This action might not be possible to undo. Are you sure you want to continue?

Welcome to Scribd! Start your free trial and access books, documents and more.Find out more

Tehranchi

Contents

Chapter 1. One-period models 1. The set-up 2. Arbitrage and the ﬁrst fundamental theorem of asset pricing 3. Contingent claims and no-arbitrage pricing 4. Pricing and hedging in an incomplete market by minimizing hedging error 5. Change of num´raire and equivalent martingale measures e 6. Discounted prices Chapter 2. Multi-period models 1. The set-up 2. The ﬁrst fundamental theorem of asset pricing 3. European and American contingent claims 4. Locally equivalent measures Chapter 3. Brownian motion and stochastic calculus 1. Brownian motion 2. Itˆ stochastic integration o 3. Itˆ’s formula o 4. Girsanov’s theorem 5. Martingale representation theorems Chapter 4. Black–Scholes model and generalizations 1. The set-up 2. Admissible strategies 3. Arbitrage and equivalent martingale measures 4. Contingent claims and market completeness 5. The set-up revisited 6. Markovian markets 7. The Black–Scholes model, PDE, and formula 8. Local and stochastic volatility models Chapter 5. Interest rate models 1. Bond prices and interest rates 2. Short rate models 3. Factor models 4. The Heath–Jarrow–Morton framework Chapter 6. Crashcourse on probability theory 1. Measures

3

7 7 8 12 16 18 20 23 23 26 29 35 37 37 40 43 47 48 51 51 52 54 55 56 58 60 63 67 67 70 73 75 81 81

4.2. 6. 8. independence Probability inequalities Characteristic functions Fundamental probability results 81 82 84 85 85 86 86 89 4 . Index Random variables Expectations and variances Special distributions Conditional probability and expectation. 7. 5. 3.

with an emphasis on the pricing and hedging of contingent claims within such models. though a crashcourse on probability theory is given in an appendix. An attempt is made to keep this course self-contained.tehranchi@statslab. expected value. Our starting point is the self-evident observation: The future is uncertain. conditional probability/expectation. including knowing the deﬁnition and key properties of the following concepts: random variable. variance. Therefore.uk. anyone with even a passing acquaintance with ﬁnance knows that it is impossible to predict with absolute certainty how the the price of an asset will ﬂuctuate. Please send all comments (including small typos and major blunders) to the author at m. Indeed. but you should be familiar with the basics of the theory. the proper language to formulate the models that we will study is the language of probability theory. Gaussian (normal) distribution.cam. Familarity with measure theoretical probability is helpful. independence.ac. etc. 5 .This course is about models of ﬁnancial markets.

.

not random. . . . 1} for single period models. the index set T will be one of the three sets • {0. Often. In this course. We will denote the collection of prices by the d + 1-dimensional column vector ¯ St = (St0 . this set-up captures most of the essential features of the general discrete-time model. .CHAPTER 1 One-period models The models we will encounter will be of form (St0 . bond. we take asset 0 0 0 to be ordinary cash. rather than simply d assets. 1. . . and • R+ = [0. S0 are constants – that is. . and large buy or sell orders tend to move the market prices. in the real world. . As we shall see later. 1} by the random variable Sti . 1.) at time t ∈ T. As simple as it seems. . . Of course. ¯ ¯ Our ﬁrst assumption on the stochastic process S = (St )t∈{0. . . We model the price of the assets. . 1}. Std )t∈T where Sti will model the price of a ﬁnancial asset (stock. ∞) when time is continuous.1} is that the time 0 prices of the assets are known at time 0. We also assume that there is no bid/ask spread – the buying price is the same as the selling price. It is. at time t ∈ {0. Then the market consists of one distinguished asset and d other assets. The reason is that in the sequel. 7 . We assume now. . therefore. 2. usually asset 0. Std ). we have the following assumption (which will be generalized in the multi-period models): 0 d Assumption. the overbar in the ¯ notation St will become apparent as this chapter proceeds. appropriate to devote a signiﬁcant portion of the course to this important special case.} when time is discrete. . we can model the cost of borrowing cash by 0 0 introducing an interest rate r ≥ 0 so that S0 = 1 and S1 = 1 + r. but we ignore these issues. d. The market consists of d + 1 assets. etc. one of the assets. The set-up In this section we describe our ﬁrst model. and we will leave it as a standing assumption throughout the course. plays a distinguished role. labelled 0. Also. . for a total of d + 1. investors face constraints when short-selling assets. Remark. so that S0 = S1 = 1. that an investor in this market can buy or sell any number (even fractional) of shares of each asset without aﬀecting the price. The random variables S0 . but not always. • Z+ = {0. much of the ﬁnancial aspects of this course already appear in oneperiod models where T = {0. . . . 1. You might wonder at this stage why we have chosen the notation such that there are d + 1 assets. Formally. Or.

. we introduce an investor. Remark. for i = 0. 2. π 1 = 3. S 1 ) . 4) qq qq qq 1/2 qq# (2. but that always has non-negative (and ¯ sometimes positive) value at time 1. our ﬁrst challenge is to ﬁnd out how to invest optimally. then the investor has borrowed |π i | units of asset i to pay back at time 1. . Again notice that we do not demand that the π i are integers. Since there is only one period. but the time-1 wealth is strictly positive in 8 . We will allow π i to be either positive. or zero with the interpretation that if π i > 0 the investor is ‘long’ asset i and if π i < 0 the investor is ‘short’ the asset. π Remark. . . .Now given this market model. negative. 1/2 www w ww ww (4. . . In particular. and the self-ﬁnancing condition says that changes in his wealth between time 0 and 1 are due only to changes in the asset prices – he does not consume or have any other source of income. 1}. We assume that the investor’s wealth and porfolio is connected by the following relationships: X0 (¯ ) = π · S0 π ¯ ¯ X1 (¯ ) = π · S1 π ¯ ¯ where the a · b = d i=0 the budget constraint the self-ﬁnancing condition ai bi is the usual Euclidean inner (or dot) product in Rd+1 . We use the following notation: • π i ∈ R denotes the (non-random) number of shares of asset i. etc. the investor only chooses his investment portfolio once. d. Arbitrage and the ﬁrst fundamental theorem of asset pricing Now that we have our market model and we’ve introduced an investor into this market. and let π = (π 0 . P(S1 = 7) = 1/2. Consider a market with two assets with prices given by (S 0 . 3) 0 1 (The above diagram should be read S0 = 3. at time 0.) Consider the portfolio π 0 = −4. . 7) (3. if π i < 0. π d ) ¯ denote the vector of portfolio weights. Example. It costs zero to buy at time 0. The budget constraint simply says that the investor’s initial wealth is the time-0 total value of his holdings. An investor’s dream is to ﬁnd a portfolio π that costs nothing to buy at time 0. • Xt (¯ ) denotes the investor’s wealth at time t ∈ {0.

s. Before we begin.s. ¯ ¯ ¯ ¯ ¯ ¯ ¯ ¯ In this section we ﬁnd a mathematical classiﬁcation of such market models. But for the sake of building realistic models. Proof. A pricing kernel (or state price density) for a market model S is a positive random variable ρ such that i i i E(ρ|S1 |) < ∞. Letting Xt = π ·St . Notice that a market has no arbitrage if and only if π · S0 ≤ 0 and π · S1 ≥ 0 a. . An arbitrage is a portfolio π ∈ Rd+1 such that ¯ • X0 (¯ ) ≤ 0. we have ρX1 ≥ 0 and hence 0 ≥ X0 = E(ρX1 ) ≥ 0 9 . Now we come to ﬁrst theorem of the course. Since ρ > 0.all states of the world: X 5 Ñ@ ÑÑ ÑÑ ÑÑ 0a aa aa 1/2 aa 1/2 1 The example above leads us to our ﬁrst deﬁnition: Definition. and one of the most important theorems in ﬁnancial mathematics. 2) is also an arbitrage in the ¯ example above. and E(ρS1 ) = S0 . and π • either X0 < 0 or P (X1 > 0) > 0 (or both). A market model has no arbitrage if and only if there exists a pricing kernel. . It is no surprise that it is often called the ﬁrst fundamental theorem of asset pricing. for all i = 0. First we prove that if there exists a pricing kernel. ¯ Definition. . ⇒ π · S0 = 0 and π · S1 = 0 a. Theorem (First fundamental theorem of asset pricing). The following proof is from Doug Kennedy’s lecture notes. then there is no arbitrage. π • X1 (¯ ) ≥ 0 almost surely. we need some vocabulary. At this point. Now suppose X0 ≤ 0 and X1 ≥ 0 almost surely. Markets with many arbitrage opportunities would be nice–we all would be a lot richer. we have by the deﬁnition of pricing kernel and the linearity of expectations ¯ ¯ ¯ E(ρX1 ) = π · E(ρS1 ) ¯ = π · S0 ¯ ¯ = X0 . d. you should check that the portfolio π = (−3. . we usually assume that markets are free of arbitrages.

s and E(Y ) = 0 then Y = 0 a. By the separating hyperplane theorem (stated and proved below) there exists a vector π such that ¯ ¯ π · (y − S0 ) ≥ 0 ¯ for all y ∈ C. (Recall the pigeonhole principle: if Y ≥ 0 a.s.) Again. We also see that ρX1 = 0 a. ¯ If S0 is contained in C. that S0 is not an element of C. we have X0 ≤ E(ρ0 X1 ) → 0 as ↓ 0. We must prove there exists a pricing kernel. In this optional subsection. for the sake of ﬁnding a contra¯ diction. we must conclude that X0 = 0 and X1 = 0 a.s. we conclude that X1 = 0 a.. we would be done. where the inequality is strict for at least one y ∈ C.1. Theorem (Separating/supporting hyperplane theorem). since ρ0 deﬁned by ρ0 = e−|S1 | certainly satisﬁes ρ0 > 0 a. suppose. Consider the set C ⊆ Rd+1 deﬁned by ¯ ¯ C = {E(ρS1 ) : ρ > 0 a.s. There are many versions of this theorem. and hence there is no arbitrage. Then there exists a λ ∈ RN such that λ · (y − x) ≥ 0 for all y ∈ C. it is easy to see that C is convex. Letting π · St = Xt the above inequality translates to ¯ ¯ X0 ≤ E(ρX1 ) for all feasible ρ. Now suppose that there is no arbitrage. First. Also. 2. Next. we must conclude X1 ≥ 0 a. 10 ¯ . a version of the separating hyperplane theorem is stated and proved. The set C is not empty. Let C ⊂ RN be convex and x ∈ RN not contained in C. and such that the above inequality is strict for at least one y ∈ C.s.. But since X0 is ﬁnite. since ρ = 0 a. by letting ρ = ρ0 with > 0 in the above inequality. but we give only the version needed to prove the ﬁrst fundamental theorem of asset pricing in one-period.s.s. and ¯ E(ρ0 |S1 |) < ∞. A separating hyperplane theorem*.so that X0 = 0 = E(ρX1 ). our desired contradiction. let ρ = ρ0 (M 1{X1 <0} + 1) so that X0 ≤ M E(ρ0 X1 1{X1 <0} ) + E(ρ0 X1 ).s. and E(ρ|S1 |) < ∞}. and we have just shown X0 ≤ 0 and X1 ≥ 0 a. But the separating hyperplane theorem says that there exists at least one ρ such that X0 < E(ρX1 ). with strict inequality for at least one ρ.s.s. Since we have assumed that there is no arbitrage. Notice that the right-hand side of the above inequality tends to −∞ as M ↑ ∞ if the event {X1 < 0} has positive probability. So. so we can conclude X0 ≤ 0.

¯ Case 2: Supporting hyperplane. Then ∗ 0 = ≤ = = |y ∗ − x|2 − λ2 |(1 − θ)y ∗ + θy − x|2 − λ2 |θ(y − y ∗ ) + λ|2 − λ2 θ2 |y − y ∗ |2 + 2θ(y − y0 ) · λ. The point x is in the closure C of C. n → ∞. where we have used the convexity of C to assert that 2 (ym + yn ) ∈ C and 1 hence 2 (ym + yn ) − x ≥ d. Deﬁne a subspace N of R by S = span{y − x : y ∈ C}. Fix a point y ∈ C and 0 < θ < 1.Case 1. and ¯ hence converges to some point y ∗ ∈ C as claimed. 11 . Now. we conclude (y − y ∗ ) · λ ≥ 0. Separating hyperplane ¯ Proof. We have established that the sequence (yn )n is Cauchy. The point x is not in the closure C of C. we establish the existence of y ∗ : Let d = inf y∈C |y − x| > 0 and let yn be a sequence in C such that |yn − x| → d. Hence (y − x) · λ = (y − y ∗ ) · λ + |λ|2 > 0 as desired. By ﬁrst dividing by θ and then taking the limit as θ ↓ 0 in the above inequality. ¯ y ∈C let λ = y ∗ − x. Case 1: Separating hyperplane. Let ¯ be the point in C closest to x. Applying the parallelogram law |a + b|2 + |a − b|2 = 2|a|2 + 2|b|2 we have 1 |ym − yn |2 = 2|ym − x|2 + 2|yn − x|2 − 4 2 (ym + yn ) − x 2 ≤ 2|ym − x|2 + 2|yn − x|2 − 4d2 → 0 1 as m. and note that the point (1 − θ)y ∗ + θy is ¯ ¯ in C since C is convex. Assuming for the moment the existence of this point.

In other words. Then λ · (y − x) = lim λn · (y − x) n = lim λn · (y − xn ) + λn · (xn − x) n ≥ 0.¯ Let (xn )n be a sequence in the complement of C such that xn − x is in S and xn → x. called the strike of the option. with limit λ ∈ S. then the owner of the option can buy the stock for the price K from the counterparty and immediately sell the stock for the price S1 to the market. but not the obligation. the payout of a contingent claim is modelled by a random variable on our probability space (Ω. Let y ∗ − xn . F. A call option gives the owner of the option the right. for us. Hence. 3. we can conclude that there exists y ∈ C such that λ·(y −x) = 0. the payout of the 12 . there exists a convergent subsequence. On the other hand. Finally. realizing a proﬁt of S1 − K. etc. One of the major triumphs of modern ﬁnance is the no-arbitrage pricing of contingent claims. we can ﬁnd the point yn ∈ C closest to xn . As Case 2. then the option is worthless to the owner since there is no point paying a price above the market price for the underlying stock. Let S1 denote the price of the stock at time 1. still denoted by (λn )n . Example (Call option). as desired. P). if K < S1 . λn = n ∗ |yn − xn | Since (λn )n is bounded. There are two cases: If K ≥ S1 . Contingent claims and no-arbitrage pricing A contingent claim is any cash payment at time 1 where the size of the payment is contingent on the realized prices of other assets. since λ ∈ S and λ = 0. to buy a given stock at time 1 at some ﬁxed price K. Supporting hyperplane ∗ ¯ in case 1.

We now study what the no-arbitrage principle has to say about the pricing of contingent ¯ claims.call option is (S1 − K)+ . The ‘hockey-stick’ graph of the function g(x) = (x − K)+ is below. So. which is just the time-0 price of this replicating portfolio. ¯ ¯ ¯ ¯0 then there exists an arbitrage. d. The conclusion of the theorem is very important. . d. and suppose this market is free of arbitrage. (1) A contingent claim with payout ξ1 is attainable. The theorem says that a contingent claim has a unique no-arbitrage price if and only if its payout can be replicated by trading in the original assets 0. . We will show that if the augmented market (S. . . ¯ (2) There exists a unique initial price ξ0 such that the augmented market (S. Our ¯ goal is to ﬁnd the time-0 price ξ0 such that the augmented market with prices (S. (1) ⇒ (2) Suppose the claim is attainable. Let φi = −π i . and φd+1 = 1. 13 . 0} as usual. . the time-0 price of such a claim. We start oﬀ with a given market model of d + 1 assets with prices S. ξ) has no arbitrage. . φd+1 ) where the ﬁrst d + 1 entries correspond to the number of shares of the assets in the underlying market. by symmetry. Proof. (3) There is a constant ξ0 such that ξ0 = E(ρξ1 ) for all pricing kernels ρ such that E(ρ|ξ1 |) < ∞. . it’s enough to show that if ξ0 < π · S ¯ Consider a portfolio φ = (φ0 . . φd . then ξ0 = π · S0 . ξ) has no-arbitrage. Following proof is a bit redundant but may help with building intuition. Introduce a contingent claim with time-1 payout of ξ1 . where a+ = max{a. . Remark. Furthermore. . Definition. ξ) is still free of arbitrage. Given an arbitrage free market model ¯ the following are equivalent: S. . so that ξ1 = π · S1 for some portfolio ¯ ¯ ¯ π . and the d + 2-th entry is for the contingent claim. There is a useful class of contingent claims for which no-arbitrage give a unique price. ¯ ¯ This set of attainable claims can be characterized: Theorem (Characterization of attainable claims). so it’s worth reprasing it for emphasis. . A contingent claim with payout ξ1 is an attainable (or replicable) claim if and only if there exists a portfolio π ∈ Rd+1 such that ¯ ξ1 = π · S1 . for i = 0. can be found by computing the expected value of the payout times a pricing kernel.

the vector ξ1 is orthogonal to µ.s. ¯ Let ρ0 be a pricing kernel for S and consider the set K = {¯ · S1 : π ∈ Rd+1 } π ¯ ¯ of attainable claims. (1) ⇒ (3) If ξ1 is attainable. Y (ωn )) ∈ Rn+1 . that asset 1 is a stock. with P{ω} > 0 for all ω ∈ Ω. Now by assumption E(ρθ ξ1 ) = E(ρ0 ξ0 ) + θE(µξ1 ) = ξ0 . and hence ρθ is a pricing kernel as well. But µ ∈ K ⊥ was arbitrary. . Viewed as a vector subspace of Rn+1 of dimension at most d + 1. ) Pick a vector µ ∈ K ⊥ and let ρθ = ρ0 + θµ. Suppose that the sample space Ω has n + 1 elements. it has an orthogonal complement K ⊥ = {µ : E(µX1 ) = 0 for all X1 ∈ K} dimension at least n − d. ξ) has no arbitrage if and only if there exists a positive random variable ρ such that ¯ ¯ E(ρS1 ) = S0 and E(ρξ1 ) = ξ0 .1} . Recall that the payout of the option is C1 = (S1 − K)+ . . but the proof is left as an exercise on the example sheet. π ¯ Hence φ is an arbitrage. Suppose that this market is free of arbitrage. Ct )t∈{0. (This is where we have used the assumption that Ω is ﬁnite. We assume that asset 0 is a riskless bond with prices B0 = 1 and B1 = 1 + r. (Put-call parity formula) Suppose we start with a market with three assets with prices (Bt . St . (Of course. then ξ1 = π · S1 for some portfolio π . . 14 . K ⊥ is allowed to be the singleton {0}. so that ξ0 = π · S0 . (3) ⇒ (1)* This is the hard direction. Since ρ0 (ω) > 0 for all ω ∈ Ω.Then X0 (φ) = −¯ · S0 + ξ0 < 0 π ¯ and X1 (φ) = −¯ · S1 + ξ1 = 0 a. a call option on that stock with strike K. and that asset 2 is a .) Furthermore. one can choose a non-zero θ small enough that ρθ (ω) > 0 for all ω ∈ Ω. (The statement is true in full generality. the augmented market (S. Then ¯ ¯ ¯ E(ρξ1 ) = π · S0 ¯ ¯ for every pricing kernel ρ. Since θ = 0.) We can identify a random variable Y with the vector (Y (ω0 ). Example. . The second equation says that only there is a unique initial price ξ0 compatible with no-arbitrage if and only if E(ρξ1 ) = ξ0 all pricing kernels ρ. i i i i E(ρθ S1 ) = E(ρ0 S1 ) + θE(µS1 ) = S0 . ¯ The ﬁrst equation above says that ρ is a pricing kernel for the underlying market S. we conclude that ξ1 ∈ K ⊥⊥ = K as desired. ¯ ¯ ¯ (3) ⇔ (2) By the ﬁrst fundamental theorem of asset pricing.

so every random variable satisﬁes the integrability property. 15 . and E( |S1 |) < ∞. C). Since the market is complete. we see that the no-arbitrage assumption implies the existence of a pricing kernel ρ with E(ρ|ξ1 |) < ∞. and the put option is attainable in the market (B. and ξ1 be an arbitrary random variable. A market is incomplete otherwise. we single out the markets for which every claim is attainable: Definition.Now we introduce another claim. every random variable is attainable. the payout of a put option is P1 = (K − S1 )+ . Since the time-0 price of an attainable claim is just the price of the replicating portfolio. π 2 ) = 1+r . so we can let ξ1 = ρ − ρ in the above equation. But there is only one pricing kernel. E( |ξ1 |) < ∞}. a miracle occurs. We will be done once we show that every ξ1 has the desired integrability property. to sell the stock for a ﬁxed strike price K at time 1. Following the argument. S. concluding the proof of the “if” direction. Assuming E(ρ|ξ1 |) < ∞. Proof. completing the proof of the “only-if” direction. π 1 . An arbitrage-free market is complete if and only if every random variable is attainable. Let ρ and ρ be two pricing kernels. and let ξ1 be an arbitrary random variable. First. suppose the market is complete. Since the integrand is non-negative. In the special case when K = K . Since ξ1 is attainable by assumption. A put option gives the owner of the option the right. All we need do is replace the set C in our proof of the ﬁrst fundamental theorem with ¯ ¯ C = {E( S1 ) : > 0 a. −1. called a put option. we have the identity P1 = (K − S1 )+ = K − S1 + (S1 − K)+ K = B1 − S1 + C1 . 1 . 1+r K so the replicating portfolio is (π 0 . we have E(ρ|ξ1 |) < ∞ and E(ρ |ξ1 |) < ∞ and E(ρξ1 ) = ξ0 = E(ρ ξ1 ) so that E[(ρ − ρ )ξ1 ] = 0. and hence ξ1 is attainable. Using a similar argument as we used for the call option.s. we see that ξ0 = E(ρξ1 ) satisﬁes our characterization of attainable claims. Indeed. Now suppose that the pricing kernel ρ is unique. we must conclude ρ = ρ almost surely. but not the obligation. We can characterize complete markets: Theorem (Second Fundamental Theorem of Asset Pricing). A market model is complete if and only if there exists a unique pricing kernel. we have just derived the famous put-call parity formula 1 K C0 − P0 = S0 − 1+r Since ﬁnding the time-0 price of an attainable claim is easy.

