Beruflich Dokumente
Kultur Dokumente
Wim Schoutens
2 Tree Models 45
2.1 The Binomial Tree . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.1.1 One Step Binomial Tree . . . . . . . . . . . . . . . . . . . . 45
2.1.2 Risk Neutral Valuation . . . . . . . . . . . . . . . . . . . . 48
3
4 CONTENTS
The aim of the course is to give a rigorous yet accessible introduction to the
modern theory of financial mathematics.
The student should already be comfortable with calculus and probability the-
ory. Prior knowledge of basic notions of finance is useful as well.
We start with providing some background on the financial markets and the
instruments traded. We will look at different kinds of derivative securities, the
main group of underlying assets and the markets where derivative securities are
traded.
The fundamental problem in the mathematics of financial derivatives is that
of pricing, risk-management and hedging. We focus in this course mainly on
the theory of no-arbitrage pricing. We start by discussing option pricing in the
simplest idealised case: the one-step Binomial tree frictionless market. Next, we
turn to more general Binomial tree models. Under these models we price European
and American derivatives and discuss pricing methods for more involved exotic
derivatives.
In addition, we also elaborate on the Monte-Carlo pricing technique, which
tries to approximate the price of a derivative by the discounted average value of
the payoff under a huge number of simulated paths.
Further, we set up general discrete-time models and look in detail at the
mathematical counterpart of the economic principle of no-arbitrage: the existence
of equivalent martingale measures. We look when the models are complete, i.e.
claims can be hedged perfectly and discuss the fundamental theorem of asset
pricing in a discrete setting.
Finally, we elaborate on the celebrated continuous time Black-Scholes model.
The model is driven by Brownian motions, which are intuitively introduced. We
further provide the basics of Ito Calculus, that learns us how to deal with stochas-
tic integrals and Stochastic Differential Equations (SDEs).
7
8 CONTENTS
Throughout the text, we often provide also some extra material, which brings
the student in contact with the more advanced establish quantitative finance the-
ory. We for example, discuss bid-ask pricing, variance swaps, jump and stochastic
volatility models.
Chapter 1
Financial Mathematics
Principles
Stocks - Equity
The basis of modern economic life are companies owned by their shareholders;
the shares provide partial ownership of the company, pro rata with investment.
Shares have value, reflecting both the value of the company’s real assets and
the earning power of the company, potentially monetized via dividends (see also
9
10 CHAPTER 1. FINANCIAL MATHEMATICS PRINCIPLES
Section 1.9). With publicly quoted companies, shares are quoted and traded
on the Stock Exchange. Stock is the generic term for assets held in the form
of shares. Stock markets date back to at least 1531, when one was started in
Antwerp, Belgium.
The value of some financial assets depends on the level of interest rates, e.g.
Treasury notes, municipal and corporate bonds. These are fixed-income securities
by which national, state and local governments and companies partially finance
their economic activity. Fixed-income securities require the payment of interest in
the form of a fixed amount of money at predetermined points in time (coupons),
as well as the repayment of the principal at maturity of the security.
Bonds are examples of such fix-income securities. A bond is an instrument of
indebtedness of the bond issuer to the bond holders. The issuer owes the holders
money (debt) and, depending on the terms of the bond, is obliged to pay them
interest (coupons) and to repay the principal at a later date, termed the maturity
date. Interest is usually payable at fixed intervals (quarterly, semi-annual, annual,
sometimes monthly). Bonds are sensitive to default/bankruptcy. When the bond
issuer no longer is solvent, he cannot pay back in full his debt to the bond holders,
which in that situation lose part or all of their initial investment. To compensate
against default risk, and depending on the probability the bond holder will default,
the bond issuers will charge a higher or lower interest rate.
Interest rates themselves are notional assets, which cannot be delivered. A
special fixed-income product is the (artificial) risk-free bank account mentioned
in the introduction. The interest rate on such a risk-free bank account is termed
the risk-free interest rate. We go into more details on interest rate assumptions
in Section 1.3.
Example 1. A EUR/USD rate of 1.0975 means that for 1 Euro, one gets 1.0975
US dollars. Note that then the USD/EUR rate equals about 0.9112.
1.1. FINANCIAL MARKETS AND INSTRUMENTS 11
Commodities
Commodities are a kind of physical products like gold, oil, cattle, fruit juice. Trade
in these assets can be for different purposes: for using them in the production
process or for speculation. Derivative instruments on these asset can be used for
hedging and speculation. Special care has to be taken with commodities because
of storage costs.
1.1.2 Indices
An index tracks the value of a basket of stocks (FTSE100, S&P500, Dow Jones
Industrial, NASDAQ Composite, BEL20, EUROSTOXX50, ...), bonds, and so
on. Derivative instruments on indices may be used for hedging if no derivative
instruments on a particular asset in question are available and if the correlation
in movement between the index and the asset is significant.
Furthermore, institutional funds (such as pension funds), which manage large
diversified stock portfolios, try to mimic particular stock indices and use derivative
on stock indices as a portfolio management tool. On the other hand, a speculator
may wish to bet on a certain overall development in a market without exposing
him/herself to a particular asset.
12 CHAPTER 1. FINANCIAL MATHEMATICS PRINCIPLES
the ask and demand forces are not in place like they are on exchanges open to
everybody on the globe.
Derivatives are omnipresent in today’s financial markets. There are various
types. One has futures and forwards which are basically contracts to buy or
sell the underlying at a predetermined price in the future on a predetermined
date. Hence the limited price stability offered. One also has swaps, that agree on
exchanging certain uncertain cash-flows over a predetermined period and finally
one has options, that are agreements in which the holder has a right to buy or
sell the underlying at specific conditions. To buy an option one has to pay the
price or the premium for the option; swaps are often initiated at zero cost and
have therefore an initial market price equal to zero. However they can be written
on a huge underlying notional and although of zero value when the deal is struck
can bare significant risks. The actual size of derivative’s markets is not easy to
estimate but at the end of 2013, the Bank for International Settlements (BIS)
estimated the total notional outstanding for OTC derivatives at USD 710 trillion
and at a USD 18.6 trillion gross market value.
The theory we develop in this book on how to determine prices and how to
deal with the risks of such derivatives, is applicable to all derivative types over all
asset classes mentioned. However, we mainly focus on equity derivative options.
The basic examples are European call and put options, often referred to as vanilla
options, because they can be regarded as the most simple type of options. Before,
we will define them and illustrate their use in Section 1.4, we first recall the
concept of the risk-free bank-account.
of r = 2% (per annum), then the value in one year’s time would be USD 1020
(= 1000(1 + 0.02)). In this calculation we have assumed that the interest is
compounded annually. This means payout after one year.
Now assume, you want to have exactly USD 2000 (N ) in the account in 3
years (T ) time from now and assume we have a (flat) interest rate of r = 2% (per
annum). How much do you have to put into the risk-free account now ? The
answer is given by the formula:
N 2000
T
= = 1884.64
(1 + r) (1 + 0.02)3
Indeed, after putting 1884.64 in the account, it grows under a 2% interest rate
after one year to 1922.33 (= 1884.64(1 + 0.02)) and after another year to 1960.78
(= 1922.33(1 + 0.02)) and finally after the last year to 2000(= 1960.78(1 + 0.02)).
The interest is compounded again annually and we call USD 1884.64 the present
value of USD 2000 received in 3 years. The ratio of both is called the discount
factor or the price of future money and equals to
1
.
(1 + r)T
just slightly more than with annual compounding. The obvious reason is that
after an interest payment is made, the investor starts to earn interest on this
payment. Note that different compounding conventions lead to different rate
quotes consistent with the same price for future money at a fixed future date.
When pricing derivatives, we usually assume continuous compounding of inter-
est rates, meaning that compounding (i.e. receiving interest) occurs continuously,
i.e. over an infinitesimally small period of time, or in other words if m → ∞. Our
discount factor then becomes
r −mT
lim 1 + = exp(−rT ).
m→∞ m
In Figure 1.1, one clearly sees the convergence if m → ∞; one can also see the
differences between yearly (m = 1), semi-annually (m = 2), quarterly (m = 4),
monthly (m = 12) and weekly (m = 52) compounding on an investment of USD
1000 at r = 2% during a period of exactly one year (T = 1).
In reality each maturity T and quoting convention has its own interest rate,
r(T ) say, reflecting the market expectation of changing interest rates over the
given period. We then speak about a yield curve and the interest rate term
structure.
The risk-free yield curve is a curve showing several yields or risk-free interest
rates across different contract lengths (maturities), known as the ”term”, for a
risk-free debt contract. One has different curves for different currencies. For
example, one has the U.S. dollar yield curve based on interest rates paid on U.S.
Treasury securities for various maturities, which are assumed to be (almost or as
close as possible to) risk-free. In Figure 1.2 the EUR yield curve is shown as of
the 30th of December 2014; this curve is calculated by the ECB and based on
”AAA-rated” Euro area central government bonds.
Interest rate curves are typically upward-sloping, with shorter-term interest
rates lower than longer-term interest rates, but also can be downward sloping
(inverted) or humped. An inverted curve for example would indicate the market
expects lower interest rates in the future. As seen in the EUR-curve of Figure
1.2, yields can be negative; meaning investors are basically paying money to park
their investments in the underlying securities.
Related to a yield curve is a discount curve. The discount curve basically
represents the discount factors for the different terms. For a given term, the
discount factor is the present value of a currency unit promised at the given term.
If r(t) is the yield (continuously compound) associated with the maturity (or
term) t, then the discount factor for that term equals D(t) = exp(−r(t)t). In
1.4. VANILLA OPTIONS 17
Figure 1.3, one finds the discount factors based on the yields of Figure 1.2. From
this one can read of that, on the 30th of December 2014 receiving EUR 100 in
30 years from then is equivalent with receiving EUR 64.16 on that day, since the
30 years interest rate was on the 30th of December 2014 at r(30) = 0.0147947.
Similar, since the one year interest rate was then r(1) = −0.0008869, investors
need to pay about EUR 100.09 on 30th of December 2014 to receiving EUR 100
back on the 30th of December 2015.
Discount factors are used to discount cash flows (at the risk free rate). Assume
for example a cash-flow consisting of N payments of EUR Ci paid out at times
ti , i = 1, . . . , N . Then the present value (P V ) of this cash-flow equals:
N
X N
X
PV = Ci exp(−r(ti )ti ) = Ci D(ti ),
i=1 i=1
with D(t) denoting here the EUR-related discount factor with term t.
The holder of an European call option will exercise his right to buy the under-
lying via the option contract only if this is beneficial to him. This happens when
the underlying at maturity (t = T ) has a market price S(T ) that is greater than
the strike price: S(T ) > K. The payoff of the option is then strictly positive and
equals the difference S(T ) − K. In the other situation, it would not be rational to
pay via the option the strike price K which is more than the price S(T ) one pays
in the market and the option contract expires worthless. The payoff is then zero.
Summarizing, the payoff of the call option can in general be given by
A similar reasoning can be made for the put option where the holder will actually
only sell the underlying via the option contract when at maturity (t = T ) its
actual market price S(T ) is lower than the strike price K: S(T ) < K. The payoff
of the put equals in general
In Figures 1.4 and 1.5 the payoff functions of a European Call and a European
put are visualized respectively.
We say a European call option is out-of-the-money (OTM), if the current stock
price is below the strike. If it is above the strike we say it is in-the-money (ITM).
When the underlying stock prices equals (or is very close to) the strike we say it is
at-the-money (ATM). For a European put, out-of-the-money (OTM) corresponds
to the situation where the current stock price is above the strike. If it is below
the strike we say it is in-the-money (ITM). When its stock prices equals (or is
very close to) the strike we again say it is at-the-money (ATM). OTM basically
means that if one would now exercise the option (which is for European options
actually not allowed in reality, but one assumes one does), the payoff would be
zero. ITM means we would have then a non-zero payoff. This amount one would
get if the option would be now exercised is called the intrinsic value of the option.
