UNIT 4 APPLICATION TO FINANCE

Structure
4.0 Objectives
4.f lntrodiiction
4.2 Financial Market .
4.2.1 Some Important Words and Concepts

4.3 Forward Contract

4.3.1 Pricilig n For\+,ard
4.3.7 Discretc I it. '- vlodels l
4.3.3 The Multi-Period Binary Model
.1.4 Option Pricing Results
4.4.1 Propel-ties of Option Pricing
4.4.2 Put Call Parity Case of E ~ ~ r o p e aOptions
n
4.4.3 Dividends-Paying Stock
4.4.4 Dolta IHedging
an Binomial Trees
4.4.5 Stochastic P~~ocesses
4.4.6 Stochastic Differential Ecluations
4.5 Black-Scholes Model
4.6 Optil~~al
Portfolios
4.6.1 Mean-Vill+ianceApproach
4.6.2 Capital Asset Pricing Model (CAPb1)
4.7 Let Us Sum Up
4.8 Key Words
4.9 Some Usef~tlBooks
4.10 Answer or I-lints to Check Your Progress

4.0 -
OBJECTIVES
After going through this unit, you will be able to:
understand the basic concepts use'd in finance; and
* appreciate price determination through stocliastic modelling.

INTRODUCTION
Operation of financial market provides an important exa~lnplethrough which we
can understand the use of stocl~asticmodelling. If you want to participate in the
stock market, the volatile lnovelnents of share prices will compel you to associate.
their realisation with a stochastic process. Modelling ilsed in finance will Lhen be
linked to insurance as source of input for decision making. Keeping this in view
'
the following discussion introduces you to basics of fir.;nci~!l derivatives. The
Quantitative Techniques
for Risk Analysis coverage of the theme however is far from adequate to understand the ongoing
work in the area of quantitative finance. You may like to refer Hull (1992) for
some of the themes presented in the unit

4.2 FINANCIAL MARKET

The stock market is a primary capital market. Firms raise finance through stock
markets by issuing new securities. It is also a secondary capital market where
securities are traded, and signals the return shareholders receive which they
provide to firms. Financial instruments can be divided i n two basic classes:
underlying and derivative. The former can be stocks, bonds or various trade
goods, while the latter are financial contracts that promise some payment to the
owner, depending on the behaviour of the " underlying" (e.g., a price at a givkn
time Tor an average price over certain time period). Derivatives are extremely
useful for risk management (apart from obvious investment properties) - we can
reduce our financial vulnerability by fixing price for a future transaction now.
A share is a type of financial underlying assets (or underlyings) dealt in financial
market. Other underlyings include bonds, commoditiks and foreign currencies. In
financial market not only underlying but also their derivatives are traded. A .
derivative is a financial instrument, which promises some payment or delivery in
the future contingent on an underlying.
4.2.1 Some Important Words and Concepts
* Stock markets. Stock is a printed-paper issued by a company to attract
investments. Usually dividends are paid to participants.
0 Bond markets. "Bond: a printed paper given by a government, city or
business company saying that money has been received and will be paid back,
usually with interest," which is fixed.
Currency markets or foreign exchange markets.
1

Commodity markets such as oil, gold, copper, wheat and electricity.

Futures and options markets. "Futures: goods and stocks bought at prices '
agreed upon at the time of the purchase but paid for and delivered afteruvards." ,
* buyer = long position; buying = going long
m
e seller = short position; selling = going short I

Short selling involves selling an asset that: is not owned with the intention of
buying it later.
I
Arbitrage = sure profit without investment
Speculator: A trader who enters the futures market in pursuit of profit, accepting
the risk.
Hedger: A trader who trades futures to reduce some preexisting risk exposure.
They are often producers or major users of a given commodity (e.g,, a farmer may
hedge by selling his anticipated harvest even before the famer plants). Thcy often
trade through a brokerage firm.
Arbitrageurs: Persons entering into several contracts in different markets to
exploit price fluctuations. If a good had two prices, a trader can get an arbitrage
 profit: a sure profit with no investment. But prices may differ because of Application toFinance
transportation costs, etc.
Spreads: A trading strategy involves taking a position in two or more options of
the same type,
portfolio = combination of several assets/securities, etc.
Types of derivatives
e Options: An option gives the holder the right (but not the obligation) to buy
or sell an asset in the future at a price that is agreed upon today.
c Futures: A futures contract is an order that one places in advance to buy or
sell an asset or commodity. The price is fixed when the order is placed but the
payment is not made until the delivery date.
c Forwards: Futures contracts are standardized products bought and sold on
organised exchanges. A forwards contract is a modified futures contract that is
not traded on organized institution. For example, a ~nultinational company
which needs to protect itself against a change in the exchange rate usually
buys or sells cunency forwards through a bank.
Swaps: Swaps are private agreements between two exchange cash flows in
the future according to a prearranged formula. Suppose that you want to swap
your sterling debt for US$ debt. In this case you can arrange for a bank to pay
you 'each year the sterling's that are needed to service your sterling debt, and
in exchange you agree to pay the bank the cost of servicing a US$ loan. Such
an arrangements is known as a currency swap.
We continue with the derivative securities introduced above and find some formal
definitions to help derive their prices.
I

4.3 FORWARD CONTRACT

This section is adopted from a discussion found in Internet. We are unabte to
determine the writer and acknowledge the contribution.
As already seen above, a forward contract is an exchange agreement to buy (or
sell) an asset on a specified future date for specified price.
Forwards are not generally traded on exchanges and it costs nothing to enter into a
forward contract. A future contract is the same as forward except that futures are
normally traded on exchanges and the exchange specifies certain standard features
of the contract.
Options are of different types. The most important of these being the American
and European options. An option is so-called because it gives the holder the right,
but not the obligation, to under take a transaction.
Definition. A call option gives the holder the right to buy whereas a put option
gives the holder the right to sell.
A European call option gives the holder the right, but not the obligation, to buy an
asset on a certain date for a certain price (the strike price).
An American option is similar except that the holder of an American call has the
right to buy the asset for the specified price at any time up until the expiration
date. It can be optional to exercise such an option prior to expiration.
Quantitative Techniques
~nal~sis
for,~isk Two main uses of options are speculation and hedging. In case of speculation,
available funds are invested opp~jrf~nisticallywith the aim of making a
Hedging, on the other hand, is usually engaged in by companies who have to deal
with risky assets such as foreign exchange, gold, oil and so on. For hedgers, the
basic purpose of an option is to spread risk.
Suppose that you know that in three months time you need a million gallons of
crude oil. However you are worried by the fluctuations in the price of oil. So you
decide to buy European call options. See that you know the maximum amount of
money needed to buy oil, Denote the price of the underlying asset (oil) at time
Twhen thd'option expires in three months time by ST.Let us assume that the
strike price is K . If, at time T , S , > K then you will exercise the option. Such
situation is called as the option is in the money. You are buying an asset worth
S, > K dollars for just Kdollars. If on the other hand, if ST < K , then you will
buy oil on the open market ignoring the option, which is worthless. Such situation
.*is 'called as the option is out of the money. The payoff from the option is, thus,