. . . in a complete market. . i. Theorem (No-arbitrage price bounds). Or looking the other way.. There is usually exists a solution only if n ≤ d. all contingent claims can be perfectly replicated. . That is. 0 d ξ1 (ωn ) = π 0 S1 (ωn ) + . we have the rule-of-thumb below: 2FTAP: Market completeness ⇔ Uniqueness of pricing kernel “ ⇔ n = d” 4. what portfolio should you buy to minimize your exposure to the unhedgeable risk? 16 .Remark. . i. one should expect to ﬁnd a solution to these equations if there are at least as many unknowns as equations. + π d S1 (ωn ) in the d + 1 unknowns π 0 . ωn } has n + 1 points (informally. Then what to do? If you are the seller of a claim.. = . Otherwise. and hence one should not expect to ﬁnd a ρ satisfying the system. . . ξ) if inf E(ρξ1 ) ≤ ξ0 ≤ sup E(ρξ1 ) ρ ρ where the inﬁmum and supremum are taken over pricing kernels ρ such that E(ρ|ξ1 |) < ∞. . the interval does not collapse to a single point.. Suppose that our sample space Ω = {ω0 . . if n ≥ d. we should expect there to be a unique pricing kernel ρ only if the number of unknowns equals the number of equations. Roughly speaking. .e. Indeed. when is the market complete? Heuristically. . d d d d S0 = ρ(ω0 )S1 (ω0 )P{ω0 } + . . we know how to compute the time-0 price of any contingent claim– just compute the expected value of the claim times the unique pricing kernel.. . . + ρ(ωn )S1 (ωn )P{ωn } = E(ρS1 ). There is no arbitrage in the augmented market ¯ (S. + π d S1 (ω0 ) . Hence. . + ρ(ωn )S1 (ωn )P{ωn } = E(ρS1 ) . the set of no-arbitrage prices is an interval. n sources of randomness). . . if n < d then there are few unknowns than equations. Since the existence of the pricing kernel is equivalent to the lack of arbitrage. but unless the claim in attainable. If ρ is a pricing kernel. Pricing and hedging in an incomplete market by minimizing hedging error In complete markets. we must solve the n + 1 equations ¯ ¯ 0 d ξ1 (ω0 ) = π 0 S1 (ω0 ) + . . = . to replicate a contingent claim ξ1 we need to ﬁnd a π such that ¯ π · S1 = ξ1 . . it must satisfy the d + 1 equations 0 0 0 0 S0 = ρ(ω0 )S1 (ω0 )P{ω0 } + . π d .e. n = d. What does the no-arbitrage principle say about the prices of contingent claims in incom¯ plete markets? Let S be a no-arbitrage market model. one has the following rule-of-thumb: 1FTAP: No arbitrage ⇔ Existence of pricing kernel “⇔ n ≥ d” Now. so their time-0 prices are just the time-0 prices of the corresponding replicating portfolio. That is.

∗ We are trying to ﬁnd the attainable claim X1 = π ∗ · S1 which minimizes the functional ¯ ¯ 2 ∗ X1 → E[(ξ1 − X1 ) ]. we have the following interpretion: • π ∗ the optimal vector of portfolio weights. In this section. ¯ We can ﬁnd the minimizer of the function F (¯ ) = E[(ξ1 − π · S1 )2 ] π ¯ ¯ by the usual means of calculus: ¯ F (π ∗ ) = E[S1 (ξ1 − π · S1 )] = 0 ¯ ¯ yielding ¯ π ∗ = E(V −1 S1 ξ1 ) ¯ ∗ ¯T ¯ X0 = E[S0 V −1 S1 ξ1 ] ¯ ¯T and V = E[S1 S1 ] is a (d + 1) × (d + 1) matrix. just as in the case of a complete market. First we show that E(S1 ρ∗ ) = S0 . One possible solution is ¯ to minimize expected square hedging error. there are many. We have the computation ¯ ¯ ¯T ¯ ¯ E(S1 ρ∗ ) = E[S1 S1 V −1 So ] = S0 . assumed positive deﬁnite. Equivalently. Consider a market S and introduce a claim with payout ξ1 at time 1. ¯ ¯ Proof. which is closest (in the least-squares sense) to the target claim ξ1 . we ﬁnd the wealth X1 . which can be attained by trading in the market. The random variable ρ∗ = S0 V −1 S1 minimizes the functional ρ → E(ρ2 ) ¯ ¯ among all random variables ρ such that E(ρS1 ) = S0 . the seller of the claim should charge at least X0 and invest the proceeds into the portfolio π ∗ . Hence if ∗ ¯ ¯ X1 = π ∗ · S1 is optimal. That is. What can we say about this solution? Note that we can express the optimal initial wealth as ∗ X0 = E(ρ∗ ξ1 ) where the random variable ρ∗ is deﬁned by ¯T ¯ ρ∗ = S0 V −1 S1 ∗ so that the pricing rule ξ1 → X0 is linear. 17 .Of course. and ¯ ∗ • π ∗ · S0 = X0 is the time-0 price of this portfolio. we suppose that the prices are square integrable 2 ¯ E(|S1 |2 ) < ∞ and E(ξ1 ) < ∞. ∗ we can think of X1 as the projection of ξ1 on the subspace of attainable claims. ¯ ¯ ∗ In particular. many answers to this question. The most important property of ρ∗ is found in the following: ¯T ¯ Theorem.

18 . Also. then it is possible that the augmented market has an arbitrage! 5. if ξ is a non-negative random variable then EQ (ξ) = EP (Zξ). Let ρ be another random variable such that ¯ ¯ ¯ E(S1 ρ∗ ) = S0 . Consider probability measures P and Q deﬁned by • P{1} = 1/2. we have the computation E(ρ2 ) = E[(ρ∗ + ∆ρ)2 ] ¯T ¯ = E[(ρ∗ )2 ] + 2E[S0 V −1 S1 ∆ρ] + E[(∆ρ)2 ] ≥ E[(ρ∗ )2 ]. we introduce the important concepts of num´raire assets and equivalent e martingale measures. The measures P and Q are equivalent. if and only if P(A) = 1 ⇔ Q(A) = 1 Example. Note that EP (Z) = 1 by putting A = Ω in the conclusion of theorem. P{2} = 1/2. completing the proof. The probability measure Q is equivalent to the probability measure P if and only if there exists a positive random variable Z such that Q(A) = EP (Z 1A ) for each A ∈ F. It turns out that equivalent measures can be characterized by the following theorem. F) be a measurable space and let P and Q be two probability measures on (Ω. Then P and Q are equivalent. 2. etc. we have discovered an explicit formula for the unique pricing kernel. and Q{3} = 0. When there are more than one probability measure ﬂoating around. and P{3} = 0 • Q{1} = 1/1000. 3} with the set F of events all subsets of Ω. Change of num´raire and equivalent martingale measures e In this section. Writing ρ = ρ∗ + ∆ρ. We need a deﬁnition: Definition. if the market is complete. Consider the sample space Ω = {1. ∗ ρ is not a pricing kernel. we use the notation EP to denote expected value with respect to P. In particular. In this case.It remains to show that ρ∗ is minimal. then ρ∗ would be the pricing kernel with the smallest L2 norm. Indeed if ρ∗ is not strictly positive and if one prices the claim by ∗ ξ0 = X0 = E(ρ∗ ξ1 ). Q{2} = 999/1000. However. Then ∆ρ = ρ∗ − ρ satisﬁes E(S1 ∆ρ) = 0. for a general market ρ∗ may well have the property P(ρ ≤ 0) > 0. Theorem (Radon–Nikodym theorem). written P ∼ Q. Let (Ω. by the usual rules of integration theory. F). Note that if ρ∗ is a positive random variable.

. P also has a density with respect to Q given by dP 1 = . let ρ be a pricing kernel. dQ Z Now let’s return to our ﬁnancial model. . then the random variable Z is called the density. we deﬁne an equivalent probability measure Qi by Qi (A) = E(Zi 1A ) for all A ∈ F. Unlike the notion of a pricing kernel. the notion of an equivalent martingale measure is ‘num´raire-dependent’. we can speak in terms e of prices relative to the num´raire. Then. dP In fact.If Q ∼ P.1} are a martingale for Qi with respect to the ﬁltration (Ft )t∈{0. the term equivalent martingale measure is appropriate since the stochastic processes (Stj /Sti )t∈{0. A num´raire is an asset with a strictly positive price at all times. As a preview of what’s to come. or the Radon–Nikodym derivative. That is. P). we need the following deﬁnition: Definition. We will elaborate on this in the multi-period case. Remark. given a pricing kernel ρ we can construct an equivalent e martingale measure corresponding to a diﬀerent num´raire. for every num´raire e e asset i. and suppose asset e i is a num´raire. we have e j Sj S0 = E Zi 1 i i S0 S1 i S1 . Hence. we can deﬁne a measure Qi by Si dQi = ρ 1. Ω} and F1 = F. i dP S0 19 j S1 i S1 = j S0 i S0 . F. That is. An equivalent martingale measure relative to the num´raire asset i is any probability e j Q i measure Q equivalent to P such that E (|S1 |/S1 |) < ∞ and EQ for all j ∈ {0. i S0 Notice that Zi > 0 almost surely and E(Zi ) = 1. Hence. e The idea is that a num´raire can be used to count money. of Q with respect to P. and is often denoted dQ Z= . The measure Qi is called an equivalent martingale measure: where Zi = ρ ¯ Definition. .1} where F0 = {∅. To start oﬀ. . In particular. The measure P is called the objective (or historical or statistical ) measure for the model. d}. Let S be a market model deﬁned on a probability space (Ω.

we count money in units of asset 0: Deﬁne the new quantities for both t ∈ {0. Definition. their riskiness being relative to the num´raire asset 0.In general. The choice of num´raire is usually arbitrary. . in which case B0 = B1 = 1 or a bank (or money market) account in which case B0 = 1 and B1 = 1 + r. 6. we can solve for φ. B0 B0 Plugging this into the self-ﬁnancing condition yields after some manipulation B1 B1 X1 = X0 + π · S1 − S0 B0 B0 To clean things up. we choose asset 0. . e If the investor’s initial wealth X0 is ﬁxed. the measures Qi constructed above are diﬀerent for diﬀerent i.) In this new notation. but this choice is arbitrary. (Of e course. Since we have assumed that asset 0 is a num´raire. X1 = φB1 + π · S1 . Discounted prices From now on. Asset 0 is a num´raire. as long as it is a num´raire. . 1} Xt St ˜ ˜ Xt = . e The remaining d assets are often called risky assets. . This brings us to a standing assumption: e Assumption. The actual identity of asset 0 is irrelevant for much of our analysis. and St = Bt Bt denoting the wealth and the risky asset prices discounted by the num´raire asset 0. and we wouldn’t need the new notation. By convention. To make the notation easier. As mentioned earlier. . if asset 0 were cash then B0 = B1 = 1 almost surely. yielding e X0 S0 φ= −π· . . we distinguish one asset and treat it diﬀerently than the d others. e but there is one case that should be mentioned. A num´raire is risk-free if its time-1 price is not random. we let 1 Bt = St0 . where r is the interest rate. B0 > 0 and B1 > 0 almost surely.e. so that X0 = φB0 + π · S0 . . asset 0 is often cash. Std ) from now on. the budget constraint and self-ﬁnancing condition neatly combine to yield ˜ ˜ ˜ ˜ X1 = X0 + π · (S1 − S0 ) . We write φ = π 0 . and π = (π 1 . even though they all correspond to the same pricing kernel ρ. i. . 20 . An equivalent e martingale measure relative to a risk-free asset is called risk-neutral. the budget constraint means that we cannot freely choose the investor’s portfolio. and St = (S1 . π d ).

so π is an arbitrage relative to asset 0. then ¯ ˜ ˜ ˜ ˜ ˜ ˜ ˜ ˜ X1 ≥ 0 ≥ X0 a. 21 . Of course. Attainable claims can be characterized in terms of equivalent martingale measures: Theorem (Characterization of attainable claims). We can now rewrite the ﬁrst fundamental theorem: Theorem (First Fundamental Theorem of Asset Pricing). once we let φ = −π · S ˜ ˜ the initial wealth is X0 = φB0 + π · S0 = 0 and X1 = B1 π · (S1 − S0 ). π) ∈ Rd+1 is ¯ ˜0 . Conversely. We can also redeﬁne the term attainable: Definition.Now it is time to consider the notion of arbitrage in this new notation. the second fundamental theorem can be rewritten: Theorem (Second Fundamental Theorem of Asset Pricing). an arbitrage by our old deﬁnition of the word. we are actually interested in the lack of arbitrage: the market model has no arbitrage if and only if ˜ ˜ ˜ ˜ π · (S1 − S0 ) ≥ 0 a. Suppose π ∈ Rd is an arbitrage relative to asset 0. in this case. where ξ1 = ξ1 /B1 . implies π · (S1 − S0 ) = 0. π) is an arbitrage according to our old deﬁnition.s. so X1 ≥ 0 a. A claim with payout ξ1 is attainable if there is a real number x and portfolio π ∈ Rd of risky assets such that ˜ ˜ ˜ ξ1 = x + π · (S1 − S0 ) ˜ where ξ1 = ξ1 /B1 . A claim with payout ξ1 is attainable if and only if there exists a constant x such that ˜ EQ (ξ1 ) = x ˜ ˜ for all equivalent martingale measure Q such that EQ (|ξ1 |) < ∞. These are the formulations that we will use for the remainder of these notes. A market model is complete if and only if the equivalent martingale measure (relative to asset 0) is unique. Note that an equivalent martingale measure Q is simply a probability measure Q ∼ P such that ˜ ˜ EQ (S1 ) = S0 . Indeed.s.s. if the portfolio π = (φ. Finally. and P(X1 > 0) > 0. Definition. An arbitrage (relative to the num´raire asset 0) is a portfolio π ∈ Rd of e risky assets such that ˜ ˜ P[π · (S1 − S0 ) ≥ 0] = 1 ˜ ˜ P[π · (S1 − S0 ) > 0] > 0 Remark. and P(X1 > X0 ) > 0. Then π = (φ. The market model has no arbitrage if and only if there exists an equivalent martingale measure (relative to asset 0). But X1 − X0 = π · (S1 − S0 ).

.

Ω}. HT } = 1/2. Pick an outcome ω ∈ Ω at random. this entails replacing the time index set {0. you can answer every question. so there are only two questions you can answer. if the ﬁrst toss is heads. but Is ω in {HH. but what if the ﬁrst toss comes up tails? So. we can proceed briskly into the natural generalization of multi-period discrete time. Then you can always answer these four questions. The set-up So we consider a market with d + 1 assets. The correct notions are that of measurablility and of a ﬁltration. T H. Is ω in Ω? Yes. but I can’t answer the question with certainty. T T }? Yes. After the second toss. At time 1. At time 0. So. T T }. in general I can’t answer the question with certainty after observing the ﬁrst ﬂip. but Is ω in {HH. Our ﬁrst goal is to ﬁnd the proper generalization to this setting of the assumption that the time-0 prices B0 and S0 are constant. Is ω in ∅? No. the ﬂow of information is modelled by the following sigma-ﬁelds • F0 = {∅. Is ω in ∅? No. HT }? I can assign a probability P{HH. Consider the experiment of tossing a coin two times. if the ﬁrst toss is heads. if the ﬁrst toss is tails. Essentially. the coin has not been tossed. of course. HT. 23 . HT }? Yes. 1. Is ω in {HH. 1} with the non-negative integers Z+ and keeping track of the resulting complications. no otherwise. Is ω in Ω? Yes. The prices of these assets are modelled by the stochastic process (B. HT. We can model this experiment on the sample space Ω = {HH. We need to formalize the concept of information being revealed as time marches forward.CHAPTER 2 Multi-period models Now that the ﬁnancial foundation has been laid in the one-period models. P). T H}? Yes. Example. F. St )t∈Z+ deﬁned on some background probability space (Ω. S) = (Bt . the coin has been tossed once. no otherwise. Is ω in {T H.

Definition. that is. {HH. For instance.• F1 = {∅. we will nearly always assume that F0 is trivial: A ∈ F0 ⇔ P(A) = 0 or P(A) = 1. since there is no information before the experiment. Definition. G-measurable) if and only if the event {ξ ≤ x} is an element of G for all x ∈ R. . the random variable Y1 must be of the form a if ω ∈ {HH.1. but the some of the following deﬁnitions are stated in more generality than we need here to avoid repetition in later chapters in which T = R+ . . ξn ) is G-measurable if and only if ξi is G-measurable for each i ∈ {1. F. . . On the other hand. The sigma-ﬁeld Ft is our model of what information is known to the market participants at time t. Now we can state the proper generalization of the assumption in one period models that the time-0 asset prices are constant: We equip the background probability space (Ω. 24 . For this chapter. . • F2 = the set of all sixteen subsets of Ω. Let T ⊆ R+ be an index set. Assumption. Notice that for all t ∈ {0. Y0 must be a constant.2} that has the property that the value of the random variable ξt is known once after t tosses of the coin. 1. n}. When dealing with ﬁltrations. A stochastic process (Yt )t∈T is adapted to a ﬁltration (Ft )t∈T iﬀ Yt is Ft measurable for each t ∈ T. Now consider a stochastic process (Yt )t∈{0. . then ξ = (ξ1 . {T H. HT } Y1 (ω) = b if ω ∈ {T H. T T }. . Y0 (ω) = a for all ω ∈ Ω. Y2 can be any function on Ω. Finally. Let G ⊆ F be a sigma-ﬁeld. St )t∈Z+ is assumed to be adapted to (Ft )t∈Z+ . In particular. we have the following deﬁnitions: Definition. The stochastic process (B. Remark. all F0 measurable random variables are almost surely constant. 2} the event {Yt ≤ x} is in Ft for every x ∈ R. If ξ is a random vector. of the form a if ω = HH b if ω = HT Y2 (ω) = c if ω = T H d if ω = T T. T T } since the only information known at time 1 is whether or not the ﬁrst coin came up heads. P) with a ﬁltration (Ft )t∈Z+ . Ω}. we have T = Z+ the positive integers. A ﬁltration (Ft )t∈T is an increasing collection of sigma-ﬁelds on Ω such that Fs ⊆ Ft if s ≤ t. . HT }. With this motivation. A random variable ξ : Ω → R is measurable with respect to G ( or brieﬂy. S) = (Bt .

Now there are multiple periods. Remark. Just as we did in the one-period case. the d-dimensional stochastic process (πt )t∈N is called the investor’s trading strategy. For a ﬁxed t ∈ N we refer to the d-dimensional random vector πt as the investor’s portfolio. e As before. and doesn’t really need further comment. 25 . • πt denotes the portfolio of risky assets held between periods t − 1 and t. The investor’s wealth and porfolio are connected by the following relationships: Xt−1 = φt Bt−1 + πt · St−1 Xt = φt Bt + πt · St the budget constraint the self-ﬁnancing condition Note that we should assume that our investor is not clairvoyant. we see that B0 and S0 are constants as before. πt ) that are Ft−1 -measurable. • Xt denotes the investor’s wealth at the beginning of period t. Since the only random variables measurable with respect to the trivial sigmaﬁeld are constants. Bt−1 Bt−1 Now count money in units of the num´raire asset 0: e B0 B0 ˜ ˜ Xt = Xt and St = St Bt Bt so that the budget constraint and self-ﬁnancing condition combine to yield ˜ ˜ ˜ ˜ Xt = Xt−1 + πt · (St − St−1 ). . Note that the time index set for a predictable process (Yt )t∈N is (usually) N = {1.}. not Z+ . we can solve for φt : φt = Xt−1 St−1 − πt · . Hence we henceforth only consider portfolios (φt . so we need to be careful about when an investor chooses his investment portfolio. • φt denotes the number of shares of asset 0 held between periods t − 1 and t. . A stochastic process (Yt )t∈N is predictable if Yt is Ft−1 -measurable for each t ∈ N. generalizes the assumption made in one period models. we introduce an investor. The ﬁnal formula to remember is then t ˜ ˜ Xt = X 0 + s=1 ˜ ˜ πs · (Ss − Ss−1 ). Assumption. Hence Y0 is not necessarily deﬁned. . Remark. Such processes have a name: Definition. Our second assumption is natural. given this market model.Remark. 2. Asset 0 is a num´raire so that Bt > 0 almost surely for all t ∈ Z+ .