OTM options have a zero intrinsic value. The difference between the current price
of the option and its intrinsic value is often referred to as time value: the extra
value the option carries because there is still time left until expiry/maturity and
the underlying asset can still move in a beneficial direction for the option holder.
OTM options have only time value.
Put and call options can be used for different purposes. We illustrate this
in the next examples. In the first example, we show how a call option can be
used for speculation. The next example shows how a put can be used as a kind
1.4. VANILLA OPTIONS 19
Example 2. An investor has USD 10000 at his disposal and is clearly convinced
the stock S will rally in the next year. The stock trades now at USD 50. Vanilla
options on the stock are trading as well. The one year at-the-money (ATM), i.e.
with strike equal to the current stock price, European call option is having a bid
price of USD 3.85 and an ask price of USD 4.00. With his USD 10000, he hence
can buy 200 stocks, or he can buy 2500 European ATM call options, (or he could
buy some of both). He decides to buy 2500 ATM calls. Note that he is buying and
hence has to pay the ask price of USD 4.00. After 6 months the stock has indeed
rallied to USD 65. The call options he bought are now deep in the money, i.e.
the current stock pice is higher than the strike. They are also closer to maturity
(6 months) and one call trades at a bid equal to USD 18.00 and has an ask price
of USD 18.25. He decides to close his position and sells his 2500 calls at the bid
price of USD 18.00. He cashes USD 45000 and actually makes a 350% profit.
Compared with a direct investment in the stock this is much better, since that only
would have given him a 30% return.
Derivatives can be used for speculation and can lead to hugely leveraged posi-
tions and hence huge gains. Of course, there is another side to the story. If the
stock would not move higher and for example closes after one year at the same
level of USD 50. The investment in call options would have led to a 100% loss,
since all the call options would have matured worthless. The direct investment
into the stock would have in that case ended flat and would have shown no loss.
Example 3. At the beginning of a new year an investor steps into the equity
market and buys 1000 stocks at a price of USD 65 each. He furthermore has
still a cash account of USD 2000. His initial total wealth is USD 67000. After
9 months, the stock trades now at USD 80. He doesn’t want to exit his position
because he still believes there is still upwards potential in the stock. However he
is also worried about the downside. The first 9 months of the year the stock has
been rallying nicely (at the beginning of the year he entered at USD 65 !) and it
is now vulnerable to potential negative market sentiment in the next months until
year end. He wants to close his year positively and seeks protection against such
averse downside market movement without exiting his position and potentially
missing a continuation of a rally in the stock. Derivatives are trading on the
underlying stock. An out-ot-the-money (OTM), i.e. with strike below the current
20 CHAPTER 1. FINANCIAL MATHEMATICS PRINCIPLES
spot, European put with maturity 3 months with strike USD 70 is trading with a bid
of USD 1.80 and an ask of USD 2.00. He decides to spent his USD 2000 cash to
buy protection and buys (at the ask price) 1000 3-month European puts with strike
USD 70. By year end the market indeed went down and the stock is now trading
at USD 60. Without any protection (and assuming no interest payments on his
cash account), he would have been down for the year by 7.46% (USD 5000/USD
67000). However due to his put options, he receives an additional payoff. At
expiration the put options gave him a payoff of USD 10 each. He received hence
USD 10000. Compared with the value of his position in the beginning of the year
(USD 65000+USD 2000), he now has stock worth USD 60000 and USD 10000 in
cash, or USD 70000 in total. He hence is up 4.48% (USD 3000/USD 67000) for
the year.
Example 4. Sales people have learnt that their retail customers are very interested
in investing their money in the medium term in the stock market, especially into
some new social media companies. However they are also worried about loosing
their investment. The structuring team therefore designs the following structured
product: a Principal Protected Note (PPN). For each USD 1000 you invest, you
receive after 4 year your initial investment (USD 1000) plus 60% of the positive
performance of a social media stock S. If the stock S would end after 4 year below
its value at initiation, the investor still gets his initial investment back. Denoting
with S(0) the initial stock price of S and with S(4) the stock price after 4 years,
the investor hence receives:
S(4) − S(0)
P P N = 1000 + 1000 × 60% × max ,0 (1.1)
S(0)
The final wealth with an initial investment of USD 1000 in either the PPN or in
the stock is compared in Figure 1.6
Assume that investing now USD 0.82 into a risk-free account would give in 4
years from now USD 1; this corresponds approximately with an interest rate of
5%. Assume S(0) = 20. Note that in this setting Equation (1.1) becomes
Further assume that an at-the-money (ATM), i.e. with strike K = 20, 4-year
European call option is available and trades at bid USD 5.25 and ask USD 5.50.
The bank implements this product therefore as follows. For each USD 1000 it
receives, it puts USD 820 on the risk-free bank account. It also buys 30 ATM call
1.5. MODELLING ASSUMPTIONS 21
options at the ask price of USD 5.50 at a total cost of USD 165. From the original
USD 1000, the bank has hence put USD 820 in a risk-free account and has bought
for USD 165 call options; it still has USD 15 (or 1.5% of the investment amount)
left which it considers as its margin. After 4 years the amount on the bank account
has grown to USD 1000; furthermore the 30 call options expire and have each as
payoff (S(4) − 20)+ . The total hence equals 1000 + 30 × max (S(4) − 20, 0), exactly
the amount it owes the structured product investor.
make the prices move and one cannot transact always the full order at the same
price.
Example 5. Consider an investor who acts in a market in which only three finan-
cial assets are traded: a (risk-less) bank account B, a stocks S and an European
Call options C with strike K = 100 and maturity T on the stock S. The investor
may invest today, time t = 0, in all three assets, leave his investment until time
t = T and gets his returns back then. We assume the current prices (in Euro,
say) of the financial assets are given by
and that at t = T there can be only two states of the world: an up-state with euro
prices
B(T, u) = 1.25, S(T, u) = 175, and therefore C(T, u) = 75,
and a down-state with euro prices
Now our investor has a starting capital of EUR 25000 from which he buys the
the portfolio (Portfolio I) given in Table 1.1. Depending of the state of the world
at time t = T the value of his portfolio will differ: In the up state the total value
of his portfolio is EUR 48750 as can be seen from Table 1.2. In the down-state
his portfolio has a value of EUR 20000 as shown in Table 1.3.
Can the investor do better ? This may sound to be a strange question - better
how ? Let us consider the restructured portfolio with initial investment of EUR
24600 (Portfolio II) as in Table 1.4. Again we compute its return in the different
possible states. In the up-state the total value of his portfolio is again EUR 48750
as can be seen in Table 1.5. In the down-state this portfolio has a value of again
EUR 20000 as shown in Table 1.6:
We see that this portfolio generates the same time t = T return while costing
only EUR 24600, a saving of EUR 400 against the first portfolio. So the investor
should use the second portfolio.
Now look what happens if he actual considers the following portfolio (Portfolio
III), which is the difference between Portfolio II and Portfolio I and is given in
Table 1.7. Here he short-sells 30 stocks for a total of EUR 3000, he puts EUR
1800 on his bank account and buys 40 calls for EUR 800. Hence he still has EUR
400 in his hands. He decides to invite his friends and have an exuberant lunch
with them for EUR 400.
Now, let us have a look what the value is of Portfolio III at time t = T in both
the up state (Table 1.8) and the down state (Table 1.9). In both cases the value
is exact zero ! Hence there is no risk any more and his lunch was actually a free
lunch !
So there is an arbitrage possibility in the above market situation, and the prices
quoted are not arbitrage-free prices. If we regard (as we shall do) the prices of the
bond and the stock (our underlying) as given, the option must be mispriced.
You can repeat the above example for a variety of different call option prices.
However there will be only one price, namely an initial call option price of EUR
30, where there is no free-lunch possible any more. Indeed, if the call option would
be EUR 10 more expensive, he needs to pay in total EUR 400 more for it, which
was exactly the price of his exuberant lunch. In absence of arbitrage, the price
of the call hence must be equal to EUR 30 and we have actually now determined
its no-arbitrage price! Let us also here emphasize that the above arguments were
independent of the preferences and plans of the investor.
In addition, if we only look at the position in bonds and stocks, we can say
that this position covers us against possible price movements of the option, i.e.
having EUR 1800 in your bank account and being 30 stocks short has the opposite
time t = T value as owning 40 call options. We say that the bond/stock position
is a hedge against the position in options.
This relation can be proven as follows. Assume that S(0) − exp(−rT )K+
EP (K, T )− EC(K, T ) > 0 then you could do the following: (short) sell the stock,
put exp(−rT )K on the risk-free account, write (i.e. sell) the put option and buy
the call option. You receive S(0) for the stock, have to put exp(−rT )K on the
bank account, you receive the put premium EP (K, T ) and you have to pay the
call premium EC(K, T ). You still then have in your hands: S(0) − exp(−rT )K+
EP (K, T )− EC(K, T ), which we assumed to be strictly positive.
The total payoff at maturity of this position equals −S(T ) + (S(T ) − K)+ −
(K − S(T ))+ + K = 0. Indeed, you have to close your short position in the stock
and buy it back (−S(T )), you receive the call payoff ((S(T )−K)+ ) but you have to
pay the put payoff (−(K − S(T ))+ ) and your initial position on the bank account
has now grown over the period of length T to the exact amount K. Hence, you
started with nothing and at the end of the trade there is no risk any more, but
after initiating the trade you had some (free) money left. This is an arbitrage
1.6. THE PUT-CALL PARITY 27
opportunity and hence not possible. Our initial assumption S(0) − exp(−rT )K+
EP (K, T )− EC(K, T ) > 0 hence can not be true.
Now assume S(0) − exp(−rT )K+ EP (K, T )− EC(K, T ) < 0 then you could
do the opposite: buy the stock, borrow exp(−rT )K from the bank, write the
call option and buy the put option. Then at time zero, you have to pay S(0)
for the stock, cash exp(−rT )K from the bank account, pay the put premium
EP (K, T ) and receive the call premium EC(K, T ). You still have in your hands:
−S(0) + exp(−rT )K − EP (K, T ) + EC(K, T ), which we assumed to be strictly
positive.
The total payoff at maturity of this position equals S(T ) − (S(T ) − K)+ +
(K − S(T ))+ − K = 0. Indeed at maturity, your stock is worth S(T ), you receive
the put payoff but you have to pay the call payoff. Finally you have to pay back
the borrowed amount plus interest rates namely the exact amount K. Hence you
started with no money and at the end of this trade there is again no risk any more,
but after initiating the trade you had some (free) money to spent. This is again
an arbitrage opportunity and hence not possible. Our initial assumption S(0) −
exp(−rT )K+ EP (K, T )− EC(K, T ) < 0 hence can not be true. In conclusion,
we must have S(0) − exp(−rT )K+ EP (K, T )− EC(K, T ) = 0.
The situation complicates in the two-price-world where the price of derivatives
depend on whether you are buying or selling. The first trade above namely, short
selling the stock, putting exp(−rT )K on the risk-free account, writing the put
option and buying the call option, gives you S(0) for the stock, you have, since
you are buying, to pay the ask price for the call, askEC(K, T ), you sell the put
and hence receive the bid price of it bidEP (K, T ) and you have to put exp(−rT )K
on the bank account. Net this leads to : S(0) − askEC(K, T ) + bidEP (K, T ) −
exp(−rT )K, which if strictly positive would lead to an arbitrage. Hence
and so
S(0) − exp(−rT )K ≤ askEC(K, T ) − bidEP (K, T ).
Similarly, for the second trade, namely buying the stock, borrowing exp(−rT )K
from the bank, buying the put option and selling the call option, you have to pay
S(0) for the stock, receive the call’s bid price bidEC(K, T ), pay the put’s ask price
askEP (K, T ) and cash exp(−rT )K from the bank account. The total amount
you have left in your hands after the trade equals : −S(0) + bidEC(K, T ) −
askEP (K, T ) + exp(−rT )K. This amount can not be strictly positive because
28 CHAPTER 1. FINANCIAL MATHEMATICS PRINCIPLES
or
S(0) − exp(−rT )K ≥ bidEC(K, T ) − askEP (K, T ).
