Why Bankers Should Learn Convex Analysis

Transcription

1 Jim Zhu Western Michigan University Kalamazoo, Michigan, USA March 3, 2011

2 A tale of two financial economists Edward O. Thorp and Myron Scholes Influential works: Beat the Dealer(1962) and Beat the Market(1967) The Black-Scholes formula(1973).

3 A tale of two financial economists Edward O. Thorp and Myron Scholes Influential works: Beat the Dealer(1962) and Beat the Market(1967) The Black-Scholes formula(1973). Place the ideas were conceived: MIT

4 A tale of two financial economists Edward O. Thorp and Myron Scholes Influential works: Beat the Dealer(1962) and Beat the Market(1967) The Black-Scholes formula(1973). Place the ideas were conceived: MIT Investment practice: Managing partner of Princeton/Newport Partners and the President of Edward O. Thorp & Associates. Annualized return of 20% over 28.5 years Partner of Long Term Capital Management: essentially bankrupted in less than two years and almost causing a crisis.

5 A tale of two financial economists Edward O. Thorp and Myron Scholes Influential works: Beat the Dealer(1962) and Beat the Market(1967) The Black-Scholes formula(1973). Place the ideas were conceived: MIT Investment practice: Managing partner of Princeton/Newport Partners and the President of Edward O. Thorp & Associates. Annualized return of 20% over 28.5 years Partner of Long Term Capital Management: essentially bankrupted in less than two years and almost causing a crisis.

6 A tale of two financial economists Edward O. Thorp and Myron Scholes Recognition: One of the story in Poundstone s 2005 book Fortune s formula: The untold story... Noble Price in Econ 1997 and main stream financial economics. What is going on? Bankers don t know convex analysis!

7 A tale of two financial economists Edward O. Thorp and Myron Scholes Recognition: One of the story in Poundstone s 2005 book Fortune s formula: The untold story... Noble Price in Econ 1997 and main stream financial economics. What is going on? Bankers don t know convex analysis!

8 Convex analysis used to play a central role in economics and finance via (concave) utility functions. A new paradigm was emerged since the 1970 s after Black-Scholes introduced the replicating portfolio pricing method for option pricing, and

9 Convex analysis used to play a central role in economics and finance via (concave) utility functions. A new paradigm was emerged since the 1970 s after Black-Scholes introduced the replicating portfolio pricing method for option pricing, and Cox and Ross developed the risk neutral measure pricing formula.

10 Convex analysis used to play a central role in economics and finance via (concave) utility functions. A new paradigm was emerged since the 1970 s after Black-Scholes introduced the replicating portfolio pricing method for option pricing, and Cox and Ross developed the risk neutral measure pricing formula. This new paradigm marginalized many time tested empirical rules.

11 Convex analysis used to play a central role in economics and finance via (concave) utility functions. A new paradigm was emerged since the 1970 s after Black-Scholes introduced the replicating portfolio pricing method for option pricing, and Cox and Ross developed the risk neutral measure pricing formula. This new paradigm marginalized many time tested empirical rules. It brought about unprecedented prosperity in financial industry yet also led to the 2008 crisis.

12 Convex analysis used to play a central role in economics and finance via (concave) utility functions. A new paradigm was emerged since the 1970 s after Black-Scholes introduced the replicating portfolio pricing method for option pricing, and Cox and Ross developed the risk neutral measure pricing formula. This new paradigm marginalized many time tested empirical rules. It brought about unprecedented prosperity in financial industry yet also led to the 2008 crisis.

13 We will show that the new paradigm is a special case of the traditional utility maximization and its dual. Overlooking sensitivity analysis in the new paradigm is one of the main problem.

14 We will show that the new paradigm is a special case of the traditional utility maximization and its dual. Overlooking sensitivity analysis in the new paradigm is one of the main problem. The recent financial crisis is a wake up call that it is time again for bankers to learn convex analysis.

15 We will show that the new paradigm is a special case of the traditional utility maximization and its dual. Overlooking sensitivity analysis in the new paradigm is one of the main problem. The recent financial crisis is a wake up call that it is time again for bankers to learn convex analysis.