Consider the random variable a if ω = HH b if ω = HT ξ(ω) = c if ω = T H d if ω = T T. . . . P(Gn ) This example relates the notion of conditional expection given a sigma-ﬁeld and that of conditional expectation given an event. Let X be a integrable random variable deﬁned on the probability space (Ω. . An arbitrage is a strategy (πt )t∈N with the property that there exists a (non-random) time T ∈ N such that T P s=1 T ˜ ˜ πs · (Ss − Ss−1 ) ≥ 0 ˜ ˜ πs · (Ss − Ss−1 ) > 0 s=1 = 1 P > 0.. Definition. {T H. T T }.}. Then there exists an integrable G-measurable random variable Y such that E(1G Y ) = E(1G X) for all G ∈ G. . F. . Suppose the coin is fair. Let X be an integrable random variable and let G ⊂ F be a sigma-ﬁeld. be a sequence of disjoint events with P(Gn ) > 0 for all n ∈ N and n∈N Gn = Ω. . written E(X|G). Furthermore. suppose Ω = {HH. G2 . HT. Let G1 . Then E(X|G)(ω) = E(X 1Gn ) if ω ∈ Gn . F. P). Theorem (Existence and uniqueness of conditional expectations). and let G ⊆ F be a sub-sigma-ﬁeld of F. Let G be the smallest sigma-ﬁeld containing {G1 . Definition.. Example. T H. P). The ﬁrst fundamental theorem of asset pricing Following the discussion from before we can deﬁne an arbitrage. is a G-measurable random variable with the property that E [1G E(X|G)] = E(1G X) for all G ∈ G. since we usually write E(X 1G ) = E(X|G). and let G = {∅. {HH. (Sigma-ﬁeld generated by a countable partition) Let X be a non-negative random variable deﬁned on (Ω. Ω} be the sigma-ﬁeld containg the information revealed by the ﬁrst toss. so that each outcome is equally likely. P(G) More concretely. 26 . The conditional expectation of X given G. if there exists another G-measurable random variable Y such that E(1G Y ) = E(1G X) for all G ∈ G. G2 . T T } consists of two tosses of a coin. then Y = Y almost surely.2. . we have to recall some results and deﬁnitions about conditional expectations and martingale theory. Before we can state the ﬁrst fundamental theorem. HT }.

• Fatou’s lemma: If Xn ≥ 0 a. it is an exercise to show that an integrable process M is a martingale only if E(Mt+1 |Ft ) = Mt for all t ≥ 0.s. • Jensen’s inequality: If f is convex. the most important concept in ﬁnancial mathematics is that of a martingale. then E(XY |G) = XE(Y |G). • linearity: E(aX + bY |G) = aE(X|G) + bE(Y |G) for all constants a and b • positivity: If X ≥ 0 almost surely. In particular. If T = Z+ . it is suﬃcient to verify the conditional expectations of the process one period ahead.s. Let all random variables appearing below be such that the relevant conditional expectations are deﬁned. The natural ﬁltration of Y is the ﬁltration (Ft )t∈T . then E(lim inf n Xn |G) ≥ lim inf n E(Xn |G) • dominated convergence theorem: If supn |Xn | is integrable and Xn → X a. E(X|G) = E(X) if G is trivial. Before listing them. that is. Below are some examples of martingales.s. and let G be a sub-sigma-ﬁeld of the sigma-ﬁeld F of all events. HT } if ω ∈ {T H. if X is G-measurable. A martingale relative to a ﬁltration (Ft )t∈T is a stochastic process M = (Mt )t∈T with the following properties: • E(|Mt |) < ∞ for all t ∈ T • E(Mt |Fs ) = Ms for all 0 ≤ s ≤ t. for all n. T T } The important properties of conditional expectations are collected below: Theorem. it is convenient to introduce a deﬁnition: Definition. • tower property or law of iterated expectations: If H ⊆ G then E[E(X|G)|H] = E[E(X|H)|G] = E(X|H) As hinted at in Chapter 1. with almost sure equality if and only if X = 0 almost surely. A martingale is simply an adapted stochastic process that is constant on average in a following sense: Definition. then we are implicitly working with the natural ﬁltration of the process. 27 . In what follows. if a stochastic process is given but a ﬁltration is not explicitly mentioned. • If X is independent of G (the events {X ≤ x} and G are independent for each x ∈ R and G ∈ G) then E(X|G) = E(X). let Ft be the smallest sigma-ﬁeld for which the random variables Ys is measurable for all 0 ≤ s ≤ t. That is. • ‘take-out-what’s-known’: If X is G-measurable. then E(Xn |G) ↑ E(X|G) a.Then E(ξ|G)(ω) = (a + b)/2 (c + d)/2 if ω ∈ {HH. then E(X|G) = X. then E(X|G) ≥ 0 almost surely.s. Remark. then E[f (X)|G] ≥ f [E(X|G)] • monotone convergence theorem: If 0 ≤ Xn ↑ X a. the smallest ﬁltration for which ξ is adapted.s. then E(Xn |G) → E(X|G) a. Given a stochastic process Y = (Yt )t∈T . In particular.

Remark. The process N in the theorem is often called a martingale transform or a discrete time stochastic integral. and let Mt = E(X|Ft ). . The process (St )t∈Z+ given by S0 = 0 and St = X1 + . . + E(|Xt |) by the triangular inequality. Theorem. X3 . First. Proof. . we have E(|Mt |) < ∞ and there exist a constant C > 0 such that |Ht | < C almost surely for all t ∈ Z+ . Then the process N deﬁned by t Nt = s=1 Hs (Ms − Ms−1 ) is a martingale. Then M = (Mt )t∈T is a martingale. . X2 . By assumption. Example. E(St+1 |Ft ) = E(St + Xt+1 |Fn ) = E(St |Ft ) + E(Xt+1 |Ft ) = St + E(Xt+1 ) = St . the random variable St is integrable since E(|St |) ≤ E(|X1 |) + . The following theorem shows how to take one martingale and build another one. . + Xt is a martingale relative to its natural ﬁltration. . E(|Mt |) = E{|E(X|Ft )|} ≤ E{E(|X||Ft )} = E(|X|) < ∞ 28 . . Indeed. . be independent integrable random variables such that E(Xi ) = 0 for all i ∈ N. Let X be an integrable random variable. Let M be a martingale and let H be a bounded predictable process.Example. Also. We now construct one of the most important examples of a martingale. Hence t E(|Nt |) ≤ s=1 t E(|Hs ||Ms − Ms−1 |) C[E(|Ms |) + E(|Ms−1 |)] < ∞ s=1 ≤ Since E(Nt+1 − Nt |Ft ) = E(Ht+1 (Mt+1 − Mt )|Ft ) = Ht+1 E(Mt+1 − Mt |Ft ) = 0 we’re done. Let X1 .

Our aim is to show X2 = 0 a. ˜ To do this. Note Q(X2 ≥ 0) = 1 because P ∼ Q.. we have X1 ≥ 0 a. called European that mature at a ﬁxed date and those. because this would imply that Q(X2 = 0) = 1 ˜ and hence P(X2 = 0) = 1. Definition. P) and adapted to a ﬁltration (Ft )t∈{0. that the existence of an equivalent martingale measure implies the lack of arbitrage. We only consider the easier direction.s. E(Mt |Fs ) = E[E(X|Ft )|Fs ] = E(X|Fs ) = Ms by the tower property. we can introduce a contingent claim. Letting n → ∞ shows X2 = 0 a. that can be exercised whenever the holder of the claim wants. ˜ is a Q-martingale. Now. An equivalent martingale measure is a measure Q equivalent to P such that ˜ is a martingale for Q. let An = {|X1 | ≤ n. which is Q-integrable since S ˜ ˜ ˜ 0 ≤ EQ [1An X2 |F1 ] = 1An X1 + 1An π2 · E(∆S2 |F1 ) ˜ = 1An X1 ˜ and letting n → ∞. F. Let π be such that ˜ ˜ ˜ ˜ Xt = t πs · ∆Ss satiﬁes X2 ≥ 0 a. s=1 ˜ Let Q be an equivalent martingale measure. St )t∈{0.T } deﬁned on a probability ˜ space (Ω. Unlike the one-period case.s.. but one has to be a bit more careful.s.s. again by equivalence of measures. |π2 | ≤ n}. note that ˜ ˜ ˜ 1An X2 = 1An X1 + 1An π2 · ∆S2 ˜ is Q-integrable. and the second term is bounded by n|∆S2 |. as desired. The idea is the same as in the T = 1 case.. Finally.s.. Furthermore. The market model has no arbitrage if and only if there exists an equivalent martingale measure. in the case where T = 2.Now. for each n.. called American contingent claims. European and American contingent claims Now given our multi-period market model. S Theorem (First Fundamental Theorem of Asset Pricing).T } . Proof.. Recall the notation St = B0 St for Bt the discounted risky asset prices. there are now essentially two types of contingent claims: those. 29 . 3. since the ﬁrst term is bounded. ¯ We now are ready to consider our model S = (Bt . Now we can compute ˜ ˜ EQ (X1 ) = π1 · E(S1 ) = 0 ˜ to conclude X1 = 0 a. since ˜ ˜ EQ [1An X2 |F1 ] = 1An X1 = 0 ˜ ˜ we conclude 1An X2 = 0 a. Note An is in F1 ..s.. We Q ˜ ˜ would be done if we could show E (X2 ) = 0.

ξt )t∈{0. Theorem. We now introduce a (European) contingent claim which matures at time T ∈ N.T } is a martingale if and only if Bt ˜ ˜ ξt = EQ (ξT |Ft ) and we’re done. Theorem (Characterization of attainable claims). We model the payout of the claim by an FT measurable random variable ξT . We concentrate on the European claims ﬁrst. T }. By the ﬁrst fundamental theorem of asset pricing.) 30 .. The following should now come as no surprise. The following theory should seem very familiar since there really is nothing new.. where ξT = Bt . S) = (Bt . Given an arbitrage free market model ¯ a contingent claim with payout ξT is attainable if and only if there a constant x such that S. but we do not do so here.. ˜ x = EQ (ξT ) ˜ for all equivalent martingale measures Q such that EQ (|ξT |) < ∞. Think of the example of a call option on a stock maturing at time T with strike K. Definition. But (ξt )t∈{0. or the set Q of equivalent martingale measures contains just one element. We just consider the case T = 2.. .3. Let (B. ˜ ˜ where ξt = B0 ξt .. The augmented market (Bt .. St )t∈Z+ be an arbitrage-free market.2} such that X2 = ξ2 a. and suppose that the claim is attainable... Proof.s. A claim maturing at time T ∈ N is attainable if and only if its payout ξT is an FT -measurable random variable such that T Bt ξ Ft BT ˜ ξT = x + s=1 ˜ ˜ πs · (Ss − Ss−1 ) ξ ˜ for some x ∈ R and predictable strategy (πt )t∈N . T The market is complete if and only if every claim is attainable. In this case ξT = (ST − K)+ . St . . The above theorem gives a lot of ﬂexibility in pricing contingent claims unless the claim is attainable.T } is a martingale for Q.... .1.T } is free of arbitrage if and only if there exists an equivalent martingale measure Q such that ξt = EQ for all t ∈ {0. European claims. (It needs to be proven that there is at least one Q with this property. so that there exists a constant x and strategy (π)t={1. there is no-arbitrage if and ˜ ˜ only if there exists an equivalent measure Q such that (St . ξt )t∈{0. .. Proof. where t ˜ Xt = x + s=1 Q ˜ ˜ πs · (Ss − Ss−1 ) ˜ ˜ We need to show x = E (ξ2 ) for all Q such that ξ2 is integrable.

. then the actual payout of the claim is modelled by the random variable ξτ .T } . the payout of the option is (K − Sτ )+ where τ ∈ {0. ˜ ˜ ˜ Now X1 = x + π1 (S1 − S0 ) is integrable and has mean x. |π2 | ≤ n} and hence ˜ ˜ EQ [1An X2 |F1 ] = 1An X1 . to rule out clairvoyance. the holder of an American claim can choose to exercise the option at any time τ before or at maturity. things are quite different.T } . Now. Unlike the European claim. . and hence so is {τ > t}c = {τ ≤ t}. .. since π1 is a constant. Theorem (Second Fundamental Theorem of Asset Pricing). t} is Ft -measurable. . For instance. Hence. Example. . in the case of an American put.. {τ > t} = {Ys ≤ a for all s = 0. Hence. T } is a time chosen by the holder of the put to exercise the option. • an adapted process (ξt )t∈{0. .. we state the characterization of complete markets. where τ is any stopping time for the ﬁltration taking values in {0. Indeed. Applying the tower law yields ˜ ˜ EQ [EQ (X2 |F1 )] = E(X1 ) = x. . . ˜ ˜ Now. the payout of an American claim is speciﬁed by two ingredients: • a maturity date T ∈ N. we can use the conditional dominated convergence theorem on the left side to let n → ∞ to get ˜ ˜ EQ [X2 |F1 ] = X1 . if an American claim matures at T ∈ N and is speciﬁed by the payout process (ξt )t∈{0. T }.˜ As before. A stopping time for a ﬁltration (Ft )t∈T is a random variable τ taking values in T ∪ {∞} such that the event {τ ≤ t} is Ft -measurable for all t ∈ T. . 3.. let An = {|X1 | ≤ n.2.. since |1An X2 | ≤ X2 and this is integrable by assumption. The market is complete if and only if there exists a unique equivalent martingale measure. The canonical example of an American claim is the American put option– a contract which gives the buyer the right (but not the obligation) to sell the underlying stock at a ﬁxed strike price K > 0 at any time between time 0 and a ﬁxed maturity date T ∈ N. We now discuss American claims. . Let (Yt )t∈Z+ be an adapted process and ﬁx a ∈ R. . However. Finally. we insist that τ is a stopping time: Definition. . American claims. Here. we may take ξt = (K − St )+ . Then the random variable τ = inf{t ∈ Z+ : Yt > a} corresponding to the ﬁrst time the process crosses the level a is a stopping time. (First passage time) Here is a typical example of a stopping time... . 31 .

St )t∈Z+ is complete. where the supremum is taken over the set T of stopping times smaller than or equal to T and where Q is the unique martingale measure.We can think of the American claim then as a family. we make the following assumption in this subsection: The market model (B. . There exists a trading strategy (πt )t∈{1. of European claims with payouts ξτ .. A martingale is a stochastic process that is both a supermartingale and a submartingale. the seller of such a claim should at time 0 charge at least the amount sup EQ τ ∈T B0 ξτ Bτ to be sure that he can hedge the option. Intuitively.. the investor can super-replicate the payout of the American claim.T } speciﬁes the payout of an American claim maturing at T ∈ N. Consider a complete market (B... .. Remark. A supermartingale relative to a ﬁltration (Ft )t∈T is an adapted stochastic process (Ut )t∈T with the following properties: • E(|Ut |) < ∞ for all t ∈ T • E(Ut |Fs ) ≤ Us for all 0 ≤ s ≤ t. and suppose that the adapted process (ξt )t∈{0. Remark. Indeed. To simplify matters. while a submartingale increases on average. Hence a supermartingale decreases on average... .. A submartingale is an adapted process (Vt )t∈Z+ with the following properties: • E(|Vt |) < ∞ for all t ∈ T • E(Vt |Fs ) ≥ Vs for all t ∈ T. T } if and only if X0 ≥ sup EQ τ ∈T B0 ξτ Bτ . We need some new vocabulary: Definition. indexed by the stopping time τ . The rest of this subsection is dedicated to proving this theorem. S) = (Bt . The theorem says that if the initial wealth is suﬃciently large.t} with corresponding discounted wealth process t ˜ ˜ Xt = X 0 + s=1 ˜ ˜ πs · (Ss − Ss−1 ) such that Xt ≥ ξt almost surely for all t ∈ {0. S). . this is the case. The unique equivalent martingale measure is denoted Q. Now we need a result of general interest: 32 . Theorem.

To show uniqueness.. E(Ut+1 |Ft )} for t ∈ {0.. ...T } by U T = YT Ut = max{Yt . . Deﬁne an adapted process (Ut )t∈{0.. let (Ut )t∈{0.T } is called the Snell envelope of (Yt )t∈{0.T } be its Snell envelope with Doob decomposition Ut = Mt − At .T } will be the process specifying the discounted payout of ˜ the given American claim Yt = ξt . 33 ... Summing up.. Thus. Then M − M is a predictable martingale. a constant. ...Theorem (Doob decomposition theorem). . Remark. The Snell envelope clearly satisﬁes both Ut ≥ Yt and Ut ≥ E(Ut+1 |Ft ) almost surely..T } be an adapted process. Theorem.. Let M0 = U0 and deﬁne Mt+1 − Mt = Ut+1 − E(Ut+1 |Ft ) for t ∈ N.. Let (Yt )t∈{0. Then there is a unique decomposition Ut = Mt − At where M is a martingale and A is a predictable non-decreasing process with A0 = 0. assume that Ut = Mt − At = Mt − At . In our application (Yt )t∈{0... t Mt − At = M0 − A0 + s=1 t (Ms − Ms−1 − As + As−1 ) = U0 + s=1 (Us − Us−1 ) = Ut . Clearly At+1 is Ft -measurable and the process A is non-decreasing process since U is a supermartingale. T } : At+1 > 0} with the convention τ ∗ = T on {At = 0 for all t}. Let U be a supermartingale indexed by Z+ . .... Let τ ∗ = min{t ∈ {0.. . .T } be a given integrable adapted process. Now let A0 = 0 and At+1 − At = Ut − E(Ut+1 |Ft ) for t ∈ N. Let (Yt )t∈{0. Then τ ∗ is a stopping time and Yτ ∗ = Mτ ∗ ... Then M is a martingale.. another way to describe the Snell envelope of a process is to say it is the smallest supermartingale dominating that process.... . that is.T } ... Definition. T − 1} The process (Ut )t∈{0.. Proof.

. Consider the event {τ ∗ = t}.. and let U be its Snell envelope. . T }.T } be the process specifying the payout of an American ˜ option.T } be the Snell envelope of the discounted payout ξ with Doob decomposition Ut = Mt − At .. T }. Let Y be an adapted process indexed by {0. ... .. [and since (πt )t∈{1. ˜ Hence Xt = Mt for all t ∈ {0.. and we’re done..T } is a martingale for Q.... . .. Finally.) But letting τ ∗ = min{t ∈ {0... we have Ut = max{Yt . Definition. T } : At+1 > 0} where U = M − A is the Doob decomposition of U . However. 34 ...T } so that T x+ s=1 ˜ ˜ πs · (Ss − Ss−1 ) = MT .. τ ∈T By the complete market assumption. ... since ˜ ˜ ξt ≤ Ut ≤ Mt = Xt almost surely for all t ∈ {0... . recalling the deﬁnition of a Snell envelope. . . where t ˜ Xt = x + s=1 ˜ ˜ πs · (Ss − Ss−1 ). ..T } is bounded because we’re in a complete market and hence any Ft -measurable random variable can only take a ﬁnite ˜ number of values] the investor’s discounted wealth X is a martingale. . we can ﬁnd a predictable trading strategy (πt )t∈{1. . Then U0 = sup E(Yτ ). we have U0 = M0 = E(Mτ ∗ ) = E(Yτ ∗ ). On this event we have At = 0 and hence Mt = Ut .T } is predictable. let (ξt )t∈{0.. At+1 > 0 so that E(Ut+1 |Ft ) = E(Mt+1 −At+1 |Ft ) < Mt = Ut ... .Proof. (See example sheet 2. .t} super-replicates the payout of the American claim. Since U is a supermartingale. That τ ∗ is a stopping time follows from the fact that the non-decreasing process (At )t∈{0. Returning to ﬁnance. T } and the trading strategy (πt )t∈{1. . τ ∈T Proof. We now will use the assumption that the market is complete: consider an investor with initial discounted wealth ˜ x = U0 = M0 = sup EQ (ξτ ). and let (Ut )t∈{0.... Theorem. E(Ut+1 |Ft )} = Yt . . U0 ≥ E(Uτ ) by the optional sampling theorem. A stopping time τ such that E(Yτ ) = U0 is called an optimal stopping time.. ˜ Since (St )t∈{0.

