IEOR E4707: Financial Engineering: Continuous-Time Models Fall 2013

c 2013 by Martin Haugh

Black-Scholes and the Volatility Surface

When we studied discrete-time models we used martingale pricing to derive the Black-Scholes formula for
European options. It was clear, however, that we could also have used a replicating strategy argument to derive
the formula. In this part of the course, we will use the replicating strategy argument in continuous time to
derive the Black-Scholes partial differential equation. We will use this PDE and the Feynman-Kac equation to
demonstrate that the price we obtain from the replicating strategy argument is consistent with martingale
pricing.
We will also discuss the weaknesses of the Black-Scholes model, i.e. geometric Brownian motion, and this leads
us naturally to the concept of the volatility surface which we will describe in some detail. We will also derive and
study the Black-Scholes Greeks and discuss how they are used in practice to hedge option portfolios. We will
also derive Black’s formula which emphasizes the role of the forward when pricing European options. Finally, we
will discuss the pricing of other derivative securities and which securities can be priced uniquely given the
volatility surface. Change of numeraire / measure methods will also be demonstrated to price exchange options.

1 The Black-Scholes PDE

We now derive the Black-Scholes PDE for a call-option on a non-dividend paying stock with strike K and
maturity T . We assume that the stock price follows a geometric Brownian motion so that

dSt = µSt dt + σSt dWt (1)

where Wt is a standard Brownian motion. We also assume that interest rates are constant so that $1 invested in
the cash account at time 0 will be worth Bt := $ exp(rt) at time t. We will denote by C(S, t) the value of the
call option at time t. By Itô’s lemma we know that

∂2C
 
∂C ∂C 1 ∂C
dC(S, t) = µSt + + σ 2 S 2 2 dt + σSt dWt (2)
∂S ∂t 2 ∂S ∂S

Let us now consider a self-financing trading strategy where at each time t we hold xt units of the cash account
and yt units of the stock. Then Pt , the time t value of this strategy satisfies

Pt = xt Bt + yt St . (3)

We will choose xt and yt in such a way that the strategy replicates the value of the option. The self-financing
assumption implies that

dPt = xt dBt + yt dSt (4)

= rxt Bt dt + yt (µSt dt + σSt dWt )
= (rxt Bt + yt µSt ) dt + yt σSt dWt . (5)

Note that (4) is consistent with our earlier definition1 of self-financing. In particular, any gains or losses on the
portfolio are due entirely to gains or losses in the underlying securities, i.e. the cash-account and stock, and not
due to changes in the holdings xt and yt .
1 It is also worth pointing out that the mathematical definition of self-financing is obtained by applying Itô’s Lemma to (3)

and setting the result equal to the right-hand-side of (4).

Black-Scholes and the Volatility Surface 2

Returning to our derivation, we can equate terms in (2) with the corresponding terms in (5) to obtain
∂C
yt = (6)
∂S
∂C 1 ∂2C
rxt Bt = + σ2 S 2 2 . (7)
∂t 2 ∂S
If we set C0 = P0 , the initial value of our self-financing strategy, then it must be the case that Ct = Pt for all t
since C and P have the same dynamics. This is true by construction after we equated terms in (2) with the
corresponding terms in (5). Substituting (6) and (7) into (3) we obtain
∂C ∂C 1 ∂2C
rSt + + σ 2 S 2 2 − rC = 0, (8)
∂S ∂t 2 ∂S
the Black-Scholes PDE. In order to solve (8) boundary conditions must also be provided. In the case of our
call option those conditions are: C(S, T ) = max(S − K, 0), C(0, t) = 0 for all t and C(S, t) → S as S → ∞.
The solution to (8) in the case of a call option is
C(S, t) = St Φ(d1 ) − e−r(T −t) KΦ(d2 ) (9)

St


log + (r + σ 2 /2)(T − t)
K
where d1 = √
σ T −t
√
and d2 = d1 − σ T − t
and Φ(·) is the CDF of the standard normal distribution. One way to confirm (9) is to compute the various
partial derivatives, then substitute them into (8) and check that (8) holds. The price of a European put-option
can also now be easily computed from put-call parity and (9).
The most interesting feature of the Black-Scholes PDE (8) is that µ does not appear2 anywhere. Note that the
Black-Scholes PDE would also hold if we had assumed that µ = r. However, if µ = r then investors would not
demand a premium for holding the stock. Since this would generally only hold if investors were risk-neutral, this
method of derivatives pricing came to be known as risk-neutral pricing.

2 Martingale Pricing
We can easily see that the Black-Scholes PDE in (8) is consistent with martingale pricing. Martingale pricing
theory states that deflated security prices are martingales. If we deflate by the cash account, then the deflated
stock price process, Yt say, satisfies Yt := St /Bt . Then Itô’s Lemma and Girsanov’s Theorem imply
dYt = (µ − r)Yt dt + σYt dWt
= (µ − r)Yt dt + σYt (dWtQ − ηt dt)
= (µ − r − σηt )Yt dt + σYt dWtQ .
where Q denotes a new probability measure and WtQ is a Q-Brownian motion. But we know from martingale
pricing that if Q is an equivalent martingale measure then it must be the case that Yt is a martingale. This then
implies that ηt = (µ − r)/σ for all t. It also implies that the dynamics of St satisfy
dSt = µSt dt + σSt dWt
= rSt dt + σSt dWtQ . (10)
Using (10), we can now derive (9) using martingale pricing. In particular, we have
h i
C(St , t) = EQt e −r(T −t)
max(S T − K, 0) (11)
2 The discrete-time counterpart to this observation was when we observed that the true probabilities of up-moves and down-

moves did not have an impact on option prices.

Black-Scholes and the Volatility Surface 3

where 2
/2)(T −t)+σ(WTQ −WtQ )
ST = St e(r−σ
is log-normally distributed. While the calculations are a little tedious, it is straightforward to solve (11) and
obtain (9) as the solution.

Dividends
If we assume that the stock pays a continuous dividend yield of q, i.e. the dividend paid over the interval
(t, t + dt] equals qSt dt, then the dynamics of the stock price satisfy

dSt = (r − q)St dt + σSt dWtQ . (12)

In this case the total gain process, i.e. the capital gain or loss from holding the security plus accumulated
dividends, is a Q-martingale. The call option price is still given by (11) but now with
2
/2)(T −t)+σWTQ
ST = St e(r−q−σ .

In this case the call option price is given by

C(S, t) = e−q(T −t) St Φ(d1 ) − e−r(T −t) KΦ(d2 ) (13)

St


log + (r − q + σ 2 /2)(T − t)
K
where d1 = √
σ T −t
√
and d2 = d1 − σ T − t.

Exercise 1 Follow the argument of the previous section to derive the Black-Scholes PDE when the stock pays
a continuous dividend yield of q.

Feynman-Kac
We have already seen that the Black-Scholes formula can be derived from either the martingale pricing approach
or the replicating strategy / risk neutral PDE approach. In fact we can go directly from the Black-Scholes PDE
to the martingale pricing equation of (11) using the Feynman-Kac formula.