From the above example, it is clear that you have to decide on how much you
should be willing to pay for such an option. You can think of it as insurance
, . premium, where you are insuring against the oil price going up.
i
Let us see how option market operates.
4.3.1 Pricing a Forward
We' have seen above that a forward contract is an agreement to buy (or sell) an
asset on a specified future date for a specified price. The pricing problem here is,
'what price for the asset should be specified in the contract?'
Example, Suppose that you enter into a long position buying on a foryard
contract and agree to buy an asset for price K at time T . Then the payoff at time
T is (S,. - K ) , where S, is the asset price at timeT, since you (must) buy the
asset S1.for price K . This pay off could be positive or negative. Since the cost of.
entering into a forward contract is zero this is also your total gain (or loss) from
the contract. What is a fair price of K ?
At the time when the contract is written, we don't know S,. . herefo fore, we can
only guess at it. Typically, we assign a probability distribution to it,
Note that stock prices should stay positive, so a,bare positive. We may think that
E [ST] would represent a fair price to write into our contract. However, the market
price will match the expected price only rarely.
We may assume that market participants can borrow money for the same risk-free
rate of interest as they do to lend money. Thus, if you borrow $1 now, your debt at
time Tit will be worth be$e r(say) and similarly if you keep$l in the bank then at
time Tit will be $er .
I) Let K > S,,er. Then the other participant in the deal can ,adopt the following
ctrategy: she borrows $Soat time zero and buys with it one unit of stock. At
time T , she must repay $So@'to the bank, but she has the stock to sell to you
for SK , leaving her a certain profit of $(seer - K ).
 2) If K < S,,e', then you can reverse this strategy. .lust sell a unit of stoclc at Applic.ation to Finance

time zero for $S,, and put the mone) in the bank. A t time T , your
accumulation Lf;S,,e1and use $ K to buy a unit of stock. which leaves you
with a certain profit of $(S,,L" - K ) .

Unless K = S,,er, dne of you will be guaranteed to make a p i f i t .

The above argument, selling of stock by you, which may not be yours, is lcnown
as short selling. An opportunity to lock into a risk free profit is called an arbitrage.
The starting point in establishing a tnodel in modern finaricc theory is to specify
that there are no arbitrage opportunities. We titid that there are people who ~nake
their living entirely from exploiting arbitrage oppol-tunities, but s ~ ~ opportunities
ch
do not exist tbr a significant length of tirne before market prices move to
eliminate these.
Remember that forwards are a very special sort of derivative. It will no be easy
for LIS to value an option. However. we must look for conditions that don't
provide eitlier party will1 an arbitrage opportunily. Such conditions will be
identified i l l the following.

I
4.3.2 Discrete Time Models I
The Single Period Binary Model
In the discussion below we try to find the 'fair' price for a European call option.
Let US take an example.
Suppose that the current price in Swiss Francs (SFR) of $100 is S,, = 100.
Consider a European call option with strike price K = 150at time T . The fair
price to be paid for this option may be obtained as follows:
We know that an option gives the buyer the right to buy $1 00 fbr ISOSFR at time
T . Hence suppose that the tr~teexchange rate at time T is not known. However it
can be modelled by a random variable. The sitnplest model is the single period
binary model where S,, takes one of two values with specified probabilities. Let's
if
I

(90 with probabilip 1 - p

Then the payoff of the option will be (1 80-1 50) SFR=30/SFR. With probability p
and 0 with probability 1 -p. Therefore has expectation 30p SFR.
To find the fair price to be paid for the option, let us assunic that interest rates are
zero and that currency is bought and sold at the same exchange rate. Remember
that in the presence of a fair price, there will be no scope of a rislc free profit for
either tlie buyer or tlie seller of the option. In tlic exaii~pleabove, wc try to arrive
at a sPecikc price by s~~pposing that p = 0.5. That is .5', talking the value 90SFR.
the fair price to pay for the option.
You claiin that if the price is I SSFR, then you can ~nstltea risk-free profit by the
following strategy: you buy the option and you borrow $33.33 and convert it
straight into 50 SFR. (You will have to pay off' your loall in dollars at Lime 'I .)
Quantitative Techniques
for Risk Analysis Your position at time zero is as follows: you have one option (to buy $100 for
15OSFR at time T), you have 35SFR (50 from the conversion of your dollar loan,.
less 15 paid for the option) and you have a debt of $33 -33.
At time T , one of two things has happened:
1) If S, = 180, then exercise the option and buy $100 for 15OSFR. Make use
$33.33 to pay off your dollar debt and that leaves a balance of $66.67.
Convert back into SFR' at the current exchange rate. This nets
2/ 3 x1800 = 120SFR. In total then, if ST = 1 80, at time T you have
35SFR-15OSFR (used to exercise the optibn)+l20SFR, which totals 5SFR
clear profit.
2) If ST = 9 0 , then you have to throw away the option (which is worthless) and
conve_rt your 35SFR into dollars, netting $ 0 . 9 35438.89.
~ You pay off your
, debt leaving a profit of $5.56.
Whatever the true exchange rate at time T , you make a profit.
, Determination of Right Price
Let's think of things from the point of view of the seller. If you are the seller of
the o,ption, then you know that at time T , you will need $(s,.- 1 5 0 ) in
~ order to
meet the claim against me. The idea is to calculate how much money you need at
time zero (to be held in a combination of dollars and Swiss Francs) to guarantee
this.
Suppose you hold x, and x,at time zero. You need this holding to be worth st
least (S,. - 150)+SFR at time T .

1) If ST = 1.80, then you will need at least 30SFR. That is, we must have *

2) On the other hand, if ST,= 90 then the payoff of the option is zero and you
just need not to be out of pocket. That is, you want

.. .
Solve the simultaneous equation to get the right price. On such a, the seller can
. exactly meet the claim if x, = -30 and x, = 30019. To purchase $300/9 at time
~ . the seller requires exactly ZOSFR at time
zero iS equal to 5 0 : 3 0 = 2 0 ~ ~Thus,
zero to construct a portfolio that will be worth exactly the payoff of the option at
time T .
The argument above can be reversed to show that for any lower price, there is a
strategy for which the buyer makes a risk-free profit.
The fair price is 20SFR.
Notice that we did not use the probability, p, of the price S, going up to 180SFR
at any point in the calculation. We just needed ihe fact that we could replicate at
. . any pbint in the calculation. We just needed the fact that we could replicate the
clai~n by this simple portfolio. The seller can hedge the contingent claim Application to ~ i n a n c e
t <

( ( S , - I SO)+-usingthe portfolio consisting of xlSFR and $x,

of 'no arbitrage'
A Cllaracterisatio~~
Tlie option pricing problem solved above through binary framework in a single
period may not help in more complex settings like multi-period situhtions. Hence
we need to take the Iielp of probability theory. We will do that after providing a
concise mathematical condition to characterize markets that have no arbitrage
opportunities. Let there be a market, which coilsist of N tradable assets. Their
prices at time zero are given by the column vector

