Greeks and Volatility Smile1, 2

Greeks:

1. A company uses delta hedging to hedge a portfolio of long positions in put and call
options on a currency. Which of the following would give the most favorable result?

(a) A virtually constant spot rate.

(b) Wild movements in the spot rate.
Explain your answer.

[Ans:] A long position in either a put or a call option has a positive gamma. From
Figure 18.8 (8th Edition), when gamma is positive the hedger gains from a large change
in the stock price and loses from a small change in the stock price. Hence the hedger
will fare better in case (b).

2. [Workshop 5] A financial institution holds the following portfolio of two calls and two
puts:

Position Delta (∆) Gamma (Γ) Vega (v)

Call 100 0.8 2 1
Call -200 0.4 1 0.5
Put 150 -0.5 0.5 1
Put -100 -0.7 2 1.5

(a) Calculate the portfolios’ delta, gamma and vega.

[Ans:] For portfolio Π

∆Π = 100 × 0.8 − 200 × 0.4 + 150 × (−0.5) − 100 × (−0.7) = −5

ΓΠ = 100 × 2 − 200 × 1 + 150 × 0.5 − 100 × 2 = −125
v Π = 100 × 1 − 200 × 0.5 + 150 × 1 − 100 × 1.5 = 0
1
The final examination will assess a wide range of topics taught in the lectures, selected and assigned
readings of the core text as well as articles, and assigned exercises and workshops. No single medium alone
can be served as a complete guide for the final examination. It is not advisable to prepare for the examination
by relying solely on e.g. workshops or past exam papers.
2
The suggested answer keys are not fully comprehensive especially the discussion parts where only the main
points are listed. You should read the original articles and recommended texts for more detailed discussion.

1
(b) Now the financial institution is considering adding the two call options below to
the portfolio above:

Delta Gamma Vega

Call 0.4 1 1
Call 0.6 2 0.5

For these two options, what position would make the final portfolio gamma and
vega-neutral simultaneously?

[Ans:] Assume that we add to the portfolio, w1 units of call 1 and w2 units of call
2. We need to choose w1 and w2 appropriately so as to make the portfolio gamma
and vega neutral:

ΓΠ + w1 Γ1 + w2 Γ2 = 0
−125 + w1 × 1 + w2 × 2 = 0

and similarly

v Π + w1 v 1 + w2 v 2 = 0
0 + w1 × 1 + w2 × 0.5 = 0

Solving this system of equations yields: w1 = −42 and w2 = 83. This means that
the financial institution needs to sell 42 units of call 1 and buy 83 units of call 2
in order to achieve simultaneous gamma and vega-neutrality.

(c) Is the gamma and vega-neutral portfolio obtained in (b) also delta-neutral? If not
what position in the underlying stock would make the portfolio delta-neutral? Is
gamma and vega-neutrality maintained after taking a position in the underlying
stock? Why or why not?

[Ans:] The portfolio obtained in (b) is: The initial portfolio plus the short position
in 42 units of call 1 and the long position in 83 units of call 2. The delta of this
portfolio:
0
∆Π = 1 × (−5) − 42 × 0.4 + 83 × 0.6 = 28
To make this delta neutral we enter into a position on w stocks. We need to pick
w so that:

1 × 28 + w × 1 = 0
w = −28

This means that to make the portfolio delta-neutral we need to sell 28 stocks.
Gamma and vega-neutrality are maintained after adding the stock because stocks
have a zero gamma and a zero vega.

2
 (d) Explain what protection is offered by making the portfolio (i) delta-neutral, (ii)
gamma-neutral and (iii) vega-neutral.

[Ans:] For protection

i. Delta-neutrality protects the portfolio against small changes in the underlying
asset price.
ii. Gamma-neutrality protects the portfolio against large changes in the underly-
ing asset price.
iii. Vega-neutrality protects the portfolio against changes in the volatility.

3. Suppose that $70 billion of equity assets are the subject of portfolio insurance schemes.
Assume that the schemes are designed to provide insurance against the value of the
assets declining by more than 5% within one year. Assuming that r = 0.06 or 6%,
σ = 0.25 or 25%, and q = 0.03 or 3% (all per annum), calculate the value of the stock or
futures contracts that the administrators of the portfolio insurance schemes will attempt
to sell if the market falls by 23% in a single day.

[Ans:] We can regard the position of all portfolio insurers taken together as a single
put option. The three known parameters of the option, before the 23% decline, are
S0 = 70, K = 66.5 (i.e. $70 × 95% = $66.5), T = 1, r = 0.06, σ = 0.25, and q = 0.03.
Then:

70


ln 66.5
+ (0.06 − 0.03 + 0.5 × 0.252 )
d1 = = 0.4502
0.25
N (d1 ) = 0.6737

The delta of the put option is

∆put = e−qT [N (d1 ) − 1] = e−0.03 [0.6737 − 1] = −0.3167

This shows that 31.67% or $22.17 billion of assets should have been sold before the
decline. After the decline, S0 = 53.9 (i.e. $70 × 77% = $53.9), K = 66.5, T = 1,
r = 0.06, σ = 0.25, q = 0.03, and
53.9


ln 66.5 + (0.06 − 0.03 + 0.5 × 0.252 )
d1 = = −0.5953
0.25
N (d1 ) = 0.2758

The delta of the option has dropped to

∆0put = e−0.03 [0.2758 − 1] = −0.7028

This shows that cumulatively 70.28% of the assets originally held should be sold. An
additional 38.61% of the original portfolio should be sold. The sales measured at pre-
crash prices are about $27.0 billion (i.e. $70 × 0.3861 = 27.03) which is wrong and
impossible to achieve. At post-crash prices they are about $20.8 billion (i.e. $53.9 ×
0.3861 = 20.81) which is what actually happened.