In this section. What the above calculation shows is that the density of the restriction Q|Ft of the measure P to the sub-sigma-ﬁeld Ft ⊆ F with respect to P|Ft is the random variable Zt . 35 . To see where we’re going. it is sometimes more convenient not to explicitly mention the horizon. To conclude this chapter. suppose that the measure Q is equivalent to P. P) be a probability space equipped with a ﬁltration (Ft )t≥T . Furthermore. In particular. However. we will deal with ﬁnite horizon problems. F. he can super-replicate the payout of the American claim. If no ﬁnal time horizon T > 0 is explicitly mentioned. The measures P and Q are locally equivalent if and only if there exists a positive martingale Z with Z0 = 1 such that dQ|Ft = Zt . By the Radon–Nikodym theorem. The measure Q is locally equivalent to P if and only if the restriction Q|Ft of the measure Q to the sigma-ﬁeld Ft is equivalent to P|Ft for each t ∈ T. S)t∈Z+ . 4. local version). Notice that if A ∈ Ft then Q(A) = = = = EP (1A Z) EP [EP (Z 1A |Ft )] EP [1A EP (Z|Ft )] EP [1A Zt ]. S) = (B. Theorem (Radon–Nikodym. let (Ω. he cannot do so with a smaller endowment. we will brieﬂy discuss a weakening of the notion of equivalence of measures that is suitable for our purposes.We have shown that if the seller of the claim has x = U0 = M0 intial discounted wealth. dP|Ft We will call the martingale Z the density process for Q with respect to P. Now we are in a position to generalize the notion of equivalent measures. Locally equivalent measures In all the examples in this course. then we will extend the deﬁnition of equivalent ˜ martingale measure to include measures Q locally equivalent to P such that S = S/B is a Q-martingale. Definition. let us consider a ﬁnancial model (B. We can associate with any equivalent measure a positive martingale given by Zt = EP (Z|Ft ). as there exists a stopping time τ ∗ such that Xτ ∗ = ξτ ∗ . there exists a positive random variable Z = dQ such that Q(A) = dP EP (Z 1A ).

.

we have considered investors whose discounted wealth process is typically of the form t ˜ ˜ Xt = X 0 + s=1 ˜ ˜ πs · (Ss − Ss−1 ). Now recall that the ﬁrst fundamental theorem of asset pricing tells us. In this chapter. the sample paths of Brownian motion are not diﬀerentiable. Brownian motion In this section. Our goals will be to deﬁne a Brownian motion. 1. we would be able to deﬁne the stochastic integral 0 αs · dWs by 0 αs · dWs ds ds and the story would be over. we will see that essentially all continuous martingales M = (Mt )t∈R+ are of the form t Mt = M0 + 0 αs · dWs where W = (Wt )t∈R+ is a Brownian motion. So although the stochastic integral behaves in some ways like a Riemann–Stieltjes integral of calculus. We will see that the above stochastic integral can. As hinted above. be deﬁned. The following chapter will provide an extremely brief introduction to this theory. in fact. then there is an equivalent measure Q such that ˜ (St )t∈Z+ is a martingale. we consider the limit as trading frequency becomes more and more frequent.CHAPTER 3 Brownian motion and stochastic calculus In the lectures up to now. 37 . and to learn the rules of the resulting calculus. in the discretetime world. our primary interest in this process is that it will be the building block for all of the continuous-time market models studied in these lectures. If the sample paths of Brownian motion were t t diﬀerentiable. Brownian motion. and hence we need to understand the continuous time generalization t ˜ ˜ X t = X0 + 0 ˜ πs · dSs . As a preview of what’s to come. to construct the Brownian stochastic integral. we introduce one of the most fundamental continuous-time stochastic processes. and so we will have to do more work to make sense of the situation. Unfortunately. that if there is no-arbitrage. a word of warning is in order: The rules of stochastic calculus are not the same as those of ordinary calculus.

does there exist a probability space (Ω.5 2. t] with vanishing norm’ to mean a collection of points 0 = t0 ≤ t1 .0 2. the Brownian motion is also often called the Wiener process.5 0. F. ≤ tN = t 38 . It is not clear that Brownian motion exists..5 3.0 −0.Definition. especially in the U.0 −1. due to L´vy. That is. . we will use the phrase ‘a sequence of partitions (N ) (N ) (N ) of [0. they are very irregular. and the distribution of Wt − Ws is N (0. Although the sample paths of Brownian motion are continuous..0 1. A Brownian motion W = (Wt )t∈R+ is a collection of random variables such that • W0 (ω) = 0 for all ω ∈ Ω. • for all 0 ≤ t0 < t1 < . of course. Below is a computer simulation of a one-dimensional Brownian motion: Sample path of Brownian motion 3. P) on which the uncountable collection of random variables (Wt )t∈R+ can be simultaneously deﬁned in such a way that the above deﬁnition holds? The answer.5 0 1 2 3 4 5 t 6 7 8 9 10 The following is a very important result.0 0. and the proof of this fact is due to Wiener in 1930. < tn the increments Wti+1 − Wti are independent. • the sample path t → Wt (ω) is continuous all ω ∈ Ω. . In this chapter.S.5 −1.5 W 1. Therefore. |t − s|). is yes. which can quantify the irregularity e of the Brownian sample path.

.. Remark. Let W and W ⊥ be independent one-dimensional Brownian motions.. the increments of Brownian motion are Gaussian randoms so that E[(Wt − Ws )2 ] = t − s and Var[(Wt − Ws )2 ] = 2(t − s)2 for every 0 ≤ s ≤ t.t] (N ) tn−1 (N ) sn (N ) tn (t(N ) − tn−1 )2 → 0 n (N ) where < < by the mean value theorem. by the independence of the increments of Brownian motion. For every sequence of partitions of [0.N } max |t(N ) n − (N ) tn−1 | n=1 (t(N ) − tn−1 ) → 0. Hence N N E n=1 (Wtn ) − Wt(N ) ) (N n−1 2 = n=1 (t(N ) − tn−1 ) = t n (N ) and.such that maxn∈{1. t] with vanishing norm we have N (Wtn ) − Wt(N ) )2 → t (N n=1 n−1 in probability and N (Wtn ) − Wt(N ) )(Wt⊥ ) − Wt⊥ ) ) → 0 (N (N (N n=1 n−1 n n−1 in probability.. For comparison.. Proof.N } |tn (N ) − tn−1 | → 0 as N → ∞. we have N N [f (t(N ) ) n n=1 − (N ) f (tn−1 )]2 = n=1 f (s(N ) )2 (t(N ) − tn−1 )2 n n N 2 n=1 (N ) ≤ max f (s) s∈[0.. n (N ) Theorem. Useful properties of such sequences is N (N ) (t(N ) − tn−1 ) = t n n=1 (N ) and N N (t(N ) n n=1 − (N ) tn−1 )2 ≤ n∈{1. By deﬁnition. Then.. We can conclude that from the above theorem that a typical Brownian sample path is not a continuously diﬀerentiable function of time. N N Var n=1 (Wtn ) − Wt(N ) )2 = 2 (N n−1 (t(N ) − tn−1 )2 → 0 n n=1 (N ) and the ﬁrst conclusion follows from Chebychev’s inequality.. 39 . consider a continuously diﬀerentiable function f : R+ → R.

Itˆ stochastic integration o We now have suﬃcient motivation to construct a stochastic integral with respect to a Wiener process. there is an integration theory that does the job. For instance. and let (Ft )t∈R+ be its natural ﬁltration. The L2 theory. There are now plenty of places to turn for a proper treatment of the subject. Definition. Markov Processes. Theorem. For simple predictable integrands we deﬁne the stochastic integral by the formula ∞ N αs dWs = 0 n=1 an (Wtn − Wtn−1 ) 40 .tn ] (t)an (ω) where an is bounded and Ftn−1 -measurable for some 0 ≤ t0 < t1 < . Testing integrability is easy since Gaussian random variables have ﬁnite exponential moments.G. Karatzas and S. Of this ﬁltration we will assume that it satisﬁes what are called the usual conditions of right-continuity Ft = >0 Ft+ and that F0 contains all P-null events. A simple predictable integrand is an adapted process α = (αt )t∈R+ of the form N αt (ω) = n=1 1(tn−1 . Now ﬁx 0 ≤ s < t. < tN < ∞..E. Proof. Rogers and D.Since E[(Wt − Ws )(Wt⊥ − Ws⊥ )] = 0 and Var[(Wt − Ws )(Wt⊥ − Ws⊥ )] = (t − s)2 the second conclusion follows similarly. E(Wt |Fs ) = E(Ws + Wt − Ws |Fs ) = Ws + E(Wt − Ws ) = Ws Note that the fact that Wt − Ws is independent of Fs is used to pass from a conditional expectation to an unconditional expectation. What follows is the briefest of sketches of the theory. Let W be a scalar Brownian motion. Then W is a martingale. We also will assume that for each 0 ≤ s < t the increment Wt − Ws is independent of Fs . let W be a scalar Brownian motion. 2. The Brownian motion is too rough to apply the Riemann–Stieltjes integration theory. It is based on the fact that Brownian motion is a martingale. Fortunately. please consult one of the following references: • L.C. 2. Shreve.1. Williams. and Martingales: Volume 2 • I. The ﬁrst building block of the theory are the simple predictable integrands.. We will assume that W is adapted to a ﬁltration (Ft )t∈R+ . Diﬀusions. Brownian Motion and Stochastic Calculus. To get things started.

Equivalently. the predictable sigma-ﬁeld is that generated by the simple. Equivalently. predictable integrands. the map deﬁned by I(α) = 0 αs dWs is an isometry from the space of simple predictable integrands to the space L2 of square-integrable random variables. The fact that L2 is complete is the key observation which allows us to build the stochastic integral of more general integrands. predictable integrands. Applying the tower property.n am an (Wtm − Wtm−1 )(Wtn − Wtn−1 ) = E n an (tn − tn−1 ) n ∞ 2 αs ds 0 = E as desired. A predictable process α is a map α : R+ × Ω → R that is P-measurable. E[(Wtm − Wtm−1 )(Wtn − Wtn−1 )|Ftn−1 ] = (Wtm − Wtm−1 )E[Wtn − Wtn−1 ] = 0 and if m = n E[(Wtn − Wtn−1 )2 |Ftn−1 ] = E[(Wtn − Wtn−1 )2 ] = tn − tn−1 since the increment Wtn −Wtn−1 is independent of Ftn−1 . adapted process is predictable. If m < n. we see Eam an (Wtm − Wtm−1 )(Wtn − Wtn−1 ) = E[E(am an (Wtm − Wtm−1 )(Wtn − Wtn−1 )|Ftn−1 )] = E[am an E((Wtm − Wtm−1 )(Wtn − Wtn−1 )|Ftn−1 )] since am and an are Ftn−1 -measurable. Now.Theorem (Itˆ’s isometry). and we have E m. Every left-continuous.n am an (Wtm − Wtm−1 )(Wtn − Wtn−1 ) Consider the terms in the sum when m ≤ n. In particular. Definition. the oﬀ diagonal terms cancel. since they are the ones that come up most in application. ∞ 2 E 0 αs dWs =E m. we have o ∞ 2 ∞ E 0 αs dWs =E 0 2 αs ds Proof. Now. suppose (α(k) )k∈N is a sequence of simple predictable integrands converging to a predictable process α in the sense that ∞ ∞ E 0 (k) (αs − αs )2 ds →0 41 . Remark. These are the examples to keep in mind. For a simple predictable integrand α. The predictable sigma-ﬁeld P is the sigma-ﬁeld on the product space R+ ×Ω generated by sets of the form (s. t]×A where 0 ≤ s < t and A is Fs -measurable. predictable processes are limits of simple.

If α is predictable and ∞ E 0 2 αs ds <∞ ∞ then 0 ∞ αs dWs = L2 − lim k 0 (k) αs dWs where α (k) is any sequence of simple. 2. This is what we take as the deﬁnition. By Itˆ’s isometry the sequence I(α(k) ) is a Cauchy sequence. In this section. 42 . and let αt (n) = αt 1{t≤τn } . Definition.2. Definition. Deﬁne the stopping times t τn = inf t ≥ 0 : 0 2 αs ds = n for each n ∈ N. 0 t ∞ αs dWs = 0 αs 1{0<s≤t} dWs whenever the right-hand side is well-deﬁned. ∞) but rather ﬁnite intervals [0. we are not really interested in integrals over the whole interval [0. which by the como 2 pleteness of L . about a process deﬁned by a stochastic integral? Theorem. Localization. This is easily handled. t]. converges to some random variable. we show how to extend the deﬁnition of stochastic integral to predictable processes α such that t 2 αs ds < ∞ almost surely 0 for all t ≥ 0. Let Mt = 0 t αs dWs where α is predictable and t E 0 2 αs ds <∞ for each t ≥ 0. where inf ∅ = ∞ as usual. predictable integrands converging in L2 to α. What can we say. Then M is a continuous martingale.as k → ∞. within the L2 -theory. But please note: The stochastic integral is not deﬁned as the almost sure limit of a sequence of Riemann–Stieltjes integrals!! Of course. The technique is called localization.

and proces I is another continuous local martingale. but in general t∈R+ there is no guarantee that L a martingale. Now ﬁx t > 0 and deﬁne the increasing sequence of events An = {ω ∈ Ω : τn ≥ t}. If you’re ambitious. An adapted process (Lt )t∈R+ is called a local martingale if and only if there exists an increasing sequence of stopping times (τn )n∈N with τn → ∞ almost surely such that the stopped process (Lt∧τn )t∈R+ is a martingale for all n ∈ N.Note that since E t (n) (αs )2 ds 0 ≤ n. t The process L = 0 αs dWs deﬁned in this way is still continuous. Since t 2 α ds < ∞ almost surely for all t ≥ 0. What makes the Itˆ stochastic integral useful is that there is a corresponding o stochastic calculus. called Itˆ’s o formula. rather than Brownian motion. so you can assume L is a square-integrable martingale. The steps are the same. M is a 2 continuous local martingale. we sketched the constructed of a stochastic integral with respect to a Wiener process. and ﬁnally. Indeed. but Itˆ’s isometry becomes o ∞ 2 ∞ E 0 αs dLs N =E 0 2 αs d L s where L t = lim N →∞ (Lt(N ) − Lt(N ) )2 . you will now try to build a stochastic integration theory starting with a general continuous local martingale L. Hence we can deﬁne 0 s the stochastic integral by the formula t t αs dWs = lim 0 n→∞ (n) αs dWs 0 where the limit is in probability. Then. you can do the L2 theory as before. Itˆ’s formula o In the last section. extend the integrands by localization.. First you localize L. the stochastic integral deﬁned by the localized version of the L2 stochastic integration theory may not be a martingale. If in addition we have E 0 αs ds < ∞ then M is a true martingale (as opposed to a being strictly local martingale). left-continuous process such that 0 αs ds < ∞ almost t surely for each t ≥ 0 then Mt = 0 αs dWs is deﬁned.s. In all cases. Once you’re done. n n=1 n−1 Then you can deﬁne the integral over ﬁnite intervals. we have P n∈N An = 1. To summarize: t 2 If α is an adapated. but by construction it is what is called a local martingale: Definition. the process t 0 αs dWs (n) is well-deﬁned by the L2 t∈R+ theory. 43 . The basic building block of this calculus is the chain rule. t Remark. ∞ a. you will have built a stochastic integration theory that has t t 2 the very nice property that the integral It = 0 αs dLs is deﬁned when 0 αs d L s . 3.

the ﬁrst as a stochastic integral and the second as a pathwise Lebesgue integral. If (Yt )t∈R+ is another semimartingale. if it exists. Y = ( X. o We are now ready for the ﬁrst version of Itˆ’s formula: o Theorem (Itˆ’s formula. This term would not appear in the chain rule of o ordinary calculus. is deﬁned by the limit in probability N X t = lim N →∞ (Xt(N ) − Xt(N ) )2 n n=1 n−1 over a sequence of partitions with vanishing norm. then the quadratic covariation process X. an Itˆ o process is a semimartingale. Fix a real number X0 and let t t X t = X0 + 0 αs dWs + 0 ks ds. (For instance. Both integrals appearing the above equation now be interpreted.1. t 0 Ws2 ds = t2 /2 < ∞. Let (αt )t∈R+ and (kt )t∈R+ be predictable real-valued processes such that t 2 αs ds < ∞ and 0 0 t |ks |ds < ∞ almost surely for all t ≥ 0. A process (Xt )t∈R+ of the above form is often called an Itˆ process.) For a semimartingale X. directly from the deﬁnition of Brownian motion. the quadratic variation process X = ( X t )t∈R+ . A semimartingale is a process X of the form Xt = At + Lt where L is a local martingale and A is a process of bounded variation. Then t t 1 2 f (Xt ) = f (X0 ) + f (Xs )αs dWs + (f (Xs )ks + f (Xs )αs )ds. that the process (Wt2 − t)t∈R+ is a martingale? We now introduce some notions which helps with computations involving Itˆ’s formula. o Definition. Let (Wt )t∈R+ be a scalar Brownian motion adapted to a ﬁltration (Ft )t≥0 satisfying our usual conditions. Let f : R → R be twice continuously diﬀereno tiable.3. n n n=1 n−1 n−1 44 . 2 0 0 Let us highlight a diﬀerence between Itˆ and ordinary calculus. But consider the example f (x) = x2 so that t Wt2 Note that since E t 0 =2 0 Ws dWs + t. Can you verify. The scalar version. Y t )t∈R+ is deﬁned by the limit N X. scalar version). the stochastic integral Ws dWs is a true t∈R+ martingale. Y t = lim N →∞ (Xt(N ) − Xt(N ) )(Yt(N ) − Yt(N ) ). by noting the mysterious o appearance of the f term in Itˆ’s formula.

then the quadratic covariation can be X. and the following table summarizes its possible values. it is customary to write out Itˆ’s formula as o 1 df (Xt ) = f (Xt )dXt + f (Xt )d X 2 45 t . Let o dXt = αt dWt + kt dt where the diﬀerential notation is shorthand for the corresponding integrals. 0 0 bs ds t = 0 0 as ds. Y ) → X. symbol d X t means dX N t 2 = αt dt. where (at )t∈R+ and (bt )t∈R+ are processes such that the relevant integrals are deﬁned. The map (X. Notice that in this diﬀerential notation. Y t = 0 (1) (1) (2) (2) (αs βs + αs βs ) ds.Theorem. 0 0 bs dWs t = 0 as bs ds 0 as dWs . suppose we have two Itˆ processes o t t (1) αs dWs(1) + 0 t 0 t (1) βs dWs(1) + 0 0 (2) (2) βs dWs(2) + 0 (2) αs dWs(2) + 0 t t Xt = X0 + Yt = Y0 + hs ds ks ds for independent Brownian motions W (1) and W computed by bilinearity: t . we are ready to see an indication for why Itˆ’s formula may be true. t] and consider the following second order Taylor approximation: f (Xt ) − f (X0 ) = n=1 N f (Xtn ) − f (Xtn−1 ) 1 f (Xtn−1 )(Xtn − Xtn−1 ) + f (Xtn−1 )(Xtn − Xtn−1 )2 2 n=1 t t ≈ ≈ f (Xs )dXs + 0 0 1 f (Xs )d X 2 s In fact. Fix a partion of [0. and where (Wt )t∈R+ and (Wt⊥ )t∈R+ are independent Brownian motions. 0 bs dWs⊥ t = 0 Example. 0 bs dWs t = 0 · · t · · as dWs . · · · · as ds. For instance. Armed with this the notion of quadratic variation. Y is bilinear.

Let (Wt )t≥0 be a d-dimensional Wiener process.2. Then g (y) = y and g (y) = − y12 . The vector version. Letting Yt = eXt . and let (kt )t≥0 be an adapted process valued in Rn . let’s check. Consider the Itˆ process given by o Xt = X0 + aWt + bt for some constants a. we would like to show that the process (Yt )t∈R+ is an Itˆ process. t = α2 Yt2 dt. where a = α and b = β − α2 /2. and also dY Again. we know from the previous example that Zt = Z0 + aWt + bt.where the diﬀerential notation should really be intrepreted as the corresponding integral equation. Let g(y) = log(y). and write down its decomo position in terms of ordinary and stochastic integrals. We insist that t 0 n d (i. Let f (x) = ex . Suppose we have the Itˆ process given implicitly as the solution of the stoo chastic diﬀerential equation dYt = Yt [βdt + α dWt ] for some constants α. We now introduce the vector version of Itˆ’s formula. Example. 1 But to be safe. Letting Zt = log(Yt ). Then f (x) = ex and f (x) = ex . but with worse notation. β ∈ R. let (αt )t≥0 be an adapted process valued in the space of n × d matrices. dXt = a dWt + b dt and d X So Itˆ’s formula says: o 1 df (Xt ) = f (Xt )dXt + f (Xt )d X 2 ⇒ dYt = Yt [(b + a2 /2)dt + a dWt ] t t = a2 dt Example. It is o basically the same as before.j) (αs )2 ds i=1 j=1 t n (i) |ks |ds < ∞ 0 i=1 < ∞ and 46 . Itˆ’s formula says: o 1 dg(Yt ) = g (Yt ) dYt + g (Yt ) d Y t 2 1 1 1 ⇒ dZt = Yt [βdt + α dWt ] + − 2 Yt 2 Yt 2 = (β − α /2) dt + α dWt . α2 Yt2 dt 3. Also. b ∈ R.

