This action might not be possible to undo. Are you sure you want to continue?
Zicklin School of Business, Baruch College
(Hull chapter: 12, 13, 14)
The Black-Scholes Model
1 / 19
The BSM model: Continuous states (stock price can be anything between 0 and ∞) and continuous time (time goes continuously). I treat all these variations as the same concept and call them indiscriminately the BSM model (combine chapters 13&14).The Black-Scholes-Merton (BSM) model Black and Scholes (1973) and Merton (1973) derive option prices under the following assumption on the stock price dynamics. BSM proposed the model for stock option pricing. Later. Liuren Wu The Black-Scholes Model Options Markets 2 / 19 . Black passed away. dSt = µSt dt + σSt dWt (explained later) The binomial model: Discrete states and discrete time (The number of possible stock prices and time steps are both ﬁnite). the model has been extended/twisted to price currency options (Garman&Kohlhagen) and options on futures (Black). Scholes and Merton won Nobel price.
but nowhere diﬀerentiable. Liuren Wu The Black-Scholes Model Options Markets 3 / 19 . It is the idealization of the trajectory of a single particle being constantly bombarded by an inﬁnite number of inﬁnitessimally small random forces. your pen would run out of ink before one second had elapsed. a Brownian motion. or else it dies. The ﬁrst who brought Brownian motion to ﬁnance is Bachelier in his 1900 PhD thesis: The theory of speculation. If you sum the absolute values of price changes over a day (or any time horizon) implied by the model. If you tried to accurately draw a Brownian motion sample path. The sample paths of a Brownian motion are continuous over time. a Brownian motion must always be moving. Like a shark. or a Wiener process. you get an inﬁnite number.Primer on continuous time process dSt = µSt dt + σSt dWt The driver of the process is Wt .
φ(0. dt). dt denotes an extremely thin slice of time (smaller than 1 milisecond). (Also referred to as Gaussian.) Everybody believes in the normal approximation. the mathematicians because they believe it is an experimental fact! The process is made of independent normal increments dWt ∼ φ(0. increments over non-overlapping time periods are independent: For (t1 > t2 > t3 ). (Wt3 − Wt2 ) ∼ φ(0. ∆t denotes a ﬁnite time step (say. Liuren Wu The Black-Scholes Model Options Markets 4 / 19 . “d” is the continuous time limit of the discrete time diﬀerence (∆). the experimenters because they believe it is a mathematical theorem. 3 months). t). t3 − t2 ) is independent of (Wt2 − Wt1 ) ∼ φ(0. t2 − t1 ). dWt = Wt+dt − Wt denotes the instantaneous increment (change) of a Brownian motion. By extension. Similarly.Properties of a Brownian motion dSt = µSt dt + σSt dWt The process Wt generates a random variable that is normally distributed with mean 0 and variance t. It is so thin that it is often referred to as instantaneous.
b 2 ). V ). Liuren Wu The Black-Scholes Model Options Markets 5 / 19 . σ 2 is the annualized variance of the instantaneous return — instantaneous return variance. If y ∼ φ(m. 1). The instantaneous return dS S = µdt + σdWt ∼ φ(µdt. Under the BSM model. b 2 V ). µ is the annualized mean of the instantaneous return — instantaneous mean return.Properties of a normally distributed random variable dSt = µSt dt + σSt dWt If X ∼ φ(0. Since dWt ∼ φ(0. σ 2 dt). then a + bX ∼ φ(a. σ 2 St2 dt). the instantaneous price change dSt = µSt dt + σSt dWt ∼ φ(µSt dt. then a + by ∼ φ(a + bm. dt). σ is the annualized standard deviation of the instantaneous return — instantaneous return volatility.
the price change is lognormally distributed. Instantaneously. (By Ito’s lemma) d ln St ∼ φ(µdt − 1 σ 2 dt. Over longer horizons. σ 2 t). µ is often referred to as the drift. σ 2 dt). σ 2 (T − t) . and σ the diﬀusion of the process. the stock price change is normally distributed. φ(µSt dt. ln St = ln S0 + µt − 1 σ 2 t + σWt 2 Liuren Wu The Black-Scholes Model Options Markets 6 / 19 . σ 2 St2 dt). The log return (continuous compounded return) is normally distributed over all horizons: 1 d ln St = µ − 2 σ 2 dt + σdWt . 2 ln St ∼ φ(ln S0 + µt − 1 σ 2 t. 2 Integral form: St = S0 e µt− 2 σ 1 2 t+σWt .Geometric Brownian motion dSt /St = µdt + σdWt The stock price is said to follow a geometric Brownian motion. 2 ln ST /St ∼ φ µ − 1 σ 2 (T − t).