Let the ~narltetbe represented by a linite number of possible states (t,2,..n) in

which it might be at time one. The security values at time one are given by an
N x n matrix b = (D,, ) , where the coefficient D,is the value of the :i security at
time one if the market is in state j .
I n this formulation, you can thinlc of a portfolio as a vector

and its market value at titne zero is the scalar product

'The payoff of the portfolio at titlie one is a vector in Rtfwhose it" entry is the '

in
value of the portfolio if the market is state i . Thus, the payoff is , I

Notation
Foravector x~liB"wewrite x2Otomean x~IW:and x>Oto mean x r O , ~ + 0 .
Notice that x > Odoes not require ' x to be strictly positive in all its coordinates. -
We write x >> 0 for vectors, which are strictly positive in all coordinates, i.e., for
vectors x E R):, . An arbitrage is then a portfolio B E RN with

'

Definition. A state price vector is a vector Ie iW:* such that so= DY . That is,
Quantitative Techniques
for Risk Analysis The vector multiplied by Y, is the security price vector if the market is in state i .
To interpret Y , we can think it to be the marginal cost of obtaining an additional
unit of wealth at the end of the time period if the system is in state i . There is no
arbitrage if and only if there is a state price vector.
The Risk Neutral Probability Measure
Recall that all the entries of Yare strictly positive. Therefore, we write
\Yo = C:, Y, such that

is taken as a vector of probabilities for being in different states. First we will see
that \Yo is the discount on riskless borrowing. Suppose that the market allows
I '

positive riskless borrowing, and for some portfolio, 8,

i.e., the value of the portfolio at time T is one, irrespective of the state of the
market. Since Y is a state price vector, the cost of such a portfolidat time zero is
S , , . G = ( D V ) . ~ = Y . ( D ~ Z,=I) = ~=I y, o
where Y,=discount on riskless borrowing, From the vector (I), the expected
value of the fhsecurity at time T is
Y'. 1 " 1
E[s;] = z D,,L=--
n
ED,,^, =-Si
\Yo \Yo , = I
151 Yo
I
where the last equality is obtained by takings, = DY That is, I

. So, under the probabilities distribution (I), the price of a security is its discounted
expected payoff. In the context of contingent claim pricing, we can say that a
claim is attainable if it can be hedged.
If there were no arbitrage, the unique time zero price of an attainable claim Cat
time T is Y,,IE[c] where the expectation is with respect to any probability
measure for which Q = \Y,,E[s;]for all iand Y ois the discount on riskless
borrowing rate.
Note that the same value is obtained if the expectation is calculated ;or any vector
of probability such that S,l = Y,B[S;]. In the absence of arbitrage, there is only
one riskless borrowing rate.
Moreover, if you find a probability vector for which the present value of each
security now is its discounted expected value at time T , then it is possible to find
 ~
I

the time zero value of any attainable contingent claim by calculating the to
expectation. Thus, if there is a por-tfolioQ, for which we get Q.S, = C , than

Example. Let us return to the problem given in the previous example.

Recall the conditiolis in that example: the current price in Swiss Francs (SFR) of
$100 is S,, = 150. We suppose that at time T , the price will have moved to either
OOSFR or 180 SFR. Therefore, the problem was to price a European call option
with strike price K = 150 at timer i.e., holder of such an option has the right, but
.
riot the obligation, to buy $100 for 1 SOSFR at time T For simplicity, interest rates
were obligation to be zero.
'l'lie fair price for such an option was worlted out to be is 20SFR, with the new
approach.
Under our assumption of zero interest rates, we have 'Y,, = I . That means, we are
seelting a probability vector such that
IE [s,] = S,,.
Let p be tlie probability that

l'hen since S , can only take values 180 and 90, it l~iustbe .that

So, for the price to behave lilte a fair game,

180p + 90(1- p) = 150, or, p=2/3.
As the claim at tin~eT is C = (S, - 150),, the fair price (in SFR) is

as hel'ore sucli that given the probability p , it is easy to value other options. For
exaniple, a European call with strike price I20SFR instead of 15OSFR woi~ldbe
valued at

For your information, it may be useful to note that the probability measure that
assigns probability 213 to S, = l8OSFRand 1/3 ,to ST = 180SFR is called
equivalent martingale probabilities for the (discounted) price process
{S,,, .
'Y,s, } The probabilities are also called the risk' neutral probabilities.

4.3.3 The Multi-Period Binary Model

Let the niarltet consists of just two securities: a bond (representing riskless
borrowing) and a stock S . Unlimited amounts of either can be bought and sold
without transaction costs. Voreover we are allowed to readjust our portfolio
instantaneously.
As in tlie single period binary model, we shall suppose that at each tick of.the
clock, the stock Inoves form its current value to one of two possible values
Quantitative Techniques
for Risk Analysis (depending on its current value). There are 2' possible states of the stock price
. afteri ticks of the 'clock, and we think of them as being arranged in tree as in
Figure 1.
I

The bond moves as bef0r.e; so over the iIh tick its value is scaled up by exp (7).
We assume that all 7 are known.
. .
Suppose again that you rlr$ pricing a'European option.with strike price Kat the
maturity t i m ? , ~ ' .We sisume now that T = n . The payoff of the option is then
. ..
, . ' (S, - K)+ at time n . ' ,. ..
i. . ... .
. ,
.
, ....
,

,
:. . '
I . . .
,
.. I. *' Let us,kse
. . backward induction onlth,etie&.:Ifwe knew the price, S,,-, of thestock
. *

. . .- .- after (n -1) ticks, then our previous analysis would tell us the valu,e, C,-, , of the
L 9

'. claim at time (n - 1). Namely,

I I6
S,,-,= Y$")En,[s;,]
and Y!) -
where the expectation is with respect ro a probability measure for'which
eYL. , .
 I

So for each state of the market at time (n - I ) , you know that you need a portfolio Application to Finance
worth Cn-,if you are to meet the claim against you at time n . You can now think
ofC',,-,as a claim at time (n - 1). In the same way then, if you know Sn-,, in order
to meet the claim against me at time(n - I ) , YOLI need to hold,a portfolio worth

where the expectation is with respect to a measure such that

S, =~ 1,
[st,-,
~ - 1 ) ~ , 1 - 2

and this in t~lrnguarantees that you can exactly meet the claim against you at time
n . Proceed following this method. You will able to calculate the cost of a'
portfolio. That is, after appropriate readj,ustmenl at each tick of the clock, but
neither with extra input of wealth not with paying dividends, you will meet the
claim against you at timen.
The strategy that readjusts the portfolio icthis way is said to be a self-financing
strategy. The probability nleasures used at each stage in the above prescribe
exactly one probdbility for each branch,in our tree of stock prices. For each vertex.
of the tree there is a unique path from the vertex, and we speci9 a probability
lneslsure on paths by declaring that the piobability of such a path is the product of
the probabilities on the branches that ~onipriseit.
Let us assume, for simplicity, that the rate of interest is everywhere zero. It is
equivalent to replacing S,by the discotinted security price

/=I

Our risk neutral probabilities then have the property the

I E[S',IS,-,]=S,-, for each k='1,2,...n . '

consider the two-step model in Figure 2

In fact much more is true. To illust~~ate,

Figure 2
Quantitative Techniques
for Risk Analysis

E[S2 I

. More generally, if j > i,

Similarly,

where the expectation is with respect to the probabilities on paths defined above.
The sequence of (discounted) prices, S o , is a stochastic process and the
probability measure that we have defined has the property that E[s,IS,] = S,.
Moreover, in this model, the stqck price 'has no memory,' so that the movement of
the stock over the next tick of the clock is not influenced by the way in which the
stock reached its current value and so for j > i ,
s[silso,s,,....s,]= s,.
A sequence of random variables (or stochastic process)
X,, XI,....,X,with E [ / X , ~<]m for each r is martingale if

E[X,IX,,X~ ,.a* X,-,] = X,-,

(3)
r =l,2,...,n. 0

We have seen in Units 2-3 that the idea comes from gambling where. if
X,denotes the capital of the gambler at time r then game is 'fair' only if (3)
holds. f

The 'information' X,,XI,X,-, is often written , In a colitinuous setting we

will be a little more careful about the definition, but the idea is the same, is
the set of events that are 'decidable' by observing the process X, up to time r - 1.
We now recast the results of the above model in this language.
Suppose that the possible values that the stocks S,,,.,..,S,can take on at times
1,2,3.., are known. We denote by i2 the set of all possible 'paths' that the stock
price vector can follow in Rf . Then the absence of arbitrage is equivalent to the
existence of a probability measure, Q and C2 that assigns strictly positive mass to
every w E SZ and
 Application to Finance

where S, is the vector of stock prices at time r .