16 Outline A tale of two financial economists The talk is divided into two parts. In the first part we discuss A discrete model for financial markets. Arbitrage and martingale (risk neutral) measure. Fundamental theorem of asset pricing. Utility functions and risk measures. Markowitz portfolio theory

17 Outline A tale of two financial economists The second part focuses on the financial derivatives. The new paradigm of financial derivative pricing. A Convex Analysis Perspective. Sensitivity and Financial Crisis. Alternative methods and an illustrative example using real historical market data.

18 Uncertainty A tale of two financial economists Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration Uncertainty is ubiquitous in the financial world Stock price is unpredictable. Financial derivatives can bring about prosperity and disaster. Bond is considered safe but that is when interest rate is stable. Cash is better if only there is no inflation. To model financial markets one has to model uncertainty.

19 Uncertainty A tale of two financial economists Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration Uncertainty is ubiquitous in the financial world Stock price is unpredictable. Financial derivatives can bring about prosperity and disaster. Bond is considered safe but that is when interest rate is stable. Cash is better if only there is no inflation. To model financial markets one has to model uncertainty.

20 Model Uncertainty Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration For problem involving only one decision such as analyzing a portfolio we need random variables. For problem involving multiple decisions such as trading we need stochastic process

21 Model Uncertainty Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration For problem involving only one decision such as analyzing a portfolio we need random variables. For problem involving multiple decisions such as trading we need stochastic process The process of information becomes available also need to be modeled.

22 Model Uncertainty Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration For problem involving only one decision such as analyzing a portfolio we need random variables. For problem involving multiple decisions such as trading we need stochastic process The process of information becomes available also need to be modeled.

23 The game of tossing a coin Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration Bet on flipping a fair coin. Head the house will double your bet. Tail you lose your bet to the house.

24 A random variable Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration Suppose we play the game only once and bet 1. Denote the outcome of the game by X. Then X is a random variable taking only 1 or 1 as its value and P(X = 1) = P(X = 1) = 1/2.

25 A discrete stochastic process Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration Play the game i times and always bet 1. Denote the outcome of the ith game by X i.

26 A discrete stochastic process Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration Play the game i times and always bet 1. Denote the outcome of the ith game by X i. Then X i is a random variable and P(X i = 1) = P(X i = 1) = 1/2.

27 A discrete stochastic process Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration Play the game i times and always bet 1. Denote the outcome of the ith game by X i. Then X i is a random variable and P(X i = 1) = P(X i = 1) = 1/2. If we start with an initial endowment of w 0 then our total wealth after the ith game is w i = w 0 +X X i. (1)

28 A discrete stochastic process Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration Play the game i times and always bet 1. Denote the outcome of the ith game by X i. Then X i is a random variable and P(X i = 1) = P(X i = 1) = 1/2. If we start with an initial endowment of w 0 then our total wealth after the ith game is w i = w 0 +X X i. (1) Now (w i ) n i=1 is an example of a discrete stochastic process.

29 A discrete stochastic process Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration Play the game i times and always bet 1. Denote the outcome of the ith game by X i. Then X i is a random variable and P(X i = 1) = P(X i = 1) = 1/2. If we start with an initial endowment of w 0 then our total wealth after the ith game is w i = w 0 +X X i. (1) Now (w i ) n i=1 is an example of a discrete stochastic process.

30 Information A tale of two financial economists Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration Will knowing X 1,...,X i, help us to play the (i +1)th game? The answer should be NO but how do we clearly describe this conclusion?

31 Information A tale of two financial economists Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration Will knowing X 1,...,X i, help us to play the (i +1)th game? The answer should be NO but how do we clearly describe this conclusion? Let us look at the game with n = 3 to get some feeling. We use H to represent a head and T, tail. The information we can get at each stage can be illustrated with the following binary tree.

32 Information A tale of two financial economists Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration Will knowing X 1,...,X i, help us to play the (i +1)th game? The answer should be NO but how do we clearly describe this conclusion? Let us look at the game with n = 3 to get some feeling. We use H to represent a head and T, tail. The information we can get at each stage can be illustrated with the following binary tree.

