page 29: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Dynamic Programming

2

,

2 2

2 2

* 2 2

( , , ) min

. . , , 1

for all such that

( , ) lim ( , , )

B

r r

n

V S q n S B

s t Se B V Se q r n

r r q

V S q V S q n

Page 30: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Conjecture

1/

* 2

1/

for 1

( , )

1 for 1

q

q

q qS S

q qV S q

q qS S S

q q

where 1, 1q qq e q e

Page 31: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Consider small q

• Original strategy-

• Optimal starting point-

• Optimal strategy:

• Black-Scholes

ˆ ˆexp[ ] 1q q

ˆ ˆexp[ / 2] 1 / 2q q

ˆ / 2q

ˆ / 2q

Page 32: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Example: 20(vs. Black-Scholes)

15%

20%

25%

30%

35%

40%

45%

50%

0.75 0.85 0.95 1.05 1.15 1.25

Strike Price K

Bla

ck

-Sc

ho

les

Imp

lied

Vo

lati

lity

of

Pri

ce

Bo

un

d

Gradient Strategy Optimal q-Bound Black-Scholes

Page 33: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Approaches to option pricing

Black and Scholes: • Continuous price paths • Constant volatility (quadratic variation)

Exact replication and pricing

With jumps & stochastic volatility, exact pricing requires :) i (A probability distribution P over price paths

) ii (A utility function, U

Our Approach:• No probability or preference assumptions• Constraints on the set of price paths (support of P)• Super-replication Upper Bound for Option Price

No probability or preference

assumptions

but

Strong assumption on allowable paths

Page 34: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Relation to Universal Portfolios:

• Cover and Ordentlich (1998)- consider the set of constant-rebalanced portfolios.

- Provide a close form tight universal regret bound (min-max). This provides a bound for the value of the derivative that pays ex-post the optimal constant-rebalanced portfolios.

- Universal means that we consider all possible scenarios (?).

Page 35: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Call (or put) options

- The relevant set of benchmarks is much simpler – buy and hold strategies.

- The min-max value is 0.5 (trivial to prove). From option pricing perspective yields an upper bound of the current stock price.

- Hence, need to consider a ‘less universal’ approach and the Blackwell strategy is useful.

Page 36: A Game Theoretic Approach to Robust Option Pricing Peter DeMarzo, Stanford Ilan Kremer, Stanford Yishay Mansour, TAU

Early empirical work (joint work with Tyler Shumway)

S&P 500 options prices from 1/96 to 4/05 from OptionMetrics.

Options with 15 to 45 calendar days to maturity.