X (j) ∂xi ∂xj t 4. as it is a stochastic integral with respect to a Brownian motion. Now consider the n-dimensional Itˆ process (Xt )t≥0 deﬁned by o t t X t = X0 + 0 αs dWs + 0 ks ds. P) be our probability space equipped with a ﬁltration (Ft )t∈R+ .j) αs dWs(j) + 0 t (i) ks ds. Xt ) d X (i) . Indeed. we aim to understand how martingales arise within the context of the Itˆ stochastic integration theory. by the Radon–Nikodym theorem there exists a density dQ|t Zt = dP|t such that (Zt )t∈R+ is a strictly positive martingale on the probability space (Ω. Then. This process is clearly positive. Furthermore. x) → f (t. Now we are ready for the statement of the theorem: Theorem (Itˆ’s formula. Xt )dt + ∂t n i=1 ∂f 1 (i) (t. Motivated by above discussion. the economic notion of an arbitrage-free market model is tied to the existence of a locally equivalent measure for which the discounted asset prices are martingales. P).almost surely for all t ≥ 0. Recall that locally equivalent measures are related to positive martingales via the Radon– Nikodym theorem. F. and let Q be locally equivalent to P in the sense that the restrictions P|t and Q|t of P and Q to Ft are equivalent for each t ≥ 0. Xt ) dXt + ∂xi 2 n n i=1 j=1 ∂ 2f (t. Then ∂f df (t. Conversely. and hence the density of a change of measure? What happens to the Brownian motion? 47 . Let f : R+ × Rn → R where (t. F. Girsanov’s theorem As we have seen in discrete time. x) o be continuously diﬀerentiable in the t variable and twice-continuously diﬀerentiable in the x variable. Xt ) = (t. notice that by Itˆ’s formula we have o dZt = Zt αt · dWt so that (Zt )t∈R+ is a local martingale. let (Ω. Consider the stochastic process (Zt )t∈R+ o given by t 1 t 2 Zt = e− 2 0 |αs | ds+ 0 αs ·dWs where (Wt )t∈R+ is a d-dimensional Brownian motion and (αt )t∈R+ is a d-dimensional adapted t process with 0 |αs |2 ds < ∞. if (Zt )t∈R+ is a positive martingale we can deﬁne a locally equivalent measure Q. vector version). What if (Zt )t∈R+ is a true martingale. interpreted component-wise as Xt = X0 + 0 j=1 (i) (i) t d (i.

Let L = (Lt )t∈R+ be a local martingale. Martingale representation theorems In the introduction to this chapter. Let Zt = e− 2 1 t 0 |αs |2 ds+ t 0 αs ·dWs and suppose (Zt )t∈R+ is a martingale. we can say a lot about continuous local martingales: Theorem (L´vy’s Characterization of Brownian Motion). Even when it’s not. it was claimed that all continuous martingales are essentially stochastic integrals with respect to Brownian motion.Theorem (Cameron–Martin–Girsanov Theorem). Q). F. Then there exists a unique adapted d-dimensional t process (αt )t∈R+ such that 0 |αs |2 ds < ∞ almost surely for all t ≥ 0 and t Lt = L0 + 0 αs · dWs . . F. |αs |2 ds+ t 0 αs ·dWs is a true martin- 5. Let (Ω. Now. L is continuous. and let (Ft )t∈R+ be a ﬁltration satisfying the usual conditions. dP|t ˆ Then the d-dimensional process (Wt )t∈R+ deﬁned by t ˆ Wt = Wt − 0 αs ds is a Brownian motion on (Ω. we hopefully clear up the situation. but a true martingale? An easy-to-check suﬃcient condition is given by: Theorem (Novikov’s criterion). you may be asking yourself: When is the process (Zt )t∈R+ not just a local martingale. Let (Ω. L(j) t = t 0 48 if i = j if i = j. P) be a probability space on which a d-dimensional Brownian motion (Wt )t∈R+ is deﬁned. In particular. F) by the density process dQ|t = Zt . and let the ﬁltration (Ft )t∈R+ be the ﬁltration generated by W . Up to now. we have assumed that the ﬁltration is generated by a Brownian motion. Then (Lt )t∈R+ is a standard d-dimensional Brownian motion. Let (Lt )t∈R+ be a continuous e d-dimensional local martingale such that L(i) . Theorem (Martingale Representation Theorem). P) be a probability space on which a d-dimensional Brownian motion W = (Wt )t∈R+ is deﬁned. F. If E e2 1 t 0 |αs |2 ds <∞ 1 t 0 for all t ≥ 0 then the process (Zt )t∈R+ deﬁned by Zt = e− 2 gale. In this section. Deﬁne the locally equivalent measure Q on (Ω.

t Wt = 0 dLss . we can attempt e a proof of Girsanov’s theorem: Proof of Girsanov’s theorem. L(n) 2 2 m=1 n=1 d d t and so (Mt )t∈R+ is a continuous local martingale. On the other hand. Let t ˆ Wt = Wt − 0 αs ds. Let (Lt )t∈R+ be a continuous real-valued local martingale. Consider . Thus for all 0 ≤ s ≤ t we have E(Mt |Fs ) = Ms which implies E(ei θ·(Lt −Ls ) |Fs ) = e−|θ| 2 (t−s)/2 . (Wt )t∈R+ is a Brownian motion such that Lt = L0 + 0 αs dWs for all t ≥ 0. and suppose (Zt )t∈R+ is a true martingale. Hence. and suppose there t 2 is an adapted process (αt )t∈R+ such that L t = 0 αs ds where αt > 0 almost surely. Let Zt = e− 2 1 t 0 |αs |2 ds+ t 0 αs ·dWs for a d-dimensional adapted process (αt )t∈R+ and a d-dimensional Brownian motion (Wt )t∈R+ . As another application of L´vy’s characterization of Brownian motion. o dMt = Mt i θ · dLt + = i Mt θ · dLt √ −1. since |Mt | = e|θ| t/2 and hence E(sups∈[0. 49 . Fix a constant vector θ ∈ Rd and let i = Mt = ei By Itˆ’s formula. (t − s)I) distribution and is independent of Fs . Note that (Wt )t∈R+ is a Brownian motion by L´vy’s characterization since e α t W t = 0 dL 2 αs t s = t.t] |Ms |) < ∞ the process (Mt )t∈R+ is a true martingale. θ·Lt +|θ|2 t/2 1 |θ|2 dt − Mt θ(m) θ(n) d L(m) .Proof. The above equation implies that the increment Lt − Ls has the Nd (0. as it is the stochastic integral with re2 spect to a continuous local martingale. Example.

ˆ Note that (Zt Wt )t∈R+ is a local martingale for P, since by Itˆ’s formula: o ˆ (i) ˆ (i) ˆ (i) ˆ d(Zt Wt ) = Wt dZt + Zt dWt + d Z, W (i) (i) ˆ (i) = Zt Wt αt · dWt + Zt dWt ˆ It now follows that (Wt )t∈R+ is a local martingale for Q, where ˆ ˆ W (i) , W (j)

t dQ|t dP|t t

= Zt . But since

=

t 0

if i = j if i = j.

ˆ then (Wt )t∈R+ is a Brownian motion of for Q.

50

CHAPTER 4

**Black–Scholes model and generalizations
**

We now return to the main theme of these lecture, models of ﬁnancial markets. We now have the tools to discuss the continuous time case, at least when the asset prices are continuous processes. 1. The set-up As before, our market model consists of a d+1-dimensional stochastic processes (B, S) = (Bt , St )t∈R+ representing the asset prices. This process will be deﬁned on a probability space (Ω, F, P) with a ﬁltration (Ft )t∈R+ satisfying the usual conditions. We will make the following now-familiar assumptions. Assumption. The stochastic process (B, S) is assumed to be is an Itˆ process adapted o to (Ft )t∈R+ . Assumption. Asset 0 is a num´raire so that Bt > 0 almost surely for all t ∈ R+ . e As before, the investor’s controls consist of the d + 1-dimensional process (φt , πt )t∈R+ , (i) where φt and corresponds to the number of shares of asset 0 held at time t, while πt corresponds to the number of shares of asset i ∈ {1, . . . , d} held at time t. We will use the d 1 notation πt = (πt , . . . , πt ). The wealth Xt at time t then satisﬁes the pair of equations Xt = φt Bt + πt · St dXt = φt dBt + πt · dSt the budget constraint the self-ﬁnancing condition

As usual, we can use the budget contraint to solve for φt , and plug this expression into the self-ﬁnancing condition to get dXt = (Xt − πt · St ) By Itˆ’s formula, we have o d Xt Bt = Xt −

d

**dBt + πt · dSt Bt dXt d X, B − 2 Bt Bt
**

t

dBt d B + 2 3 Bt Bt St St Bt

(i)

t

+

t

=

i=1

πt

(i)

−

dBt d B + 2 3 Bt Bt

+

dSt d S (i) , B − 2 Bt Bt

t

= πt · d

.

51

**As usual, we express everything in terms of asset 0 prices Xt St ˜ ˜ and Xt = St = Bt Bt so that
**

t

˜ ˜ Xt = X0 +

0

˜ πs · dSt

2. Admissible strategies In order to make sense of the stochastic integral deﬁning the discounted wealth, we need to impose the integrablity condition

t 0 d i ˜ (πs )2 d S i i=1 s

<∞

almost surely for all t ≥ 0. However, in moving from discrete to continuous time, we have to be careful. We will now see that this condition isn’t strong enough to make our economic analysis interesting. Example. (A doubling strategy.) This example is intended to provide motivation for restricting the class of trading strategies that we will consider in these lectures. The problem with continuous time is that events that will happen eventually can be made to happen at any ﬁnite time by speeding up the clock. In particular, we will now construct T 2 a real-valued adapted process (αt )t∈[0,T ] such that 0 αs ds < ∞ almost surely, but

T

αs dWs = K

0

almost surely, where (Wt )t∈[0,T ] is a standard scalar Brownian motion, and T > 0 and K > 0 are real constants. Let f : [0, T ) → R+ be a strictly increasing, continuous function such that f (0) = 0 and limt→T f (t) = ∞. Note in particular that we assume that f (t) > 0 for t ∈ [0, T ) and there exists an inverse function f −1 : R → [0, T ) such that f ◦ f −1 (u) = u. For instance, to be t Tu explicit, we may take f (t) = T −t and f −1 (u) = 1+u . Now deﬁne a process (Zu )u∈R+ by

f −1 (u)

Zu =

0

(f (s))1/2 dWs

**Note that the quadratic variation is
**

f −1 (u)

Z

u

=

0

f (s)ds

**= f (f −1 (u)) − f (0) = u so by L´vy’s characterization (Zu )u∈R+ is a Brownian motion. Deﬁne the stopping time τ by e τ = inf{u ≥ 0, Zu = K}.
**

52

we usually insist that the investor cannot go arbitrarily far into debt. employing a doubling strategy (with borrowed money) at a quicker and quicker pace. we usually do not impose this extra integrability condition because in ﬁnance we are often computing expectations under an equivalent measure. Indeed. Hence u≥0 1To (n) √ nZu/n . In this case. The strange fact is that (Mt )t∈[0. if such strategies were a good model for investor behaviour. However. S) be a market model. This situation is rather unrealistic.Since (Zu )u∈R+ is a Brownian motion.1 Now let αs = (f (s))1/2 1{s≤f −1 (τ )} and Mt = 0 t αs dWs Note that since T 2 αs ds 0 f −1 (τ ) = 0 f (s)ds = τ < ∞ the stochastic integral is well-deﬁned. we all could be much richer by just spending some time trading over the internet. the stochastic integrals against Brownian motions would be martingales. 53 . We see that integrand (αs )s∈[0. we have τ < ∞ almost surely. u≥0 u≥0 (n) P(sup Zu > K) = P(sup u≥0 But it is easy to see that P(supu≥0 Zu > 0) = 1. The above discussion shows that the natural integrability condition t 0 d i ˜ (πs )2 d S i i=1 s < ∞ almost surely is not really suﬃcient for our needs. (Can you ﬁnd an example?) So instead. Let (B.T ] corresponds to an gambler starting at noon with £0. One can check that that the proces (Zu )u∈R+ is also a standard √ √ nZu/n > K) = P(sup Zu > K/ n) → P(sup Zu > 0). and in particular. we would be able to use the L2 theory. Definition. We might choose to impose the stronger condition t d i ˜ (πs )2 d S (i) 0 i=1 s E <∞ for all t ≥ 0. but MT = Zτ = K almost surely. it doesn’t necessarily follow that EQ (Y ) < ∞. even if P and Q are equivalent and Y is a positive random variable such that EP (Y ) < ∞.T ] is a local martingale with M0 = 0. That is. A strategy (πt )t∈R+ is admissible if and only if there is a constant C > 0 such that t ˜ πs · dSs > −C almost surely 0 see why. until ﬁnally he gains £K almost surely before the clock strikes one o’clock. particularly since the gambler must go arbitrarily far into debt in order to secure the £K winning. let Zu = Brownian motion.

˜ for all t ≥ 0, where St =

St . Bt

Note that the doubling strategy is not admissible, since the investor now has only a ﬁnite credit line. However, a suicide strategy, that is, a doubling strategy in which the object is to lose a ﬁxed amount K by time T , is admissible. 3. Arbitrage and equivalent martingale measures To see that our restriction to admissible strategies is reasonable, let’s now consider continuous-time arbitrage theory. Definition. An admissible strategy (πt )t∈R+ is an arbitrage if there is a (non-random) time T > 0 such that

T

P

0 T

˜ πs · dSs ≥ 0 ˜ πs · dSs > 0

0

= 1 > 0.

P

˜ ˜ Definition. A probability measure Q, locally equivalent to P, such that S = (S)t∈R+ is a local martingale is called a equivalent martingale measure. The following theorem will serve the role of the ﬁrst fundamental theorem of asset pricing in continuous time. Theorem. If there exists an equivalent martingale measure, then the market model has no arbitrage. Remark. Note that the theorem doesn’t say that no-arbitrage implies the existence of an equivalent martingale measure. Indeed, our notion of arbitrage is too strong. The ‘correct’ notion of arbitrage is called ‘free-lunch-with-vanishing-risk,’ but it is outside the scope of these lectures. See the recent book of Delbaen and Schachermayer The Mathematics of Arbitrage for an account of the modern theory. Proof. Suppose (πt )t∈R+ is admissible and let

t

˜ Xt =

0

˜ πs · dSs .

˜ By the deﬁnition, there exists a constant C > 0 such that Xt > −C P− almost surely. ˜ t > −C Q− almost surely. But under Q, the Since P and Q are locally equivalent, then X ˜ process (St )t∈R+ is a local martingale. But a local martingale that is bounded from below is necessarily a supemartingale. (See Problem 3.5) In particular, we have the following inequality ˜ ˜ EQ (Xt ) ≤ X0 = 0. for all t ≥ 0. ˜ ˜ Now suppose that there is a T > 0 such that XT ≥ 0 P-almost surely. Then XT ≥ 0 Q˜ ˜ almost surely. But since EQ (XT ) ≤ 0, we conclude that XT = 0 Q-almost surely, which ˜ implies XT = 0 P-almost surely.

54

4. Contingent claims and market completeness As before, given a market model (B, S), we can introduce a contingent claim. Recall that a European contingent claim maturing at a time T > 0 is modelled as random variable ξ that is FT -measurable. Let Q be the set of equivalent martingale measures, and which we shall assume is not empty. Hence, there is no arbitrage in the market. Theorem. Let (ξt )t∈[0,T ] be a process such that ξT = ξ. The augmented market (Bt , St , ξt )t∈[0,T ] is free of arbitrage if there exists an equivalent martingale measure Q ∈ Q such that Bt ξt = EQ ξ Ft BT for all t ∈ [0, T ]. ˜ ˜ Proof. There is no arbitrage if there exists an equivalent measure Q such that (St , ξt )t∈[0,T ] ξt ˜ ˜ is a local martingale for Q, where ξt = Bt . But (ξt )t∈[0,T ] is a martingale by construction. And just as before, we are interested in replicating contingent claims by trading in the market. Definition. A (European) contingent claim with payout ξT maturing at time T > 0 is attainable if and only if there exists a constant x ∈ R and an admissible strategy (π)t∈[0,T ] such that T ˜ ˜ πs · dSs ξT = x +

0 ξ ˜ where ξT = BT . T ˜ A market is complete if for every bounded FT -measurable random variable ξT , there exists x, π such that t

˜ Xt = x +

0

˜ πs · dSs

˜ deﬁnes a bounded discounted wealth process X such that XT = ξT . The following is a version of the second fundamental theorem of asset pricing for this continuous time setting. Theorem. Suppose that there exists an equivalent martingale measure Q and the market is complete. Then Q is the unique equivalent martingale measure. ˜ Proof. Let Q be another equivalent martingale measures and let ξT be an arbitrary bounded FT -measurable random variable. By assumption ξT is attainable by a wealth process ˜ X, and since X is a bounded, it is a martingale for both Q and Q . ˜ ˜ EQ (ξT ) = EQ (XT ) ˜ = x = EQ (ξT ) ˜ ˜ Hence, for all bounded random variables ξT , the inequality EP [(ZT −ZT )ξT ] ≥ 0 holds, where dQT dQT ˜ ZT = dPT and ZT = dPT . Letting ξ = 1{ZT <ZT } shows that P(ZT ≥ ZT ) = 1. Symmetry completes the argument shows ZT = ZT a.s. At this stage we’re a bit too general to say anything interesting. Hence, it is wise to introduce more notation to ﬂesh out the story...

55

5. The set-up revisited In this section, we revisit the set-up of the continuous time market model (B, S). We will make one extra assumption, which is standard, but not really necessary: Assumption. Asset 0 is risk-free in the sense that B

t

= 0 almost surely.

The point of this section is to set notation that will be used for the remaining lectures, and to interpret the existence of an equivalent martingale measure via Girsanov’s theorem. Asset 0, which we now assume is risk-free, will be intepreted as a bank account. Its dynamics are given by dBt = rt Bt dt for an adapted process (rt )t≥0 which can be interpreted as the spot interest rate. The above random ordinary diﬀerential equation has the solution Bt = B0 e

t 0 rs ds

.

**Assets 1, . . . , d are assumed to have dynamics given by
**

n (i) dSt

=

(i) St

(i) µt

dt +

j=1

σt

(i,j)

dWt

(j)

(i) for some adapted process (µt )t∈R+ (j) (i) (Wt )t∈R+ . The random variable µt

**and and independent Brownian motions can be considered the instantaneous drift of the return
**

n (i,j) (σt )2 j=1 1/2

(i,j) (σt )t∈R+ ,

on stock i, while the random variable

is its instantaneous volatility. In vector notation, these equations can be written as dSt = diag(St )(µt dt + σt dWt ) where St = (St , . . . , St ), and the Rd -valued random variables µt and Wt , and d × n matrix-valued random σt deﬁned similarly. Here we are using the notation s1 0 · · · 0 ... 0 0 s diag(s1 , . . . , sd ) = . . 2 . .. .. . . . . . 0 0 · · · sd The dynamics of the discounted stock price are given by ˜ ˜ dSt = diag(St )((µt − rt 1)dt + σt dWt ) where 1 = (1, . . . , 1) ∈ Rd . Now we come to the two theorems of this section: The ﬁrst is a reasonably easy-to-check suﬃcient condition to ensure no-arbitrage.

56

(1) (d)

The following theorem is a suﬃcient condition for completeness: Theorem. and let ˜ ˜ ξt = EQ (ξT |Ft ). Notice then that the dynamics of the discounted stock price are given by ˜ ˜ ˆ dSt = diag(St )σt dWt . we see that t ˜ ξt = x + 0 ˜ πs · dSs . then the market is complete. so that λt = σt (µt −rt 1). Proof. and further suppose −1 ˆ n = d and σt (ω) is invertible for all (t. Take as given the hypotheses of the previous theorem. deﬁnes a martingale (for instance if Novikov’s criterion t 0 |λs |2 ds )<∞ holds for all t ≥ 0). If the process λ is such that Zt = e− 2 EP (e 2 1 1 t 0 |λs |2 ds− t 0 λs ·dWs . ˜ so that S is a local martingale for Q.Theorem.T ] such that t ˜ ˜ ξt = EQ (ξT ) + 0 ˆ αs · dWs .T ] is a bounded Q-martingale. ˜ ˆ so that (ξt )t∈[0. ˜ Fix a bounded FT -measurable random variable ξT . 57 . Let dWt = dWt +λt dt. the martingale representation theorem asserts the existence an adapted process (αt )t∈[0. The n-dimensional random vector λt is a generalization of the Sharpe ratio. then the market has no arbitrage. The process λ = (λt )t∈R+ is often called the market price of risk. Hence Q is an equivalent martingale measure and there is no arbitrage in this market. Let Q be the equivalent martingale measure in the proof of the previous theorem. we can write the discounted stock dynamics as ˜ ˜ ˆ dSt = diag(St )σt dWt . On the other hand. the locally equivalent measure Q whose density process ˆ is Z is such that the process W given by t ˆ Wt = Wt + 0 λs ds is a Brownian motion for Q. since it measures in some sense the excess return of the stocks per unit of volatility. By Girsanov’s theorem. T ˜ By taking x = EQ (ξT ) and πt = diag(St )−1 (σt )−1 αt . Proof. ω). Remark. Suppose there exists a predictable process λ such that σt λt = µt − rt 1 almost surely for all t ≥ 0. ˆ If the process W generates the ﬁltration. and generates the ﬁltration. Since W is a Brownian motion under Q.

com Furthermore. Graph of the Standard & Poor’s 500 stock index 1950-2008. Markovian markets Now that we have our two main structural theorems in the context of a market with continuous asset prices. 6. but we do not yet know how to actually compute it. This problem is the subject of the next section. A good model should give a reasonable statistical ﬁt to the actual market data. Figure 1. this is not theorem.and hence the claim with payout ξT = XT is attainable. exactly one solution if n = d. We now have a suﬃcient condition that the market model is complete. the market is complete.yahoo. and many solutions if n > d. and the discounted wealth process ˜ ˜ X = ξ is bounded. one expects from the rules of linear algebra for there to be no solution if n < d. See ﬁgure 1 below. a useful model is one in which the prices and hedges of contingent claims can be computed reasonably easily. Data taken from http://finance. Financially. we are still left with the question: How do you price and hedge contingent claims? The ﬁrst step is to pose a model for the asset prices (Bt . the rule of thumb becomes: n < d ‘⇒’ n = d ‘⇒’ n > d ‘⇒’ The market has arbitrage. at this stage we can only assert the existence of a replicating strategy for a given claim. If we consider the equation σt λt = µt − rt 1 where σt is an d × n matrix. However. The market has no arbitrage and is incomplete. Of course. In this section. Hence. St )t∈R+ . just a rule of thumb. The market has no arbitrage and is complete. we will study models in which the asset 58 .