If, as above, we consider the discounted stock prices, then
-- - -
s; [srls,,.......,s,-,] = s,:,.
In other words, the discounted stock price vector is a Q -n~artingale.
Recalling that two probabilities measure P a n d Q on a space C2 are said to be
equivalent i T for events A c R
( $ ( A ) = 0 if and only if P(A) = 0 we ,nay observe the following:

Suppose$hen that we have a market model in which the stock price vector can
follow on of a finite number of paths ~2through*R:. We may even have our own
.belief as to how the will evolve, encoded in a probability measure,P, on
iZ .We can say, there is no arbitrage if and only i7 there is an equivalent
martingale lneasilre Q . That is, there is a measure, Q , equivalent to P,such that
the discounted price process is a Q -martingale.
In that case, the market price of all attainable claim C to be delivered at time n at
time zero is i~niquea~idis given by
E .. [ y , C ] ,
where

is the discount factor over n periods.

'I'he same statement applies in the continuoi~ssetting.
The Cox Ross Rubinstein Model
'['he Cox Ross Rubinstein (CRR) model is a special case of the multi-period
binary model in which in each time interval the stock price moves from its current
value. S . to one of S,,, St,,where u apd d are fixed constants with d <rrA' < u.
It is referred to as the binomial model. Thus, at time k , there are kpossible
values that the price can take and

TIie recombining of the tree makes this highly numerically efficient.

'T'lic values of zr and d lnust be calibrated with the market. The usual assumption
is that zr = I l d and p can then be determined by rislc neutrality as

Finally, u is fitted using the variance of the stock price.

Quantitative Techniques
for Risk Analysis
Check Your Progress 1
I ) What do you mean by financial derivatives?

..............................................................................................
2) Write the meaning of the terms: futures and forwards.

.............................................................................................
3) List the models used for determining option price.

........................................................................... ....
..........,,,,#
..............................,..,,,........................................................,
4) How do'you formulate a binary model?

.................,.,.....a,..(...,..,, .II ....

................................I.............

5) What is the meaning of no-arbitrage?'

6) Remember the result of multi-peiiod binary model and translate these in the
language of a martingale,
................".,.'..,..'.'....(...
..........................................................
 Application to Finance
4.4 OPTION PRICING RESULTS
4.4.1 Properties of Option Pricing
~aciors-affecting option prices
e the current stock price and the strike price (call options are more valuable if
.the stock price increases and less valuable if the strike price increases)
0 the time to expiration (put and call American options are more valuable as
time to expiration increases; European options are not necessarijy more
valuable)
i the volatility of the stock price, 0,so that a& is the standard deviation of
the stock price in a short length of time At (the owner of a call benefits from
price increases but has limited downside risk; the owner of a put benefits
from price decreases but has limited risk if price increases; therefore, the
value of both calls and puts increases as volatility increases)
the 'risk-free interest rate (if the rate increases, the expected growth rate of
the stock price tends to increase, however, the present value of any future
cash flows decreases, therefore the value of a put option decreases; for call
options the first effect tends to increase the price, while the second tends to
decrease it, but the first always dominates, so the value increases)
the dividends expected during the life of an option (reduce the stock price on
the ex-dividerid date, decrease the value of a call option, but increase the
value of a put option)
Assume that there are no transaction costs; all trading profits (losses) are subject
to the same tax rate; borrowing and lending at the risk-free interest rate is
possible,
Socurrent stock price
S, stock price at time t
Xstrike price of option
%

T time of expiration of option

I
r risk-free rate of interest at time T
C value of American call option.to buy one share
I' value of.American put option to sell one share
c value of European call option to buy one share
p value of European put option to sell one share
, Upper bounds .
c 5 so, C r; So,

Lower bouuds
. . . Lower bound for Eurapean calls on non-dividend paying stock
Quantitative Techniques
for Risk Analysis Example. So = 20, X = 18, r = 10% p.a., T = 1 year. Then So- ~ e -" 3-71. ~
Assume that the option costs $3. Then an arbitrageur buys the call and shorts the
stock. This provides cash 20 - 3 = 17, invested for one year risk-free, it gives
1 7e0.'= 18.79. Then the option expires. If the stock price is greater than $1 8, the
arbitrageur exercises the option ,and makes a profit of 18.79 - 18 = 0.79, If the
stock price is less than $18, say $17, then the stock is bought in the market and
the profit is 18.79 T 17 = 1.79.
Proof of the lower bound.
Consider two portfolios. A (one European call option and cash ~ e " and ~ ) B (one
share). At time T, portfolio A is worth max (SZ X), while portfolio B is worth ST.
Thus A must be worth more than B today, meaning that c + 2 So.
,
Lower bound for European puts on non-diyidend paying stock

For the proof consider two portfolios. A (one put option and one share) and B
(cash
4.4.2 Put Call Parity Case of European Options
Consider two portfolios.
A -one call option plus an amount of cash equal to xemrT;
B -one put option plus one share.
At time T both portfolios are worth max(Sr, 4, so (because they are European)
they must have identical values today.

This is known as put-call parity.

Case of American options
Example. A trader owns a call option with strike price $40 and the current stock.
. price $50 with one month to expiry. It is better not to exercise early, because $40
call earn interest for a month, and also the option is a protection against falling
price of a stock. It is better to sell the option (or keep the option and short the
stock).
1
!

Thus
C r SO - xemrT.
In contrast to call options, it can be optimal to exeroise a put option early. For
example, let X = 10 and let the stack. price be zero. Then it is better to exercise
immediately to get the maximal profit and invest it. T ~ U S ,
P r x --so. -.
There is no put-call parity for American options.
4.4.3 Dividends-paying Stock
Consider two portfolios.
A - one European call option,and cash D xe?;
-
B one, share.
..
Then Application to Finance

Similarly,

13ut-callparity for European options

For dividends. paying stock, it may be optimal to exercise a call option early.

