This document discusses risk-neutral valuation of options in discrete time using binomial trees. It introduces key concepts like put-call parity and pricing American options by working backwards. It also covers estimating the up and down factors in the binomial tree based on volatility, and discusses pricing of options on currencies, futures contracts, and using state prices. The document compares using true probabilities versus risk-neutral probabilities for valuation.
This document discusses risk-neutral valuation of options in discrete time using binomial trees. It introduces key concepts like put-call parity and pricing American options by working backwards. It also covers estimating the up and down factors in the binomial tree based on volatility, and discusses pricing of options on currencies, futures contracts, and using state prices. The document compares using true probabilities versus risk-neutral probabilities for valuation.
This document discusses risk-neutral valuation of options in discrete time using binomial trees. It introduces key concepts like put-call parity and pricing American options by working backwards. It also covers estimating the up and down factors in the binomial tree based on volatility, and discusses pricing of options on currencies, futures contracts, and using state prices. The document compares using true probabilities versus risk-neutral probabilities for valuation.
Recalling Put-Call Parity: Pricing American Options
Call(St , K, T ) − P ut(St , K, T ) Work backwards, step by step; The end-nodes are the = St e−δ(T −t) − Ke−r(T −t) same as that of European options. Each non end-node = P Ft,T − Ke−r(T −t) value Ck is equal to: = e−r(T −t) [Ft,T − K] max earlypayoff i , e−rh [p∗ Sk u + (1 − p∗ )Sk d]
Binomial Trees (one period)
Estimating u and d Given the volatility of the underlying stock σ (the stan- dard deviation of the stock return) we can calculate u and d as follows: Cu − Cd Cu − Cd ∆ = e−δh = e−δh Forward tree: S0 (u − d) Su − Sd √ √ u = exp{(r−δ)h+σ h}, d = exp{(r−δ)h−σ h} uCd − dCu Su Cd − Sd Cu B = e−rh = e−rh √ u−d Su − Sd ⇒ p∗ = (1 + eσ h −1 ) (r−δ)h u − e(r−δ)h
e −d Cox-Ross-Rubinstein binomial tree: C0 = e−rh Cu + Cd √ √ u−d u−d u = exp{σ h}, d = exp{−σ h} = e−rh [p∗ Cu + (1 − p∗ )Cd ] = e−rh E∗ [Ch ] Jarrow-Rudd binomial tree: p∗ is the risk-neutral probability of an increase in the 1 √ u = exp{(r − δ − σ 2 )h + σ h} stock price in one step h. 2 1 2 √ e(r−δ)h − d S0 e(r−δ)h − Sd d = exp{(r − δ − σ )h − σ h} p∗ = = 2 u−d Su − Sd The stock expected future value under the risk-neutral measure is the forward price: Options On Currencies E∗ (Sh ) = S0 e(r−δ)h It is the same treatment, where δ = rf the foreign cu- rrency risk free interest rate. √ The price of the replicating portfolio and the derivati- u = exp{(r − rf )h + σ h}; ve: √ C0 = ∆S0 + B = e−rh E∗ [Ch ] d = exp{(r − rf )h − σ h} σ is the volatility of the exchange rate. Multiperiod Binomial Trees Call(K, T ) − P ut(K, T ) = x0 e−rf T − Ke−rT ;
C0 = e−rT E∗ [CT ]; T = nh x0 = exchange rate at time t = 0 (dmstc/frng).
For a tree with n steps, there are n+1 end nodes.
Options on Future Contracts The k th end-node has value S0 uk dn−k ; It is the same treatment with δ = r. k = 0, 1, ..., n. √ √ u = exp{σ h}, d = exp{−σ h}, p∗ = 1−d u−d The risk-neutral probability that the k th end no- de will be reached is (nk )(p∗ )k (1 − p∗ )n−k Call(k, T ) − P ut(k, T ) = F0 e−rT − ke−rT
1 Module 1: Risk-Neutral Valuation in Discrete-time
Pricing with true probabilities State Prices
Let α be the expected return on a stock in the real Consider a Stock, world, i.e., E(Sh ) = S0 exp{(α − δ)h}. H is the security that pays 1 dollar at time h if the stock price goes up after h years. e(α−δ)h − d Let QH be the price of H p= u−d L is the security that pays 1 dollar at time h if is the true probability that the stock price will go up the stock price goes down after h years. after a time period of h. Let QL be the price of L Let γ be the appropriate continuously compounded, For any derivative that pays CH in the up state and annualized discount rate for the option: CL in the down state S0 ∆ B ⇒ eγh = eαh + erh ⇒ price = QH CH + QL CL S0 ∆ + B S0 ∆ + B Then derivative price is To obtain QH , QL we replicate the bond and the pre- C0 = exp(−γh)[pCu + (1 − p)Cd ] paid forward contract of the stock. . QH + QL = e−rh
Using true probabilities to price an option is not SH QH + SL QL = S0 e−δh useful. ⇒ When extended to multi-period tree, we have to calculate ∆, B and γ at each node. S0 e−δh − SL e−rh Sh e−rh − S0 e−δh QH = ; QL = Unless we are given p and γ values, we shouldn’t SH − SL SH − SL use the P measure for pricing. Notice that to com- pute γ, we need to obtain ∆ and B, but if ∆ and Relation between state prices and other valuation met- B are available, we can calculate de option price hods. as ∆S0 + B QH p∗ = QA ; p∗A = QA +Q B +QL QH + QL QH = pUH , QL = (1 − p)UL
e dx) e dxdy = e rdrdθ − 3) i − z j + (cos z) k is conservative. ≤ t ≤ 2π then x ds = 40π. · dr = 0. ∂ (x, y, z) ∂ (u, v, w) − 1, y = 3v − 4, z= − 4) is 3