In conclusion we have:
The forward price is the strike K ∗ for which EC(K ∗ , T ) = EP (K ∗ , T ). One can
easily see from the put-call parity that this is the case when K ∗ = exp(rT )S(0).
1.8. SPREAD INEQUALITIES 29
(S(T ) − K1 )+ − (S(T ) − K2 )+ .
As can be seen from Figure 1.7, since K1 < K2 , this payoff is always non-negative
and has a region where it pays out some strictly positive amount.
Hence the price for this position at time zero needs to be strictly positive. In
other words:
EC(K1 , T ) > EC(K2 , T ), if K1 < K2 .
Furthermore, we also have always that (S(T ) − K1 )+ − (S(T ) − K2 )+ ≤ K2 − K1 .
Therefore,
EC(K1 , T ) − EC(K2 , T ) ≤ exp(−rT )(K2 − K1 ),
or
0 ≤ exp(−rT )(K2 − K1 ) + EC(K2 , T ) − EC(K1 , T ).
Similar arguments, lead to the following inequalities for European puts on the
same stock and with he same maturity T and with strikes K1 < K2 .
And,
EP (K2 , T ) − EP (K1 , T ) ≤ exp(−rT )(K2 − K1 ),
or
0 ≤ exp(−rT )(K2 − K1 ) + EP (K1 , T ) − EP (K2 , T ).
30 CHAPTER 1. FINANCIAL MATHEMATICS PRINCIPLES
then one could buy the long dated call and sell the short dated call and still have
due to the assumed inequality some positive money left at time zero. One then
waits until T1 and if the shorted dated call would end out of the money, the short
1.9. DIVIDENDS 31
dated call is expiring worthless and we still hold the long dated which can only
give us a non-negative payoff. If the short dated call ends in the money, we have
a negative payoff equal to −(S(T1 ) − K exp(−r(T2 − T1 ))), since we sold the call
and it is now ending in the money. At that point, we short one stock for the price
of S(T1 ). All the cash is put on the bank account and totals K exp(−r(T2 − T1 )).
Indeed, we got S(T1 ) from shorting the stock and −(S(T1 ) − K exp(−r(T2 − T1 )),
the (negative) payoff from the short dated call. So now, we are short a stock and
have on our bank account K exp(−r(T2 −T1 )). We then wait until T2 , at which our
bank account has grown to the value K. If the long dated call ends in the money
(S(T2 ) > K), we exercise it and receive the stock for which we pay the price K
and hence our short position in stock (initiated at time T1 ) is closed and the bank
account is back to zero. If the long dated call ends OTM (S(T2 ) ≤ K), we buy a
stock in the market for a value less than K and hence close our short position in
stock and still have money on the bank account left since S(T2 ) ≤ K. In all the
above cases we started with nothing, had some free money after we bought and
sold the two calls. In none of the cases we had a loss at the end, in some we even
had some extra profit. This is a clear arbitrage and hence our starting assumption
is false. We must have that EC(K exp(−r(T2 − T1 )), T1 ) − EC(K, T2 ) ≤ 0.
A similar argument can be made to derive the calender spread inequality for
a put. The prices of two European put options on the same underlier but with
different maturities 0 < T1 < T2 satisfy
1.9 Dividends
Holding a stock entitles the owner of it to receive dividend payments. Usually, the
stock price decreases by approximately the dividend amount after it is paid out
(the ex-dividend date). Since dividend payments hence influence the stock price,
the pricing of an equity derivative instrument should take into account dividends
as well and for this one needs to make an estimation of future dividend payments.
This is however not straightforward, since dividends are decided at firm level and
will depend on the future performance and strategy of the company. Dividends are
usually expressed in a currency amount per stock. As is the case for interest rates,
a continuously compounded dividend rate or yield q can be estimated. Using such
32 CHAPTER 1. FINANCIAL MATHEMATICS PRINCIPLES
1.10 Extra
1.10.1 Bid - Ask Pricing Inequalities
Above we have derived several price relations most of the time in a one price
world. The situation in a two-price-framework is typically a bit more involved.
Similarly, if you sell the call and buy the put option - the former is done at bid
price bidEC(K, T ) and the later at the ask price askEP (K, T ) - and in addition
go long, i.e. buy, the forward at askF orward(K, T ), the total cash received, i.e.
bidEC(K, T )− askEP (K, T ) − askF orward(K, T ), can for the same no-arbitrage
reason not be strictly positive. Therefore
Similar arguments establish the following inequalities for bid and ask prices
for put options, or the put spread:
and
0 ≤ exp(−rT )(K2 − K1 ) + askEP (K1 , T ) − bidEP (K2 , T ).
is the value of this forward contract and for what value of K has the contract a
zero value (the forward price of the USD) ? It will turn out that price is now
K = exp((rd − rf )T )S0 .
Indeed, suppose K > exp((rd − rf )T )S0 . An investor can then do the fol-
lowing (at time 0): Borrow exp(−rf T )N S0 euros at rate rd ; use this cash to buy
exp(−rf T )N USD and put this on an USD bank account at rate rf and sell/short
the forward contract. Then the holding of the foreign currency grows to N USD
because of the interest (rf ) earned. Under the terms of the contract this holding is
exchanged for N × K Euro at time T . An amount exp(rd T ) exp(−rf T )N S0 is re-
quired to repay the borrowing. Hence a net profit of N (K−S0 exp((rd −rf )T )) > 0
is, therefore, made at time T .
In case K < exp((rd − rf )T )S0 you can do the following: borrow exp(−rf T )N
USD at rate rf ; use this cash to buy exp(−rf T )N S0 euros and put this on an euro
denominated bank account at rate rd ; take a long (buy) position in the forward
contract. Then the domestic currency grows to N S0 exp((rd − rf )T )), you pay
N × K Euro to receive N USD and uses these dollars to pay the loan. In total
you earned N (S0 exp((rd − rf )T )) − K) euro which in this case was assumed to
be positive.
1020,20134
1020.20
1020.18
1.020,15 1020.15
1.020,1 1020.10
1.020,05
1.020
5 10 15 20 25 30 35 40 45 50 55
m
1.4
1.2
1
interest rate (%)
0.8
0.6
0.4
0.2
−0.2
0 5 10 15 20 25 30
maturity
Discount Factors based on risk−free EUR yield curve (30 december 2014)
1.05
0.95
0.9
discount factor
0.85
0.8
0.75
0.7
0.65
0.6416
0 5 10 15 20 25 30
t
45
40
35
30
payoff
25
20
15
10
0
0 50 100 150
Stock price at maturity
90
80
70
60
payoff
50
40
30
20
10
0
0 50 100 150
Stock price at maturity
PPN investment
1800
stock investment
1600
1400
1200
wealth
1000
800
600
400
200
0
0 20 40 60 80 100 120 140 160 180 200
final stock price in percentage of spot
Figure 1.6: Comparing the Principal Protected Note (PPN) with a direct stock
investment. Initial investment is USD 1000; participation rate equals 60 %.
42 CHAPTER 1. FINANCIAL MATHEMATICS PRINCIPLES
30
25
20
payoff
15
10
0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Stock price at maturity
Butterfly
12
10
8
Payoff
0
0 50 100 150
Stock price at maturity
Figure 1.8: Butterfly payoff function (K1 = 90, K2 = 100 and K3 = 105)
44 CHAPTER 1. FINANCIAL MATHEMATICS PRINCIPLES
Chapter 2
Tree Models
Example 6. Let our financial market consist of two financial assets, a risk-less
bank account (or bond) B and a risky stock S, with today’s price S0 = 20 Euro.
We look at a single-period model and assume that starting from today (t = 0) the
world can only be in one of two states at time t = T : the stock price will either
be ST = 22 Euro or ST = 18 Euro. We are interested in valuing a European call
option to buy the stock for 21 Euro at time t = T . At time t = T , this option can
have only two possible values. It will have value 1 Euro, if the stock price is 22
Euro; if the stock price turns out to be 18 Euro at time t = T , the value of the
option will be zero. The situation is illustrated in Figure 2.1.
It turns out that we can price the option by the assumption that no arbitrage
opportunities exist. We set up a portfolio of the stock and the option in such a way
that there is no uncertainty about the value of the portfolio at the time of expiry,
t = T . We then argue that, because the portfolio has no risk, the return earned
on it must equal the risk-free interest rate of the bank account. This enables us
to work out the cost of setting up the portfolio and, therefore, the option’s price.
45
46 CHAPTER 2. TREE MODELS
22∆ − 1 = 18∆
or
∆ = 0.25.
A risk-less portfolio is, therefore long 0.25 shares and short 1 option. If the
stock price moves up to 22 Euro, the value of the portfolio is 22×0.25−1 = 4.5. If
the stock price moves down to 18 Euro, the value of the portfolio is 18×0.25 = 4.5.
Regardless of whether the stock price moves up or down, the value of the portfolio
is always 4.5 Euro at the end of the life of the option.
Risk-less portfolios must, in the absence of arbitrage opportunities, earn the
risk free rate of interest. Suppose that in this case the risk-free rate is 12 percent
per annum and that T = 0.5, i.e. six months. It follows that the value of the
portfolio today must be the present value of 4.5 Euro, or 4.5 exp(−0.12 × 0.5) =
4.238 The value of the stock today is known to be 20 Euro. Suppose the option
price is denoted by f . The value of the portfolio today is 20 × 0.25 − f = 5 − f .
It follows that
5 − f = 4.238
or
f = 0.762.
This shows that, in the absence of arbitrage opportunities, the current value of the
option must be 0.762. If the value of the option were more than 0.762 Euro, the
portfolio would cost less than 4.238 Euro to set up and would earn more than the
risk-free rate. If the value of the option were less than 0.762 Euro, shorting the
portfolio would provide a way of borrowing money at less than the risk-free rate.
In other words, if the value of the option were more than 0.762 Euro, for example
1 Euro, you can borrow for example 42380 Euro and buy 10000 times the above
portfolio at a cost of 10000(0.25 × 20 − 1) = 40000 Euro.
2.1. THE BINOMIAL TREE 47
You pocket 2380 Euro and after 6 months, you sell 10000 portfolio and cashes
in 45000, because the value of one portfolio is always 4.5 Euro. With this money
you pay back the bank for the money you borrowed plus the interests on it, i.e. you
pay the bank an amount of 42380 × exp(−0.12 × 0.5) = 45000 Euro. At the end
of all this you earned 2380 Euro without taking any risk and without an initial
capital. If the value of the option were less than 0.762 Euro, you do the opposite.
S0 u∆ − fu = S0 d∆ − fd ,
or
fu − fd
∆= . (2.1)
S0 u − S0 d
In this case, the portfolio is risk-less and must earn the risk-less interest rate. If
we denote the risk-free interest rate by r, the present value of the portfolio is
S0 ∆ − f.
It follows that
(S0 u∆ − fu ) exp(−rT ) = S0 ∆ − f,
or
f = S0 ∆ − (S0 u∆ − fu ) exp(−rT ).
48 CHAPTER 2. TREE MODELS
Substituting from Equation 2.1 for ∆ and simplifying, this equation reduces to
where
exp(rT ) − d
p= . (2.3)
u−d
Note, that if we assume that u > exp(rT ) > d ≥ 0, one can easily show
that the value of p given in Equation 2.3 satisfies 0 < p < 1 and hence can
be interpreted as a probability. Note that it is actual natural to assume that
u > exp(rT ), because it means that after a time T , you can gain more (a factor
u) by investing in the risky stocks, than you can earn with a risk-less investment
in bond (a factor exp(rT )). If this was not the case no one would invest in stocks.
Of course, you can also lose money (d factor) by investing in stocks.
Also note that Equation 2.1 shows that ∆ is the ratio of the change in the
option price to the change in the stock price.
Finally, remark that the option pricing formula in (2.2) does not involve the
actual probabilities of the stock moving up or down; p can be interpreted as an
artificial probability (see later). This is surprising and seems counter-intuitive.