Exercise 2 Derive the same PDE as in Exercise 1 but this time by using (12) and applying the Feynman-Kac
formula to an analogous expression to (11).

While the original derivation of the Black-Scholes formula was based on the PDE approach, the martingale
pricing approach is more general and often more amenable to computation. For example, numerical methods for
solving PDEs are usually too slow if the number of dimensions are greater3 than three. Monte-Carlo methods
can be used to evaluate (11) regardless of the number of state variables, however, as long as we can simulate
from the relevant probability distributions. The martingale pricing approach can also be used for problems that
are non-Markovian. This is not the case for the PDE approach.

The Black-Scholes Model is Complete

It is worth mentioning that the Black-Scholes model is a complete model and so every derivative security is
attainable or replicable. In particular, this means that every security can be priced uniquely. Completeness
follows from the fact that the EMM in (10) is unique: the only possible choice for ηt was ηt = (µ − r)/σ.

3 The Black-Scholes PDE is only two-dimensional with state variable t and S. The Black-Scholes PDE is therefore easy to

solve numerically.
Black-Scholes and the Volatility Surface 4

3 The Volatility Surface

The Black-Scholes model is an elegant model but it does not perform very well in practice. For example, it is
well known that stock prices jump on occasions and do not always move in the smooth manner predicted by the
GBM motion model. Stock prices also tend to have fatter tails than those predicted by GBM. Finally, if the
Black-Scholes model were correct then we should have a flat implied volatility surface. The volatility surface is a
function of strike, K, and time-to-maturity, T , and is defined implicitly

C(S, K, T ) := BS (S, T, r, q, K, σ(K, T )) (14)

where C(S, K, T ) denotes the current market price of a call option with time-to-maturity T and strike K, and
BS(·) is the Black-Scholes formula for pricing a call option. In other words, σ(K, T ) is the volatility that, when
substituted into the Black-Scholes formula, gives the market price, C(S, K, T ). Because the Black-Scholes
formula is continuous and increasing in σ, there will always4 be a unique solution, σ(K, T ). If the Black-Scholes
model were correct then the volatility surface would be flat with σ(K, T ) = σ for all K and T . In practice,
however, not only is the volatility surface not flat but it actually varies, often significantly, with time.

Figure 1: The Volatility Surface

In Figure 1 above we see a snapshot of the5 volatility surface for the Eurostoxx 50 index on November 28th ,
2007. The principal features of the volatility surface is that options with lower strikes tend to have higher
implied volatilities. For a given maturity, T , this feature is typically referred to as the volatility skew or smile.
For a given strike, K, the implied volatility can be either increasing or decreasing with time-to-maturity. In
general, however, σ(K, T ) tends to converge to a constant as T → ∞. For T small, however, we often observe
an inverted volatility surface with short-term options having much higher volatilities than longer-term options.
This is particularly true in times of market stress.
It is worth pointing out that different implementations6 of Black-Scholes will result in different implied volatility
surfaces. If the implementations are correct, however, then we would expect the volatility surfaces to be very
4 Assuming there is no arbitrage in the market-place.
5 Note that by put-call parity the implied volatility σ(K, T ) for a given European call option will be also be the implied
volatility for a European put option of the same strike and maturity. Hence we can talk about “the” implied volatility surface.
6 For example different methods of handling dividends would result in different implementations.
Black-Scholes and the Volatility Surface 5

similar in shape. Single-stock options are generally American and in this case, put and call options will typically
give rise to different surfaces. Note that put-call parity does not apply for American options.
Clearly then the Black-Scholes model is far from accurate and market participants are well aware of this.
However, the language of Black-Scholes is pervasive. Every trading desk computes the Black-Scholes implied
volatility surface and the Greeks they compute and use are Black-Scholes Greeks.

Arbitrage Constraints on the Volatility Surface

The shape of the implied volatility surface is constrained by the absence of arbitrage. In particular:
1. We must have σ(K, T ) ≥ 0 for all strikes K and expirations T .
2. At any given maturity, T , the skew cannot be too steep. Otherwise butterfly arbitrages will exist. For
example fix a maturity, T and consider put two options with strikes K1 < K2 . If there is no arbitrage then
it must be the case (why?) that P (K1 ) < P (K2 ) where P (Ki ) is the price of the put option with strike
Ki . However, if the skew is too steep then we would obtain (why?) P (K1 ) > P (K2 ).
3. Likewise the term structure of implied volatility cannot be too inverted. Otherwise calendar spread
arbitrages will exist. This is most easily seen in the case where r = q = 0. Then, fixing a strike K, we can
let Ct (T ) denote the time t price of a call option with strike K and maturity T . Martingale pricing implies
that Ct (T ) = Et [(ST − K)+ ]. We have seen before that (ST − K)+ is a Q-submartingale and now
standard martingale results can be used to show that Ct (T ) must be non-decreasing in T . This would be
violated (Why?) if the term structure of implied volatility was too inverted.
In practice the implied volatility surface will not violate any of these restrictions as otherwise there would be an
arbitrage in the market. These restrictions can be difficult to enforce, however, when we are “bumping” or
“stressing” the volatility surface, a task that is commonly performed for risk management purposes.

Why is there a Skew?

For stocks and stock indices the shape of the volatility surface is always changing. There is generally a skew,
however, so that for any fixed maturity, T , the implied volatility decreases with the strike, K. It is most
pronounced at shorter expirations. There are two principal explanations for the skew.
1. Risk aversion which can appear as an explanation in many guises:
(a) Stocks do not follow GBM with a fixed volatility. Markets often jump and jumps to the downside
tend to be larger and more frequent than jumps to the upside.
(b) As markets go down, fear sets in and volatility goes up.
(c) Supply and demand. Investors like to protect their portfolio by purchasing out-of-the-money puts and
so there is more demand for options with lower strikes.
2. The leverage effect which is due to the fact that the total value of company assets, i.e. debt + equity,
is a more natural candidate to follow GBM. If so, then equity volatility should increase as the equity value
decreases. To see this consider the following:

Let V , E and D denote the total value of a company, the company’s equity and the company’s debt,
respectively. Then the fundamental accounting equations states that
V = D + E. (15)
Equation (15) is the basis for the classical structural models that are used to price risky debt and credit
default swaps. Merton (1970’s) recognized that the equity value could be viewed as the value of a call
option on V with strike equal to D.

Let ∆V , ∆E and ∆D be the change in values of V , E and D, respectively. Then

V + ∆V = (E + ∆E) + (D + ∆D) so that
V + ∆V E + ∆E D + ∆D
= +
V V V
Black-Scholes and the Volatility Surface 6
   
E E + ∆E D D + ∆D
= + (16)
V E V D

If the equity component is substantial so that the debt is not too risky, then (16) implies

E
σV ≈ σE
V
where σV and σE are the firm value and equity volatilities, respectively. We therefore have
V
σE ≈ σV . (17)
E

Example 1 (The Leverage Effect)

Suppose, for example, that V = 1, E = .5 and σV = 20%. Then (17) implies σE ≈ 40%. Suppose σV remains
unchanged but that over time the firm loses 20% of its value. Almost all of this loss is borne by equity so that
now (17) implies σE ≈ 53%. σE has therefore increased despite the fact that σV has remained constant.

