Why Does the Mathematics of Probability Have Applicability and Explanatory/Predictive Power for Financial Derivatives Markets? IPPP School on Physics and Techniques of Event Generators Durham, United Kingdom 18-20 April 2007 Bryan Lynn Managing Director Merrill Lynch Global Markets and Investment Banking 1 Summary: Comparison of Applicability, Explanatory Power and Predictive Power of Mathematics of Probability Statistical Physics (Natural Law) Gambling, Financial Derivatives Markets (Binding Contracts) Criminology (Sociology/Psychology) Finance Industry Standard Practice: Hedging of Derivatives Risk Implied Volatility Conclusions Post Script: Finance Industry Standard Practice: Diversification of Cash Portfolio Risk Value at risk (VaR) Finance Innovation: Diversification of NonLinear Derivatives Risk Finance Industry Standard (Quadratic Approximation) An Unsolved Monte Carlo Financial Mathematical problem 2 Statistical Physics Applicability and Explanatory/Predictive Power of Probability Conservation of Probability Current Heat Equation Brownian motion in non-relativistic gases Normal distribution emerges from conservation of Energy-Momentum during molecular collisions Probability predicts relationships among thermodynamic (statistically averaged) quantities 3 Gambling (Dice or Roulette) Probably (historically) invented mathematics of probability Can analyze transparent complete-information games: Applicability and explanatory/predictive power comes from ultra-strong constraints in gaming rules and consequent financial agreements. Best Strategy: use probability to drive all betting 4 Gambling (Poker) Non-transparent incomplete-information games: Still significant explanatory/predictive power because of weaker constraints due to agreements/rules Probability still compellingly useful for early (and most later) betting But human behaviour (e.g. bluffing, mindreading, body language) is very important and the best hand does not always win If you don’t know the odds you will certainly lose, but you must have human skill in order to win “The race is not always to the swift, nor is the victory to the strong. But that’s the way to bet!”…Sergeant Bilko (~1960) 5 Criminology (Sociology/Psychology) Predictions of recidivism for individual criminals chosen from among a large set of prisoners (e.g. murderers) up for parole. Could use fancy mathematics (e.g. probability), but this approach is known to fail: i.e. by physics, gambling and financial derivatives valuation/risk standards Instead, examine (and try to evaluate) the character and behaviour of individuals Mathematics is not applicable and has no explanatory or predictive power 6 Financial Derivatives derivable - by selfconsistent “no-arbitrage” - from existing Contracts/Agreements (i.e. analogous with Dice/Roulette) Building blocks $ Bonds $ interest rate (compounding) discount factor P0$T exp( r $T ) Yen Bonds Yen interest rate (compounding) discount factor Yen P0Yen exp( r T) T r $ , r Yen are annualized continuous interest rates $ / Yen S Spot FX: t 0 the number of $ bought for 1 Yen at t=0 (now) Pricing is driven by self-consistency (noarbitrage) with related contracts $-Yen FX Forwards Forward T$ / Yen S 0$ / Yen PTYen / P0$T Obscure remark: Neglected $-Yen basis 1000’s on self-consistent derivatives contracts across all asset classes and regions Typical pricing accuracy ~ 10 4 (comparable to best LEP physics accuracies) 7 Market Risk (i.e. analogy with odds-based Poker) Statistical fluctuations in Spot FX over differentially small time period Ito process for Spot FX: dt 0 dS t$ / Yen / S t$ / Yen (r $ r Yen )dt $ / Yen dWt $ / Yen dWt $ / Yen (0,1) dt (0,1) NormallyDi stributed r $ , r Yen , $ / Yen Cons tan ts( Simplicity ) dt dt dt 3 / 2 Connection to IPPP School: (0,1) introduces random event generation together with a conserved probability current. 8 Delta Hedging and Option Pricing (i.e. analogy with odds-based Poker) FXOption(t , St$ / Yen ;T , K $ / Yen ; r $ , r Yen , $ / Yen ) t T DeltaHedge dFXOptionPortfolio $ / Yen FXOption FXOption St St$ / Yen d ( DeltaHedge dFXOptionPortfolio ) FXOption 1 2 FXOption $ / Yen 2 dt dSt 2 $ / Yen t 2 St O (dt 3 / 2 ) $ / Yen ~ dW t Statistical Fluctuations have cancelled! 9 Remember the Heat (Probability) Kernal (with slightly transformed variables) xt ln S t$ / Yen ; xT ln ST$ / Yen ; t T Kolmogorov( Backward ) Equation : 1 $ / Yen 2 2 1 $ / Yen 2 $ Yen ( r r ) Kernal 0 2 2 xt xt t 2 Fok ker Planck ( KolmogorovForward ) Equation : 1 $ / Yen 2 2 1 $ / Yen 2 $ Yen (r r ) Kernal 0 2 T 2 2 x xT T Kernal (t T , xt ; T , xT ; r $ , r Yen , $ / Yen ) ( xt xT ) Kernal (t , xt ; T , xT ; r $ , r Yen , $ / Yen ) 0 Kernal (t , xt ; T , xT ; r $ , r Yen , $ / Yen ) 0 Kernal (t , xt ; T , xT ; r $ , r Yen , $ / Yen ) 1 $ / Yen 2 $ Yen ik x x ( r r )( T t ) t T dk 2 exp 2 1 2 $ / Yen 2 k ( T t ) 2 10 Black-Scholes-Merton (1973) European Vanilla Option Pricing Equation = Heat (Probability) Equation: xt ln S t$ / Yen ; xT ln ST$ / Yen