Let r be the risk-free interest rate. Then the present value of the portfolio should
be equal to the cost setting up the portfolio

After substituting the value of A

where
er' -d
q = .
ZI - d..
'I'his option price J'does not refer to the probabilities of the stock moving up or
down. The probabilitiesdareincorporated into the price of the stock. Then q can be
interpreted as the probability of up niovement, ( I - q) is the probability of a down
movenient, and qj;,+ (1 -q)Aj is the expected payoff from the option.
.
,
I lien the expected stock price
E ( s ~ ) qsou+ ( I - y) ,yod = soerT,
meaning that the price grows, on average, risk-free*
The expectation with respect to the risk-free probabilities will be denoted by EQ,
. I
while the expectation in the real world i s Ep.
Matching volatility &h u and L
Let p be the expected return of the stock and abe its volatility. Assume So = 1.
Then the probability p of up-movement can be found from

so that

The variance of the stock price aAer time A1 is c? At. On the other hand, this
variance is I

prZ + ( 1 -p)$ - [pu + (1 -p)dJ2

,

From the equality (if higher powers of At are ignored),

123
- Quantitative Techniques ~ f id = , - o f i ..
for Risk Analysis u=e , -
In a risk-neutral world, the probability of upward movement will be
e 'A' - d
p " .
u- d - -. -. .
--.

If q is used instead o f p in the formula for the variance, then the variance remains
the same. The change of risk refemces(ca@also the change of measure) may
lead to change of the expected return, but does not change the variance (this is
Girsanov's theorem of stochastic calculus).
4.4.4 Delta Hedging
Delta is the ratio of the change in the price of the stock option to the change in the
price for the underlying stock, or, b

Delta hedging requires buying A x (the value) shares of the stock.

4.4.5 Stochastic Processes on Binomial Trees
Main concepts
r The set of possible stock values is a stochastic process S, it depends on time
as Si(discrete time) or St (continuous time).
e The set of probabilities pi or qi associated with nodes of the tree is called a
measure P or Q . They describe how likely is to jump up or down.
A filtration is the history of the stock up until tick-time t = i on the tree.
A claim X is a function of the nodes at a claim time-horizon equivalently,
it is an FT - measurable function. The claim is only defined on the nodes at
time T.
The conditional expectation operator Ep(. 6). I
For example, E ~ C 6) Y ~ is
the expectation of X along the latter position of .paths which have initial
segment 3;. This expectation depgnds on 3; and so is a random variable
itself. The claim X can be converted into a process EP(X1 3; )or E ~ ( x (q )if
the measure P or Q is given.
A previsible process y, is a process on the same tree whose value at any
given node at time tick i depends only on the history up to one time-tick
earlier q-, . Previsible processes play the part of trading strategies where we
cannot tell in advance where prices are going to go.
A process S is a martingale with respect to a measure Q and a filtration Ft if
~~($1.6)
=&for all i s j , -
This means that S has no drift under Q. Then Q is called a martingale measure
forB and S is called a Q-martingale.
For any claim X, the process EQ(xI3;) is a martingale.
Binomial Representation Theorem Application to Finance
Theorem. If S is Q-martingale and E is any other Q-qrtingale, then there exists
a previsible process p such that

where ASk = Sk - Sk-I.

Proof. Consider a typical node. As there are two values only, the random
variables can be transformed to each other by scaling and offset

for p, and k known by Pi- 1 . The conditional expectation given Fiil of ASi and AEi
should be zero, as they are martingales. Thus, k = 0.
The bond process
'The bond process B; represents the value of $1 at 'time i. It is a previsible
and positive process, BO= 1.
The process B,-' is another previsible process called the discount process.
e Zi = BT'S; is the discounted stock process.
BF'Xis the discounted claim.
Self-financing strategies
Let the discounted stock process Zi= B;'s~ be a Q-maahgale. Another Q-
martingale is El = Eo (B;' XI
F;). Then there exists a previsible process p such
that

At time i buy .the portfolio II,with

* $,,, units of stock;
-
v / ; +=~(Ei p i + ~B~-'s/) units of the cash bond.

At time i, the portfolio is worth q= p i. p i+lBi.At time zero, portfolio no is

worth
~7ISO+ p $0 r ~ ~ ( 1 3 f - I 4.
After one time tick nois worth

by the binomial representation theorem. On the other hand, the portfolio

scheduled to be acquired at time 1 costs also BI El, so we can cash no to buy n ~ .
This strategy is called self-financing, At the end the portfolio will be worth BI BI-
X = X, as required. Thus, the claim price is E~(BT'X).
~ n t tn V,= p i Sif tyj Bi. Then
For FI u e n ~ r a lc t ~ ni t h i i a r v ~ ~ r n lead
I
Quantitative Techniques
for Risk Analysis
Self-financing.Hedging Strategy (g, i , P J
o both g,and vare previsible
e the change in value V of the portfolio defined by the strategy obeys the
- Vi = A 6 " P) PI Mi + Yi+l Mi -
difference equation 6.1.~
0 9, ST+ yr BT is identically equal to the claim X
Option Price Formula (discrete case)
The value at time i of a claim X maturing at date T is
Bi Ea (B;' x I F;).
A martingale measure Q always exists in the binomial model.
The real measure P (which S follows) is irrelevant.
Continuous Processes
Heuristic Arguments
Let BT= er'. Assume that over a small time interval At the stock moves to either
value ~ & " " + ~(iff iup) or to ~ e p (if ~down). ~ ~if n tIAr,
- Then,

where X,, is the total number of up-jumps.

In the risk-neutral'world, the probability of up-movement is

- - n)/&
Then Xn Bi(n, q), and (Xn has mean
1.
-&@+ -a 2
- r ) / ~ r a n dthe variance 1,
2
1 o2- r ) 1b ,I).
and SO converges to N(-&(,LL+ -
2
- + (r - 1 - - 01 2 ) t , - 01' 1 )
Finally,, log St N (log'~o
2 2
or

-
where z N(0,l) under Q.
Thus log($!) has the normal distribution and so S, has the log-normal distribution.

where ,u is called the expected return pn the stock,

Let us discuss the stock price process. It can be assumed that if follows a
generalised Weiner process. That means it has constant expec~eddriA rate and
constant variance rate. Such an assumption however, fails to satisfy a key feature
observed incase of stock prices. That is, the expected percentage return required
 by investors from a stock is independent of stock price. So we introduce a to Fwnce
~pplicitloa
modification and assume that the expected drift be expressed as a proportion .
of
the stock price is constant. In the real world the stock price is distributed as
. .
Stock Price Process
Assume that the volatility vanishes, i.e., cr= 0. In this case .,

where p is the expected return on the stock. Then

and

Then the volatility of stock prices is wadelled using the Brownian Motion (also
called Wiener process). The standard Brownian~motionW, is a stochastic process,
such that
w,=o;
Wt- W,is normally distributed with mean zero and the variance t - s for t r
'

8;
0 W, has independent increments, i.e. Wtn-V,;_',......,, q j , y , a r e jointly .

independent for'any time moments t, s; tz 6 .,.. 1,.