3
4. [Workshop 5] Does a forward contract on a stock index have the same delta as the
corresponding futures contract? Explain your answer.

[Ans:] Assume the strike price as K which was determined some time ago when the
contract was initialised. Today forward/futures price will be F0,T = S0 e(r−q)T . The
value of the forward contract is

ff orward = (F0,T − K)e−rT = S0 e−qT − Ke−rT (1)

In practrice, the value of the futures contract is not discounted when it is marked to
market. So the value of the futures contract is

ff utures = (F0,T − K) = S0 e(r−q)T − K (2)

Hence, the delta for the forward contract is:

∂
∆f orward = (S0 e−qT − Ke−rT ) = e−qT (3)
∂S
and the delta for the futures contract is:
∂
∆f utures = Se(r−q)T − K = e(r−q)T = erT e−qT = erT ∆f orward


(4)
∂S
We conclude that the deltas of a futures and forward contract are not the same. The
delta of the futures is greater than the delta of the corresponding forward by a factor
of erT > 1 for positive interest rate, r.

5. A bank’s position in options on the dollar–euro exchange rate has a delta of 30,000 and
a gamma of -80,000. Explain how these numbers can be interpreted. The exchange
rate (dollars per euro) is 0.90. What position would you take to make the position
delta neutral? After a short period of time, the exchange rate moves to 0.93. Estimate
the new delta. What additional trade is necessary to keep the position delta neutral?
Assuming the bank did set up a delta-neutral position originally, has it gained or lost
money from the exchange-rate movement?

[Ans:] The delta indicates that when the value of the euro exchange rate increases
by $0.01, the value of the bank’s position increases by 0.01 × 30, 000 = $300. The
gamma indicates that when the euro exchange rate increases by $0.01 the delta of the
portfolio decreases by 0.01 × 80, 000 = $800. For delta neutrality 30,000 euros should be
shorted. When the exchange rate moves up to 0.93, we expect the delta of the portfolio
to decrease by (0.93 − 0.90) × 80, 000 = 2, 400 so that it becomes 27,600. To maintain
delta neutrality, it is therefore necessary for the bank to unwind its short position 2,400
euros so that a net 27,600 have been shorted. As shown in the figure below (y-axis is
the change in portfolio value Π and the x-axis is the change in S), when a portfolio
is delta neutral and has a negative gamma, a loss is experienced when there is a large
movement in the underlying asset price. We can conclude that the bank is likely to have
lost money.

4
Volatility Smile:

1. What volatility smile is likely to be observed when:

(a) Both tails of the stock price distribution are less heavy than those of the lognormal
distribution?

[Ans:] A hump shape ‘smile’ similar to that in Figure 19.7 (8th Edition) is ob-
served.

(b) The right tail is heavier, and the left tail is less heavy, than that of a lognormal
distribution?

[Ans:] An upward sloping volatility smile is observed.

2. What volatility smile is likely to be caused by jumps in the underlying asset price? Is
the pattern likely to be more pronounced for a two-year option than for a three-month
option?

[Ans:] Jumps tend to make both tails of the stock price distribution heavier than
those of the lognormal distribution. This creates a volatility smile similar to that in
Figure 19.1 (8th Edition). The volatility smile is likely to be more pronounced for the
three-month option.

3. [Workshop 5] Explain carefully why a distribution with a heavier left tail and less
heavy right tail than the lognormal distribution gives rise to a downward sloping volatil-
ity smile.

5
 [Ans:] Consider a deep out of the money (DOTM) put, i.e. a put with a small strike
price K1 . This put pays off only when S < K1 . The probability that S < K1 is lower
for the lognormal distribution than under a distribution that has a heavier left tail (lets
name this ’implied distribution’). This means that the put price obtained from the
lognormal distribution (i.e. the Black-Scholes-Merton price) is lower than the price ob-
tained from the implied distribution. Hence, the implied volatility under the lognormal
distribution is lower than the implied volatility under the implied distribution. This is
because there is a one-to-one to relation between option price and implied volatility.
Consider a deep out of the money (DOTM) call, i.e. a call with a large strike price
K2 . This call pays off only when S > K2 . The probability that S > K2 is greater
for the lognormal distribution than under a distribution that has a less heavy right tail
(lets name this ’implied distribution’). This means that the call price obtained from the
lognormal distribution (i.e. the Black-Scholes-Merton price) is greater than the price ob-
tained from the implied distribution. Hence, the implied volatility under the lognormal
distribution is greater than the implied volatility under the implied distribution. This
is because there is a one-to-one to relation between option price and implied volatility.
To sum up, the heavier left tail should lead to high prices, and therefore high implied
volatilities, for out-of-the-money (low-strike-price) puts. Similarly the less heavy right
tail should lead to low prices, and therefore low volatilities for out-of-the-money (high-
strike-price) calls. A volatility smile where volatility is a decreasing function of strike
price results.
(See also Section 19.3 in Hull’s book, 8th edition)