The key reason is that the probabilities of future up or down movements are
already incorporated into the price of the stock.
showing that the stock price grows, on average, at the risk-free rate r. Setting
the probability of an up movement equal to p is therefore, equivalent to assuming
that the return on the stock on average equals the risk-free rate. In a risk-neutral
world the expected return on all securities is the risk-free interest rate. Equation
2.4 shows that we are assuming a risk-neutral world when we set the probability
of an up movement to p. Equation 2.2 shows that the value of the option is
its expected payoff in a risk-neutral world discounted at the risk-free rate. This
result is an example of an important general principle in option pricing known as
risk-neutral valuation. The principle states that it is valid to assume the world
is risk neutral when pricing options. The resulting option prices are correct not
just in a risk-neutral world, but in the real world as well.
or
20 exp(0.12 × 0.5) − 18
p= = 0.8092.
4
At the end of the six months, the call option has a 0.8092 (risk-neutral) probability
of being worth 1 Euro and a 0.1908 (risk-neutral) probability of being worth zero.
Its expected value is, therefore,
To arrive to the option price, this should be discounted at the risk-free rate. The
value of the option today is, therefore,
This is the same value as the value obtained earlier, illustrating that no-arbitrage
arguments and risk-neutral valuation give the same answer.
Example 7. Assume the stock price starts at 20 Euro and in each of the two time
steps may go up by 10 percent or down by 10 percent. We suppose that each time
step is six months long and the risk-free interest rate is 12 percent per annum.
We consider an European call option with a strike price of 21 Euro.
Figure 2.3 shows the tree with both the stock price and the option price at
each node. The stock price is the upper number and the option price is the lower
number. The option prices at the final nodes of the tree are easily calculated. They
are the payoffs from the option. At node D, the stock price is 24.2 Euro and the
option price is 24.2 − 21 = 3.2 Euro; at nodes E and F, the option is out of the
money and its value is zero.
At node C, the option price is zero, because node C leads to either node E or
node F and at both nodes the option price is zero. Next, we calculate the option
price at node B. Using the notation introduced earlier in the chapter, u = 1.1,
d = 0.9, r = 0.12, and T = 0.5 so that p = 0.8092. Equation 2.2 gives the value
of the option at node B as
We can generalize the case of two time steps by considering the situation in
Figure 2.4.
The stock price is initially S0 . During each step, it either moves up to u times
its value or moves down to d times its value. The notation for the value of the
option is shown on the tree. For example, after two up movements, the value of
the option is fuu . We suppose that the risk-free interest rate is r and the length
of the time step is ∆t years.
Repeated application of Equation 2.2 gives
and
exp(r∆t) − d
p=
u−d
Substituting the first two equations in the last one, we get
• The expected return from all traded securities is the risk-free interest rate.
• Future cash flows can be valued by discounting their expected values at the
risk-free interest rate.
We make use of this when using a binomial tree. The tree is designed to
represent the behaviour of a stock price in a risk-neutral world. In this risk-
neutral world the probability of an up movement will be denoted by p. The
probability of a down movement is 1 − p; as seen above in Equation 2.3:
exp(r∆t) − d
p= . (2.9)
u−d
Figure 2.5 illustrates the tree of stock prices over 5 time periods that is consid-
ered when the binomial model is used. At time zero, the stock price S0 is known.
At time ∆t there are two possible stock prices, S0 u and S0 d; at time 2∆t , there
are three possible stock prices, S0 u2 , S0 ud, and S0 d2 ; and so on. In general, at
time i∆t , i + 1 stock prices are considered. These are
S0 uj di−j , j = 0, ..., i.
European options are evaluated by starting at the end of the tree (time T )
and working backward. The value of the option is known at time T . For example,
an European put option is worth (K − ST )+ and an European call option is worth
(ST − K)+ , where ST is the stock price at time T and K is the strike price.
Because a risk-neutral world is being assumed, the value at each node at time
T − ∆t can be calculated as the expected value at time T discounted at rate r
for a time period ∆t. Similarly, the value at each node at time T − 2∆t can be
calculated as the expected value at time T − ∆t discounted for a time period ∆t
2.2. MATCHING TREES WITH A GIVEN VOLATILITY 53
at rate r, and so on. Eventually, by working back through all the nodes, the value
of the option at time zero is obtained. This procedure is illustrated in Figure 2.6.
Another way of calculating the option prices is by directly taking the dis-
counted value of the expected payoff of the option in the risk-neutral world. For
example the European put, with strike price K and maturity T has a value:
N
X + N j
exp(−rT ) K − S0 uj dN −j p (1 − p)N −j . (2.10)
j
j=0
For more complex options, but where the payoff only depends on the final stock
price, i.e. the payoff is a function of ST , g(ST ) say, a similar expression can be
derived; the current value of the option is then given by:
N
X
j N −j N j
exp(−rT )Ep [g(ST )] = exp(−rT ) g(S0 u d ) p (1 − p)N −j . (2.11)
j
j=0
where Ep denotes the expectation in the risk-neutral world, i.e. with a probability
p of an one-step up-move of size u given by Equation 2.9. Then, probability of an
one-step down-move of size d equal 1 − p. Therefore, we have a probability
N j
p (1 − p)N −j , j = 0, 1, . . . , N. (2.12)
j
where S0 is the current level of the stock and S1 denotes the level of the stock after
one year.
Note, that if we would assume (as is approximately the case in the Black-
Scholes model and in a binomial tree model with many step (due to the CLT))
that (S1 − S0 )/S0 follows a Normal distribution with mean µ and variance σ 2 , we
have
To see how one can match the choice of u and d with a given volatility, suppose
that the expected return on a stock in the real world is µ: The expected stock
price at the end of the first time step is S0 (1 + µ∆t). The volatility of a stock
price, σ, is defined so that σ 2 ∆t is the variance of the return in a short period of
time of length ∆t.
Note that ∆t is here typically assumed small and expressed in year terms.
For example we can have ∆t = 1/52, and then it is representing one week. If we
assume independence over each period (in the example a week), we have that due
to linearity of the variance for independent variables, that the variance over one
year is the sum of the variances over each period. If each week is the same, then
52 weeks with each a variance of σ 2 ∆t add up to a yearly variance of σ 2 .
Let us denote the probability of an up movement in the real world by q. In
order to match the expected return on the stock, we must therefore, have
or
(1 + µ∆t) − d
q= . (2.13)
u−d
2.2. MATCHING TREES WITH A GIVEN VOLATILITY 55
In order to match the real world stock price volatility we must therefore have
or equivalently
q(1 − q)(u − d)2 = σ 2 ∆t. (2.14)
Substituting from Equation (2.13)into Equation (2.14) we get
Another setting is
√
u = exp(σ ∆t) (2.17)
√
d = exp(−σ ∆t) (2.18)
where pu , pm and pd are respectively the probabilities to move to the up-, middle-
or down-state. One can show that these probabilities are providing us an almost
(but not exact) risk-neutral setting, i.e.
Therefore, one often refers to these probabilities as the risk-neutral ones and the
current price f of this derivative can be obtained, or better approximated, as the
discounted expected payoff under the ”risk-neutral” measure.
f ≈ exp(−r∆t) (pu fu + pm fm + pd fd ) .
After these are calculated we can move one-step backward to the initial state and
calculate the initial price as
f ≈ exp(−r∆t) (pu fu + pm fm + pd fm ) .
Suppose that we divide the life of the option into five intervals of length one month
(= 0.0833 year) for the purposes of constructing a binomial tree.
Then
∆t = 0.0833
√
u = exp(σ ∆t) = 1.1224
√
d = exp(−σ ∆t) = 0.8909
p = (exp(r∆t) − d)/(u − d) = 0.5073
58 CHAPTER 2. TREE MODELS
Figure 2.10 shows the related binomial tree. At each node there are two num-
bers. The top one shows the stock price at the node; the lower one shows the value
of the option at the node. The probability of an up movement is always 0.5073;
the probability of a down movement is always 0.4927. The stock price at the jth
node (j = 0, 1, ..., i) at time i∆t, i = 0, 1, 2, 3, 4, 5 is calculated as S0 uj di−j . The
option prices at the final nodes are calculated as (K − ST )+ = max{K − ST , 0}.
The option prices at the penultimate nodes are calculated from the option prices
at the final nodes. First, we assume no exercise of the option at the nodes. This
means that the option price is calculated as the present value of the expected option
price one step later. For example at node C, the option price is calculated as
The option should, therefore, not be exercised at this node, and the correct option
value at the node is 10.36 Euro. Working back through the tree, we find the value
of the option at the initial node to be 4.49 Euro. This is our numerical estimate
for the option’s current value. In practice, a smaller value of ∆t, and many more
nodes, would be used. It can be shown that with 30, 50, and 100 time steps we get
values for the option of 4.263, 4.272, and 4.278.
node at time i∆t as the (i, j) node. Define fi,j as the value of the option at the
(i, j)node. The stock price at the (i, j) node is S0 uj di−j . Because the value of an
American put at its expiration date is (K − ST )+ = max{K − ST , 0}, we know
that
fN,j = max{K − S0 uj dN −j , 0}, j = 0, 1, ..., N.
There is a probability, p, of moving from the (i, j) node at time i∆t to the (i +
1, j + 1) node at time (i + 1)∆t, and a probability 1 − p of moving from the (i, j)
node at time i∆t to the (i + 1, j) node at time (i + 1)∆t. Assuming no early
exercise, risk-neutral valuation gives
Note that, because the calculations start at time T and we work backward, the
value at time i∆t captures not only the effect of early exercise possibilities at time
i∆t, but also the effects of early exercise at subsequent times.
In the limit as ∆t tends to zero, an exact value for the American put is
obtained. In practice, N = 50 usually gives reasonable results.
Proof. That C ≥ 0 is obvious, otherwise buying the call would give a riskless
profit now and no obligations later. To prove the remaining lower bound, we set
up an arbitrage table (see Table 2.1) to examine the cash flows of the following
portfolio:
sell 1 stock short, buy 1 call, invest in bank account exp(−rT )K.
Suppose now that the American call is exercised at some time t strictly less
than expiry T , i.e. t < T . The financial agent thereby realises a cash-flow St − K.
From the above proposition we know that the value of the call must be greater
or equal to St − exp(−r(T − t))K, which is greater than St − K, if r ≥ 0. Hence
selling the call would have realised a higher cash-flow and the early exercise of the
call was suboptimal. In conclusion the price of an Amercian call equals the price
of an European call:
AC = EC.
Proof yourself that if r ≤ 0, it is never optimal to exercise an American put
option on a non-dividend paying stock : EP = AP .
For a fixed maturity. T and taking N time steps, i.e. ∆t = T /N , we have that
the risk-neutral probability of moving upwards equals:
p
exp(rT /N ) − exp(−σ T /N )
pN = p p .
exp(σ T /N ) − exp(−σ T /N )
Let us now investigate the risk-neutral limiting distribution of ST :
YN p p XN
ST = S0 exp(Zj σ T /N ) = S0 exp σ T /N Zj
j=1 j=1
where Zj are independent random variables taking the values −1 and 1, with
probabilities 1 − pN and pN respectively, for j = 1, . . . , N . In other words:
p N
X
log ST = log S0 + σ T /N Zj
j=1
E[Zj ] = 2pN − 1;
V ar[Zj ] = 4pN (1 − pN ).
62 CHAPTER 2. TREE MODELS
leads to √
log ST →D log S0 + σ T N + (r − σ 2 /2)T
when N → +∞. The distribution of the logarithm of the stock price thus follows
a Normal distribution with mean (r −(1/2)σ 2 )T and variance σ 2 T ; the stock price
itself is thus lognormally distributed.
The price of the derivative in the limit will be given by
√
lim exp(−rT )EpN [g(ST )] = exp(−rT )E[g(S0 exp((r − (1/2)σ 2 )T + σ T N ))].
N →+∞
In case of the European call option with strike K and time to maturity T , one
can with a little effort show that its initial price is given by:
where
log(S0 /K) + (r + σ 2 /2)T
d1 = √
σ T
log(S0 /K) + (r − σ 2 /2)T √
d2 = √ = d1 − σ T
σ T
and N (x) is the cumulative probability distribution function for a variable that
is standard Normal distributed. This is the famous Black-Scholes formula. This
lognormal model (the Black-Scholes model), will be studied in Chapter 5.