It is interesting to note that there was little or no skew in the market before the Wall street crash of 1987. So it
appears to be the case that it took the market the best part of two decades before it understood that it was
pricing options incorrectly.

What the Volatility Surface Tells Us

To be clear, we continue to assume that the volatility surface has been constructed from European option prices.
Consider a butterfly strategy centered at K where you are:
1. long a call option with strike K − ∆K
2. long a call with strike K + ∆K
3. short 2 call options with strike K

The value of the butterfly, B0 , at time t = 0, satisfies

B0 = C(K − ∆K, T ) − 2C(K, T ) + C(K + ∆K, T )

≈ e−rT Prob(K − ∆K ≤ ST ≤ K + ∆K) × ∆K/2
≈ e−rT f (K, T ) × 2∆K × ∆K/2
= e−rT f (K, T ) × (∆K)2

where f (K, T ) is the probability density function (PDF) of ST evaluated at K. We therefore have

C(K − ∆K, T ) − 2C(K, T ) + C(K + ∆K, T )

f (K, T ) ≈ erT . (18)
(∆K)2

Letting ∆K → 0 in (18), we obtain

∂2C
f (K, T ) = erT .
∂K 2
The volatility surface therefore gives the marginal risk-neutral distribution of the stock price, ST , for any time,
T . It tells us nothing about the joint distribution of the stock price at multiple times, T1 , . . . , Tn .
This should not be surprising since the volatility surface is constructed from European option prices and the
latter only depend on the marginal distributions of ST .
Black-Scholes and the Volatility Surface 7

Example 2 (Same marginals, different joint distributions)

Suppose there are two time periods, T1 and T2 , of interest and that a non-dividend paying security has
risk-neutral distributions given by
2 √
ST1 = e(r−σ /2)T1 +σ T1 Z1
(19)
2 √ √

(r−σ /2)T2 +σ T2 ρZ1 + 1−ρ2 Z2
ST2 = e (20)

where Z1 and Z2 are independent N(0, 1) random variables. Note that a value of ρ > 0 can capture a
momentum effect and a value of ρ < 0 can capture a mean-reversion effect. We are also implicitly assuming
that S0 = 1.
(A) (B)
Suppose now that there are two securities, A and B say, with prices St and St given by (19) and (20) at
times t = T1 and t = T2 , and with parameters ρ = ρA and ρ = ρB , respectively. Note that the marginal
(A) (B)
distribution of St is identical to the marginal distribution of St for t ∈ {T1 , T2 }. It therefore follows that
options on A and B with the same strike and maturity must have the same price. A and B therefore have
identical volatility surfaces.
But now consider a knock-in put option with strike 1 and expiration T2 . In order to knock-in, the stock price at
time T1 must exceed the barrier price of 1.2. The payoff function is then given by

Payoff = max (1 − ST2 , 0) 1{ST1 ≥1.2} .

Question: Would the knock-in put option on A have the same price as the knock-in put option on B?
Question: How does your answer depend on ρA and ρB ?
Question: What does this say about the ability of the volatility surface to price barrier options?

4 The Greeks
We now turn to the sensitivities of the option prices to the various parameters. These sensitivities, or the
Greeks are usually computed using the Black-Scholes formula, despite the fact that the Black-Scholes model is
known to be a poor approximation to reality. But first we return to put-call parity.

Put-Call Parity
Consider a European call option and a European put option, respectively, each with the same strike, K, and
maturity T . Assuming a continuous dividend yield, q, then put-call parity states

e−rT K + Call Price = e−qT S + Put Price. (21)

This of course follows from a simple arbitrage argument and the fact that both sides of (21) equal max(ST , K)
at time T . Put-call parity is useful for calculating Greeks. For example7 , it implies that Vega(Call) = Vega(Put)
and that Gamma(Call) = Gamma(Put). It is also extremely useful for calibrating dividends and
constructing the volatility surface.

The Greeks
The principal Greeks for European call options are described below. The Greeks for put options can be
calculated in the same manner or via put-call parity.
Definition: The delta of an option is the sensitivity of the option price to a change in the price of the
7 See below for definitions of vega and gamma.
Black-Scholes and the Volatility Surface 8

underlying security.

The delta of a European call option satisfies

∂C
delta = = e−qT Φ(d1 ).
∂S
This is the usual delta corresponding to a volatility surface that is sticky-by-strike. It assumes that as the
underlying security moves, the volatility of the option does not move. If the volatility of the option did move
then the delta would have an additional term of the form vega × ∂σ(K, T )/∂S. In this case we would say that
the volatility surface was sticky-by-delta. Equity markets typically use the sticky-by-strike approach when
computing deltas. Foreign exchange markets, on the other hand, tend to use the sticky-by-delta approach.
Similar comments apply to gamma as defined below.

(a) Delta for European Call and Put Options (b) Delta for Call Options as Time-To-Maturity Varies

Figure 2: Delta for European Options

By put-call parity, we have deltaput = deltacall − e−qT . Figure 2(a) shows the delta for a call and put option,
respectively, as a function of the underlying stock price. In Figure 2(b) we show the delta for a call option as a
function of the underlying stock price for three different times-to-maturity. It was assumed r = q = 0. What is
the strike K? Note that the delta becomes steeper around K when time-to-maturity decreases. Note also that
delta = Φ(d1 ) = Prob(option expires in the money). (This is only approximately true when r and q are
non-zero.)
In Figure 3 we show the delta of a call option as a function of time-to-maturity for three options of different
money-ness. Are there any surprises here? What would the corresponding plot for put options look like?

Definition: The gamma of an option is the sensitivity of the option’s delta to a change in the price of the
underlying security.

The gamma of a call option satisfies

∂2C φ(d1 )
gamma = = e−qT √
∂S 2 σS T
where φ(·) is the standard normal PDF.
In Figure 4(a) we show the gamma of a European option as a function of stock price for three different
time-to-maturities. Note that by put-call parity, the gamma for European call and put options with the same
Black-Scholes and the Volatility Surface 9

Figure 3: Delta for European Call Options as a Function of Time-To-Maturity

(a) Gamma as a Function of Stock Price (b) Gamma as a Function of Time-to-Maturity

Figure 4: Gamma for European Options

strike are equal. Gamma is always positive due to option convexity. Traders who are long gamma can make
money by gamma scalping. Gamma scalping is the process of regularly re-balancing your options portfolio to be
delta-neutral. However, you must pay for this long gamma position up front with the option premium. In Figure
4(b), we display gamma as a function of time-to-maturity. Can you explain the behavior of the three curves in
Figure 4(b)?