St ) then there is no arbitrage in the augmented market (Bt . St (ω)). T ) × Rd . Solving this equation may be diﬃcult to do by hand. d (t. µ : R+ ×Rd → Rd and σ : R+ ×Rd → Rd×n are given. but it can usually be done by computer if the dimension d is reasonably small. Theorem. And most importantly for the banker selling such a contingent claim: the replicating portfolio πt can be calculated as the 2sometimes called the Feynman–Kac PDE. . and where all functions in the PDE are evaluated at the same point (t. µt (ω) = µ(t. ξt )t∈[0. St (ω)). S0 ) and the replicating strategy ∂V ∂V πt = grad V (t. St . though there seems to be some controversy over how well they ﬁt actual market data. Furthermore. as now rt (ω) = r(t. Notice that this is a special case of the set-up of the last section.prices are Markov processes. the PDE reduces to the (backward) Kolmogorov 59 equation. S) = g(S) where a = σσ T . and σt (ω) = σ(t. St )dt + σ(t. St ).j S S ∂S i ∂S j j=1 i j d V (T. . then the claim with payout ξ = g(ST ) is attainable with initial capital ξ0 = V (0. St )dWt ) where the nonrandom functions r : R+ ×Rd → R. St ) dt dSt = diag(St )(µ(t. . St (ω)). S) ∈ [0. The above theorem says that if the market model is Markovian. .T ] . If ξt = V (t. the price of a claim contingent on the future risky asset prices can be written as a deterministic function V of the current market prices. Assume there exists an equivalent martingale measure. Now suppose that the d + 1 assets have Itˆ dynamics which can be expressed as o dBt = Bt r(t. 1 ∂S ∂S Remark. . St ) = (t. If r = 0. T ] × Rd → R satiﬁes the partial diﬀerential equation ∂V + ∂t d i=1 1 ∂V rS + ∂S i 2 i d i=1 ∂ 2V = rV ai. Furthermore. These models are useful in the above sense. if V is non-negative. the asset prices (St )t∈R+ are a d-dimensional Markov process. In this special situation. St ) . The next theorem says how to ﬁnd the no-arbitrage price and the replicating strategy for a contingent claim maturing at time T with payout ξ = g(ST ) for some non-random function g : Rd → R. the pricing function V can be found by solving a certain linear parabolic partial diﬀerential equation2 with terminal data to match the payout of the claim. Suppose the function V : [0.

gradient of the pricing function V with respect to the spatial variables. we will consider the simplest possible Markovian model of the type studied in section 6. ˜ Hence (ξt )t∈[0. we have o ˜ dξt = d 1 = Bt 1 = Bt + 1 Bt V (t. St ) dBt 2 Bt dt Sti i=1 ∂V 1 + i ∂S 2 σ ik σ jk Sti Stj i=1 j=1 k=1 ∂ 2V − rV ∂S i ∂S j i=1 j=1 ∂V i ij S σ dWtj ∂S i n d = i=1 ∂V S i (µi − r)dt + ∂S i Bt σ ij dWtj j=1 ˜ = grad V · dSt .e. As we’ve seen. We are interested in pricing and hedging a European contingent claim with payout ξT = g(ST ). We’ve seen before that since λ = (µ − r)/σ is constant. ξ can be attained by the announced admissible trading strategy πt = grad V (t. This is often called the Black–Scholes model. the model is complete. the market has no arbitrage. there are two ways of doing this. 7. if V is non-negative. We will assume that all coeﬃcients are constant. PDE. there exists a locally equivalent ˆ measure Q such that the process deﬁned by Wt = Wt +λt is a Brownian motion. µ. so the price dynamics are given by the pair of equations dBt = Bt r dt dSt = St (µ dt + σdWt ) for real constants r. Furthermore. i. ˜ Proof. a measure under which S is a local martingale. σ where σ > 0. St ) Bt ∂V + ∂t ∂V + ∂t d n d i=1 d ∂V 1 dSti + i ∂S 2 µ i d d i=1 j=1 d d ∂ 2V d S i. . St ). if the ﬁltration (Ft )t∈R+ is the natural ﬁltration of the asset prices (Bt . Pricing by expectations. evaluated at time t and current price St . St )t∈R+ . there is no arbitrage if ξt = EQ Bt g(ST )|Ft BT 60 . From Section 4. S j ∂S i ∂S j n t − V (t. Furthermore. Consider the case of a market with two assets. and formula In this section.1. 7.T ] is a local martingale for Q. Let Q be an equivalent martingale measure. In particular. The Black–Scholes model. By Itˆ’s formula.

which describe the sensitivities of a pricing formula with respect to some of its parameters.assuming the expectation exists.. the quantity delta. vega.. however. of the claim. etc. St ) is an admissible replicating portfolio ∂S T e −rT g(ST ) = V (0. St ) = EQ [e−r(T −t) g(ST )|Ft ]. there is a whole list if quantities. assuming the ﬁltration is generated by S. Ss )dSs . 7. S) ∈ [0. theta. name delta is inspired by the notation of the original Black–Scholes paper. In nearly all cases of interest. From the general results of section 6. we do not know how to compute π. S)| ≤ C(1 + S p ) for all (t. g St e = e 2π −∞ In section 5. 61 3The . the prices of traded assets can be written explicitly: Bt = B0 ert and St = S0 e(µ−σ and hence ξt = EQ e−r(T −t) g S0 e(r−σ = EQ e−r(T −t) g St e(r−σ ∞ −r(T −t) 2 /2)T +σ W ˆT 2 /2)t+σW t = S0 e(r−σ |Ft T −Wt ) 2 /2)t+σ W ˆ t 2 /2)(T −t)+σ(W ˆ ˆ |Ft 2 e−z /2 √ dz.T ] is a local martingale. These quantities are collectively knowns as the greeks. T ] × R+ . the Remark. gamma. St ) then ξT = g(ST ) and (ξt )t∈[0. It is worth pointing out. then V (t. the delta.3 is given a special name. ∂S ∂V ∂S Because of its central importance in the theory. S0 ) + 0 ∂V ˜ (s. assuming V is bounded from below. Furthermore. that the above suﬃcient condition is speciﬁc to the Black– Scholes model. In ﬁnance. the martingale representation theorem asserts the existence of a process π such that √ (r−σ 2 /2)(T −t)+σ T −tz T e−rT g(ST ) = EQ [e−rT g(ST )] + 0 ˜ πs dSs Unfortunately. hence ξt is a noarbitrage price for the claim. we can solve the Black–Scholes PDE ∂V ∂V 1 ∂ 2V + rS + σ 2 S 2 2 = rV ∂t ∂S 2 ∂S V (T. Here is a suﬃcient condition: Suppose there exists positive contants C and p such that |V (t.2. we see that πt = ∂V (t. the two approaches yield the same answer. In this simple case. we argued that. S) = g(S) ˜ If ξt = V (t. Pricing and hedging by PDE..

the delta. K) = EQ [e−r(T −t) (ST − K)+ |Ft ] yielding the Nobel-prize-winning Black–Scholes formula: Ct (T. then one could insert the value σ 2 into the Black–Scholes formula to obtain the price of a call option. σ T −t x 2 7. The payout is (ST − K)+ . Volatility estimation.e. the option’s maturity time T . . the option’s strike K. −e−r(T −t) K Φ σ T −t where Φ(x) = −∞ √1 e−y /2 dy is the standard normal distribution function. and a volatility parameter σ. There is no arbitrage if the time t price Ct (T. . The maximum likelihood estimators are then 1 ν= ˆ tn − t0 and 1 ˆ σ2 = n n n Yi i=1 i=1 (Yi − ν ∆ti )2 ˆ . not the equivalent martingale measure Q. ∆ti If one was to truly believe that the stock price was a geometric Brownian motion. Then the random variables Yi = log St ti then has distribution i−1 Yi = (µ − σ 2 /2)(ti − ti−1 ) + σ(Wti − Wti−1 ) ∼ N (ν∆ti .3. Notice that we have done the statistics under the objective measure P.We can easily apply this result to a speciﬁc payout function g. the hedging portfolio. K) = St Φ √ log(St /K) √ + (r/σ + σ/2) T − t σ T −t √ log(St /K) √ + (r/σ − σ/2) T − t . One possibility is to is to collect n + 1 historical price observations S at times t0 < . that is. 62 . the spot interest rate r. < tn . Notice that in 2π this case. K) of the call is given by Ct (T. To use the Black–Scholes formula to ﬁnd the price of real call options. only the volatility parameter is neither speciﬁed by the option contract nor quoted in the market. σ 2 ∆ti ) where ν = µ − σ 2 /2 and ∆ti = ti − ti−1 . the underlying stock’s price St at time t. of ˆ the form St = S0 eνt+σWt . is given by πt = Φ √ log(St /K) √ + (r/σ + σ/2) T − t . For instance. one must ﬁrst estimate the volatility σ. Of these six numbers. introduce a European call option maturing at T with strike K. i. What made this formula so popular after its publication in 1973 is the fact that the right-hand-side depends only on six quantities: the current calendar time t.

Let the nonrandom function BS : R+ × R+ → [0. given the market price Ct (T. Hence for ﬁxed m. m) is strictly increasing and continuous since √ ∂BS 1 log m v (t.7. . The Black–Scholes formula gives an explicit representation of the prices Ct (T. is is usually the case that the implied volatility surface (T. in real markets. K) of call options in this model in terms of the calendar time t. and a volatility parameter σ. A completely diﬀerent approach to ﬁnd the volatility parameter is to observe the price Ct (T. or that the Black–Scholes model does not quite match reality. 63 2 −r(T −t) . spot interest rate r. 8. 1) be deﬁned by BS(v. K) = σ for all 0 ≤ t < T and K > 0. However. Then Black–Scholes formula says that in the context of a Black–Scholes model the call price is given by Ke−r(T −t) Ct (T. One could either conclude Black–Scholes model is the true model of the stock price and that the the market is mispricing options. we can ﬁnd a number Σt (T. K) is not ﬂat. So. called the implied volatility of the option. K)2 . the option maturity T and strike K.4. the implied volatility is now used as a common language to quote option prices. we have considered the Black–Scholes model–a two asset market model in which the risky asset price is a geometric Brownian motion. thanks to the enormous inﬂuence of the Black–Scholes theory. such that Ct (T. m) can be inverted. K) → Σt (T. m) = √m Φ − log v + (1 − m)+ √ v 2 − mΦ − log m √ v √ − v 2 if v > 0 if v = 0. the function 2π v → BS(v. If the market was still pricing call options by Black–Scholes formula. m) = √ φ − √ + >0 ∂v 2 2 v v where φ(x) = √1 e−x /2 is the standard normal density. Implied volatility. Now notice that v → BS(v. However. and then try to work out which σ to put into the Black–Scholes to get the right price. the current stock price St . The second approach is more prudent. Ke−r(T −t) /St ). why even consider implied volatility? As Rebonato famously put it: Implied volatility is the wrong number to put into wrong formula to obtain the correct price. K) = St BS((T − t)Σt (T. K) of the option. K) = St BS (T − t)σ 2 . K). K) from the market. then there would exist one parameter σ such that Σt (T. St appearing in the above formula is often called the moneyness of the The quantity Ke St option. Then. Local and stochastic volatility models In the section 7.

K) + rK ∂C0 (T. We will assume that σ is smooth and bounded from below and above. That is. K) = EQ [e−rT (ST − K)+ ] Then ∂C0 ∂C0 σ(T. In general. We consider a model given by dBt = Bt r dt dSt = St (µ dt + σ(t. The next theorem in the present context is usually attributed to Dupire’s 1994 paper. the idea is replace the constant volatility parameter in Black–Scholes model with a local volatility function σ : R+ × R+ → R+ . K)2 2 ∂ 2 C0 (T. St )dWt ). K) + K (T. K) = 2[ ∂C0 (T. however. It can be shown that for all t > 0. K) = −rK (T.However. ∞ Hence. 64 . ∞ C0 (T. K) = S0 BS(σ0 T. K) + rK ∂C0 (T. Proof. K). K) ∂K 2 1/2 then the no-abitrage prices of call options in this model exactly match the observed prices. the existence of a Markovian martingale with a given marginal distribution was proven in 1972 by Kellerer. that is. We now consider another Markovian model which can match a given implied volatility surface exactly. Theorem. 2 Of course. K) = e−rT 0 fST (y)(y − K)+ dy = e−rT K fST (y)(y − K)dy. if C0 (T. practitioners and researchers have proposed various generalizations of the Black– Scholes model to better match the observed implied volatility surface. ∂T ∂K 2 ∂K 2 Remark. St ). K > 0} is observed from the market. since the implied volatility surface Σt (T. As always. the local volatility surface need not be ﬂat. K)] ∂T ∂K 0 K 2 ∂ C2 (T. the random variable St has a continuous density with respect to Lebesgue measure. If one chooses the local volatility function σ by Dupire’s formula σ(T. K) of real-world option prices is usually not ﬂat. K) : T > 0. let Q be the equivalent martingale measures with density process deﬁned by dZt = −Zt λt dWt where λt = (µ − r)/σ(t. K) ∂K 2 1/2 = σ0 . Suppose that C0 (T. The point of the above theorem is this: Suppose that today’s call price surface {C0 (T. K)] ∂T ∂K 0 K 2 ∂ C2 (T. However. Ke−rT /S0 ) for some constant σ0 then 2[ ∂C0 (T. there exists a continuous function fSt : R+ → R+ such that x Q(St ≤ x) = 0 fSt (y) dy.

K)2 erT ∂T 2 K and the result follows from noting rT + ∞ ∞ ∞ T ∞ fST (y) y dy = K 0 fST (y)(y − K)+ dy + K K fST (y)dy and applying the appropriate identities.∞) (x). K) + rC0 (T. σt = σ(t. K) = e−rT fST (K) ∂K 2 [The second equality shows that the risk-neutral density of the stock price at time T can be recovered from the prices of the calls of maturity T . other models have been proposed. the Dirac delta ‘function’. St ) • CEV model. K)2 dt 2 0 0 K and then diﬀerentiating both sides with respect to T yields ∞ ∂C0 1 (T. and o g (x) = δK (x). Here are a few popular ones: • Local volatility model. Now computing expected values of both sides 1 T fSt (y)y r dy dt + (1) e C0 (T. K) = (S0 − K) + fSt (K)K 2 σ(t. K) = −e−rT ∂K K ∂ 2 C0 (T. g (x) = 1[K. σt = γStβ−1 2 2 • Heston model. can actually be o rigorously stated in terms of a quantity called local time. dσt = λ(¯ 2 − σt )dt + γσt dZt σ version of Itˆ’s formula for non-smooth convex functions. It turns out that although a local volatility model can be made to exactly match all European call option prices.] To outline the argument.The fundamental theorem of calculus then implies ∞ ∂C0 fST (y) dy (T. These so-called stochastic volatility models are of the form dBt = Bt r dt ˆ dSt = St (r dt + σt dWt ) where (σt )t∈R+ is a given stochastic process. generally the model fails to correctly price path-dependent options. This result was proven by Breeden and Litzenberger in 1978. Therefore. St )dWt where we have appealed to Itˆ’s formula4 with g(x) = (x − K)+ . St )2 dt + 0 ˆ 1{St ≥K} St σ(t. Local volatility models are not the end of the story. we proceed formally T (ST − K)+ = (S0 − K)+ + 0 T 1{St ≥K} dSt + 1 2 1 2 T δK (St )d S 0 t = (S0 − K) + 0 T + 1{St ≥K} St r + δK (St )St2 σ(t. 65 4A . called Tanaka’s formula. K) = fST (y)y r dy + fSt (K)K 2 σ(t.

etc. σt = γt St and dγt = αγt dZt ˆ t + 1 − ρ2 W ⊥ for an independent Brownian motion (W ⊥ )t∈R+ for Q. a practioner’s choice of model must be made on a combination of issues: how well the model ﬁts data. 66 . dσt = λ(¯ 2 − σt )dt + γσt dZt β−1 • SABR model. stochastic volatility models are incomplete since there more Brownian motions than risky assets.2 2 2 σ • GARCH model. Notice. how easy the model is to calibrate. how quickly a computer can calculate exotic option prices with the model. where Zt = ρW t t As the economic notion of no-arbitrage is too weak to pin down the precise functional form of a stochastic volatility model. however. that aside from the local volatility model (including CEV).

where ˜ P (t. etc. But since this number is very large. Recall that we can write the dynamics of the bank account as dBt = Bt rt dt where the adapted process (rt )t∈R+ is called the spot interest rate or the short interest rate. Now. < Td .T ] is a local martingale for all T > 0. rather than a discrete set of points.T1 ] has no arbitrage. . T1 ). we explore models for the interest rate term structure. Definition. T ) the price at time t ∈ [0. there is no arbitrage if P (t. assume that the bond issuer is absolutely credit worthy. Treasury bond prices in Figure 1. We also assume that there is a risk-free num´raire process (Bt )t∈R+ . . T ] of the bond. Note that on 22 November 2008. which we will think e of as a bank or money market account. the market model will be speciﬁed by a family of processes {(P (t.S. government. Of course. . and there is zero probability of default. Therefore. T ))t∈[0. the market (Bt . The basic ﬁnancial instruments in this setting are the zero-coupon bonds. T )/Bt .CHAPTER 5 Interest rate models 1. T ) = EQ (e− for all 0 ≤ t ≤ T .S. T ) = P (t. Bond prices and interest rates In this last chapter. P (t. there are only a ﬁnite number of maturities of bonds traded on the the ﬁxed income market. . In particular. Theorem. We will work in market model where there are bonds of all maturities T available to trade. 67 1We T t rs ds |Ft ) . they are virtually riskless and will serve as a convenient example. we formulate a condition so that for any collection of maturities T1 < .S. consider the graph of U. mortgage-backed securities. . We denote by P (t. To get a feel for how we should model the bond prices. A (zero-coupon) bond with maturity T is a European contingent claim that pays exactly1 one unit of currency at time T . However. . we are not discussing corporate bonds. Of course. Treasury are backed by the ‘full faith and credit’ of the U. Td ))t∈[0. There is no arbitrage if there exists a locally equivalent measure Q such that ˜ the discounted bond price process (P (t. Therefore. it is common practice to represent the zero-coupon bond prices as a continuous curve. the above diﬀerential equation has the solution Bt = B0 e t 0 rs ds . since bonds issued by the U.T ] : T > 0}. This is the typical situation. T ))t∈[0. the map T → Pt (T ) was decreasing. P (t.

and r is suitably well behaved.S. Indeed. T ) = e−(T −t) 68 y(t. Graph of the U.treasury. Data taken from http://www. The yield curve and the bond price curve contain the same information. T ) = EQ (e− t rs ds |Ft ). then ∂ Pt (T ) − ∂T = lim E T =t T ↓t Q T 1 − e− t rs ds |Ft T −t T = rt by the dominated convergence theorem. t).S. Treasury zero-coupon bond price curve on 22 November 2008. it seems that we should like to model the interest rate (rt )t∈R+ as a non-negative process. .T ) . we can diﬀerentiate the bond price with respect to maturity to recover the spot rate. since P (t.gov/ Notice that if P (t. Treasury yield curve on 22 November 2008. Note that spot interest rate is just the left hand point of the yield curve rt = y(t.Figure 1. if the spot rate process is bounded and continuous. However. it is often easier to speak of interest rates. We have already deﬁned the short interest rate. T ) at time t of a bond maturing at time T deﬁned by the formula 1 y(t. if rt ≥ 0 almost surely for all t ≥ 0 then the map T → Pt (T ) is decreasing almost surely. T ) = − log P (t. Indeed. From common experience. the interest rate is often modelled by a Gaussian process which might become negative with positive probability. in actual practice. Rather than speak of bond prices. T ) T −t Figure 2 is a graph of the U. A popular interest rate is the yield y(t.