2 0 −1 −0.02 0.018 0.002 0 50 100 150 St 200 250 The earliest application of Brownian motion to ﬁnance is Louis Bachelier in his dissertation (1900) “Theory of Speculation.8 1.8 0.016 0.4 0.012 0.004 0.5 1 PDF of St 1.008 0.01 0. 2 1.5 0 ln(St/S0) 0. S0 = 100.2 0. σ = 20%.4 µ = 10%.6 0. 0.006 0.6 1.” He speciﬁed the stock price as following a Brownian motion with drift: dSt = µdt + σdWt Liuren Wu The Black-Scholes Model Options Markets 7 / 19 . t = 1.Normal versus lognormal distribution dSt = µSt dt + σSt dWt .014 S0=100 PDF of ln(St/S0) 1 0.
0.03 −0.01 140 120 100 −0.05 0.02 −0. √ Convert them into daily log returns: Rd = (µ − 1 σ 2 )∆t + σ ∆tε.02 Daily returns Stock price 50 100 150 Days 200 250 300 0. 8 / 19 Options Markets . 2 Discretize to daily intervals dt ≈ ∆t = 1/252. 2 Convert returns into stock price sample paths: St = S0 e Liuren Wu The Black-Scholes Model P252 d=1 Rd .Simulate 100 stock price sample paths dSt = µSt dt + σSt dWt .04 0 80 60 0 50 100 150 days 200 250 300 Stock with the return process: d ln St = (µ − 1 σ 2 )dt + σdWt . S0 = 100.04 0.01 0 −0. t = 1. σ = 20%. Draw standard normal random variables ε(100 × 252) ∼ φ(0. 200 180 160 0.03 µ = 10%. 1).
Reversely. the binomial tree converges to the distribution behavior of the geometric Brownian motion. if we cut ∆t small enough and add enough time steps. Only σ matters. The Brownian motion dynamics imply that if we slice the time thin enough (dt).The key idea behind BSM The option price and the stock price depend on the same underlying source of uncertainty. The portfolio is riskless (under this thin slice of time interval) and must earn the riskfree rate. we can combine the option with the stock to form a riskfree portfolio. Recall our hedging argument: Choose ∆ such that f − ∆S is riskfree. it behaves like a binominal tree. Magic: µ does not matter for this portfolio and hence does not matter for the option valuation. We do not need to worry about risk and risk premium if we can hedge away the risk completely. Liuren Wu The Black-Scholes Model Options Markets 9 / 19 . Under this thin slice of time interval.
.Partial diﬀerential equation The hedging argument mathematically leads to the following partial diﬀerential equation: ∂f ∂f 1 ∂2f + (r − q)S + σ 2 S 2 2 = rf ∂t ∂S 2 ∂S At nowhere do we see µ. BSM can derive analytical formulas for call and put option value. Similar formula had been derived before based on distributional (normal return) argument. subject to the terminal payoﬀ condition of the derivative (e. Plus. The realization that option valuation does not depend on µ is big. fT = (ST − K )+ for a European call option). but µ (risk premium) was still in. The only free parameter is σ (as in the binominal model). Solving this PDE.g. Liuren Wu The Black-Scholes Model Options Markets 10 / 19 . it provides a way to hedge the option position.
σ T −t Recall: Ft = St e (r −q)(T −t) .2 = ln(Ft /K )± 1 σ 2 (T −t) √ 2 . Once I know call value. 1 ln(St /K )+(r −q)(T −t)+ 2 σ 2 (T −t) √ . σ T −t ln(St /K )+(r −q)(T −t)− 1 σ 2 (T −t) 2 √ = σ T −t √ d1 − σ T − t. with d1. Liuren Wu The Black-Scholes Model Options Markets 11 / 19 .The BSM formulae ct pt where d1 d2 = = = St e −q(T −t) N(d1 ) − Ke −r (T −t) N(d2 ). I can obtain put value via put-call parity: ct − pt = e −r (T −t) [Ft − Kt ]. Black derived a variant of the formula for futures (which I like better): ct = e −r (T −t) [Ft N(d1 ) − KN(d2 )]. = −St e −q(T −t) N(−d1 ) + Ke −r (T −t) N(−d2 ).
as St becomes very small or K becomes very large. pt = e −r (T −t) [−Ft + K ]. Properties of the BSM formula: As St becomes very large or K becomes very small. ln(Ft /K ) ↑ ∞.2 = See tables at the end of the book for its values. d1. Most software packages (including excel) has eﬃcient ways to computing this function. 1)) is less than x. Liuren Wu The Black-Scholes Model Options Markets 12 / 19 . N(d1 ) = N(d2 ) = 1. ct = e −r (T −t) [Ft N(d1 ) − KN(d2 )] . ln(Ft /K ) ↑ −∞. N(−d1 ) = N(−d2 ) = 1.Cumulative normal distribution 1 ln(Ft /K ) ± 2 σ 2 (T − t) √ σ T −t N(x) denotes the cumulative normal distribution. Similarly. ct = e −r (T −t) [Ft − K ] . which measures the probability that a normally distributed variable with a mean of zero and a standard deviation of 1 (φ(0.