Definition: The vega of an option is the sensitivity of the option price to a change in volatility.

The vega of a call option satisfies

∂C √
vega = = e−qT S T φ(d1 ).
∂σ
In Figure 5(b) we plot vega as a function of the underlying stock price. We assumed K = 100 and that
r = q = 0. Note again that by put-call parity, the vega of a call option equals the vega of a put option with the
same strike. Why does vega increase with time-to-maturity? For a given time-to-maturity, why is vega peaked
near the strike? Turning to Figure 5(b), note that the vega decreases to 0 as time-to-maturity goes to 0. This is
Black-Scholes and the Volatility Surface 10

(a) Vega as a Function of Stock Price (b) Vega as a Function of Time-to-Maturity

Figure 5: Vega for European Options

consistent with Figure 5(a). It is also clear from the expression for vega.

Question: Is there any “inconsistency” to talk about vega when we use the Black-Scholes model?

Definition: The theta of an option is the sensitivity of the option price to a negative change in
time-to-maturity.

The theta of a call option satisfies

∂C σ
theta = − = −e−qT Sφ(d1 ) √ + qe−qT SN(d1 ) − rKe−rT N(d2 ).
∂T 2 T

(a) Theta as a Function of Stock Price (b) Theta as a Function of Time-to-Maturity

Figure 6: Theta for European Options

In Figure 6(a) we plot theta for three call options of different times-to-maturity as a function of the underlying
Black-Scholes and the Volatility Surface 11

stock price. We have assumed that r = q = 0%. Note that the call option’s theta is always negative. Can you
explain why this is the case? Why does theta become more negatively peaked as time-to-maturity decreases to
0?
In Figure 6(b) we again plot theta for three call options of different money-ness, but this time as a function of
time-to-maturity. Note that the ATM option has the most negative theta and this gets more negative as
time-to-maturity goes to 0. Can you explain why?

Options Can Have Positive Theta: In Figure 7 we plot theta for three put options of different money-ness
as a function of time-to-maturity. We assume here that q = 0 and r = 10%. Note that theta can be positive for
in-the-money put options. Why? We can also obtain positive theta for call options when q is large. In typical
scenarios, however, theta for both call and put options will be negative.

Figure 7: Positive Theta is Possible

The Relationship between Delta, Theta and Gamma

Recall that the Black-Scholes PDE states that any derivative security with price Pt must satisfy
∂P ∂P 1 ∂2P
+ (r − q)S + σ 2 S 2 2 = rP. (22)
∂t ∂S 2 ∂S
Writing θ, δ and Γ for theta, delta and gamma, we obtain
1
θ + (r − q)Sδ + σ 2 S 2 Γ = rP. (23)
2
Equation (23) holds in general for any portfolio of securities. If the portfolio in question is delta-hedged so that
the portfolio δ = 0 then we obtain
1
θ + σ 2 S 2 Γ = rP (24)
2
It is clear from (24) that any gain from gamma is offset by losses due to theta. This of course assumes that the
correct implied volatility is assumed in the Black-Scholes model. Since we know that the Black-Scholes model is
wrong, this observation should only be used to help your intuition and not taken as a “fact”.

Delta-Gamma-Vega Approximations to Option Prices

A simple application of Taylor’s Theorem says
∂C 1 ∂2C ∂C
C(S + ∆S, σ + ∆σ) ≈ C(S, σ) + ∆S + (∆S)2 2 + ∆σ
∂S 2 ∂S ∂σ
Black-Scholes and the Volatility Surface 12

1
= C(S, σ) + ∆S × δ + (∆S)2 × Γ + ∆σ × vega.
2
where C(S, σ) is the price of a derivative security as a function8 of the current stock price, S, and the implied
volatility, σ. We therefore obtain
Γ
P&L = (∆S)2 + vega ∆σ
δ∆S +
2
= delta P&L + gamma P&L + vega P&L

When ∆σ = 0, we obtain the well-known delta-gamma approximation. This approximation is often used, for
example,in historical Value-at-Risk (VaR) calculations for portfolios that include options. We can also write
2
ΓS 2
  
∆S ∆S
P&L = δS + + vega ∆σ
S 2 S
= ESP × Return + $ Gamma × Return2 + vega ∆σ

where ESP denotes the equivalent stock position or “dollar” delta.

5 Delta Hedging
In the Black-Scholes model with GBM, an option can be replicated exactly by delta-hedging the option. In
fact the Black-Scholes PDE we derived earlier was obtained by a delta-hedging / replication argument. The idea
behind delta-hedging is to re-balance the portfolio of the option and stock continuously so that you always have
a new delta of zero. Of course it is not practical in to hedge continuously and so instead we hedge periodically.
Periodic or discrete hedging then results in some replication error. Consider Figure 8 below which displays a
screen-shot of an Excel spreadsheet that was used to simulate a delta-hedging strategy.

Figure 8: Delta-Hedging in Excel

8 The price may also depend on other parameters, in particular time-to-maturity, but we suppress that dependence here.
Black-Scholes and the Volatility Surface 13

Mechanics of the Excel spreadsheet

In every period, the portfolio is re-balanced so that it is delta-neutral. This is done by using the delta of the
options portfolio to determine the total stock position. This stock position is funded through borrowing at the
risk-free rate and it accrues dividends according to the dividend yield. The timing of the cash-flows is ignored
when calculating the hedging P&L. Let Pt denote the time t value of the discrete-time self-financing strategy
that attempts to replicate the option payoff and let C0 denote the initial value of the option. The replicating
strategy is then given by
P0 := C0 (25)


Pti+1 = Pti + (Pti − δti Sti ) r∆t + δti Sti+1 − Sti + qSti ∆t (26)
where ∆t := ti+1 − ti which we assume is constant for all i, r is the annual risk-free interest rate (assuming
per-period compounding) and δti is the Black-Scholes delta at time ti . This delta is a function of Sti and some
assumed implied volatility, σimp say. Note that (25) and (26) respect the self-financing condition. Stock prices
are simulated assuming St ∼ GBM(µ, σ) so that
2
√
St+∆t = St e(µ−σ /2)∆t+σ ∆tZ

where Z ∼ N(0, 1). Note the option implied volatility, σimp , need not equal σ which in turn need not equal the
realized volatility (when we hedge periodically as opposed to continuously). This has interesting implications for
the trading P&L which we may define as
P&L := PT − (ST − K)+
in the case of a short position in a call option with strike K and maturity T . Note that PT is the terminal value
of the replicating strategy in (26). Many interesting questions now arise:

Question: If you sell options, what typically happens the total P&L if σ < σimp ?
Question: If you sell options, what typically happens the total P&L if σ > σimp ?
Question: If σ = σimp what typically happens the total P&L as the number of re-balances increases?