33 Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration F 0 F 1 F 2 F 3 HHH HH HHT H HTH HT HTT {Ω} THH TH THT T TTH TT TTT

34 Filtration for 3 coin tosses Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration All the information are represented by F 3 = 2 Ω,Ω = {HHH,HHT,HTH,HTT,THH,THT,TTH,TTT}. Similarly, after 2 tosses F 2 = 2 {HH,HT,TH,TT}, where {HH,HT,TH,TT} = {{HHH,HHT},{HTH,HTT},{THH,THT},{TTH,TTT}}. F 2 has less information than F 3. Similarly, F 1 = 2 {H,T}, where {H,T} = {{HHH,HHT,HTH,HTT},{THH,THT,TTH,TTT}}. F 0 = {,{Ω}}.

35 Filtration for 3 coin tosses Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration The sequence is a filtration for (w i ) 3 i=0. F : F 0 F 1 F 2 F 3 For each i, F i is a set algebra, i.e., its elements as sets are closed under union, intersection and compliment.

36 General filtration Uncertainty Model Uncertainty An example A random variable A discrete stochastic process Information Example of Filtration Definition of filtration Let Ω be a sample space (representing possible states of a chance event). A sequence of algebra (σ-algebra when Ω is infinite) F : F i,i = 0,1,...,n satisfying is called a filtration. F 0 F 1 F 2... F n (2) If F 0 = {Ω} and F n = Ω then F is called an information structure.

37 Random variable Random variable Information system Market Portfolio Trading strategy All possible economic states is represented by a finite set Ω. Probability of each state is described by a probability measure P on 2 Ω. Let RV(Ω) be the space of all random variables on Ω, with inner product ξ,η = E[ξη] = = ΩξηdP ξ(ω)η(ω)p(ω). ω Ω 0 < ξ RV(Ω) means ξ(ω) 0 for all ω Ω and at least one of the inequality is strict.

38 Information system Random variable Information system Market Portfolio Trading strategy Suppose that actions can only take place at t = 0,1,2,... Use F = {F t t = 0,1,...} to represent an information system of subsets of Ω, that is, σ({ω}) = F 0 F 1... F t... and t=0 F t = σ(ω). Here, algebra F t, represents available information at time t. Implied in the definition is that we never loss any information and our knowledge increases with time t. If action is finite t = 0,1,...,T, we assume F T = σ(ω). The triple (Ω, F, P) models the gradually available information.

39 Market A tale of two financial economists Random variable Information system Market Portfolio Trading strategy Let A = {a 0,a 1,...,a M } be M +1 assets. a 0 is reserved for the risk free assets. The prices of these assets are represented by vector stochastic process S := {S t } t=0,1,..., where S t := (S 0 t,s 1 t,...,s M t ) is the discounted price vector of the M +1 assets at time t. Using the discounted price, we have St 0 = 1 for all t. Assume S t is F t -measurable, i.e. determined up to the available information. We say such an S is F-adapted. A described above is a financial market model.

40 Portfolio A tale of two financial economists Random variable Information system Market Portfolio Trading strategy Portfolio A portfolio Θ t on the time interval [t 1,t) is a F t 1 measurable random vector Θ t = (Θ 0 t,θ1 t,...,θm t ) where Θm t indicates the weight of asset a m in the portfolio. A portfolio Θ t is always purchased at t 1 and liquidated at t. The acquisition price is Θ t S t 1 and the liquidation price is Θ t S t.

41 Trading strategy Random variable Information system Market Portfolio Trading strategy Trading strategy A trading strategy is a F-predictable process of portfolios Θ = (Θ 1,Θ 2,...), where Θ t denotes the portfolio in the time interval [t 1,t). A trading strategy is self-financing if at any t Θ t S t = Θ t+1 S t, t = 1,2,... We use T(A) to denote all the self-financing trading strategies for market A. Θ = (Θ 1,Θ 2,...) is F-predictable means that Θ t is F t 1 measurable.