Data taken from http://www.s) ds . the price data encoded in either of the functions T → y(t. deﬁned by f (t. say bounded and continuous. notice that the spot rate is the left hand end point of the forward rate curve rt = f (t. Graph of the U. The term structure of interest rates refers the function T → P (t. T ) = e− T t f (t. T ) such that the no-arbitrage price at time t of the claim that pays rT − f (t. Treasury yield curve on 22 November 2008. T ). then f (t.treasury. T ) at time T is zero. or equivalently.gov/ For us. Note that if (rt )t∈R+ is suitably regular. 69 . T ). T ) or T → f (t. T ) = − ∂ log P (t. ∂T Again. t). T ) at time t for maturity T . note that the forward rates contain the same information as the bond pricese since P (t. T ).S. Again. T ) = EQ (rT e− EQ (e− T t rs ds |Ft ) T t rs ds |Ft ) so that the forward rate can be interpreted as the Ft -measurable random variable f (t. a more useful interest rate is the forward rate f (t.Figure 2.

. no such choice is possible since the short rate is not traded. r) (t. so (P (t. Since we are interested in pricing and hedging. rt )dWt ∂r ˜ Since the drift vanishes by assumption. rt )drt + ∂t ∂r 2 ∂r2 t ∂V ∂V 1 ∂ 2V = e− 0 rs ds (t.T ] is a local martingale. r) = rV (t. the price of contingent claims can be expressed in terms the solution of a PDE: Theorem. T ))t∈[0. there is no need to model the processes (at )t∈R+ and (λt )t∈R+ separately. T ) = V (t.2. However. rt ) dt + (t. . and a Brownian motion (Wt )t∈R+ for P. . rt )2 2 (t. r) ∂t ∂r 2 ∂r V (T. rt ) + α(t. We will consider an Itˆ o process short interest rate model of the form drt = at dt + βt dWt for adapted process (at )tR+ and (bt )t∈R+ . . r)2 2 (t. rt )β(t. rt )d r t (t. r) + β(t. rt ) + β(t. rt )dt + β(t. We will assume that the market price of risk is given by the process (λt )t∈R+ so that ˆ drt = αt dt + βt dWt ˆ where dWt = dWt +λt dt deﬁnes a Brownian motion for the measure Q whose density process is given by dZt = −Zt λt dWt . rt )dt ∂V ∂V 1 ∂ 2V (t. Proof. Note that while in a complete stock market model there was only one way to switch to an equivalent martingale measure. r) + α(t. rt ) dt + e− 0 rs ds (t. Itˆ’s formula implies o d e− t 0 rs ds t 0 rs ds V (t. However. As we have learned for Markovian stock models. t 0 rs ds 70 . rt ) ∂t ∂r 2 ∂r t ∂V ˆ −rt V (t. r) = 1 If P (t. T ] × R → R satisﬁes the PDE ∂V 1 ∂ 2V ∂V (t. Short rate models We begin with a market that has just the bank account B. rt ) (t. rt ) = −rt e− +e− V (t. rtn . rt )dWt for some non-random functions α : R+ × R → R and β : R+ × R → R. we know that there is no arbitrage if the market somehow picks an equivalent martingale measure Q to price the bonds. and where αt = at − βt λt deﬁnes the risk-neutral drift. we must be careful to realize that is impossible to estimate the distribution of the random variable αt directly from a time series rt1 . rt ) then there is no arbitrage. We now study the case when the short rate is Markovian. Assume that ˆ drt = α(t. Fix T > 0 and suppose V : [0.

In 1977. Indeed. r) we have (−A (t)r − B (t))V (t. the advantage of this type of model is that it is relatively easy to compute prices. In particular. we have ¯ 2λ 1 T T rs ds → r Q − almost surely. However. ﬁx T > 0 and consider the PDE ∂V ∂V 1 ∂ 2V (t. r) ∂t ∂r 2 ∂r V (T. Substituting this into the PDE yields σ2 A(t)2 V (t. We can also use the above theorem to compute bond prices. σ . r) = e−rA(t)−B(t) for some functions A and B which satisfy the boundary conditions A(T ) = B(T ) = 0. r) + r A (t) = λA(t) − 1 B (t) = −λ¯A(t) + r 2A σ2 A(t)2 2 continuous Gaussian Markov process is often called a Ornstein–Uhlenbeck process. r). r) + σ 2 2 (t. 2 Since this is supposed to be an identity for all (t. 71 .1. both positive and negative. Vasicek proposed the following model for the short rate: ˆ drt = λ(¯ − rt )dt + σdWt r for a parameter r > 0 interpreted as a mean short rate. explicitly. We can make the ansatz V (t. Since the short rate rt is Gaussian. unless we have a model for the market price of risk.2. r) = 1. Note that for each t ≥ 0 the random variable rt is Gaussian2 under the measure Q with t E (rt ) = e Q −λt r0 + (1 − e −λt )¯ and Var (rt ) = r 0 Q e −2λ(t−s) 2 σ2 σ ds = (1 − e−2λt ). ¯ 0 Please note. that in the present framework we can say absolutely nothing about the distribution of rt for the objective measure P. A disadvantage of this model is that there is a chance that rt < 0 for some time t > 0. a mean-reversion parameter λ > 0. one can show that the process is ergodic and converges to the invariant distri2 bution N r. r) − λ(¯ − r)A(t)V (t. for instance of bonds. Vasicek model. 2λ Moreover. Recall that a normal random variable can take any real value. however. r) + λ(¯ − r) r (t. ¯ and a volatility parameter σ > 0. the Q-probabilty of the event {rt < 0} is pretty small. for sensible parameter values. This stochastic diﬀerential equation can be solved explicitly to yield t rt = e−λt r0 + (1 − e−λt )¯ + r 0 ˆ e−λ(t−s) σdWs . r) = rV (t. r) = rV (t.

ﬁx T > 0 and consider the PDE ∂V ∂V 1 ∂ 2V (t. one can show that the process is ergodic and its invariant disribution is a gamma distribution with mean r. T ) = B(t. ¯ An advantage of this model over the Vasicek model is that the short rate rt is non-negative for all t ≥ 0.The solution to this pair of equations is A(t. the forward rates are more manageable: f (t. The process (rt )t∈R+ satisfying the above stochastic diﬀerential equation is often called a square-root diﬀusion or CIR process.) 2. As before we can make the ansatz V (t. In 1985. r). explicit formula are still available for the bond prices. r) = rV (t. Furthermore. 2 . Cox–Ingersoll-Ross model. ¯ and a volatility parameter σ > 0. r) ∂t ∂r 2 ∂r V (T. T ) = exp −rt (1 − e−λ(T −t) ) −r ¯ λ (1 − e−λu ) du + 0 σ2 2λ2 (1 − e−λu )2 du . r) = 1. Ingersoll. r) + σ 2 r 2 (t. T ) = t 1 (1 − e−λ(T −t) ) λ T σ2 λ¯A(s) − A(s)2 ds r 2 T −t T −t so that the bond price is given by P (t. r) + λ(¯ − r) r (t. the forward rates at time t are an aﬃne function of the short rate at time t. Cox. r) = e−rA(t)−B(t) for some functions A and B which satisfy the boundary conditions A(T ) = B(T ) = 0. Substituting this into the PDE yields (−A (t)r − B (t))V (t. Althought the stochastic diﬀerential equation cannot be solved explicitly. For instance. one can say quite a lot about this process. r) − λ(¯ − r)A(t)V (t. (An aﬃne function is of the form g(x) = ax + b. the CIR model is another example of an aﬃne term structure model: Again. r) = rV (t. r) + r 72 σ2 rA(t)2 V (t. t + x) = rt e−λx + r(1 − e−λx ) − ¯ σ2 (1 − e−λx )2 2λ2 This formula says that for the Vasicek model. though this stochastic process was studied as early as 1951 by Feller. a mean-reversion parameter λ > 0.2. However. Indeed. and Ross proposed the following model for the short rate: √ ˆ drt = λ(¯ − rt ) + σ rt dWt r for a parameter r > 0 interpreted as a mean short rate. that is. its graph is a line. 0 This is a mess.

Zt )dWt ˆ where a : R+ × Rd → Rd and b : R+ × Rd → Rd×d are given functions and (Wt )t∈R+ is a d-dimensional Brownian motion for the equivalent martingale measure Q. there is very little ﬂexibility in the possible shapes of the forward rate curve. The idea is to assume that there are d underlying economic ‘factors’ in the market. t + x) = 4γ 2 eγx 2λ¯(eγx − 1) r rt + . However. r) is aﬃne. r The equation for A is a Riccati equation. the function r → g(t. f (t. T.This time we have σ2 A (t) = λA(t) + A(t)2 − 1 2 B (t) = −λ¯A(t). We model these factors as the solution (Zt )t∈R+ of a stochastic diﬀerential equation ˆ dZt = a(t. T ) = B(t. Zt )dt + b(t. and R(t. we consider more general factor models. in the Vasicek and CIR models. T. r) = r. In this section. whose solution is A(t. but the forward rates are given by f (t. γx + (γ − λ)]2 [(γ + λ)e (γ + λ)eγx + (γ − λ) In particular. the forward rates for the CIR model are again given by an aﬃne function of the short rate. so that the correlation coeﬃcient ρ(rt . In particular. T ) = t 2(eγ(T −t) − 1) (γ + λ)eγ(T −t) + (γ − λ) T λ¯A(s)ds r √ where γ = λ2 + 2σ 2 . especially the Vasicek and CIR models in which formulas for the bond prices are available in closed form. a possible shortcoming of these models is that they predict a very rigid term structure. 73 . rt ) where D = {(t. T )) = 1 for all 0 < t ≤ T . The main theorem is below. since there exists a deterministic function g : D × R → R such that f (t. Zt ). 3. We then assume that the short rate is given by a function rt = R(t. The bond prices are too messy to write down. The short rate models considered in the last section have d = 1. Factor models The Markovian short rate models are popular in practice. Indeed. Zt = rt . T ) : 0 ≤ t ≤ T }. T ) = g(t. of which the short rate models are only a special case.

+ Zt . The analysis of such a stochastic diﬀerential equation is very simple when all the δi ’s are zero. .Theorem. We let √ γ1 + δ1 · Zt 0 ··· 0 √ . γ1 . √ γd + δd · Zt 0 ··· Here. the solution is a d-dimensional Ornstein–Uhlenbeck process. T ) = V (t. and in fact. Remark. . and β is a d × d constant matrix.T )·Zt −B(t. . T ) = e−A(t. for deterministic functions A : D → R → Rd and B : D → R. . T ] × Rd → R+ satiﬁes the partial diﬀerential equation ∂V + ∂t d i=1 1 ∂V + ai ∂Zi 2 d d Bij i=1 j=1 ∂ 2V = RV ∂Zi ∂Zj V (T. Suppose the short rate is given by rt = Zt + . . Fix T > 0. . We now consider a special case of the above theorem. As before. Existence and uniqueness of solutions of such stochastic diﬀerential equations is not guaranteed. . δd are d-dimensional constant vectors.. γd are d positive real constants. α and δ1 . subject to the boundary conditions Ak (T. T ) 74 . ﬁrst studied by Duﬃe and Kan in 1996. 0 γ2 + δ2 · Zt · · · . the situation is much more delicate when some of the δi ’s are non-zero. Z) = 1 where B(t. we can make the ansatz that the bond prices can be written in the exponential aﬃne form P (t. z))T . and analyzing the properties of this non-Gaussian diﬀusion is not easy because of the random volatility created by the non-zero δi ’s. In this case. . Zt ) then the market consisting of the bank account and bond maturity T has no arbitrage. . ˆ dZt = (α + βZt )dt + dWt . . As in the case of the Vasicek and CIR models. If P (t. T ) = 0 = B(T. . z) = b(t. Suppose V : [0. . z)b(t. the functions A and B can be found by solving the following system of d + 1 coupled Riccati equations ∂Ak = − ∂t ∂B = − ∂t d (1) (d) i=1 d 1 βik Ai + δk A2 − 1 k 2 1 αi Ai + 2 d γi A2 . there is existence and uniqueness of a solution. i i=1 i=1 These equations can be solved numerically.T ) . However. . . . for instance.

T ). More concretely.. . Td and consider the d benchmark rates f (t. ti ) and the self-ﬁnancing condition is Bti = πti since P (t. . there exists an equivalent martingale ˜ measure Q such that all discounted bond prices (P (t.. the budget constraint is Bti−1 = πti P (ti−1 . where the discounted bond price at time t for maturity T is given by ˜ P (t. T )|T =t ∂T 75 By taking the limit as ti − ti−1 → 0. Motivation. suppose at time 0 the investor has B0 units of wealth..T ] are local martingales. The Heath–Jarrow–Morton framework Starting from a factor model. . f (t. the derived bond prices are necessarily Itˆ processes. . We can construct the bank account by considering an investor holding his wealth in just-maturing bonds. . but without the bank account. the rate of change of the wealth is given by Bti − Bti−1 ti − ti−1 = Bti−1 1 − P (ti−1 .j∈{1.. Hence. T ) = Γt = ∂Aj (t. ∂T ∂T Fix d dates T1 . . Indeed. If the matrix f (t. by construction. . . d}. and Morton in 1992 was that we can change perspectives by modelling the bond prices directly. of times and suppose that during the interval (ti−1 . . Ti ) − (t. then we can recover the factors as linear combinations of the benchmark forward rates: ∂B Zt = Γ−1 f (t. suppose we start out with just the bond market. T ))t∈[0. T1 ). Fix a sequence 0 ≤ t0 < t1 < . ti ] the investor holds all of his wealth in the bond which matures at time ti . Given the exponential aﬃne bond prices. The insight of Heath. Ti ) ∂T i. we can deﬁne the spot rate by rt = − so that dBt = Bt rt dt as before. . .. There o is no arbitrage in a factor model since. and the number of shares of the just-maturing bond by πt . Td ). note that ∂A ∂B (t.. ti ) Pti−1 (ti ) ti − ti−1 ∂ P (t. t) = 1 for all t.d} 4. Jarrow. T ) · Zt + (t. . T ) = e− t 0 rs ds Pt (T ). . Ti ) t ∂T i∈{1.d} is invertible. .for all k ∈ {1. . If the investor’s wealth at time t is denoted by Bt ...

P (t. T ) · t ˆ σ(t. Indeed.The usual formulation of the HJM idea is in terms of the forward rates. (T ) = i=1 σ (t. T1 ). T )dt + i=1 σ (i) (t. . s)ds . T ) = σ(t. P (t. T ))t∈[0. as in Section 2. ˆ For this measure. T ). T ) = a(t. Notice that this drift/volatility contraint is not present in models in which only the dynamics of the short rate are speciﬁed. If there exists a d-dimensional bounded adapted process (λt )t∈R+ such that d a(t. note that any of the short rate or factor models can be put into the HJM framework. just by choosing the initial forward rate curve to match the one predicted by the model. T ))t∈[0. we can initialize the model with any initial forward rate curve T → f (0. . . Td ))t∈ has no arbitrage. T ) (i) (i) λt T + t σ (i) (t. T ) = e− T t f (t. Theorem. s)ds dt + σ(t.T ] and (σ (i) (t.T ] has dynamics n df (t. The upshot of the HJM result is that the drift and the volatilty of the forward rate dynamics cannot be prescribed independently. . Finally. . T ))t∈[0. 76 . the market model with prices (Bt . Let the short rate be given by rt = f (t. Remark. We can rewrite the forward rate dynamics as T df (t. T ) · λt + t σ(t. in the HJM framework. usually called the HJM drift condition. then. the process deﬁned by dWt = dWt + λt dt is a Brownian motion. they must be related by the famous formula T a(t. t) and the bank account dynamics by dBt = Bt rt dt. Nevertheless. let the bond prices be given by P (t. As usual. < Td .s) ds .T ] . Proof. T )dWt (i) for some suitably regular adapted processes (a(t. Deﬁne a locally equivalent measure Q by the density process dZt = −Zt λt · dWt . the foward rate process (f (t. T ) · dWt . we put ourselves in the context of a probability space (Ω. T ) = σ(t. P) on which we can deﬁne a d-dimensional Brownian motion (Wt )t∈R+ . Suppose for each T . The diﬀerence with the short rate models is that we are now trying to model the dynamics of the whole term structure. s)ds . Indeed. F. for any set of d maturities 0 < T1 < . .

Now applying some formal manipulations t T T d 0 rs ds + t f (t.s)ds is a local martingale. t))dt + t df (t. S] × [0. and o S T T T 2 SE 0 0 (g(s. t)2 ds dt < ∞ a. t)dWt ds? First note that the equality holds for g ∈ S where S is the set n S= i=1 ki 1(si−1 . An example which will occur frequently is when g is not random and continuous (or at least Riemann integrable) on [0. We conclude this section with some examples. T ]. and hence are vulnerable to the criticism that there is a positive probability that the rates become negative. t)]2 ds dt =E 0 0 [g(s.s. s)ds · dWt . by localization. ki is bounded andFti−1 measurable . t)]2 dt ds → 0.e− ˜ It is enough to show that for each T > 0. s) ds = (rt − f (t. Fubini’s theorem. since T S T S 2 T 0 T 0 S 2 dt E 0 0 g(s. t) − gn (s. T ) = −P (t. Now suppose there exists a sequence (gn )n∈N in S such that T S S T E 0 0 [g(s. The question is: if we ﬁx S. t) − gn (s. t)2 ds dt < ∞ with 0 0 g(s. t))ds S = by Itˆ’s isometry and the Cauchy–Schwarz inequality. t) − gn (s. we can replace the condition o T S T S E 0 0 g(s. t) − gn (s. t) − gn (s. t))2 ds dt → 0 S E 0 S 0 T T 2 dt E 0 0 g(s. t))dWt S = SE 0 0 (g(s. the discounted bond price process P (t. t))2 dt ds → 0 by the Cauchy–Schwarz inequality. and Itˆ’s isometry. t)dWt ds − 0 0 gn (s. s)ds dt + t ˆ σ(t. Then g satisﬁes the exchange of order of integration equality. T ) = T t 3 0 rs ds− t f (t. t)dWt ds ≤ (g(s. we have o 1 2 T t 2 σ(t. T ≥ 0. T ˜ ˜ dP (t. 0 ≤ ti−1 < ti ≤ T. T ) t ˆ σ(t. by Itˆ’s formula. the forward rates are Gaussian under the measure Q.sj ]×(ti−1 . t) − gn (s.ti ] : 0 ≤ si−1 < si ≤ S. would like to appeal to a stochastic Fubini theorem in order to exchange the order of integration in the double integral. t)ds dWt = 0 0 g(s. Finally. t)ds dWt − 0 0 gn (s. when do we have the equality T 0 0 S S T 3We g(s. s)ds · dWt and we’re done. In these examples. s) ds T = Hence. 77 . t)ds dWt = E (g(s.

mean rate r(t). Then ˆ df (t. (1986) This model is the simplest possible model HJM model. Vasicek–Hull–White. T ) under the risk-neutral measure Q is Gaussian with mean t T EQ [f (t. T ) · dWu .2. λ rt = f (0. s)ds du and covariance CovQ [f (s. (1994) Note that for the HJM models discussed above. the forward rates are given by t T t f (t. 0 Hence the Ho–Lee model corresponds to the following short rate model: ˆ drt = (f (t) + σ 2 t)dt + σ0 dWt .1. T ) = The short rates are given by t 2 σ0 −λ(T −t) ˆ e (1 − e−λ(T −t) )dt + σ0 e−λ(T −t) dWt . If σ is not random. T ) + 0 σ(u. T )] = 0 σ(u. then for positive times t the forward rates f (0. 78 . T ) = σ 2 (T − t) dt + σ0 dWt . t) + σ 2 t2 /2 + σ0 Wt . the Hull–White extension of the Vasicek essentially replaces the mean interest rate r ¯ with a time-varying. T )du. s)ds du + 0 ˆ σ(u. S). T ) = σ0 be constant. T ) + σ0 (T t − t2 /2) + σ0 Wt . T ) is bounded from below. t) + 2 σ0 (1 2λ2 − e−λt )2 + 0 ˆ σ0 e−λ(t−s) dWs The short rate dynamics are given by t 2 σ0 −λt ˆ ˆ e (1 − e−λt ) dt + σ0 dWt − λ σ0 e−λ(t−s) dWs dt λ 0 2 σ0 ˆ = f0 (t) + λf0 (t) + (1 − e−2λt ) − λrt dt + σ0 dWt 2λ Hence. T ) · u σ(u. Ho–Lee. T ) = σ0 e−λ(T −t) for positive constants σ0 and λ. T ) = f (0. Kennedy. 0 0 4. Here is an unusual feature of this model: if the initial forward rate curve T → f (0. T ) → ∞ as T → ∞. T ) · u s∧t σ(u. S) · σ(u. T ) = f (0. Let d = 1 and σ(t. then the distribution of f (t. t) + 0 2 σ0 −λ(t−s) e (1 − e−λ(t−s) )ds + λ t t ˆ σ0 e−λ(t−s) dWs 0 = f (0. ¯ drt = f0 (t) + 4.4. f (s. (1990) Again let d = 1 but now σ(t. Then df (t. 0 or 2 ˆ f (t.3. T )] = f0 (T ) + 0 σ(u. The short rate is then given by ˆ rt = f (0. but non-random.

the increments of (f (t. An advantage of this formulation of the Gaussian HJM model is that one is no longer restricted to ﬁnite dimensional Brownian motions. T ) so that. and. 79 .’ This can be achieved by taking c to be t ct (S. T ) : 0 ≤ t ≤ T } with mean µ(t. T ))t∈[0. T ) = cs∧t (S. T ) and covariance C(s. one choice is to have the correlation of the increments decay exponentially in the diﬀerence of the maturities: ‘ρ(dft (S). Then there is no arbitrage if the mean is given by T µ(t. for each ﬁxed T > 0. For instance. and considered a Gaussian random ﬁeld {f (t. Suppose that covariance has the special form C(s. T )ds. dft (T )) = e−β|T −S| . T ) = e−β|T −S| 0 αu (S)αu (T )du for real valued functions αu . S. S. T ) + 0 ct∧s (s. t. T ) = f (0.T ] are independent. T ). there is much more ﬂexibility to specify the correlation of the increments. t.Kennedy reversed this logic. therefore.