Some Answers to Delta-Hedging Questions

Recall that the price of an option increases as the volatility increases. Therefore if realized volatility is higher
than expected, i.e. the level at which it was sold, we expect to lose money on average when we delta-hedge an
option that we sold. Similarly, we expect to make money when we delta-hedge if the realized volatility is lower
than the level at which it was sold.
In general, however, the payoff from delta-hedging an option is path-dependent, i.e. it depends on the price
path taken by the stock over the entire time interval. In fact, we can show that the payoff from continuously
delta-hedging an option satisfies
Z T 2 2
St ∂ Vt 2
− σt2 dt


P&L = 2
σimp (27)
0 2 ∂S
where Vt is the time t value of the option and σt is the realized instantaneous volatility at time t.
S2 2
The term 2t ∂∂SV2t is often called the dollar gamma, as discussed earlier. It is always positive for a call or put
option, but it goes to zero as the option moves significantly into or out of the money.

Returning to self-financing trading strategy of (25) and (26), note that we can choose any model we like for the
security price dynamics. In particular, we are not restricted to choosing geometric Brownian motion and other
diffusion or jump-diffusion models could be used instead. It is interesting to simulate these alternative models
and to then observe what happens to the replication error in (27) where the δti ’s are computed assuming
(incorrectly) a geometric Brownian motion price dynamics.
Black-Scholes and the Volatility Surface 14

6 Pricing Exotics
In this section we will discuss the pricing of three exotic securities: (i) a digital option (ii) a range accrual and
(iii) an exchange option. The first two can be priced using the implied volatility surface and so their prices are
not model dependent. We will price the third security using the Black-Scholes framework. While this is not how
it would be priced in practice, it does provide us with an opportunity to practice change-of-measure methods.

Pricing a Digital Option

Suppose we wish to price a digital option which pays $1 if the time T stock price, ST , is greater than K and 0
otherwise. Then it is easy9 to see that the digital price, D(K, T ) is given by
C(K, T ) − C(K + ∆K, T )
D(K, T ) = lim
∆K→0 ∆K
C(K + ∆K, T ) − C(K, T )
= − lim
∆K→0 ∆K
∂C(K, T )
= − .
∂K
In particular this implies that digital options are uniquely priced from the volatility surface. By definition,
C(K, T ) = CBS (K, T, σBS (K, T )) where we use CBS (·, ·, ·) to denote the Black-Scholes price of a call option
as a function of strike, time-to-maturity and volatility. The chain rule now implies
∂CBS (K, T, σBS (K, T ))
D(K, T ) = −
∂K
∂CBS ∂CBS ∂σBS
= − −
∂K ∂σBS ∂K
∂CBS
= − − vega × skew.
∂K

Example 3 (Pricing a digital)

Suppose r = q = 0, T = 1 year, S0 = 100 and K = 100 so the digital is at-the-money. Suppose also that the
skew is 2.5% per 10% change in strike and σatm = 25%. Then
 σ  σ  −.025
atm atm
D(100, 1) = Φ − − S0 φ ×
2 2 .1S0
 σ  σ 
atm atm
= Φ − + .25 φ
2 2
≈ .45 + .25 × .4
= .55

Therefore the digital price = 55% of notional when priced correctly. If we ignored the skew and just the
Black-Scholes price using the ATM implied volatility, the price would have been 45% of notional which is
significantly less than the correct price.

Exercise 3 Why does the skew make the digital more expensive in the example above?

9 This proof is an example of a static replication argument.

Black-Scholes and the Volatility Surface 15

Example 4 (Pricing a Range Accrual)

Consider now a 3-month range accrual on the Nikkei 225 index with range 13, 000 to 14, 000. After 3 months
the product pays X% of notional where

X = % of days over the 3 months that index is inside the range

e.g. If the notional is $10M and the index is inside the range 70% of the time, then the payoff is $7M .
Question: Is it possible to calculate the price of this range accrual using the volatility surface?
Hint: Consider a portfolio consisting of a pair of digital’s for each date between now and the expiration.

Example 5 (Pricing an Exchange Option)

Suppose now that there are two non-dividend-paying securities with dynamics given by
(y)
dYt = µy Yt dt + σy Yt dWt
(x)
dXt = µx Xt dt + σx Xt dWt
(x) (y)
so that each security follows a GBM. We also assume dWt × dWt = ρ dt so that the two security returns
have an instantaneous correlation of ρ.
Let Zt := Yt /Xt . Then Itô’s Lemma (check!) implies

dZt (y) (x)

µy − µx − ρσx σy + σx2 dt + σy dWt − σx dWt .


= (28)
Zt
The instantaneous variance of dZ/Z is given by
 2 2
dZt 
(y) (x)
= σy dWt − σx dWt
Zt
σx2 + σy2 − 2ρσx σy dt


=

Now define a new process, Wt as

σy (y) σx (x)
dWt = dWt − dWt
σ σ


where σ 2 := σx2 + σy2 − 2ρσx σy . Then Wt is clearly a continuous martingale. Moreover,
!2
(y) (x)
2 σy dWt − σx dWt
(dWt ) =
σ
= dt.

Hence by Levy’s Theorem, Wt is a Brownian motion and so Zt is a GBM. Using (28) we can write its dynamics
as
dZt
= µy − µx − ρσx σy + σx2 dt + σdWt .


(29)
Zt

Consider now an exchange option expiring at time T where the payoff is given by

Exchange Option Payoff = max (0, YT − XT ) .

We could use martingale pricing to compute this directly and explicitly solve

P0 = EQ
 −rT 
0 e max (0, YT − XT )
Black-Scholes and the Volatility Surface 16

for the price of the option. This involves solving a two-dimensional integral with the bivariate normal
distribution which is possible but somewhat tedious.
Instead, however, we could price the option by using asset Xt as our numeraire. Let Qx be the probability
measure associated with this new numeraire. Then martingale pricing implies
 
P0 max (0, YT − XT )
= EQ 0
x

X0 XT
= EQ
0 [max (0, ZT − 1)] .
x
(30)

Equation (29) gives the dynamics of Zt under our original probability measure (whichever one it was), but we
need to know its dynamics under the probability measure Qx . But this is easy. We know from Girsanov’s
Theorem that only the drift of Zt will change so that the volatility will remain unchanged. We also know that
Zt must be a martingale and so under Qx this drift must be zero.
But then the right-hand side of (30) is simply the Black-Scholes option price where we set the risk-free rate to
zero, the volatility to σ and the strike to 1.
Pricing Other Exotics
Perhaps the two most commonly traded exotic derivatives are barrier options and variance-swaps. In fact at this
stage these securities are viewed as more semi-exotic than exotic. As suggested by Example 2, the price of a
barrier option cannot be priced using the volatility surface as the latter only defines the marginal distributions of
the stock prices. While we could use Black-Scholes and GBM with some constant volatility to determine a price,
it is well known that this leads to very inaccurate pricing. Moreover, a rule employed to determine the constant
volatility might well lead to arbitrage opportunities for other market participants.
It is generally believed that variance swaps can be priced uniquely from the volatility surface. However, this is
only true for variance-swaps with maturities that are less than two or three years. For maturities beyond that, it
is probably necessary to include stochastic interest rates and dividends in order to price variance swaps
accurately. Variance-swaps will be studied in detail in the exercises.