42 Gain A tale of two financial economists Random variable Information system Market Portfolio Trading strategy For a trading strategy Θ, the initial wealth is w 0 = Θ 1 S 0. (3) The net wealth at time t = T is T w T = Θ t (S t S t 1 )+w 0. (4) Random variable G T (Θ) = is the net gain. t=1 T Θ t (S t S t 1 ) = w T w 0 (5) t=1

43 Arbitrage A tale of two financial economists Arbitrage Fair game and martingale Examples of martingales Martingale for a financial market Martingale characterization of no arbitrage Arbitrage A self-financing trading strategy is called an arbitrage if G t (Θ) 0 for all t and at least one of them is strictly positive. Intuitively, an arbitrage trading strategy is a risk free way of making money. We note that we may always assume G T (Θ) > 0.

44 No Arbitrage Principle Arbitrage Fair game and martingale Examples of martingales Martingale for a financial market Martingale characterization of no arbitrage No Arbitrage Principle There is no arbitrage in a competitive financial market.

45 Fair game and martingale Arbitrage Fair game and martingale Examples of martingales Martingale for a financial market Martingale characterization of no arbitrage Toss a fair coin is a fair game in the sense that no player has an advantage. In other words, restricted to information at (i 1)th game, the expectation of w i and w i 1 are the same. Mathematically, E P [w i F i 1 ] = w i 1. (6) A F-adapted stochastic process satisfying (6) is called a F-martingale. We will omit P and/or F if it is clear in the context.

46 Examples A tale of two financial economists Arbitrage Fair game and martingale Examples of martingales Martingale for a financial market Martingale characterization of no arbitrage 1 Let X i be independent with E[X i ] = 0 for all i. Then, S 0 = 0, S i = X X i defines a martingale. 2 Let X i be independent with E[X i ] = 0 and Var[X i ] = σ 2 for all i. Then, M 0 = 0, M i = S 2 i iσ 2 gives a martingale. 3 Let X i be independent random variables with E[X i ] = 1 for all i. Then, M 0 = 0, M i = X 1... X i gives a martingale with respect to F i.

47 Martingale for a financial market Arbitrage Fair game and martingale Examples of martingales Martingale for a financial market Martingale characterization of no arbitrage 1 Let A be a financial market. 2 We say that a probability measure P is a martingale of A if P(ω) > 0 for all ω Ω and all the price process,m = 0,1,...,M are martingales with respect to P. S m t 3 We use M(A) to denote the set of all martingale measures of A.

48 Arbitrage Fair game and martingale Examples of martingales Martingale for a financial market Martingale characterization of no arbitrage Let A be a financial market model with finite period T. Then the following are equivalent (i) there are no arbitrage trading strategies; (ii) M(A).

49 Proof (ii) (i) A tale of two financial economists Arbitrage Fair game and martingale Examples of martingales Martingale for a financial market Martingale characterization of no arbitrage Let Q M(A). If Θ T (A) is an arbitrage, then, for some t, G t (Θ) > 0 and consequently E Q (G t (Θ)) > 0, a contradiction.

50 Proof (i) (ii) A tale of two financial economists Arbitrage Fair game and martingale Examples of martingales Martingale for a financial market Martingale characterization of no arbitrage Observe that, G T (T (A)) intrv(ω) + =. Since G T (T (A)) is a subspace, by the convex set separation theorem G T (T (A)) contains a vector q with all components positive. We can scale q to a probability measure Q. Then it is easy to check Q M(A).

51 Remark A tale of two financial economists Arbitrage Fair game and martingale Examples of martingales Martingale for a financial market Martingale characterization of no arbitrage (i) No arbitrage principle does not say one cannot make more than the risk free rate. (ii) It says to do that one has to take risk. (iii) Martingale probability measure is not the same as the real probability of economic events.

52 A tale of two financial economists Utility functions Risk Measure To discuss beating the risk free rate by taking risks we need measures for risk and reward. The preference of different market participants are different. Common way of modeling the preference are (i) Utility functions; (ii) Risk measures; and (iii) The combination of the two.

53 Utility functions Utility functions Risk Measure Experience tells us that mathematical expectation is often not what people use to compare payoffs with uncertainty. Lottery and insurance are typical examples. Economists explain this using utility functions: people are usually comparing the expected utility. Utility function is increasing reflecting the more the better and Concave: the marginal utility decreases as the quantity increases. Concavity is also interpreted as the tendency of risk aversion: the more we have the less we are willing to risk.