.

A subset of Ω which is an element of F is called an event. Let Ω be a set and let F be a sigma-ﬁeld on Ω. Let Ω be a set. Measures Definition. and if P(A) = 0 then A is called a null event. and P a probability measure on (Ω. i=1 i=1 Theorem. and an element of Ω is called an outcome. The set Ω is called the sample space. F). Let (Ω.CHAPTER 6 Crashcourse on probability theory These notes are a list of many of the deﬁnitions and results of probability theory needed to follow the Advanced Financial Models course. these notes should be used only as a reference. There exists a unique measure Leb on (R. A sigma-ﬁeld is called trivial if each of its elements is either almost sure or null. 1. F) is a measure such that P(Ω) = 1. F) is a µ : F → [0. Let A ∈ F be an event.s. F. A random variable is a function X : Ω → R such that the set {ω ∈ Ω : X(ω) ≤ t} is an element of F for all t ∈ R. and since no proofs are given for any of the theorems. if Ω is a topological space. A measure µ on the measurable space (Ω. for instance Rn . A2 . ∈ F are disjoint then µ( ∞ Ai ) = ∞ µ(Ai ). 81 . Definition. . F a sigma-ﬁeld on Ω. 2. If P(A) = 1 then A is called an almost sure event. B) such that Leb(a. i=1 The terms sigma-ﬁeld and sigma-algebra are interchangeable. A table of notation is in the appendix. F. ∈ F then ∞ Ai ∈ F. This measure is called Lebesgue measure. ∞] such that (1) µ(∅) = 0 (2) if A1 . the Borel sigma-ﬁeld on Ω is the smallest sigma-ﬁeld containing every open set. . The Borel sigma-ﬁeld B on R is the smallest sigma-ﬁeld containing every open interval. More generally. . P) be a probability space. The phrase ‘almost surely’ is often abbreviated a. Since they are free from any motivating exposition or examples. . The triple (Ω. b] = b − a for every b > a. . A probability measure P on (Ω. P) is called a probability space. (2) if A1 . A2 . Let Ω be a set. A sigma-ﬁeld on Ω is a non-empty set F of subsets of Ω such that (1) if A ∈ F then Ac ∈ F. . Definition. Random variables Definition.

takes only a ﬁnite number of values x1 . . The expected value of X is denoted by E(X) and is deﬁned as follows • X is simple. Definition. . . and mean are interchangeable. 1 0 if ω ∈ A if ω ∈ Ac 3. and let X be a random variable. E(X) = E(X + ) − E(X − ) • X is vector valued and E(|X|) < ∞. Xd )] = (E[X1 ]. Y ). . written Cov(X. The variance of an integrable random variable X.} Note that the expected value of a non-negative random variable may take the value ∞. In particular. is Cov(X. E(X) = sup{E(Y ) : Y simple and 0 ≤ Y ≤ Xa. we call a function X : Ω → Rn a random variable or random vector if X(ω) = (X1 (ω). the event {X ≤ t} denotes {ω ∈ Ω : X(ω) ≤ t}. xn . . 1] deﬁned by FX (t) = P(X ≤ t) for all t ∈ R.Let A be a subset of R. . The covariance of square-integrable random variable X and Y . . • X ≥ 0 almost surely. expectation. Let X be a random variable on (Ω. . . . n}. Xn (ω)) and Xi is a random variable for each i ∈ {1. . E[Xd ]) A random variable X is integrable iﬀ E(|X|) < ∞ and is square-integrable iﬀ E(X 2 ) < ∞. written ρ(X. Y ) . The distribution function of X is the function FX : R → [0. . i. is Var(X) = E{[X − E(X)]2 } = E(X 2 ) − E(X)2 . . E[(X1 . Let A be an event in Ω. Y ). Y ) = E{[X − E(X)][Y − E(Y )]} = E(XY ) − E(X)E(Y ). 1} deﬁned by 1A (ω) = for all ω ∈ Ω. Y ) = Cov(X. F. then their correlation. The indicator function of the event A is the random variable 1A : Ω → {0. written Var(X). • Either E(X + ) or E(X − ) is ﬁnite. . If neither X or Y is almost surely constant. .s. For instance. . is ρ(X. Var(X)1/2 Var(Y )1/2 82 .e. P). Expectations and variances Definition. . The terms expected value. We also use the term random variable to refer to measurable functions X from Ω to more general spaces. . n E(X) = i=1 xi P(X = xi ). . We use the notation {X ∈ A} to denote the set {ω ∈ Ω : X(ω) ∈ A}. .

Let the function g : R → R be such that g(X) is integrable. the above inequality is strict unless X is constant. i. The case when p = q = 2 is called the Cauchy–Schwarz inequality. Let X and Y be integrable random variables. If X is discrete. Theorem. Let X be a random variable and g : R → R be a convex function. is the function fX : Rn → [0. the space Lp is the collection of random variables such that E(|X|p ) < ∞. Then E[g(X)] ≥ g(E[X]) whenever the expectations exist. t∈S If X is an absolutely continuous random variable with density function fX then ∞ E(g(X)) = −∞ g(x) fX (x) dx. A random variable X is called discrete if X takes values in a countable set. For p ≥ 1. If g is strictly convex. ∞) such that P(X ∈ A) = A fX (x)dx for all Borel subsets A ⊆ Rn . there is a countable set S such that X ∈ S almost surely. Let X and Y be random variables and let p. Theorem (Jensen’s inequality). If X is a random vector taking values in Rn . The random variable X is absolutely continuous (with respect to Lebesgue measure) if and only if there exists a function fX : R → [0. 1] deﬁned by pX (t) = P(X = t) is called the mass function of X. in which case the function fX is called the density function of X. if it exists. 83 . Then E(X) ≥ 0 with equality if and only if X = 0 almost surely. If X ∈ Lp and Y ∈ Lq then E(XY ) ≤ E(|X|p )1/p E(|Y |q )1/q with equality if and only if X = 0 almost surely or |Y | = a|X|p−1 almost surely for some constant a ≥ 0. q > 1 with o 1 + q = 1. The space L∞ is the collection of random variables which are bounded almost surely.Random variables X and Y are called uncorrelated if Cov(X. the function pX : R → [0. Definition. 1 p Theorem (H¨lder’s inequality). If X is a discrete random variable with probability mass function pX taking values in a countable set S then E(g(X)) = g(t) pX (t). Definition.e. ∞) such that t P(X ≤ t) = −∞ fX (x)dx for all t ∈ R. Y ) = 0. Theorem. then the density of X. • linearity: E(aX + bY ) = aE(X) + bE(Y ) • positivity: Suppose X ≥ 0 almost surely.

written X ∼ unif(a. if pX (k) = n k p (1 − p)n−k for all k ∈ {0. 2. • geometric with parameter p if pX (k) = p(1 − p)k−1 for all k = 1. Then E(X) = np and Var(X) = np(1 − p). . b). Then E(X) = a+b . σ 2 ). Then E(X) = p and Var(X) = p(1 − p). . 4. • binomial with parameters n and p. . Then E(X) = λ. p). Then E(X) = µ and Var(X) = σ 2 . 2 • normal or Gaussian with mean µ and variance σ 2 . 2. . The random variable X is called • uniform on the interval (a. Let X be a continuous random variable with density function fX . where 0 < p < 1. 1.More generally. . Let X be a discrete random variable taking values in Z+ with mass function pX . if fX (t) = 1 for all a < t < b b−a for some a < b. written X ∼ bin(n. . if X is a random vector valued in Rn with density fX and g : Rn → R then E(g(X)) = Rn g(x) fX (x) dx. 3. Definition. 84 . The random variable X is called • Bernoulli with parameter p if pX (0) = 1 − p and pX (1) = p. if fX (t) = √ 1 (x − µ)2 exp − 2σ 2 2πσ for all t ∈ R for some µ ∈ R and σ 2 > 0. • exponential with rate λ. • Poisson with parameter λ if pX (k) = λk −λ e for all k = 0. n} k where n ∈ N and 0 < p < 1. if fX (t) = λe−λt for all t ≥ 0 for some λ > 0. . written X ∼ N (µ. k! where λ > 0. . . where 0 < p < 1. b). Special distributions Definition. Then E(X) = 1/λ. Then E(X) = 1/p. 1. .

{X2 ≤ t2 }. A2 . 5. P(B) Theorem (The law of total probability). Definition. Xj ) = Vij . If X and Y are independent and integrable. independence Definition. be events. P(B) The conditional expectation of X given B. are called independent if the events {X1 ≤ t1 }. written P(A|B). Random variables X1 . . . Let A1 . Let B be an event with P(B) > 0. Then σ2 P(|X − µ| ≥ ) ≤ 2 for all > 0. B2 . The phrase ‘independent and identically distributed’ is often abbreviated i. The conditional probability of an event A given B. then X is said to have the ndimensional normal (or Gaussian) distribution with mean µ and variance V . . V ). are independent. 85 . Then E(X) P(X ≥ ) ≤ for all > 0.i.If X is a random vector valued in Rn with density 1 fX (x) = (2π)−n/2 det(V )−1/2 exp − (x − µ) · V −1 (x − µ) 2 for a positive deﬁnite n × n matrix V and vector µ ∈ Rn . . Conditional probability and expectation. . . Let B1 . is P(A ∩ B) P(A|B) = . X2 . non-null events such that ∞ Bi = Ω. . Then i=1 ∞ P(A) = i=1 P(A|Bi )P(Bi ) for all events A. . Then E(Xi ) = µi and Cov(Xi . be disjoint. 6.d. Probability inequalities Theorem (Markov’s inequality). . Let X be a positive random variable. . If P( i∈I Ai ) = i∈I P(Ai ) for every ﬁnite subset I ⊂ N then the events are said to be independent. Corollary (Chebychev’s inequality). written X ∼ Nn (µ. . then E(XY ) = E(X)E(Y ). written E(X|B). . is E(X 1B ) E(X|B) = . Let X be a random variable with E(X) = µ and Var(X) = σ 2 . Theorem.

[The phrase ‘inﬁnitely often’ is often abbreviated i. Let X1 . . Let φX and φY be the characteristic functions of X and Y .o.] Theorem (The ﬁrst Borel–Cantelli lemma). where i = −1. 86 . More generally. if r ≥ p ≥ 1 then Xn → X in Lr ⇒ Xn → X in Lp . . . If ∞ P(An ) < ∞ n=1 then P(An inﬁnitely often) = 0. p ≥ 1 Furthermore. Let A1 . Definition. Fundamental probability results Definition (Modes of convergence). and X be random variables. Characteristic functions Definition. . for p ≥ 1. Theorem (Uniqueness of characteristic functions). 8. Let A1 . The following implications hold: Xn → X almost surely or ⇒ Xn → X in probability ⇒ Xn → X in distribution Xn → X in Lp .7. Let X and Y be real-valued random variables with distribution functions FX and FY . . X2 . if X is a random vector valued in Rn then φX : Rn → C deﬁned by φX (t) = E(eit·X ) is the characteristic function of X. • Xn → X almost surely if P(Xn → X) = 1 • Xn → X in Lp . A2 . if E|X|p < ∞ and E|Xn − X|p → 0 • Xn → X in probability if P(|Xn − X| > ) → 0 for all > 0 • Xn → X in distribution if FXn (t) → FX (t) for all points t ∈ R of continuity of FX Theorem. be events. . The characteristic function of a real-valued random variable X is the function φX : R → C deﬁned by φX (t) = E(eitX ) √ for all t ∈ R. Then φX (t) = φY (t) for all t ∈ R if and only if FX (t) = FY (t) for all t ∈ R. . be a sequence of events. A2 . The term eventually is deﬁned by {An eventually} = N ∈N n≥N An and inﬁnitely often by {An inﬁnitely often} = N ∈N n≥N An . . .

Let X1 .. If ∞ P(An ) = ∞ n=1 then P(An inﬁnitely often) = 1. . n↑∞ n↑∞ Theorem (Dominated convergence theorem). . . . . X2 . If E(supn≥1 |Xn |) < ∞ then E(Xn ) → E(X). . σ n Then Zn → Z in distribution. + Xn − nµ √ Zn = . . Then E(lim inf Xn ) ≤ lim inf E(Xn ). be independent and identically distributed with E(Xi ) = µ and Var(Xi ) = σ 2 for each i = 1. . Let X1 . . Then X 1 + . Let X1 . A2 . and X be random variables such that Xn → X almost surely. Theorem (Fatou’s lemma). be positive random variables with Xn ≤ Xn+1 almost surely for all n ≥ 1. Let X1 . where Z ∼ N (0. X2 . . . . . Let X1 . 1). and let X = supn∈N Xn . X2 . Theorem (Monotone convergence theorem).Theorem (The second Borel-Cantelli lemma). + Xn → µ almost surely. and let X1 + . Theorem (A strong law of large numbers). be a sequence of independent events. . 2. . X2 . be positive random variables. X2 . . . . n Theorem (Central limit theorem). . 87 . . . Let A1 . . . be independent and identically distributed integrable random variables with common mean E(Xi ) = µ. Then Xn → X almost surely and E(Xn ) → E(X). . .

R R+ N C Z Z+ Ac FX pX fX φX E(X) Var(X) Cov(X.} set of non-negative integers {0. . Ac = {ω ∈ Ω. −1. . 2. . x2 .} complement of a set A. Y ) E(X|B) a∧b a∨b a+ lim supn↑∞ xn lim inf n↑∞ xn a·b |a| X∼ν the the the the the the the the the the the the the the the set of real numbers set of non-negative real numbers [0. b) the set of random variables X with E|X|p < ∞ Table 1. 1. . b} max{a. . b} max{a. . 1. . p) unif(a. . 0. . . . ω ∈ A} / distribution function of a random variable X mass function of a discrete random variable X density function of an absolutely continuous random variable X characteristic function of X expected value of the random variable X variance of X covariance of X and Y conditional expectation of X given the event B min{a.} set of complex numbers set of integers {. 0} the limit superior of the sequence x1 . . Euclidean inner (or dot) product in Rn . . Notation 88 . ∞) set of natural numbers {1. . |a| = (a · a)1/2 n i=1 ai b i 1A N (µ. −2. b) Lp the random variable X is distributed as the probability measure ν the indicator function of the event A the normal distribution with mean µ and variance σ 2 the n-dimensional normal distribution with mean µ ∈ Rn and variance V ∈ Rn×n the binomial distribution with parameters n and p the uniform distribution on the interval (a. a · b = Euclidean norm in Rn . 2. . σ 2 ) Nn (µ. V ) bin(n. . the limit inferior of the sequence x1 . . . 2. x2 .

51 multi-period. 87 Breeden–Litzenberger formula. 38 budget constraint continuous time. 9. 55 multi-period. 20 discrete random variable.s. 15. 8 call option. 83 density process. 83 central limit theorem. 84 binomial random variable. 26 one-period. 21 Bernoulli random variable. 86.Index 1FTAP one-period. 72 density function. 24 admissible trading strategy. 60 Black–Scholes PDE. 62 Black–Scholes model. 13. 53 almost sure event. 67 Borel sigma-ﬁeld. 19 equivalent measures. 21 attainable claim continuous time. 21 second multi-period. 9. 9 2FTAP one-period. 21 characterization. 59 ﬁltration. 52 Dupire’s formula. 29 one-period. 31 one-period. 12 Cameron–Martin–Girsanov theorem. 54 multi-period. 81 absolutely continuous random variable. 87 Feynman–Kac PDE. 83 dominated convergence theorem. 26 given a sigma-ﬁeld. 54 multi-period. 84 Black–Scholes formula. 15 conditional expectation existence and uniqueness. 21 . 65 Brownian motion. 30 one-period. 61 bond. 25 one-period. 31 arbitrage continuous time. 55 one-period. 65 Chebychev’s inequality. 83 adapted process. 85 Cox–Ingersoll–Ross model. 26 given an event. 35 one-period. 18 European contingent claims. 13. 81 Borel–Cantelli lemmas.. 87 Doob decomposition. 84 Fatou’s lemma. 72 complete market continuous time. 85 CIR model. 48 Cauchy–Schwarz inequality. 24 forward rate. 64 equivalent martingale measure continuous time. 69 fundamental theorem of asset pricing ﬁrst multi-period. 29 no time horizon. 33 doubling strategy. 30 characterization. 30 exponential random variable. 15 a. 35 discounted prices. 81 American contingent claims. 87 89 CEV model.

81 sigma-ﬁeld. 48 H¨lder’s inequality. 83 probability mass function. 41 pricing kernel one-period. 14 quadratic covariation. 83 Kennedy model. 19 e objective probability measure. 81 num´raire. 25 one-period. 78 Hull–White extension of Vasicek. 84 Gaussian random vector. 34 Poisson random variable. 82 interest rate term structure. 84 Girsanov’s theorem. 39 quadratic variation. 28 . 85 independent random variables. 83 put option. 19 stochastic Fubini theorem. 82 Snell envelope. 83 measurable with respect to a sigma-ﬁeld. 67 sigma-algebra. 35 replicable claim one-period. 39 Radon–Nikodym derivative. 66 self-ﬁnancing condition continuous time. 87 multivariate Gaussian. 81 probability. 51 multi-period. 47 Itˆ’s isometry. 70 Itˆ process. 44 of Brownian motion.i. 27 Lebesgue measure. 85 indicator function. 78 i. 20 risk-neutral measure one-period.GARCH model. 27 normal random variable. 44 separating hyperplane theorem. 35 Markov’s inequality. 24 measure. 19 optimal stopping time. 85 natural ﬁltration of a process. 84 normal random vector. 15 put-call parity. 44 o Itˆ’s formula o scalar version. 82 state price density. 27 martingale representation theorem. 9 statistical probability measure. 79 Kolmogorov equation. 59 law of iterated expectations. 43 locally equivalent measure. 63 incomplete market one-period. 76 Ho–Lee model. 85 martingale. 9 probability density function. 44 of independent Brownian motions. 85 90 Novikov’s criterion. 82 integrable random variable.d. 67 square-integarable random variable. 8 semimartingale. 10 short interest rate. 28 mass function. 85 geometric random variable. 66 Gaussian random variable. 76 Heston model.. 81 monotone convergence theorem. 19 Radon–Nikodym theorem. 41 discrete time. 40 simple random variable. 15 independent events. 84 predictable process continuous time. 85 implied volatility. 81 simple predictable integrand. 65 historical probability measure. 85 multivariate normal distribution. 19 HJM drift condition. 20 SABR model. 83 o Heath–Jarrow–Morton drift condition. 48 martingale transform. 18 multi-period. 44 vector version. 41 o Jensen’s inequality. 33 spot interest rate. 77 stochastic integral discrete time. 81 local martingale. 25 predictable sigma-ﬁeld. 48 null event. 13 risk-free asset one-period.

32 supporting hyperplane theorem. 84 usual conditions. 40 Vasicek model. 54 super-replicatation of American option.stopping time. 10 term structure of interest rates. 69 zero-coupon bond. 27 trading strategy. 25 trivial sigma-ﬁeld. 71 yield curve. 87 suicide strategy. 67 91 . 31 strong law of large numbers. 70 tower property. 81 uniform random variable.

- stochint
- Reading and Understanding Economics
- Beyond Value at Risk
- Quality of Earnings
- Quantitative Methods in Derivatives Pricing
- Column 24 Understanding the Kelly Criterion 2
- Column Understanding the Kelly Criterion
- Pairs Trading--Quantitative Methods and Analysis (Wiley Books)
- Putting Volatility to Work
- Understanding Fortunes Formula
- Modeling Derivatives in C Tqw Darksiderg
- blatt11
- TilsonBehavioralFinance
- Number Play - Gyles Brandreth
- ekonomikos knyga
- Quantitative Methods in Mgmt Ref1
- The Mathematics of Gambling Edward O Thorpe
- The Quants by Scott Patterson - Excerpt
- Gatarek D. Bachert P. Maksymiuk R. the LIBOR Market Model in Practice
- Rational Choice
- Risk Neutral+Valuation Pricing+and+Hedging+of+Financial+Derivatives
- OR
- Curtis Arnold - The PPS Trading System.pdf
- Rachev S.T. Financial Eco No Metrics (LN, Karlsruhe 2006)(146s)_FL
- A Map and Simple Heuristic to Detect Fragility,Antifragility, and Model Error NNT
- 3642125972
- Complete Newbie s Guide to Online Forex Trading
- LP_driver
- Lecture Notes
- The Future Has Thicker Tails Than The Past

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue listening from where you left off, or restart the preview.

scribd