7 Dividends, the Forward and Black’s Model

Let C = C(S, K, r, q, σ, T − t) be the price of a call option on a stock. Then the Black-Scholes model says

C = Se−q(T −t) Φ(d1 ) − Ke−r(T −t) Φ(d2 )

where
log(S/K) + (r − q + σ 2 /2)(T − t)
d1 = √ ,
σ T −t
√
d2 = d1 − σ T − t. Let F := Se(r−q)(T −t) so that F is the time t forward price for delivery of the stock at
time T . Then we can write

C = F e−r(T −t) N(d1 ) − Ke−r(T −t) N(d2 ) (31)

−r(T −t)
= e × Expected-Payoff-of-the-Option

where
log(F/K) + (T − t)σ 2 /2
d1 = √ ,
σ T −t
√
d2 = d1 − σ T − t.
Note that the option price now only depends on F, K, r, σ and T − t. In fact we can write the call price as

C = Black(F, K, r, σ, T − t).
Black-Scholes and the Volatility Surface 17

where the function Black(·) is defined implicitly by (31). When we write option prices in terms of the forward
and not the spot price, the resulting formula is often called Black’s formula. It emphasizes the importance of the
forward price in establishing the price of the option. The spot price is only relevant in so far as it influences the
forward price.

Dividends and Option Pricing

As we have seen, the Black-Scholes formula easily accommodates a continuous dividend yield. In practice,
however, dividends are discrete. In order to handle discrete dividends we could convert them into dividend yields
but this can create problems. For example, as an ex-dividend date approaches, the dividend yield can grow
arbitrarily high. We would also need a different dividend yield for each option maturity. A particularly important
problem is that delta and the other Greeks can become distorted when we replace discrete dividends with a
continuous dividend yield.
Example 6 (Discrete dividends)
Consider a deep in-the-money call option with expiration 1 week from now, a current stock price = $100 and a
$5 dividend going ex-dividend during the week. Then
Black-Scholes delta = e−qT Φ(d1 )
≈ e−qT
= e−(.05×52)/52
= 95.12%
But what do you think the real delta is?

Using a continuous dividend yield can also create major problems when pricing American options. Consider, for
example, an American call option with expiration T on a stock that goes ex-dividend on date tdiv < T . This is
the only dividend that the stock pays before the option maturity. We know the option should only ever be
exercised at either expiration or immediately before tdiv . However, if we use a continuous dividend yield, the
pricing algorithm will never “see” this ex-dividend date and so it will never exercise early, even when it is optimal
to do so.
There are many possible solutions to this problem of handling discrete dividends. A common solution is to take
X0 = S0 − PV(Dividends) as the “basic” security where
PV(Dividends) = present value of dividends going ex-dividend between now and option expiration.
This works fine for European options (recall that what matters is the forward). For American options, we could,
for example, build a binomial lattice for Xt . Then at each date in the lattice, we can determine the stock price
and account properly for the discrete dividends, determining correctly whether it is optimal to early exercise or
not. In fact this was the subject of a question in an earlier assignment.

8 Extensions of Black-Scholes
The Black-Scholes model is easily applied to other securities. In addition to options on stocks and indices, these
securities include currency options, options on some commodities and options on index, stock and currency
futures. Of course, in all of these cases it is well understood that the model has many weaknesses. As a result,
the model has been extended in many ways. These extensions include jump-diffusion models, stochastic
volatility models, local volatility models, regime-switching models, garch models and others.
One of the principal uses of the Black-Scholes framework is that is often used to quote derivatives prices via
implied volatilities. This is true even for securities where the GBM model is clearly inappropriate. Such securities
include, for example, caplets and swaptions in the fixed income markets, CDS options in credit markets and
options on variance-swaps in equity markets.
Black-Scholes and the Volatility Surface 18

Exercises
1. Show that a stock that has a continuous dividend yield of q has risk-neutral dynamics
dSt = (r − q)St dt + σSt dWt where Wt is a Q-Brownian motion corresponding to the cash account as
the numeraire.

2. Consider a stock that has a continuous dividend yield of q with risk-neutral dynamics
dSt = (r − q)St dt + σSt dWtQ where WtQ is a Q-Brownian motion corresponding to the cash account as
the numeraire. Show that, as expected,
"Z #
T
S0 = EQ
0 e−rt qSt dt + e−rT ST . (32)
0

(You can assume that exchanging the order of integrations in (32) is justified.)

3. Derive the Black-Scholes PDE when the underlying stock has a constant dividend yield of q.

4. Derive the same PDE as in Exercise 3 but this time by using (12) and applying the Feynman-Kac formula
to an analogous expression to (11).

(T )
5. (a) Use martingale pricing to derive the time t price, Ft , of a futures contract for delivery of a stock at
time T . You can assume that the Black-Scholes model holds and that the stock pays a dividend yield of q.
(You can use your knowledge of futures prices from discrete-time models to justify your answer.)

(b) Compute the fair price of an option on a futures contract in the Black-Scholes model. You should
assume that the futures contract expires at time T and that the option expires at time τ < T . (This is
straightforward using the original Black-Scholes formula and your answer from part (a).)

(c) Confirm directly that the option price you derived in part (b) satisfies the Black-Scholes PDE of
Exercise 3.

6. (Variation on Q3.11 in Back) Suppose an investor invests in a portfolio with price S and constant
dividend yield q. Assume the investor is charged a constant expense ratio α (which acts as a negative
dividend) and at date T receives either his portfolio value or his initial investment, whichever is higher.
This is similar to a popular type of variable annuity. Letting D denote the number of dollars invested in
the contract, the contract pays
De(q−α)T ST
 
max D, (33)
S0
at date T . We can rearrange the expression (33) as

De(q−α)T ST De(q−α)T ST
   
max D, = D + max 0, −D
S0 S0
eqT ST
 
= D + e−αT D max 0, − eαT .
S0

Thus the contract payoff is equivalent to the amount invested plus a certain number of call options
written on the gross holding period return eqT ST /S0 . Note that Zt := eqt St /S0 is the date-t value of the
portfolio that starts with 1/S0 units of the asset (i.e., with a $1 investment) and reinvests dividends.
Black-Scholes and the Volatility Surface 19

Thus, the call options are call options on a non-dividend paying portfolio with the same volatility as S and
initial price of $1.

(a) Compute the date-0 value of the contract to the investor assuming Black-Scholes dynamics for the
portfolio. (In practice such a contract would be priced using the appropriate implied volatility surface.)

(b) Create a function (in Matlab, VBA or whatever you prefer) to compute the fair expense ratio; i.e. find
α such that the date-0 value of the contract is equal to D. (Hint: You can use α = 0 as a lower bound.
Because the value of the contract is decreasing as α increases, you can find an upper bound by iterating
until the value of the contract is less than D.)

(c) How does the fair expense ratio vary with the maturity, T ?