2.5. LIMITS OF TREE MODELS 63
stock derivative
1
22
f=?
20
0
18
u S0 fu
S0 f
d S0 fd
u S0 fu
f fm
S0 m S0
d S0 fd
u2 S0
u S0 u m S0
S0 m S0 m2 S0
d S0 d m S0
d2 S0
fuu
fu fum
f fm fmm
fd fmd
fdd
Mathematical Finance in
Discrete Time
Any variable whose value changes over time in an uncertain way is said to follow
a stochastic process. Stochastic processes can be classified as discrete-time or
continuous-time. A discrete-time stochastic process is one where the value of the
variable can change only at certain fixed points in time, whereas a continuous-time
stochastic process is one where changes can take place at any time. Stochastic
processes can also be classified as continuous-variables or discrete-variables. In a
continuous-variable process, the underlying variable can take any value within a
certain range, whereas in a discrete-variable process, only certain discrete values
are possible. Binomial tree models belong to the discrete-time, discrete-variable
stochastic processes.
In this chapter we study so-called finite markets, i.e. discrete-time models of
financial markets in which all relevant quantities take a finite number of values.
We specify a time horizon T , which is the terminal date for all economic activities
considered. For a simple option pricing model the time horizon typically corre-
sponds to the expiry date of the option. We thus work with a finite probability
space (Ω, P ), with a finite number |Ω| of possible outcomes ω, each with a positive
probability: P ({ω}) > 0.
73
74 CHAPTER 3. MATHEMATICAL FINANCE IN DISCRETE TIME
the most important determinant of success in financial life. We shall confine our-
selves to the situation where agents take decisions on the basis of information in
the public domain, available to all. We shall further assume that information once
known remains known and can be accessed in real time.
Our financial market contains two financial assets. A risk-free asset (the bond)
with a deterministic price process Bi , and a risky assets with a stochastic price
process Si . We assume B0 = 1 (we reckon in units of the initial value of the bond)
and Bi > 0; we say it is a numeraire. 1/Bi is called the discounting factor for
time i.
As time passes, new information becomes available to all agents. There exists
a mathematical object to model this information flow, unfolding with time: fil-
trations. The concept filtration is not that easy to understand. The full theory
will lead us too far.
In order to clear this out a bit, we explain the idea of filtration in a very
idealized situation. We will consider a stochastic process X which starts at some
value, zero say. It will remain there until time t = 1, at which it can jump with
positive probability to the value a or to a different value b. The process will stay
at that value until time t = 2 at which it will jump again with positive probability
to two different values: c and d say if it was at time t = 1 at a and f and g say if
the process was at time t = 1 at state b. From then on the process will stay in the
same value. The set of outcomes of the probability space consists of all possible
paths the process can follow, i.e. all possible outcomes of the experiment. We
will denote the path 0 → a → c by ω1 , similarly the paths 0 → a → d, 0 → b → f
and 0 → b → g are denoted by ω2 , ω3 and ω4 respectively. So we have
Ω = {ω1 , ω2 , ω3 , ω4 }.
In this situation we will take the following flow of information, i.e. filtrations:
Ft = {∅, Ω}, 0 ≤ t < 1;
Ft = {∅, {ω1 , ω2 }, {ω3 , ω4 }, Ω}, 1 ≤ t < 2;
Ft = D(Ω) = F, 2 ≤ t.
We set here D(Ω), the set of all subsets of Ω. To each of the filtrations given
above, we associate resp. the following partitions (i.e. the finest possible one):
P0 = {Ω}, 0 ≤ t < 1;
P1 = {{ω1 , ω2 }, {ω3 , ω4 }} 1 ≤ t < 2;
P2 = {{ω1 }, {ω2 }, {ω3 }, {ω4 }} 2 ≤ t.
3.1. INFORMATION AND TRADING STRATEGIES 75
is the change in the market value due to changes in security prices which occur
between time i − 1 and i. We call Gφ = G = {Gφi , i = 1, . . . , T }, where
i
Gφi =
X
βj (Bj − Bj−1 ) + ζj (Sj − Sj−1 )
j=1
V0 = β1 B0 + ζ1 S0 , Vi = βi+1 Bi + ζi+1 Si , i = 1, . . . , T.
• X is F-adapted
EQ [|Xi |] < ∞ i = 1, . . . , T.
EQ [Xi |Fj ] = Xj , 0 ≤ j ≤ i ≤ T.
X̃i = Xi /Bi .
S̃i = Si /Bi .
where the third line is because S̃ is a P ∗ -martingale and the fifth line is because
of the self-financing property of φ.
EP ∗ [ṼTφ ] = V0φ .
Together with the fact that for each ω ∈ Ω, we have P ∗ ({ω}) > 0, this leads to
a contradiction. Indeed, we can not have a non-negative random variable with a
zero mean and positive mass on the positive real numbers.
VTφ = VTψ = X.
So
82 CHAPTER 3. MATHEMATICAL FINANCE IN DISCRETE TIME
Bi
πiX = Viφ = EP ∗ [X|Fi ], 0≤i≤T
BT
and call πiX the no-arbitrage price of the contingent claim X at time i. For, if an
investor sells the claim X at time i for πiX , he can follow strategy φ to replicate X
at time T and clear the claim; an investor selling this value is perfectly hedged.
To sell the claim for any other amount would provide an arbitrage opportunity.
We note that, to calculate prices as above, we need to know only:
• the filtration F,
• P ∗.
We do not need to know the underlying probability measure P (only its null
sets, to know what ’equivalent to P ’ means and actually in our finite model there
are no non-empty null-sets, so we do not need to know even this). Now pricing of
contingent claims is our central task, and for pricing purposes P ∗ is vital and P
itself irrelevant. We thus may - and shall - focus attention on P ∗ , which is called
the risk-neutral probability measure.
To summarize, we have:
3.5.2 Examples
The One-step Binomial Model
We return to model given in Section 2.1.1. There exists only two possible out-
comes. There is an up-state u in which the stock price after one time step equals
S1 = uS0 and a down-state d if the stock price changes to S1 = dS0 :
Ω = {u, d}.
f = (1/b)(p∗ fu + (1 − p∗ )fd ).
84 CHAPTER 3. MATHEMATICAL FINANCE IN DISCRETE TIME
In order to hedge or replicated this claim one has to solve the equations
βb + ζuS0 = fu
βb + ζdS0 = fd
Note that this system of equations has a unique solution if and only if S0 6= 0,
b 6= 0, and u 6= d (all which are ruled out).
or equivalently
Unfortunately this equation has more than one solution as can be easily been
seen after a simple rewriting:
(b − d) − (m − d)q ∗
p∗ =
u−d
For every 0 < q ∗ < 1 there is a corresponding p∗ . If we then take also into
account that the values of p∗ and q ∗ must give rise to a probability distribution,
i.e. 0 < p∗ , q ∗ < 1 and p∗ + q ∗ < 1, there still are infinitely many solutions. In
conclusion there exist more then one martingale measure for the discounted stock
3.6. THE PHYSICAL AND THE RISK-NEUTRAL WORLD 85
price. So the one-step trinomial model is arbitrage free, but is not complete. If
we have a contingent claim with payoff fu in the up-state, fm in the middle state
and fd in the down state it can only be replicated if there exists a solution to the
equations
βb + ζuS0 = fu
βb + ζmS0 = fm
βb + ζdS0 = fd
So only contingent claims which payoff function satisfies the above condition are
attainable and can be replicated and priced in an arbitrage-free way.
typical example is found in insurance where the forward premium paid for life,
fire or car insurance by a particular individual far exceeds the probability of the
insured event. Given the personal damage caused by the event induces people in
markets to pay more than probability for coverage. Competitive pressures do not
reduce prices to probability assessments as sufficiently large pools of identical risks
may not be available, leaving sellers exposed to risk that must be compensated in
the premium. In these two worlds, one looks differently to the stochastic behaviour
of the assets under investigation; the prices of events seen as probabilities (pricing
world) differ from the probabilities of the events happening in the real-world.
Most of the times, the probability measuring how things happen in the real-
world is denoted with a P ; therefore also this real-world is often named the P -
world. This probability measure is measuring how things actually happen in
reality and one refers often to it as the physical measure. Therefore, one is typ-
ically estimating distributions in this world on the basis of historically observed
real data of the underlying asset, like daily log-returns of a stock. In contrast,
the risk-neutral world is the pricing world, created by financial markets trading
event risk and modelled by financial engineers. Most of the times, the probability
measure of how things happen in the pricing world is denoted with a Q and one
refers to it as the pricing measure. One can prove that in absence of arbitrage,
under this measure, traded assets all behave in a ”risk-neutral” way, meaning that
there expected return is equal to the return of the risk-free account. For example,
for a stock with dividend yield q, we have that
since the price of the forward stock net of intermediate dividends must equal the
cost of buying it in the spot market on borrowed money and repaying the loan.
One then has a rate of return q from the dividends and since r is the rate of
return on the risk-free account, the rate of return on a stock needs to be r − q,
otherwise there would be the spot forward arbitrage opportunity. The pricing
world or Q-world is hence also often referred to as the risk-neutral world.
The reason for the existence of the particular risk-neutral world is the trading
of risks in markets thus leading to the existence of forward prices. As indicted
in this Chapter, fundamental theory actually shows, that under the no-arbitrage
assumption (and some additional technical assumptions), the price of a derivative
is given by the discounted expected payoff of the derivative, with expectations
taken under the risk neutral (Q) measure. One hence has for a European option
3.6. THE PHYSICAL AND THE RISK-NEUTRAL WORLD 87
these market quotes are discounted expectations under one particular Q of the re-
lated payoffs of the derivatives, one can try to estimate this underlying probability
measure Q from these market quotes.
Chapter 4
Derivatives with more complicated payoffs than the standard European or Amer-
ican calls and puts are referred to as exotics options. Most exotics options are
traded in the OTC market and have been designed to meet particular needs of
investors.
In this chapter we describe different types of exotic options and discuss their
valuation. Options of an European nature can typically be priced by Monte-Carlo
simulation (see Chapter 6).
89
90 CHAPTER 4. EXOTIC OPTIONS AND STRUCTURED PRODUCTS
• The down-and-out barrier call (DOBC) is worthless unless the asset price
remains during its lifetime, i.e. until T , above some low barrier H, in which
case it retains the structure of a European call with strike K. Staying above
a certain barrier means that the miniumum asset prices never goes below
the barrier. Its initial price is given by:
DOBC = exp(−rT )EQ [(ST − K)+ 1( min St > H)]
0≤t≤T
• The down-and-in barrier call (DIBC) is a standard European call with strike
K, if the asset went below some low barrier H before T . If this barrier was
never reached during the life-time of the option, the option is worthless. Its
initial price is given by:
DIBC = exp(−rT )EQ [(ST − K)+ 1( min St ≤ H)].
0≤t≤T
• The up-and-in barrier call (UIBC) is worthless unless the asset price went
above some high barrier H, in which case it retains the structure of a Eu-
ropean call with strike K. Its price is given by:
U IBC = exp(−rT )EQ [(ST − K)+ 1( max St ≥ H)].
0≤t≤T
• The up-and-out barrier call (UOBC) is worthless unless the asset price re-
mains below some high barrier H, in which case it retains the structure of
a European call with strike K. Its price is given by:
U OBC = exp(−rT )EQ [(ST − K)+ 1( max St < H)].
0≤t≤T
Example 22. An investor buys a DIBC on the S&P 500 beginning of February.
The option expires the 3rd week of February. So the call is struck at the money
and knocks in when the index falls 10% from its initial level. Assume that this is
offered for 0.94%. With the index at 950, this puts the strike at 950, the barrier
at 855 and the premium is 8.93 index points.