The BSM model says that IV = σ. In reality. Traders and brokers often quote implied volatilities rather than dollar prices. Liuren Wu The Black-Scholes Model Options Markets 13 / 19 . The only unknown (and hence free) parameter is σ. (standard deviation calculation). Alternatively. we can choose σ to match the observed option price — implied volatility (IV). There is a one-to-one correspondence between prices and implied volatilities. (K . d1. ct = e −r (T −t) [Ft N(d1 ) − KN(d2 )] . the implied volatility calculated from diﬀerent options (across strikes. dates) are usually diﬀerent. maturities.2 = We can estimate σ from time series return. t.Implied volatility 1 ln(Ft /K ) ± 2 σ 2 (T − t) √ σ T −t Since Ft (or St ) is observable from the underlying stock or futures market. T ) are speciﬁed in the contract.
and other types of carrying costs may complicate the pricing formula a little bit. foreign interest rates. as long as we observe the forward price (or we can derive the forward price). Know how to price a forward. A simpler approach: Assume that the underlying futures/forwards price (of the same maturity of course) process follows a geometric Brownian motion. Dividends. we can use (some versions) of the BSM formula to price European options. and use the Black formula. Liuren Wu The Black-Scholes Model Options Markets 14 / 19 .Options on what? Why does it matter? As long as we assume that the underlying security price follows a geometric Brownian motion. we do not need to worry about dividends or foreign interest rates — They are all accounted for in the forward pricing. Then.
we can also assume that there exists such an artiﬁcial risk-neutral world. Liuren Wu The Black-Scholes Model Options Markets 15 / 19 . dSt /St = (r − q)dt + σdWt . Risk premiums (recall CAPM) are hidden in the risk-neutral probabilities. The stock price dynamics under the risk-neutral world becomes. If in the real world. in which the expected returns on all assets earn risk-free rate. Otherwise. we derive a set of risk-neutral probabilities such that we can calculate the expected payoﬀ from the option and discount them using the riskfree rate. people are indeed risk-neutral. Under the BSM model.Risk-neutral valuation Recall: Under the binomial model. Simply replace the actual expected return (µ) with the return from a risk-neutral world (r − q) [ex-dividend return]. they are diﬀerent. the risk-neutral probabilities are the same as the real-world probabilities.
under risk-neutral probabilities. investing in all securities make the riskfree rate as the total return. Liuren Wu The Black-Scholes Model Options Markets 16 / 19 . then the risk-neutral expected return from stock price appreciation is (r − q). In the risk-neutral world. such as the total expected return is: dividend yield+ price appreciation =r . If a stock pays a dividend yield of q. Regard q as rf and value options as if they are the same. Investing in a currency earns the foreign interest rate rf similar to dividend yield. Hence. the risk-neutral expected currency appreciation is (r − rf ) so that the total expected return is still r .The risk-neutral return on spots dSt /St = (r − q)dt + σdWt .
the forward price dynamics has zero mean (driftless) under the risk-neutral probabilities: dFt /Ft = σdWt . we do not make any excess return. Liuren Wu The Black-Scholes Model Options Markets 17 / 19 . The carrying costs are all hidden under the forward price.The risk-neutral return on forwards/futures If we sign a forward contract. Under the risk-neutral world. Hence. making the pricing equations simpler. we do not pay anything upfront and we do not receive anything in the middle (no dividends or foreign interest rates). Any P&L at expiry is excess return.
estimated based on historical data and other insights about the company. Q is not about getting close to the actual probability. Q P: Actual probabilities that cashﬂows will be high or low. Q: “Risk-neutral” probabilities that we can use to aggregate expected future payoﬀs and discount them back with riskfree rate. It is related to real-time scenarios. Valuation is all about getting the forecasts right and assigning the appropriate price for the forecasted risk — fair wrt future cashﬂows and your risk preference. regardless of how risky the cash ﬂow is. We just care about what all the possible scenarios are and whether our hedging works under all scenarios. Liuren Wu The Black-Scholes Model Options Markets 18 / 19 . Since the intention is to hedge away risk under all scenarios and discount back with riskfree rate. but it has nothing to do with real-time probability.Readings behind the technical jargons: P v. but about being fair wrt the prices of securities that you use for hedging. we do not really care about the actual probability of each scenario happening.
Liuren Wu The Black-Scholes Model Options Markets 19 / 19 . Can go back and forth with the put-call parity conditions. Understand forward pricing and link option pricing to forward pricing.Summary Understand the basic properties of normally distributed random variables. lower and upper bonds. either in forward or in spot notation. Map a stochastic process to a random variable. Understand the link between BSM and the binomial model. Memorize the BSM formula (any version).
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 listening from where you left off, or restart the preview.