(1) (n)
7. Consider a market with n risky securities with price processes, (St , . . . , St ) := St , say. Suppose that
the P -dynamics of these securities are driven by m independent Brownian motions so that

dSt = µt St dt + Σt dWt

where Wt is an m × 1 standard Brownian motion that generates the filtration Ft , µt is an Ft -adapted

n × n diagonal matrix and Σt is an Ft -adapted n × m matrix. Assume that there is a cash-account that
earns interest at a constant continuously compounded risk-free rate of r. Use Girsanov’s Theorem to
determine the conditions under which (i) this market is arbitrage-free and (ii) arbitrage-free and complete
and (iii) arbitrage-free and incomplete. (Note that there is no loss of generality in assuming that the m
Brownian motions are independent. If they were dependent, we could use the Cholesky Decomposition to
work with independent Brownian motions. Finally, note that we could also easily allow r to be an adapted
process driven by the same set of m Brownian motions.)

8. The current index price is $100 and the term structure of interest rates is constant at 3%. European call
and put option prices of various strikes and maturities are presented below.

T .25 .5 1 1.5
Strike
60 40.2844 42.4249 50.8521 59.1664 Call Prices
70 30.5281 33.5355 42.6656 51.2181
80 21.0415 24.9642 34.4358 42.9436
90 12.2459 16.9652 26.4453 34.7890
100 5.2025 10.1717 19.4706 27.8938
110 1.3448 5.4318 14.4225 23.3305
120 0.2052 2.7647 11.2103 20.7206
130 0.0216 1.4204 9.1497 19.1828
140 0.0019 0.7542 7.7410 18.1858

T .25 .5 1 1.5
Strike
60 0.0858 2.1546 10.6907 19.3603 Put Prices
70 0.2548 3.1164 12.2087 20.9720
80 0.6934 4.3962 13.6833 22.2575
90 1.8232 6.2483 15.3972 23.6629
100 4.7050 9.3060 18.1270 26.3276
110 10.7725 14.4171 22.7834 31.3243
120 19.5582 21.6012 29.2757 38.2744
130 29.2999 30.1080 36.9195 46.2965
140 39.2055 39.2929 45.2152 54.8595
Black-Scholes and the Volatility Surface 20

(a) Use put-call parity to determine a piece-wise constant dividend yield implied by the option prices.
Does your dividend yield depend on the strikes you choose?

(b) Write a piece of code to determine the Black-Scholes implied volatility for each option and plot the
volatility surface.

(In practice, only the bid and ask prices of options are available in the market place and some
pre-processing will be necessary to build the volatility surface and calibrate the implied dividends. For
example, some options will have very wide bid-offers and are therefore less informative. Moreover, because
these options are less liquid it is also the case that these bid-offers may not have been updated as recently
as the more liquid options. It is often preferable then to ignore them when building the volatility surface.)

9. Write a computer program that simulates the delta-hedging of a long position in a European option in the
Black-Scholes model. Your code should take as inputs the initial stock price S0 , option expiration T ,
implied volatility σimp , risk-free rate r, dividend yield q and strike K as well as whether the option is a call
or put. Your code should also take as inputs: (i) the number of re-balancing periods N and (ii) the drift
and volatility, µ and σ respectively, of the geometric Brownian motion used to simulate a path of the
underlying stock price. Note that σimp and σ need not be the same. At the very least your code should
output the option payoff and the total P&L from holding the option and executing the delta-hedging
strategy. (See Figure 8 in the Black-Scholes and the Volatility Surface lecture notes for an example where
the code was written in VBA with the output in Excel.) Once you have tested your code answer the
following questions:

(a) When σimp = σ how does the total P&L behave as a function of N ? What happens on average if
σimp < σ? If σimp > σ?

(b) For a fixed N , how does the total P&L behave as σimp = σ increases?

(c) How does the drift, µ, affect the total P&L?

(d) Run your code repeatedly for σimp = 20% and σ = 40% with S0 = K = $50. Why does the total P&L
move about so much? How does the variance of the total P&L depend on the money-ness of the option?

10. Referring to Example 5, suppose asset X pays a continuous dividend yield of qx . Show that only the strike
in (30) needs to be changed in order to obtain the correct option price.

11. Referring again to Example 5, suppose asset Y pays a continuous dividend yield of qy . Show that (30) is
still valid but that we must now assume Zt pays the same dividend yield.

12. (Call on the Maximum of 2 Assets)

(1) (2)
Suppose there are two assets with price processes St and St , respectively, that satisfy
(1) (1) (1) (1)
dSt = (r − q1 )St dt + σ1 St dWt
(2) (2) (2) (2)
dSt = (r − q2 )St dt + σ2 St dWt
(1) (2)
where Wt and Wt are Q-Brownian motions with correlation coefficient, ρ. Q is the EMM
corresponding to taking the cash account as numeraire and q1 and q2 are the respective dividend yields of
the stocks. A call-on-the-max option with strike K and expiration T has a payoff given by
     
(1) (2) (1) (2)
max 0, max ST , ST − K = max 0, ST − K, ST − K .
Black-Scholes and the Volatility Surface 21

(a) Show that the value of the option at maturity T may be written as
(1) (2)
xST + yST − zK

where x, y and z are binary random variables taking the values 0 or 1.

(1) (1) (2) (2)
(b) By considering numeraires Vt := eq1 t St , Vt := eq2 t St and Rt := ert , show that the time 0
value of the option, C say, is given by
(1) (2)
C = e−q1 T S0 ProbV1 (x = 1) + e−q2 T S0 ProbV2 (y = 1) − e−rT K ProbR (z = 1) (34)

where ProbZ (·) denotes a risk-neutral probability corresponding to the numeraire Z.

(c) Compute the probabilities in (b) and therefore determine the price of the option.

13. (Variation on Back Q7.2) In the Black-Scholes framework, determine the delta-hedging strategy for a
call option on a futures contract where the the futures contract is written on a stock that has a
continuous dividend yield of q. Without doing any calculations you should be able to tell that as part of
the delta hedge you always invest Ct in the cash account at time t where Ct is the time t value of the
option. Why is this the case? (This assumes that you use the futures contract and the cash account as
your hedging securities rather than the underlying stock and the cash account. The answer to Exercise 5
should be useful in determining the time t position in the underlying futures contract.)