4. Assume an option on a non-dividend paying stock when the stock price is £45, the strike
price is £47, the risk-free interest rate is 10% per annum, and the time-to-maturity is
three months. The option sells in the market for a price of £3.

(a) Assuming that the option is a European call option, is its implied volatility equal
to 0.1552, 0.3762 or 0.4522?

[Ans:] The Black-Scholes formula for European call√options is: c = S√ 0 N (d1 ) −

−rT 2
Ke N (d2 ), with d1 = [ln(S0 /K) + (r + σ /2)T ] /σ T and d2 = d1 −σ T . Plug
−rT
the known values into the Black Scholes formula: q 3 = 45×N (d1 )−47×e
q N (d2 ),
with d1 = ln(45/47) + (0.10 + σ 2 /2) (T ) 14 σ 14 and d2 = d1 −σ 14 . We want



to choose σ in such a way that the price of the European call option equals £3.
One problem that we have is that, unfortunately, we cannot solve the Black-Scholes
formula analytically for σ. You could instead use trial-and-error, that is, try all of
the above values. You would then that the implied volatility of the option is equal
to 0.3762.

(b) Assuming that the option is a European put option, is its implied volatility equal
to 0.2826, 0.3426 or 0.3936?

6
 [Ans:] The Black-Scholes formula for European put options is: p = Ke−rT N (−d2 )−
S0 N (−d1 ), with d1 and d2 defined as in part (a). Use the same approach as above
to find that implied volatility is equal to 0.2826.

5. [Workshop 5] A European call option on a certain stock has a strike price of $30, a
time to maturity of one year, and an implied volatility of 30%. A European put option
on the same stock has a strike price of $30, a time to maturity of one year, and an im-
plied volatility of 33%. What is the arbitrage opportunity open to a trader? Does the
arbitrage work only when the lognormal assumption underlying Black–Scholes–Merton
holds? Explain carefully the reasons for your answer.

[Ans:] Put–call parity implies that European put and call options have the same implied
volatility. If a call option has an implied volatility of 30% and a put option has an
implied volatility of 33%, the call is priced too low relative to the put. The correct
trading strategy is to buy the call, sell the put, short the stock and put the spare cash
in deposit (i.e. c + Ke−rT = p + S, we long the left hand side and short the right hand
side). This does not depend on the lognormal assumption underlying Black–Scholes–
Merton. Put–call parity is true for any set of assumptions.

6. [Workshop 5] “The Black–Scholes–Merton model is used by traders as an interpolation

tool.” Discuss this view.

[Ans:] When plain vanilla call and put options are being priced, traders do use the
Black-Scholes-Merton model as an interpolation tool. They calculate implied volatilities
for the options whose prices they can observe in the market. By interpolating between
strike prices (or moneyness) and between times to maturity, they estimate implied
volatilities for other options. These implied volatilities are then substituted into Black-
Scholes-Merton to calculate prices for these options. In practice much of the work
in producing a table such as Table 19.2 in the over-the-counter market is done by
brokers. Brokers often act as intermediaries between participants in the over-the-counter
market and usually have more information on the trades taking place than any individual
financial institution. The brokers provide a table such as Table 19.2 to their clients as
a service.

7. Answer the following two questions on implied volatility and volatility smiles:

(a) Define the following terms in no more than three sentences: (1) Black-Scholes im-
plied volatility and (2) volatility smile.

[Ans:] (1) Implied volatility: In the Black-Scholes model the only unobservable
parameter is volatility. Implied volatility sets the option price observed in the
market equal to the theoretical Black-Scholes option price. We obtain implied

7
 volatility numerically.

(2) Volatility smile: Volatility smile refers to the relationship between im-plied
volatility and strike price. According to Black-Scholes implied volatility should be
constant across K and T, i.e. the relationship should be flat. This is not observed
in real life where we observe a smile.

(b) Is the volatility smile derived from market put price the same as the one derived
from market call price with the same strike price K and time-to-maturity T ? Ex-
plain your answer.

[Ans:] Consider the put call parity for observed market prices:

cmkt + Ke−rT = pmkt + S0 (5)

Now consider the put call parity for theoretical Black-Scholes prices:

cBS + Ke−rT = pBS + S0 (6)

Subtracting (6) from (5) yields:

cmkt − cBS = pmkt − pBS

This equation implies the dollar pricing error for calls is the same to the dollar
pricing error from puts and hence, the volatility smile derived from market call
prices is the same as the volatility smile derived from marker put prices.