In Figure 4.1, one sees two path where the barrier is crossed. If for example
the stock price would crosses the barrier at 855, end of June, and at expiration,
the index is at 876.05, the option expires worthless, since the strike is at 950; if
however at expiration, the index would be at 1030.78 (and the barrier is crossed
in August), the option pays out 80.78 = 1030.78 - 950.00 index points.
In case the stock price would never crosses the barrier at 855 during its lifetime,
the option expires always worthless.
Example 23. An investor buys a UIBP on the S&P 500 on the S&P 500 beginning
of February.The option expires the 3rd week of February, is struck at the money
and knocks in when the index rises 5% from its initial level.
The DIBP is offered for 2.60%. With the index at 950, this puts the strike at
950, the barrier at 997.50 and the premium is 24.70 index points.
In figure 4.2, the first path never reaches a level greater than 952.53, which is
much less than the knock in barrier at 997.50. Hence, the option expires worthless
even though the index at expiration is below the strike. The second path does knock
in, crossing above the 977.50 level in mid June. At expiration, the level of the
index is 911.40, so the option pays out 38.60 = 950.00 - 911.40 index points.
4.2 Lookbacks
Lookback options, are a type of exotic options where the payoff depends on the
optimal (maximum or minimum) underlying asset’s price occurring over the life
of the option. The option allows the holder to ”look back” over time to determine
the payoff.
There exist two kinds of lookback options: with floating strike and with fixed
strike.
The floating strike lookback call has a payoff
LC = ST − min St ,
0≤t≤T
giving one essentially the right to buy at the low over [0, T ], and the lookback put
with payoff
LP = max St − ST
0≤t≤T
and +
LP f ix = K − min St .
0≤t≤T
94 CHAPTER 4. EXOTIC OPTIONS AND STRUCTURED PRODUCTS
For the put version just switch the sum and the strike price :
Pn +
k=1 Stk
AsianP = K− .
n
Average price options are typically less expensive than regular options and
are arguably more appropriate than regular options for meeting some of investors
needs. Asian options are widely used in practice - for instance, for oil and foreign
currencies. The averaging complicates the mathematics, but e.g., protects the
holder against speculative attempts to manipulate the asset price near expiry.
2 2
N (σrealized − σstrike ),
n
2 AX 2
σrealized = log(Sti /Sti−1 )
n
i=1
downside risk, a potential higher bonus in case markets move sideways and an
unlimited upside potential.
1. the FTSE-100 return (over the lifetime of the product) if it is the FTSE-100
ends above 140% of the initial level (at the start of the SP), or
2. a return of 40%, if the final level of the FTSE-100 is between 75% and 140%
of the initial level, unless the index level has fallen below 75% of the initial
level during the lifetime of the certificate in which case ...
Other products can have a early redemption before the maturity date. An
example of such a note that can have early redemption, is the auto-callable. If a
certain trigger is activated, e.g. if stock is below some barrier, then there is early
redemption.
This chapter overviews the most basic and well-known continuous-time, continuous-
variable stochastic model for stock prices: The Black-Scholes(-Merton) model. An
understanding of this is the first step to the understanding of the pricing of options
in a more advanced setting.
In the early 1970s, Fischer Black, Myron Scholes, and Robert Merton made
a major breakthrough in the pricing of stock options by developing what has
become known as the Black-Scholes model. The model has had huge influence
on the way that traders price and hedge options. In 1997, the importance of the
model was recognized when Myron Scholes and Robert Merton were awarded the
Nobel prize for economics. Sadly, Fischer Black died in 1995, otherwise he also
would undoubtedly have been one of the recipients of this prize.
99
100 CHAPTER 5. THE BLACK-SCHOLES MODEL
5.1.2 Martingales
A stochastic process X = (Xt , 0 ≤ t ≤ T ) is a martingale relative to (P, F) if
• X is F-adapted;
A martingale is ’constant on average’, and models a fair game. This can be seen
from the third condition: the best forecast of the unobserved future value Xt
based on information at time s, Fs , is the at time s known value Xs .
(x − µ)2
2 1
fN ormal (x; µ, σ ) = √ exp − . (5.1)
2πσ 2 2σ 2
In Figure 5.1, one sees the typical bell-shaped curve of the density of a standard
normal (Normal(0, 1)) density.
We will denote by
Z x
N(x) = fN ormal (u; 0, 1)du (5.2)
−∞
5.2. THE NORMAL DISTRIBUTION AND THE BROWNIAN MOTION 101
the cumulative distribution function (cdf) for a variable that is standard normally
distributed (Normal(0, 1)). This special function is available in most mathematical
software packages. The function is shown in Figure 5.2.
The Normal(µ, σ 2 ) distribution is symmetric around its mean, and has always
a kurtosis equal to 3:
Normal(µ, σ 2 )
mean µ
variance σ2
skewness 0
kurtosis 3
The Normal distribution arises from the Central Limit Theorem (CLT) (see
Theorem 11). Intuitively, it tells us that the suitable normalized sum of many
independent random variables is approximately normally distributed.
Definition
A stochastic process X = {Xt , t ≥ 0} is a standard Brownian motion on some
probability space (Ω, F, P ), if
1. X0 = 0 a.s.
Properties
Next, we look at some of the classical properties of Brownian motion.
Path Properties One can proof that Brownian motion has continuous paths,
i.e. Wt is a continuous function of t. However the paths of Brownian motion
are very erratic. They are for example nowhere differentiable. Moreover, one can
prove also that the paths of Brownian motion are of infinite variation, i.e. their
variation is infinite on every interval.
Another property is that for a Brownian motion W = {Wt , t ≥ 0}, we have
that
P (sup Wt = +∞ and inf Wt = −∞) = 1.
t≥0 t≥0
This result tells us that the Brownian path will keep oscillating between positive
and negative values.
insight that stochastic integrals can be defined for a suitable class of integrands.
We only show how these integrals can be defined for some simple integrands X.
Indicators
If Xt = 1[a,b] (t), i.e. it equals 1 between a and b and is zero elsewhere, we define
Z t 0 if t ≤ a
It (X) = Xs dWs = Wt − Wa if a ≤ t ≤ b
0
Wb − Wa if t ≥ b
we define
Z t n
X Z t
It (X) = Xs dWs = ci 1[ai ,bi ] (s)dWs .
0 i=1 0
n−1
X
Xt = ζ0 10 (t) + ζi 1(ti ,ti+1 ](t), 0 ≤ t ≤ T.
i=0
Then if tk ≤ t ≤ tk+1 , k = 0, . . . , n − 1,
Z t k−1
X
It (X) = Xs dWs = ζi (Wti+1 − Wti ) + ζk (Wt − Wtk ).
0 i=0
5.3. ITÔ’S CALCULUS 105
Properties
It is not so hard to prove some simple properties of the stochastic integrals defined
so far:
• Linearity: It (aX + bY ) = aIt (X) + bIt (Y );
• It (X) is a martingale;
Rt
• Itô’s isometry: E[(It (X))2 ] = 0 E[Xu2 ]du.
Rt
The Itô isometry above suggests that the stochastic integral 0 Xu dWu should
be defined only for processes with
Z t
E[Xu2 ]du < ∞, for all t ≥ 0
0
and this is indeed the case. Each such X may be approximated by a sequence
of simple stochastic processes and the stochastic integral may be defined as the
limit of this approximation. Furthermore the three above properties remain true.
We will not include the technical and detailed proofs of this procedure in this
book. Note that one also can construct a closely analogous theory for stochastic
integrals with the Brownian integrator W above replaced by a more general (semi-
)martingale integrator M .
by using
F (t, x) = x2 .
We have
Ft (t, x) = 0
Fx (t, x) = 2x
Fxx (t, x) = 2
Hence Z t Z t Z t
Wt2 − 0 = 2 Wu dWu + du = 2 Wu dWu + t
0 0 0
So that
t
Wt2
Z
t
Wu dWu = − .
0 2 2
Note the contrast with ordinary calculus ! Itô calculus requires the second term
on the right - the Itô correction term.
where the coefficient functions b(t, x) and σ(t, x) satisfy the so-called following
Lipschitz and growth conditions:
for all t ≥ 0, x, y ∈ R, for some constant K > 0. These are some technical
conditions such that existence and uniqueness of the solutions is guaranteed ; all
SDEs that we will encounter satisfy these conditions.
To see that the SDE (5.5) has a solution, we first define recursively
Z t Z t
(0) (n+1)
Xt = x, Xt =x+ b(u, Xu(n) )du + σ(u, Xu(n) )dWu .
s s
(n)
One can then prove that Xt converges (in some sense), to Xt say; Xt will then
be the unique (strong) solution to SDE (5.5), i.e.
Z t Z t
Xs = x, Xt = x + b(u, Xu )du + σ(u, Xu )dWu .
s s
or
dXt = b(t, Xt )dt + σ(t, Xt )dWt , Xs = x
for short.
Consider a function F (t, x) ∈ C 1,2 . Then we have the following extension of Itô’s
Lemma:
108 CHAPTER 5. THE BLACK-SCHOLES MODEL
In the Black-Scholes model, one models the time evolution of a stock price
S = {St , t ≥ 0} as follows. Consider how S will change in some small time
interval from the present time t to a time t + ∆t in the near future. Writing
∆St for the change St+∆t − St , the relative return over this time step is then
∆St /St . It is economically reasonable to expect this return to decompose into
two components, a systematic part and a random part.
Let us first look at the systematic part. One assumes that the stock’s expected
return over a period is proportional to the length of the period considered. This
means that in a short interval of time [t, t + ∆t] of length ∆t, the expected relative
return in S is proportional to the length of the time step, namely ∆t, and is
therefore given by µ∆t, where µ is some parameter representing the mean rate of
the return of the stock.
A stock price fluctuates stochastically, and a reasonable assumption is that
the variance of the return over the interval of time [t, t + ∆t] is proportional to
the length of the interval. Furthermore, taking into account the Central Limit
Theorem and viewing the stock return as the sum on many random effects, it
makes sense to model the random part of the return by a Normally distributed
random variable with variance σ 2 ∆t, with σ > 0 a parameter which describes
how much effect this random part has - or how volatile the return is; σ is called
the volatility of the stock. Putting this together and recognizing that σ∆Wt =
σ(Wt+∆t − Wt ) has a normal distribution with the required variance σ 2 ∆t, we
could write
∆St = St (µ∆t + σ∆Wt ), S0 > 0.
In the limit, as ∆t → 0, we have the Stochastic Differential Equation (SDE)
Using Itô’s Lemma (see next example), one can prove that this stochastic
differential equation has the unique solution
σ2
St = S0 exp µ− t + σWt .
2
σ2
log St − log S0 = µ − t + σWt
2
110 CHAPTER 5. THE BLACK-SCHOLES MODEL
We have
F (t, Wt ) − F (s, Ws ) =
Z t Z t
1 t 2
Z
2
(µ − σ /2)F (u, Wu )du + σF (u, Wu )dWu + σ F (u, Wu )du
s s 2 s
or
σ2
dF = σF dWt + (µ − σ 2 /2)F dt + F dt = µF dt + σF dWt
2
for short. Hence
In Figure 6.2, one sees the realization of the geometric Brownian motion based
on the sample path of the standard Brownian motion of Figure 6.1.
on the way that traders price and hedge options. In 1997, the importance of the
model was recognized when Myron Scholes and Robert C. Merton were awarded
the Nobel prize for economics. Sadly, Fischer Black died in 1995, otherwise he
also would undoubtedly have been one of the recipients of this prize as well.
Next, we detail the Black-Scholes model and employ it for valuing European
call and put options on a stock.
nomically meaningful. A lot of effort has been put into solving this question, and
several alternatives have been proposed, but the details will lead us to far.
By the risk-neutral valuation principle one can prove that the price Vt at time
t, of a contingent claim with payoff function G({Su , 0 ≤ u ≤ T }) is given by
Vt = exp(−(T − t)r)EP ∗ [G({Su , 0 ≤ u ≤ T })|Ft ], 0≤t≤T (5.8)
where P ∗ is an equivalent martingale measure. In a general setting their is not
a unique martingale measure (incomplete market models). Roughly speaking
incompleteness means that a general contingent claim can not be perfectly hedged.