The Brownian motion has continuous but nowhere differentiable paths,
Let W,be the standard Brownian motion. Then \

, or with iime as subscript

dSt = N,dt 4- +s$lw,.
4.4.6 Stochastic Differential Equations
'
In a more general case X, is a general stochastic process (not necessrily stock
price process) such that

The drift p, and the volatility cr, may be random, but they must be adapted io the .
same filtration, so depend on the events up to the current time (i.e. Ft), but not on
the future. The drift and volatility determine uniquely the underlying stock
process and can be determined uniquely from Xi.
Stochastic differential equation for X,
a , , [ = p ( X , , r ) d t + o ,( X , , t ) d T
Example (Due ta Hull 1992) A stock pays dividends at 15%p.a, with continuous
compounding and has a volatility of 30%p.a. Then
Quantitative Techniques
for Risk Analysis Then

-
where Z N(0, 1).
It6 Formula
If = fidt -I-D ~ and
K f is a deterministic twice continuously differentiable
function, then Y, =AX)
satisfies

More generally,'let Y , = F(&, t) for a twice differentiable function F. Then

This formula transforms a stochastic differential equation for .& into a stochastic
differential equation for Y,.
Example. X = W, Y = x2,
Then
dY, = d(w2*)= dt + 2WdWl.
This can be used to [wIdW,.,since

whence i(

Example. Let dS, = @dl + cr SdW, for the (non-dividend paying) stock price.
Find the differential for the forward price F,.= ~ , e ' ( ~ - ' ) .
Geometric Brownian motion
<
Consider , t

dSt = j&dt + +StdWl . ;

Its solution using Ito's formula is
S I = S o e x p { a W r + ( p - ! h cr2)t) and
called the geometric (or exponential) Brownian motion.
This formula gives the stock price in the real world if the stock follows the
Geometric Brownian motion with given ,u and a
Note: W,has the normal distribution with mean zero and the variance t, so that
where 7 has the normal distribution with mean p - !A c? and the variance 2 It. Application to Finance
Estimating Volatility
Take a sample of daily returns

for i = 1, . . . , n and then the standard deviation of this sample will estimate
o f i , where T is the length of time interval in years. The choice of n is crucial,
more observations do not necessarily mean a better estimate as the volatility may
change over time.
When dealing with u, it is appropriate to give latter observations more weights,
for example, use exponentially decreasing weights. It is possible to use various
models from Time Series.
Martingales
A process M, is called a Q-martingale if

Examples:
I) A constant process is a martingale.
2) Q-Brownian motion insa Q-martingale.
3) Mt = exp{ o W,- % G 2 t } is a martingale.
4) For any claim X (with Ep < m), the process Nt = Ep 1
1x1 XI F,) is a P-
martingale.
For pricing options, the central point was to ensure that the process is a martingale
with respect to some measure. The price of derivatives then becomes the
expectation with respect to this martingale measure.
Self-financing Portfolios
Portfolio is a pair of processes y3, and ~ which describe the number of units of
security and of the bond which we hold at time t , The processes can take positive
or'negative values (short-selling is alloived). The security component p should be
F-previsible. The value of the portfolio is K = pat+ wBt.
A portfolio is self-financing if and only if the change in its value,only depends on
the change of the asset prices, i.e.
dv,= #,dS, +ly,dB,
Examples: If we assume that St 7 Wt and B, = 1 then
1) = ly, = 1 is self-financing

2) ,I~ = - t WY2is self-financing

p, = 2 W,,U
If X is a claim that depends on events up to time T, then a replicating strategy for
X is a self-financing portfolio (p;t,i) such that [ ~ , ? ~ jcdwand
t X = VT= @T +
yrBr. 'Then the price of X at time t must be
Quantitative Techniques
fm Risk Analysis Change of Measure
If y is a constant, and Wl is a P-Brownian motion, then there exists a measure Q
such that

is a Q-Brownian motion. This change of measure theorem is a particular case of

the Cameron-Martin-Girsanov theorem from stochastic calculus.
The change of measure changes only the drift; the volatility remains the same.
Example. X, = aW,+p!,where W,is a P-Brownian motion. Then, with y, = p / q
there exists a measure Q such that % = W, +(,u/a)t
is a Q-Brownian motion up to
time T, i.e., 4 = a- Wt

4.5 BLACK-SCHOLES MODEL

Black and Scholes derived a differential equation that must be satisfied by the
price of any derivative security dependent on a non-dividend paying stock. The
fundamental insight gained from their result is, the option is implicitly priced if
the stock is traded.
Assumptions: .
The assumptions of the Black-Scholes models are: The price of the underlying
instrument S, follows a geometric Brownian motion with constant driftpand
vo tati lity cr in the equation dS, = ,uS,dt+ aS,dw,.
Short-selling is permitted, no transaction costs or taxes, all securities are perfectly
divisible, there are no dividends, there are no riskless arbitraqe opportunities,
security trading is continuous, the risk-free interest rate is constant and the same
for all maturities. .
The PDE
Given the assumptions of the Black-Scholes models, the price V,of a derivative
written on a stock with price process S, evolves according to the following partial
1
differential equation (PDE):

is notable that the P DE does not contain p , the drift of the stock. An informal
. It
derivation of the PDE,explaining this result, is given below.
The Formula
The above discussions lead to the following formula for the price of a call option
with exercise price Kon a stock currently trading at price S, i.e., the right to buy
a share of the stock at price Kafter Tyears. The consant interest is r , and the
constant stock volatility is a .
T )= SO(d,) - ~ e - " L( 4 )
C (8,
where . Appllcatron to Finance

Here@is the standard normal cumulative distribution function.

The price of a put option may be computed from this by put-call parity and
si~nplifiesto P(S, T ) = ~ e - ' ~ @ ( - d ,-) ~ @ ( - d , ) ,
Assume first that the interest rate is zero, r = 0.
Step 1. Find a measure Q under which S, is a martingale. Take St as a geometric
Brownian motion, so that

To convert it into a martingale we must get rid of ,u

= Sd exp{a P,- '/Z 02t),

where ,?l is the Q-Brownian motion (see change of measure theorem).
Step 2. Put El = E ~ (F,)
x/
Step 3. dE, = q d S t
Thus, the replicating strategy is to hold qtunits of stock at time t a d hold
= E, = ~ t s t
units of the bond at time I .
Assume now that the interest rate is not zero.
-'
Take the discount process Bt and form a discounted stock Zt = B, St and a I'
discounted claim BT -'x.Then operate with them as though interest rates were
zero, remember that B' = e r r.
Step 1. Make 2, into a martingale.
Note that

for a Q-Brownian motion .W,, so that Z is a Q-martingale. Then

r
S, = B, Z = e t Z, and so

determines the stock price process in the risk-neutral world. The corresponding
stochastic differential equation is
Quantitative Techniques
for Risk Analysis step 2. E, = EQ(Bt -' x I F; )
Step 3. By martingale representation theorem, dEt = (odZ,.
Replicating strategy.
a hold p, units of the stock at time t
-
hold Y; El = ptZt units of the bond
A strategy (4,+lr/,) of holding in a stock S,and a non-volatile cash bond B, has
value +
= +,S, lr/,Bland discounted value E, = plZl +y,

The strategy is self-financing if either

dV, = pt dSt + % B, or, equivalently,
dEt = p&I.
All claims X, knowable up to some horizon, T have associated replicating .
strategies. The arbitrage price of such a claim is given by
-'
V, = B, EQ(Bl X I F,) = e 'r(T-l) = EQ(XI F,)
where Q is a measure that makes the discounted itock a martingale.
Example: Pricing of a Call Option
Consider a call option with the exercise date T and strike price k. Then the claim
is X = max(Sr - k, . Find VO, the value of the replicating strategy (and thus the
option) at time zero as
Vo = e - r T ~ a(max(ST- k, 0)).
We should find the marginal distribution of ST under Q
ST= SO exp{a W,+ (r - !4 /t2)f)
-
Since WT N(0, T ) under Q,

where Z - N(% cr2 T c27). Then

ST = s& Z + r T
so that the claim price is
Yo= emrT~((s0e
z + r T - k)+)

1
a f to
Change of variables v =- ( x + - ~ ~ ~ ) /leads i
2
 1 Application to Finance
where ri = (log(s1k) + (r - -02~)lcrll?;.
Then writing
2

yields ' ,

This leads to Black-Scholm formula

We only need that the drift of the stock price is constant, its exact value is
immaterial. .
The Black-Scholes formula may be written as
.. c = So0 (dl)- kQ, (dz)
: Note that d2 = - o f i . 6
::, m e cdrresponding price formula for a European put is
/...