FXCallOption(t , S t$ / Yen ;T , K $ / Yen ; r $ , r Yen , $ / Yen ) Martingale FXCallOption(t , S t$ / Yen ;T , K $ / Yen ; r $ , r Yen , $ / Yen ) Martingale FXCallOption(T , ST$ / Yen ;T , K $ / Yen ; r $ , r Yen , $ / Yen ) Max ( ST$ / Yen K $ / Yen ,0) FXCallOption(t , S t$ / Yen 0;T , K $ / Yen ; r $ , r Yen , $ / Yen ) 0 FXCallOption(t , S t$ / Yen ;T , K $ / Yen ; r $ , r Yen , $ / Yen ) S t$ / Yen BlackScholesMertonCall (t , S t$ / Yen ;T , K $ / Yen ; r $ , r Yen , $ / Yen ) FXCallOption(t , S t$ / Yen ;T , K $ / Yen ; r $ , r Yen , $ / Yen ) dxT Kernal (t , xt ;T , xT ; r $ , r Yen , $ / Yen ) Max ( ST$ / Yen K $ / Yen ,0) Vanilla Options with European Exercize 11 Derivatives are Traded on Wall Street, City of 13 London, Hong Kong, Tokyo ~ $10 Notional Asset Classes in all G11 Currencies Foreign Exchange Interest Rates (e.g. Swaps) Commodities Equities Credit Derivatives Mortgage-backed Asset Backed (e.g. Automobiles) All Asset Classes in Emerging Markets and Local Currencies Derivative Classes Exchange-traded Vanilla Over-the-counter (OTC) Vanilla OTC Vanilla Exotics (e.g. Barriers) OTC Exotics Exercize Style American Bermudan European Asian 12 Why does probability mathematics have explanatory and predictive power? Financial Contracts and Derivatives Agreements (i.e. analogy with odds-based Poker) Financial Derivatives Markets Financial contracts are (very) binding agreements Industry-wide agreements: No-arbitrage self-consistency: certain new contracts are super-positions of existing contracts Industry-wide agreement to use probability to drive a large part of (vanilla and exotic) derivatives pricing Black Scholes pricing of vanilla European options Price Risk (Greeks) Industry-wide agreement to quote “Implied Volatility” severely constrains possible methods of option pricing (Credit-based) Implied Volatility Skew: human behaviour generates independent markets (like poker) 13 What special skills do physicists have for Financial Derivatives? Partial differential equations Very complicated boundary conditions specified to absorb proprietary and customer risk Analytic (especially Interest Rates) Lattice methods Monte-Carlo methods Choice of Monte-Carlo vs. lattice integration techniques is governed by the usual criteria for PDEs Remember: ~ $1013 Notional 14 Implied Volatility (i.e. analogy with human skill in Poker): European exercize VanillaFXC allOption (t , St$ / Yen ;T , K $ / Yen ; r $ , r Yen , $ / Yen ) Max ( ST$ / Yen K $ / Yen ,0) VanillaFXP utOption(t , St$ / Yen ;T , K $ / Yen ; r $ , r Yen , $ / Yen ) Max ( K $ / Yen ST$ / Yen ,0) BlackScholesMerton(t , S t$ / Yen ;T , K $ / Yen ; r $ , r Yen , $ / Yen ) is not quite the right price when volatility taken from historical data) is used Calls are less valuable than BlackScholesMerton formula Puts are more valuable than BlackScholesMerton formula Supply and demand Credit-based $ / Yen historical (i.e. 15 Implied Volatility (i.e. analogy with human skill in Poker) Price-quoting convention: $ / Yen BlackScholesMerton(t , S t$ / Yen ; T , K $ / Yen ; r $ , r Yen , Im plied ) with Implied Volatility $ / Yen Im plied (T , K $ / Yen ) This price-quoting convention Imposes a severe constraint (i.e. by agreement!) on options markets Restores mathematical applicability, explanatory power and predictive power Allows market-making experts to harvest $ 16 Conclusions Comparison of Applicability, Explanatory Power and Predictive Power of Mathematics of Probability Finance Industry Standard Practice: Hedging of Derivatives Risk Implied Volatility Physicists special skill: solve partial differential equations 17 Post Script: Financial Industry Standard Practice: Diversification of Risk for Cash Portfolios Portfolios of Cash Secutities Industry Agreement: trade securities by Markowitz’ “factor decomposition” Value at Risk (VaR): 1-day portfolio loss with 95% probability Delta Risk is “Diversified” (i.e. statistical fluctuations 0 but smaller, due to correlations) Barra, BlackRock, Wilshire, Morgan Stanley, ML, etc. build large covariance matrices Factors are used to buy securities: e.g. Industries Credit Quality Country Interest Rates Foreign Exchange Mortgage pre-payment Factor Covariance ~ 800 X 800 6 6 Asset Covariance ~ 10 X 10 Large covariance matrix calculations are possible only because Analytic result: Normal Distribution + Normal Distribution = Normal Distribution 18 Financial Industry Innovation: Diversification of Non-linear Derivatives risk VaR for portfolios of cash and derivatives Quadratic Approximation Gamma Risk Normal Distribution + Chi-Square Distribution = Analytic Integral Distribution VaR for portfolios of cash and derivatives: Full Revaluation Monte-Carlo event generation in very high dimensions 19 An Unsolved Monte Carlo Financial Mathematical Problem Finance Industry Innovation: Diversification of Nonlinear Derivatives Risk Beyond Quadratic Approximation Unsolved Monte Carlo problem in mathematical finance Monte Carlos in very high dimensions 20