Most models are not complete, and most practitioners believe the actual market
is not complete. we have to choose an equivalent martingale measure in some
way and this is not always clear. Actually, the market is choosing the martingale
measure for us.
In the Black-Scholes world however, one can prove (Girsanov Theorem) that
there is a unique equivalent martingale measure and we do not have to deal with
choosing an appropriate one. It is not hard to see that under P ∗ , the stock price
is following a Geometric Brownian motion again. This risk-neutral stock price
process has the same volatility parameter σ, but the drift parameter µ is changed
to the continuously compounded risk-free rate r.
The Black-Scholes models is hence a complete model, that is, every contingent
claim can be replicated by a dynamic self-financing trading strategy.
Next, we will calculate European call option prices under this model.
where
σ2
log(S0 /K) + (r − q + 2 )T
d1 = √ , (5.11)
σ T
σ2 √
log(S0 /K) + (r − q − 2 )T
d2 = √ = d1 − σ T , (5.12)
σ T
and N(x) is the cumulative probability distribution function for a variable that is
standard normally distributed (Normal(0, 1)).
From this, one can also easily (via the put-call parity) obtain the price EP (K, T )
of the European put option on the same stock with same strike K and same ma-
turity T :
For the call, the risk-neutral probability (Q) of finishing in the money corre-
sponds with N(d2 ). Similarly, the delta (i.e. the change in the value of the option
compared with the change in the value of the underlying asset of the option cor-
responds with N(d1 ) ; see also Section 5.7) .
5.7 Hedging
If we have an investment in a portfolio of a stock, a risk-free account and a
derivative, all 3 asset’s value can change from day to day. The option price changes
because the underlying stock price moved (Delta effect) or the volatility of the
underlying has changed (Vega effect). Moreover because time has passed, we are
closer to maturity and this has an effect on the value (Theta effect). Furthermore
other inputs like the interest rate (Rho effect) or dividends can change and lead to
5.7. HEDGING 115
a change in value of the option price. Finally, second order effects can moreover
come into play (like for example the Gamma effect).
Consider an option with option price O. In the Black-Scholes world, O depends
on the current stock price S, the risk-free rate, r, the dividend yield q and the
stock’s volatility σ and of course of the option characteristics, like the strike K
and time to maturity τ in for example the case of an European Call or Put. We
write O = O(S, r, q, σ, τ ).
We call delta
∂O(S, r, q, σ, τ )
∆= ,
∂S
the rate of change of the option price with respect to the price of the underlying
asset. Suppose that the delta of a call option on as stock is ∆ = 0.6. This means
that when the stock price changes by a small amount h, the option price changes
by about 60 percent of that amount.
Indeed for h small, we have
∂O(S, r, q, σ, τ ) O(S + h, r, q, σ, τ ) − O(S, r, q, σ, τ )
∆= ≈
∂S h
and hence
O(S + h, r, q, σ, τ ) − O(S, r, q, σ, τ ) ≈ h∆. (5.13)
Suppose for example that the stock price is 100 Euro and the option price is
10 Euro. Imagine an investor who has sold 2000 option contracts - that is, he sold
options to buy 2000 shares. The investor’s position could be hedged by buying
0.6 × 2000 = 1200 shares. The gain (loss) on the option position would then tend
to be offset by the loss (gain) on the stock position. For example, if the stock
goes up by 1 Euro (producing a gain of 1200 Euro on the shares purchased), the
option price will tend to go up by 0.6 × 1 = 0.60 Euro (producing a loss of 2000
× 0.6 = 1200 Euro on the options written); if the stock price goes down by 1 Euro
(producing a loss of 1200 Euro on the stock position), the option price will tend
to go down by 0.60 (producing a gain of 1200 Euro on the option position).
It is important to realize that, because delta changes (with time and stock
price movements), the investor’s position remains delta-hedged (or delta neutral)
for only a relatively short period of time. In order to have a perfect hedge, the
positions have to be adjusted continuously. In practice however one can only
adjust periodically. This is known as rebalancing. For example, suppose that an
increase in the stock leads to an increase in delta, say from 0.60 to 0.65. An extra
of 0.05 × 2000 = 100 shares would then have to be purchased to maintain the
hedge.
116 CHAPTER 5. THE BLACK-SCHOLES MODEL
∂O(S, r, q, σ, τ )
v= ,
∂σ
as the rate of change of the option price with respect to the volatility of the
underlying asset; one defines theta as
∂O(S, r, q, σ, τ )
Θ= ,
∂τ
as the rate of change of the option price with respect to the time to maturity; one
defines rho as
∂O(S, r, q, σ, τ )
ρ= ,
∂r
as the rate of change of the option price with respect to the risk-free rate.
An example of a second order effect is Gamma, the rate of change of delta
with stock price:
∂ 2 O(S, r, q, σ, τ )
Γ= .
∂S 2
An extension of the Taylor Expansion (5.13) incorporating the other effects
can be easily derived:
1
O(S+δS, r+δr, q, σ+δσ, τ +δτ )−O(S, r, q, σ, τ ) ≈ ∆δS+ρδr+Θδτ +vδσ+ Γ(δS)2 .
2
(5.14)
Now, suppose we are long an option, i.e. we have bought derivative for price
O and we delta hedge the derivative, i.e. we sell ∆ stocks for price S. The money
we put on bank account is (and if ∆ is negative this amount can also be negative)
: ∆S − O. After a small time step δt, we receive (or have to pay) interest. The
interest rate received is approximately: r(∆S − O) × δt.
Assume now that the stock moves (no other changes happen). Then Equation
5.14 becomes
1
Γ(δS)2
2
Now, under the Black-Scholes setting:
√ √
(δS) ≈ µSδt + σS δt ≈ σS δt,
with epsilon a standard normal random variable. So the portfolio increase in value
by
1 √ 1
Γ(σS δt)2 = Γσ 2 S 2 δt2 .
2 2
Since E[2 ] = 1, we have that expected increase due to stock moving is
1 2 2
Γσ S δt.
2
In addition, we have also a time effect and over our step δt, the price of the option
moved by Θδt.
In total, we have due to the interest payment, the time effect (Θδt) and due
to the stock movement, an expected change in value of our portfolio of:
1
Θδt + Γσ 2 S 2 δt + r(∆S − O)δt,
2
which can be rewritten as
1 2 2
Θ + Γσ S + r∆S − rO δt.
2
Going back to the Black-Scholes PDE (which itself was derived from Itô’s lemma),
we know that
1
Θ + Γσ 2 S 2 + r∆S − rO = 0
2
In conclusion, the total change in the value of a delta-hedged portfolio is equal to
zero ... on average.
118 CHAPTER 5. THE BLACK-SCHOLES MODEL
1 2 2 2 1 2 2 1
Γσ S δt − Γσ S δt = Γσ 2 S 2 δt(2 − 1)
2 2 2
Note that the square of a standard normal random variable, 2 , follows a chi-
squared distribution. Further note that, the error is proportional to the gamma
(often resulting in the fact that the hedging error gets worse close to the money,
close to expiry). In addition, the hedging error is proportional to δt and the
shorter the periods in between rebalancing the better. Finally, remark that the
hedging errors are independent of each other from each day to the next and that
the total hedging error (up to expiration) is path dependent.
where {Wt , t ≥ 0} is standard Brownian Motion and σ > 0 is the usual volatility,
turned out to be very popular. One should bear in mind however, that this
elegant theory hinges on several crucial assumptions. One assumes that there
are no market frictions, like taxes and transaction costs or constraints on the
stock holding, etc. Moreover, empirical evidence suggests that the classical Black-
Scholes model does not describe the statistical properties of financial time series
very well.
Summarizing we could say that the Black-Scholes framework has several short-
comings which can have a serious impact on the modelling of financial assets and
the corresponding pricing and hedging of financial derivatives. The most common
critiques are:
• paths of the stock process under the Black-Scholes model are continuous
and show no jumps. However in reality asset prices do jump and the more
pronounced jumps typically have the most impact on the derivative pricing.
• the volatility parameter (the only model parameter of relevance for the
pricing of derivatives) is assumed to be constant. However, it has been
observed that the volatilities or the parameters of uncertainty estimated (or
more generally the environment) change stochastically over time.
for σ. We start with some initial value we propose for σ; we denote this starting
value with σ0 . It turns out that a σ0 around 0.20 performs very well for most
common stocks and indices. In general, if we denote by σn the value obtained
120 CHAPTER 5. THE BLACK-SCHOLES MODEL
where the function in the denominator, ECBS 0 , refers to the differential with
respect to σ of the call price function. This quantity is also referred to as the
vega. For the European call option (under Black-Scholes) we have:
σn2 !
√ √ log(S0 /K) + (r − q +
0
√ 2 )T
ECBS (σn ) = S0 T N(d1 ) = S0 T N ,
σn T
where S0 is the current stock price, d1 as in Equation (5.11) and N(x) is the cumu-
lative probability distribution of a Normal(0, 1) random variable as in Equation
(5.2).
In Figure 5.7 and Figure 5.8 one sees the so-called volatility surface of the S&P
500 on the 12th of December 2014 and the 17th of July 2015 respectively based
on closing mid-prices. Under the Black-Scholes model, all σ’s should be the same;
clearly we observe that there is a huge variation in this volatility parameter both
in strike as in time to maturity. One says often there is a volatility smile or skew
effect. Again this points to the fact that the Black-Scholes model is not appropri-
ate and the traders already count in this deficiency into their prices. Note further,
that both surfaces also differ quite significantly. Volatility or in other words the
parameters of uncertainty estimated (or more generally the environment) change
stochastically over time.
Great care has to be taken by using implied volatilities to price options. Fun-
damentally, using implied volatilities is wrong. Taking different volatilities for
different options on the same underlying asset, give rise to different stochastic
models for one asset. Moreover, the situation worsens in case of exotic options.
One can show that if one tries to find the implied volatilities coming out of exotic
options like barrier options (see Section 4.1), there are cases where there are two
or even three solutions to the implied volatility equation. Implied volatilities are
thus not unique in these situations. More extremely, if we consider an up-and-out
put barrier option, where the strike coincides with the barrier and the risk-free
rate equals the dividend yield, the Black-Scholes price (for which there is a for-
mula in closed form available) is independent of the volatility. So if the market
price happens to coincide with the computed value, you can have any implied
volatility you want. Otherwise there is no implied volatility.
5.10. EXTRA 121
From this, it should be clear that great caution has to be taken by using
European call option implied volatilities for exotic options with apparently similar
characteristics (like the same strike price for example). There is no guarantee that
the obtained prices are reflecting true prices.
5.10 Extra
5.10.1 Drawbacks of the Black-Scholes Model
Normal Returns
In Table 5.1 we summarize i.a. the empirical mean, standard deviation, skewness
and kurtosis for a set of popular indices. The first data set (SP500 (1970-2001))
contains all daily log-returns of the SP500 index over the period 1970-2001. The
second data set (*SP500 (1970-2001)) contains the same data except the excep-
tional log-return (-0.2290) of the crash on the 19th of October 1987. All other
data sets are over the period 1997-1999.
E[(X − µX )3 ]
Var[X]3/2
For a symmetric distribution (like the Normal(µ, σ 2 )), the skewness is zero. If a
distribution has a longer tail to the left than to the right, it is said to have negative
skewness. If the reverse is true, then the distribution has a positive skewness. If
we look at the daily log-returns of the different indices, we observe typically some
significant (negative) skewness.
Tail behavior and peakedness are measured by kurtosis, which is defined by
E[(X − µX )4 ]
.
Var[X]2
For the Normal distribution (mesokurtic), the kurtosis is 3. If the distribution
has a flatter top (platykurtic), the kurtosis is less than 3. If the distribution has
a high peak (leptokurtic), the kurtosis is greater than 3.