Remark: Here’s the way to think about the self-financing condition with futures as the underlying: Let Pt
denote the value of the self-financing replicating strategy at time t and assume you hold xt units of the
cash account, Bt , and yt units of the futures contract at time t. Then

Pt = x t B t + y t A t (35)

where At is the time t value of the futures contract. (Yes, At is in fact identically 0 as we know but let’s
leave it as At for now.) Also let Ft be the time t futures price. If you think about it for just a couple of
seconds you’ll see that the correct way to view a futures contract is that it is a security that is always
worth 0 but that pays out a continuous dividend yield of dFt . Therefore the self-financing condition
applied to (35) yields:
dPt = xt dBt + yt (dAt + dFt ) (36)
which is just the self-financing condition for a dividend paying security. Of course At = 0 for all t so
dAt = 0 so (36) becomes dPt = xt dBt + yt dFt which from an economic perspective is clearly correct!
14. (Idealized Variance-Swap) A variance-swap with maturity T is a derivative security whose time T
payoff is a function of annualized realized variance between t = 0 and t = T . In particular, the purchaser
of the variance-swap will receive
2 2


Nvar × σRealized − σStrike (37)
2 2
upon expiration at time T where σRealized is the annualized realized variance, σStrike is the strike and
Nvar is the variance notional or number of variance units. (The vega notional of the variance-swap is
2
defined by Nvega = 2 × Nvar × σStrike .) The strike, σStrike , is chosen at time 0 so that the initial value
of the variance-swap is zero. Suppose now that the risk-neutral dynamics of a stock price, St , satisfy

dSt = (r − q)St dt + σt St dWt (38)

where Wt is a Q-Brownian motion. Then the annualized continuous realized variance is given by
Z T
2 1
σRealized = σt2 dt.
T 0
Black-Scholes and the Volatility Surface 22

(a) Show that

Z T Z T  !
1 2 dSt ST
σt2 dt = − ln . (39)
T 0 T 0 St S0
Equation (39) implies that the realized leg of a var-swap can be replicated by taking a short position in
the log contract and by following a dynamic trading strategy that at each time t holds 1/St shares of the
underlying stock.

(b) Use equations (37), (38) and (39) to show that the fair strike of a variance swap satisfies
     
2 x ST
2
σStrike = (r − q)T − ln − EQ 0 ln
T S0 x

where x > 0 is any constant.

(c) Use the mathematical identity

x ∞
(y − x)
y Z Z
1 1
− ln = − + max (0, k − y) dk + max (0, y − k) dk
x x 0 k2 x k2
to express the fair strike of the variance-swap in terms of call and put options with expiration T . Simplify
your answer by taking x equal to the time T forward value of the stock.

(You have now shown that the fair strike of a variance swap only depends on the prices of vanilla
European options with expiration T . Our only model assumptions were (i) that the stock price follow a
diffusion as in (38) and that (ii) interest rates were constant. Note that we did not assume σt in (38) was
constant or even deterministic. Of course we did assume that the realized variance was observed
continuously. In practice realized variance is accumulated at discrete time intervals, usually via daily or
weekly observations. We will address this in the next question.)

15. (A Variance-Swap in Practice) In Exercise 14 we considered the continuous-time version of a

variance swap. In practice, variance-swap returns are based on discrete (typically daily or weekly)
observations. As before, the payoff of a short position in a variance swap satisfies

Payoff = N K 2 − σrealized
2

2
but now σrealized is calculated as
v
u PM  2
Si
t A × i=1 ln Si−1
u
σrealized = 100 ×
M
where M is the number of observation periods and A is the annualization factor. For example, if daily
returns are used then we typically have A = 252. (There are approximately 252 business days in a year.
See the sample term sheet that is posted on the course web-site.)
(a) Use (i) the approximation ln(1 + x) ≈ x for small x and (ii) Taylor’s Theorem to show that
M  2 M  
X Si X 1 ST
ln ≈ 2 (Si − Si−1 ) − 2 ln .
i=1
Si−1 i=1
Si−1 S0

Show that risk-neutral pricing therefore implies that the fair strike satisfies
   
2 2 Q ST
K ≈ 10, 000 × × rT − E0 ln (40)
T S0

where r is the risk-free interest rate and T = M/A is the time-to-maturity. (In practice variance-swaps on
indices are not “dividend-adjusted” whereas variance-swaps on single stocks are. If a variance-swap is
Black-Scholes and the Volatility Surface 23

dividend-adjusted then, assuming a stock goes ex-dividend on date i − 1 with dividend di−1 , the variance
 2  2
is calculated with a contribution of ln Si−1S+d
i
i−1
instead of ln S
Si
i−1
. Recall also from Question 4
that the log contract in (40) can be replicated using call and put options with maturity T . In particular a
variance-swap can be priced using the volatility smile for the maturity of the variance-swap.)

(b) Suppose we want to mark-to-market a variance swap that expires at time T and that was initiated at
time 0 with a total of M observations and strike K. Assume today is date t and that exactly m
2
observations have already occurred. Let σexpected, t := expected realized variance given the returns up-to
time t. Show that
2 m 2 M −m 2
σexpected, t = σ0,t + Kt,T
M M
2
where Kt,T is the fair strike at time t for a new variance swap expiring at time T and σ0,t is the realized
variance to date. Hence show that the time t value of the variance-swap, Vt , satisfies

Vt = e−r(T −t) N K 2 − σexpected,

2


t
 
m (M − m)
= e−r(T −t) N K 2 − σ0,t
2
K 2 − Kt,T
2



+N
M M
= Realized P&L + Implied P&L.

This result is not surprising since it is easily seen that a variance-swap is simply a sum of 1-period variance
swaps. Note that this is not true of volatility swaps which are therefore much harder to price. (Note also
that in practice many-variance-swaps are traded with caps on the realized variance. These caps are deep
out-of-the-money options on variance.)

(c) Greeks for variance-swaps can be calculated either analytically or by “bumping”, i.e. shifting the
parameter by a small amount, recomputing the value of the variance-swap and computing the derivative
numerically.

(i) In a Black-Scholes world with a flat volatility surface, does a variance-swap have any delta exposure at
the beginning of an observation? Does it have a delta exposure in the middle of an observation?

(ii) What is the daily theta of a variance swap, i.e. the amount you will lose or earn over the next day if
the stock price does not move? (You can assume the variance-swap is based on daily observations.)

(iii) Describe at least 2 different methods by which the vega of a variance-swap could be calculated.

(iv) Assuming a constant volatility surface, compute the gamma of a variance-swap at the beginning of
an observation. What is the dollar-gamma of the variance-swap.

16. Download the Excel spreadsheet [Link] from the course website where the details of a long
variance-swap position can be found. The position was initiated on May 10th 2010 at a strike of 27.2%
and with a vega-notional of $200k, i.e. “2N K = 200”. The maturity of the variance-swap is May 12th
2011. Today’s date is July 2nd 2010 and the closing price of the underlying index today is 3772.59. The
fair value today for a variance-swap that expires on May 12th 2011 is 35.3%. We wish to unwind the
variance-swap immediately. What is the realized P&L of the variance swap after we unwind it?

17. Use the volatility surface you computed in Exercise 8 to estimate the price of a digital option that pays $1
in the event that that the stock price in exactly 1 year from now is greater than or equal to $120. (In order
to do this you will need some way of estimating the volatility at non-traded strikes. A convenient way to
do this is by fitting a spline. This is easy to do in Matlab using just one or two lines of code.)
Black-Scholes and the Volatility Surface 24

18. Read the Dividend Swap Primer that can be downloaded from the course web-site. Your goal should be to
understand everything in the primer!