. ,

. p = ke-rr@ (-d2) - So @ (-4).

, 7 ,

, '
,,
. . .. ., ..

Finding the Replicating Strategy

The value of the claim X at time t is equal to V($, t) as in Black-Scholes formula
with S, (stock price at time t) instead of Soand (1"- t) (the remaining time) instead
of T. Since 4= d d W t + rS&, Ito's formula yields

ontheother hand, the self-financing condition gives

dVt = pdSI + Y d B I.
By equating coe.fficients of dSI, we get

meaning that the amount of stock at any stage is the derivative of the option price
with respect to the' stock price,
Compare the values of the replicating strategy V (S,, I ) = p, S, + Y: Bt with option
price given by the Black-Scholes formula
Quantitative Techniques
for Risk Analysis to see that the replicating st@tegy is given by

1
1
log(s1/k)+ ( r k - a 2 ) ( T - t )
2
Vl =Q,
D J ' T g

and

.The value of pt is always between zero and one. The amount of borrowing is
bounded by the exercise price k.
4.5.1 Black-Scholes Differential Equation
The Black-Schotes differential equation is derived in the following through no-
arbitrage or delta-hedging argument.
The Blac k-Scholes PDE
As per the model assumption above, we assume that the underlying (typically the
stock) follows a geometric Brownian motion. That is,
dS, = @',dl -t crSldW,
. ,
where W; is Brownian. .. .

Now let V be some sort of option on S -mathematically V is a function of Sand

t . v ( A , t) is the value of the option at time t if the price of the underlying stock at
time r is S . The value of the option at the time that the option matures is known.
To determine its value at an earlier time we need to know how the value evolves
as we go backward 'in time.
Another derivation
Consider a portfolio with one long option and short position in the asset

Then dl7 = dV- Ads and using It6 formula for Vt = V(S,, t )

Therefore

The coefficient before dS is zerb if

. '.
as in Delta hedging. with such A,
 Application to Finance

This change is riskless and must be equal. to the risk-free growth (to avoid
arbitrage possibilities)

and we obtain the Black-Scholes equation

Dividend Paying Stock

Example (European call option on a dividend paying stock).
Ex-dividend dates in two months and five months. The dividend on each ex-
dividend date is expected to be $0.50. The current share price is $40, the volatility
is 30%p.a., the risk-free rate of interest is 9%p.a. The time of maturity is 6
months. The present rate of the dividends is
09 = 0.974 1.
o.5e-2/12x009 + 0,5~-2/12~0

The option price can be calculated from the Black-Scholes formula with So=
39.0259;k = 40, r = 0.09, o= 0.3, and T = 0.5. The result is $3.67.
Stocks paying dividends continuously. If the stock is paying dividends at rate q
(continuously), then the Black-Scholes formula is applicable with So replaced by
~ ~ , is also applicable for options on
the dividend-adjusted stock price ~ o e - This
stock indices.
'The dividend payment is assumed to be a t d l over a small interval d,. Then the
process S, is not the value of the asset as a whole..To.make it tradable, we can
assume that dividends are used to purchase more stock, so at time t the number of
stoclc units will be St. Then the worth of such a portfolio will be
st= So exp { owt+ ( r - 6- 'h a 2 ) f }
Then the standard slrgulnents are applicable, and, under the martingale measure,
S, =So exp(crW, + ( r - 6- 1/2 c2)t}
is log-normally distributed. Then the forward price is F = e (r-s)TSo,

and the hedge is p, = e -&T- t) $( dJ)units of the stock and have a negative holding
of kr-''$( dz) units of the bond.
Note The same result is obtained if one adjusts So for the dividends and uses the
standard Black-Scholes formula on non-dividend paying stock with SOreplaced by
~ o e. - ~ ~
Quantitative Techniques
for Risk Analysis 4.6 OPTIMAL PORTFOLIOS
In portfolio theory we study investors optimal decision to diversifL their portfolio,
as well as pricing of risky assets. Usually the portfolio theory models the returns
of an asset as random variable.
If properly hedged, derivatives provide riskless returns. This feature is used by
banks who sell their products for a bit more than it is worth. Fund managers buy
and sell assets (and also derivatives) with the aim of beating the bank's rate of
return. This often involves taking a degree of risk.
4.6.1 ern-variance Approach,
The mean-variance approach was developed by H. Markowitz. To see its
formulation let us examine the following presentation.
Let P,jt) be the prick of security i at time t.
The returns Rt = P,{t)IP,jO)are modelled as random variables with
= E(Rj), ~ j =j COV(R,,A/),
The variance-covariance matrix o[ returns 'is denoted by Z. If the first asset is risk-
free, then the corresponding variance and covariances are zeros.
/
Let q be the fraction of initial wealth X of the investor invested in security i at
time t = 0
/

where yi is the number of shares of security i held by investor at time t = 0,

... , %lT i s called the portfolio wctor. Note that
r = (n,,
4 + ni+ ... + ?qr: 1 .
If 'shot&selling is not allowed, portfolios should have all non-negative
components. Such portfolios are called admissible. 'The total return is given by

Then

Two Assets Portfolios , ..

Take n = (n,,n,)' with q + n2 = 1 . For simplicity, denote nl

Then
- x and 4 = I - x.
E(R) = ~j +p(1 - X) , ~ 2 ,
 V ( R ) = x 2 ~4-: ( 1 - ~ ) ~+a2x(1
f -- . x ) p q c ~ , : Application to Finance

where p i s the correlation coefficient between the two individual returns. If short
sales are forbidden, then 0 I 3 r I , otherwise x can be an arbitrary number.
Now V(R)is a quadratic function of E(R), namely,

If the short sales are allowed, then the miriimuln variance is achieved at

Note that the corresponding x may be negative or greater than 1 , If short sales are
forbidden, then the optimal x can be found as either 0 or 1 or x* given above
(if x* E (0, 1)).
The objective function may be written as \

fi)= -AE (H)+ V(R),

where A is a risk aversion index. 1fA = 0, the portfolio with the lowest variance of
return will be selected. As A increases, the investor becolnes more willing to
accept risk in order to achieve a higher expected return, and ifA = oo the portfolio
with the highest expected return will be opli~i~al.
111application to two asset

portfolios, this gives

) A ( x ~+, ( I - x);L., ) + 2x0; 4-(I - x - ) ~ 9- ~ X (- Ix ) P D ~ o - ~ .
, f ( ~ .=