122 CHAPTER 5. THE BLACK-SCHOLES MODEL
We clearly see that our data always gives rise to a kurtosis clearly bigger than
3, indicating that the tails of the Normal distribution go much faster to zero than
the empirical data suggests and that the distribution is much more peaked. So
large asset price movements occur more frequently than in a model with Normal
distributed increments. This feature is often referred to as excess kurtosis or fat
tails; it is one of the main reasons for considering asset price processes with jumps.
Table 5.1: Mean, standard deviation, skewness and kurtosis of major indices
do not move that much; there is a considerable amount of mass around zero.
Also in Figure 5.9 we plotted the Normal density with mean µ = 0.0003112 and
σ = 0.0099 corresponding to the empirical mean and standard deviation of the
daily log-returns.
Semi-Heavy Tails
Density plots focus on the center of the distribution, however also the tail behavior
is important. Therefore, we show in Figure 5.9 the log densities, i.e. log fˆh (x)
and the corresponding log of the Normal density. The log-density of a Normal
distribution has a quadratic decay, whereas the empirical log-density seems to
have a much more linear decay. This feature is typical for financial data and is
often referred to as the semi-heaviness of the tails. We say that a distribution or
its density function f (x) has semi-heavy tails, if the tails of the density function
behave as
(x − µ)2 √ 1
log fNormal (x; µ, σ 2 ) = − − log σ 2π ∼ − 2 x2 (5.16)
2σ 2 2σ
as x → ±∞. In conclusion, we clearly see that the Normal distribution leads to
a very bad fit.
Statistical Testing
All the above is confirmed by statistical tests on the Normal hypotheses. A
standard and straightforward way of testing goodness of fit of a distribution can
124 CHAPTER 5. THE BLACK-SCHOLES MODEL
be done with the so-called χ2 -test. The χ2 -test counts the number of sample
points falling into certain intervals and compares them with the expected number
under the null hypothesis.
More precisely, suppose we have n independent observations x1 , . . . , xn from
the random variable X and we want to test whether these observations follow
a law with distribution D, depending on h parameters which we all estimate by
some method. First, make a partition P = {A1 , . . . Am } of the support (in our
case R) of D. The classes Ak can be chosen arbitrarily; we consider classes of
equal width.
Let Nk , k = 1, . . . , m be the number of observations xi falling into the set
Ak ; Nk /n is called the empirical frequency distribution. We will compare these
numbers with the theoretical frequency distribution πk , defined by
πk = P (X ∈ Ak ), k = 1, . . . , m,
If necessary we collapse outer cells, such that the expected value nπk of observa-
tions becomes always greater than five.
We say a random variable χ2j follows a χ2 -distribution with j degrees of free-
dom if it has a Gamma(j/2, 1/2) law (see Chapter 5):
E[exp(iuχ2j )] = (1 − 2iu)−j/2 .
General theory says that the Pearson statistic χ̂2 follows (asymptotically) a χ2 -
distribution with m − 1 − h degrees of freedom.
The P -value of the χ̂2 statistic is defined as
In words, P is the probability that values are even more extreme (more in the
tail) than our test-statistic. It is clear that very small P -values lead to a rejection
of the null hypotheses, because they are themselves extreme. P -values not close
to zero indicate that the test statistic is not extreme and lead not to a rejection
of the hypothesis. To be precise we reject if the P -value is less than our level of
significance, which we take equal to 0.05.
5.10. EXTRA 125
Next, we calculate the P -value for the same set of indices. Table 5.2 shows
the P -values of the test-statistics.
We see that the Normal hypothesis is always rejected. Basically we can con-
clude that the Normal distribution, is not sufficiently flexible to capture all fea-
tures of the data. We need at least four parameters: a location parameter, a
scale (volatility) parameter, an asymmetry (skewness) parameter and a (kurtosis)
parameter describing the decay of the tails.
We have just seen that the well-mannered bell curve of the Gaussian distri-
bution isn’t so normal at all. Next, we focus a bit more on the impact of this on
the extreme events and the corresponding implications of more fatter tails.
Table 5.3: Ten largest down moves of the Dow since 1954)
behaves stochastically – another point we will come back to shortly. In the figure,
volatility is typically below 25%. Let us calculate for a 25% vol the frequency of a
negative log-return of -0.0582 or even worse. Under the assumption of Normality,
it happens just once every 35 years. In reality, we have witnessed ten in the last
50 years! If the mathematician Thales (c.624–c.546 BC) – one of the ancient
derivatives traders – would have been granted eternal live, he would according
to the Normal distribution have seen only one down move of -0.0716 or worse up
to now. In the last fifty years we had five! A Homo Sapiens would likely have
witnessed only one down move of -0.0838 or worse up to now. In a particularly
bad month, October 1987, there were two! What is the probability of a down
move of -0.25 or worse: It is of the order once in the 1053 years (in US language:
100 sexdecillion years, UK language: 100000 octillion years). In contrast, the Big
Bang only happened around 15 × 109 years ago. The present generation must be
really exceptional that God allowed the Dow to crash in October 1987.
Expected Shortfall
Let us focus a bit more on the modeling of extreme values and the tale a tail
has to tell. One of the main developer of the theory was the German mathemati-
cian, pacifist, and anti-Nazist Emil Julius Gumbel who described the Gumbel
distribution in the 1950s . Extreme value theory is by now a well-developed area
of statistics and finds applications in many areas of research: besides finance,
5.10. EXTRA 127
5.10.3 No Jumps
Brownian motion has continuous sample paths, whereas in reality prices are driven
by jumps. The Brownian motion needs a substantial amount of time to reach
a low barrier, whereas in reality jumps can cause an almost immediate move
over the barrier. This has serious impact for example on the pricing of barrier
products. Because the probability that on the short-term Brownian motion will
hit a barrier far away from its current position is almost zero, prices of down-
and-in and up-and-in type of barrier options with short maturities are completely
underestimated. Indeed since under Black-Scholes there is almost no possibility
that n the short-term the Barrier is hit and thus the options becomes ”in”, the
price of the product will be extremely low. In reality however, we have seen above
that even in one day extreme movements are possible and that actually the hitting
of the barrier is much more likelier. Processes with jumps incorporate this effect
and actually make it possible that even in the next instance the Barrier is trigger.
Another important feature which the Black-Scholes model is missing is the fact
that volatility or more generally the environment is changing stochastically over
time.
128 CHAPTER 5. THE BLACK-SCHOLES MODEL
Historic Volatility
It has been observed that the volatilities estimated (or more general the parame-
ters of uncertainty) change stochastically over time. This can be seen for example
by looking at historic volatilities. Historical volatility is a retrospective measure of
volatility. It reflects how volatile the asset has been in the recent past. Historical
volatility can be calculated for any variable for which historical data is tracked.
For the SP500 index, we estimated for every day from 1971 to 2001 the stan-
dard deviation of the daily log-returns over a one year period preceding the day.
In Figure 5.12, we plot, for every day in the mentioned period, the annualized
standard deviation, i.e. we multiply the stimulated standard deviation with the
square root of the number of trading days in one calendar year. Typically, there
are around 250 trading days in one year. This annualized standard deviation
is called the historic volatility. In Figure 5.10 the historical volatility estimated
(using a three-years window) was already given for the Dow Jones Industrial Av-
erage. Clearly, we see fluctuations of this historic volatility. Moreover, we see a
kind of mean-reversion effect. The peak in the middle of the figures comes from
the stock market crash on the 19th of October 1987; windows including this day
(with an extremal down-move), give rise to very high volatilities.
Volatility Clusters
Moreover, there is evidence for volatility clusters, i.e. there seems to be a suc-
cession of periods with high return variance and with low return variance. This
can be seen for example in Figure 5.13, where the absolute log-returns of the
SP500-index over a period of more than 30 years is plotted. One clearly sees that
there are periods with high absolute log-returns and periods with lower absolute
log-returns. This is in contrast with the picture in Figure 5.14, where similarly the
absolute value of simulated normal random variables (with the empirical standard
deviation of the SP500) are graphed. Here one sees a more homogeneous picture,
often referred to as white noise. Large price variations are more likely to be fol-
lowed by large price variations. These observations motivate the introduction of
models for asset price processes where volatility is itself stochastic.
5.10. EXTRA 129
0.45
0.4
0.35
0.3
f(x)
0.25
0.2
0.15
0.1
0.05
0
−5 −4 −3 −2 −1 0 1 2 3 4 5
x
1
N(x)
0.5
0
−5 −4 −3 −2 −1 0 1 2 3 4 5
x
Brownian Motions
2.5
1.5
0.5
Wt
−0.5
−1
−1.5
−2
−2.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
t
0.8
0.6
0.4
0.2
−0.2
−0.4
−0.6
−0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
k=1 k=3
10 10
5 5
0 0
−5 −5
−10 −10
0 10 20 30 40 50 0 10 20 30 40 50
k=10 k=50
10 10
5 5
0 0
−5 −5
−10 −10
0 10 20 30 40 50 0 10 20 30 40 50
220
200
180
160
St
140
120
100
80
60
40
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
t
0.35
0.3
0.25
0.2
0.15
T=0.0959
T=0.1726
0.3 T=0.2493
T=0.4219
T=0.4986
T=0.6712
Black Scholes implied volatility
0.25 T=0.9205
0.2
0.15
0.1
1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500
Strike
40
30
20
10
0
−0.04 −0.03 −0.02 −0.01 0 0.01 0.02 0.03 0.04
−1
−2
−0.04 −0.03 −0.02 −0.01 0 0.01 0.02 0.03 0.04
Figure 5.9: Normal density and Gaussian Kernel estimator of the density of the
daily log-returns of the SP500 index
138 CHAPTER 5. THE BLACK-SCHOLES MODEL
0.24
0.22
0.2
0.18
historic vol
0.16
0.14
0.12
0.1
0.08
0.06
1954 time 2004
4.5
0.15 4
expected shortfall
expected shortfall
3.5
0.1 3
2.5
0.05 2
1.5
0 1
0 0.02 0.04 0.06 0.08 0.1 0.005 0.01 0.015 0.02 0.025 0.03 0.035
negative logreturn negative logreturn
(a) (b)
Figure 5.11: (a) Expected shortfall over 1000 largest negative daily log-return of
the Dow (1954–2004). (b) Similarly for a Normal random sample with the same
mean and variance.
5.10. EXTRA 139
0.35
0.3
0.25
volatility
0.2
0.15
0.1
0.05
1970 time 2001
0.2
0.15
absolute log return
0.1
0.05
0
1970 time 2001
White Noise
0.2
0.15
0.1
0.05
141
142 CHAPTER 6. MONTE CARLO SIMULATION
1. Sample a random path for S in a risk-neutral world under the give parameter
setting.
3. Repeat steps one and two to get many sample values of the payoff from the
derivative in a risk neutral world.
4. Calculate the mean of the sample payoff to get an estimate of the expected
payoff in a risk-neutral world.
5. Discount the estimated expected payoff at the risk-free rate to get an esti-
mate of the value of the derivative.
Up to now, we only have seen either tree models, where simulation is very
easy. For example, in the binomial tree, we have to determine in each time step,
whether we go up or down and know the (risk-neutral) probability of such possible
moves.
Next, we briefly show how, one can simulate a standard Brownian motion
W = {Wt , t ≥ 0}. Since this process is driving the Black-Scholes model, once we
have a path of such a Brownian motion, we can easily obtain a path of the stock
price process, which is essential a geometrical Brownian motion:
In Figure 6.1, we depict 10 paths of standard Brownian motions over the time
interval [0, 1]. Paths of Brownian motion are then easily transformed into paths
6.2. SIMULATION OF A BROWNIAN MOTION 143
St = S0 exp((r − q − σ 2 )t + σWt ), t ≥ 0,
for the relevant risk-free rate r, dividend yield q and volatility σ > 0.
In Figure 6.2, one sees the realization of the geometrical Brownian motion
based on the sample path of the standard Brownian motion of Figure 6.1.
144 CHAPTER 6. MONTE CARLO SIMULATION
Brownian Motions
2.5
1.5
0.5
Wt
−0.5
−1
−1.5
−2
−2.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
t
220
200
180
160
St
140
120
100
80
60
40
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
t