54 Examples of Utility functions Utility functions Risk Measure Several frequently used utility functions are Log utility u(x) = ln(x) goes back to Bernoulli and the St. Petersburg wager problem. Power utility functions (x 1 γ 1)/(1 γ),γ > 0 and Exponential utility functions e αx,α > 0. We note that ln(x) = lim γ 1 (x 1 γ 1)/(1 γ).

55 Common properties Utility functions Risk Measure The following is a collection of conditions that are often imposed in financial models: (u1) (Risk aversion) u is strictly concave, (u2) (Profit seeking) u is strictly increasing and lim t + u(t) = +, (u3) (Bankruptcy forbidden) For any t < 0, u(t) = and lim t 0+ u(t) =, (u4) (Standardized) u(1) = 0 and u is differentiable at t = 1.

56 Risk Measure A tale of two financial economists Utility functions Risk Measure An alternative to maximizing utility functions is to minimize risks. Pioneering work: Markowitz s portfolio theory measures the risks using the variation. Modeling the risk control of market makers of exchanges, Artzner, Delbaen, Eber and Heath introduced the influential concept of coherent risk measure.

57 Common properties Utility functions Risk Measure Here are some common properties of risk measures (r1) (Convexity) for X 1,X 2 RV(Ω) and λ [0,1], ρ(λx 1 +(1 λ)x 2 ) λρ(x 1 )+(1 λ)ρ(x 2 ), diversification reduces the risk. (r2) (Monotone) X 1 X 2 RV(Ω) + implies ρ(x 1 ) ρ(x 2 ). a dominate random variable has a smaller risk. (r3) (Translation property) ρ(y +c 1) = ρ(y) c for any Y RV(Ω) and c R, one may measure the risk by the minimum amount of additional capital to ensure not to bankrupt. (r4) (Standardized) ρ(0) = 0.

58 Convexity and diversity Utility functions Risk Measure Convexity is essential in characterizing the preference. Diverse in choosing particular preference is intrinsic.

59 Portfolio problem Portfolio problem Dual problem Markowitz bullet Use Ŝ and ˆΘ to denote risky part of the price process and the portfolio. Giving the expected payoff r 0 and an initial wealth w 0, Markowitz s problem is minimize Var(ˆΘ Ŝ 1 ) subject to E[ˆΘ Ŝ 1 ] = r 0 (7) ˆΘ Ŝ0 = w 0.

60 Equivalent form A tale of two financial economists Portfolio problem Dual problem Markowitz bullet The portfolio problem is equivalent to the entropy maximization problem minimize f(x) := 1 2 x Σx subject to Ax = b. (8) Here x = ˆΘ, Σ = (E[(S i 1 E[Si 1 ])(Sj 1 E[Sj 1 ])]) i,j=1,...,m and [ E[ Ŝ A = 1 ] Ŝ 0 ] [ r0, and b = w 0 ].

61 Dual problem A tale of two financial economists Portfolio problem Dual problem Markowitz bullet Assuming Σ positive definite the dual problem is maximize b y 1 2 y AΣ 1 A y. (9)

62 Dual problem A tale of two financial economists Portfolio problem Dual problem Markowitz bullet Solving the dual problem we derive the following relationship γr0 2 σ(r 0,w 0 ) = 2βr 0w 0 +αw0 2 αγ β 2, (10) where α = E[Ŝ1]Σ 1 E[Ŝ1], β = E[Ŝ1]Σ 1 Ŝ 0 and γ = Ŝ0Σ 1 Ŝ 0. The corresponding minimum risk portfolio is Θ(r 0,w 0 ) = E[Ŝ1](γr 0 βw 0 )+Ŝ0(αw 0 βr 0 ) αγ β 2 Σ 1 (11)

63 Markowitz bullet Portfolio problem Dual problem Markowitz bullet Draw this function on the σµ-plan we get u s which is commonly known as a Markowitz bullet for its shape.

64 Markowitz portfolio theory became popular largely due to its simple linear -quadratic problem with explicit solutions. Well known extensions and applications include Capital Asset Pricing Model and Sharpe ratio for mutual fund performances. Are there other risk - utility function pairings can lead to convenient explicit solutions?