Then

The derivative is zero if

If short sales are forbidden, then the optimal value of x is either 0 or 1 or x* above
2 I)).
( i f x * ~(0,
Risk-free Assets and the Capital Allocation
Assume the portfolio R,, has two assets, a risk-free and another risky asset. The
capital allocation line connects all portfolios that can be formed using a risky
.:r(8($setand riskless asset. The second asset R also has several risky assets with
return E(R) and variance V (R). Let x be the share of the fund investe.d in the first
(risky) asset. Assume that the risk-free asset provides return RF. Then the asset of
the combined portfolio R, will be
Quantitative Techniques
for Risk Analysis Then, if o ( R )= JV(R)is the standam deviation of tlie risky portfolio and

Thus, the set of portfolios comprising a combination of the risk-free asset and a
risky portfolio is a straight line. The tangent 'point M to the set of admissible
portfolios represents the market portfolio or the best combination of risky assets.
The optimal position on the capital market line corresponds to the degree of risk
preferred. If continued beyorid My it represents a 'borrowing' portfolio.
a

If the borrowing rate RB is different from the lending rate RF, then two tangent
points are t o be determined.
4.6.2 Capital Asset Pricing Model (CAPM) #

The CAPM is based on investor's decision of not to invest in an asset which does
not improve the risk-return charactertics (more risk- more return) of her existing
portfolio. Thus, CAPM derivei the return for an asset in a market given the risk-
free rate available to investor arid the risk of the market as a whole.
Assume that the return of a security is determined entirely by the market index
and random factors

where RM is the return of the market portfolio and Ej is a random error term. Then
E(Bi ) = r + P, [E(RM)- r],
where r is the risk-free rate: E(Rl ) is th,e expected return of asset j; E(RM)is the
expected return on the market portfolio; P, is the "beta" of assetj. The beta's can
be estimated from the following linear model.

where r,,, is the return of asset j in the tth period, md,, is the return of the market
portfolio in the rth perlod, aj is the intercept, Ei,, is the residual error forperiod I .
The beta of an asset Ri can be also defined as

Further,
Var(Rj ) = 4: Var(Rbf)+ V a r ( ~ ),
and, for i ;r: , j , Cov(Ri,R,,) = P,Pl Var(Rni).
The beta of a portfolio can be found as the weighted sum of beta's of individual
assets
Cheek Your-Progress 2 Application to Finance
1) What do you mean by delta hedging?

2) What is the main finding of Black -Scholes model?

3) What do you mean by self-financing strategy?

4) I-Iow do you get the replicating strategy in Black-Scholes model?

...*..,.I...,..IL
..............I...

5) How do you account for risk in a mean-variance approach?

6) What is the main finding of capital asset pricing model?

,
Quantitative Techniques
for Risk Analysis' 4.7 LET US SLIM UP
In this unit we have discussed the basic concepts of finance that help understand
modelling pricing stochastically. The financial terms frequently used in derivates
have been documented before discussing the lnodelling on option prices in
arbitrage free conditions. Binary modelling technique has been introduced with
single and multi-period framework. The results have been further seen by
translating these in terms of martingale language. The Black-Scholes model
which showed that option is ilnplicitly priced if the stock is traded, has been
highlighted. The basic fortnulation of Black-Scholeg model which is built on the
assumption of option price following a geometric Brownian motion is deliberated
upon and tools of stochastic calculi~sused have been pointed oilt. The last part of
the unit gives the themes on optilnal portfolios and discusses the t11ean variance
approach and capital asset pricing model to show the determination of tlie return
of an asset in risky environnient.

4.8 KEYWORDS
Contingent Claims: Assets or securities whose prices depend on the values of
other assets or nilmerical indices,
Derivative: A conditional instrument used bymarket participants to trade or
manage an asset. There are essentially two categories of financial instruments that
are grouped under the term derivatives: options/futures and swaps.
Financial Marliet: A mechanism which allows people to trade money for
securilies.
Financial Risk: Any risk associated with money.
Forward: An agreement between two parties to buy or sell an asset at a pre-
agreed future point in time .
Futures: A contract to buy or sell a certain underlying instrutnent at a certain date
in the fittitre, at a pre-set price. The filture date is called the delivery date or final
settlement date. The pre-set price is called the futures price.
Hedge: An investment made to reduce the risk in another investment.
Hedging: A strategy designed to mininiise exposure to an unwanted business risk,
while still allowing the business to profit from an invegt~nentactivity.
Long Position: A treader buys an option contract that she has not already written
(i.e. sold). Here buyers are referred as tlie long.
Market Risk: Tlie potential to experience financial losses due to fluctuations in i
the prices at which equities, foreign currencies, 'interest rate linked securities, and
com~noditiescan be bought or sold.
Option: A derivative type of tradeable contract in securities markets, whereby
one party buys the right to exercise a feature of the contract on or before a future
date.
Position: A colnmitlnent to buy or sell a given amount of securities or
commodities.
Put Option (or Put): A financial contract between two parties, the buyer and the
writer of the option. The put allows the buyer the right but not the obligation to
 sell a commodity or financial instrument (the underlying instrument) to the writer Apl;lieafion to Finance
of the option for a certain time for a certain price (the strike price).
Put-call Parity: A relationship between the price of a call option and a put option
- both with tlie identical strike price and expiry.
Replication: An option trading strategy used to ensure at a certain date the payoff
without trading it.
Self-Financing: A strategy when a company is able to release liquidity in order to
finance operating charges and development. Lt is equal to the net profit to which
you reintegrate all the entries that are not associated with tlie movements of cash
such as depreciation and provisions, appreciation and depreciation of asset
disposals.
Short Position: A trader selling an option contract that she does not already own.
Here sellers are referred as the short.
Spot Price or spot rate: A price of a security quoted for immediate (spot)
settlement (payment and delivery).
Stock: Capital raised by a firm through the'issuance and distribution of shares.
Swap: A derivative where two counterparts exchange one stream.of cash flows
against another stream. Often used to hedge certain risks, such as interest rate risk
or for speculation.

4.9 SOME USEFUL BOOKS

Boshizen, F., A.W. Vaat, H. Zanten 'and K. Banachewtcz (2005), Stochastic
..Process for Finance Risk Management Tools (see internet)
Hull, John ( 1 992), Options, Futures, and other Derivative Securities, Prentice-Hall
International, Inc., New Jersey
Chalasami, P.,Somesh Jha (1 997), Steven Shreve: Stochastic Calculus Finance
Carregie Mellon University (see internet)

4.110 ANSWER OR HINTS TO CHECK YOUR

Check Your Progress 1

1) See Section 4.2 and answer
2) , See Sub-section 4.2.1 s t ~ danswer
3) See Section 4.3 and answer
4) See Section 4.3.2and answer
5) See Sub-section 4.3.2and answer
6) See Sub-section 4.3.3and answer

Check Your Progress 2

I) See Sub-section 4.4.5and answer
2 See Sectio~i4.5 and answer
Quantitative Techniques
for Risk Analysis 3) See Sub-section 4.4.5 and answer
4). See Section 4.5 and answer
5) See Sub-section 4.6.1 and answer
6) See Sub-section 4.6.2 ahd answer