Stochastic Finance:

Applications of Stochastic
Calculus in Finance

1
 Dr. Babita Goyal
Department of Statistics, Ramjas College
University of Delhi
Delhi-110007

2
Market Volatility and Stochastic
Calculus

3
Stochastic Calculus Vs Mathematical
Calculus

4
Brownian motion

5
 Financial Market Jargon
Financial products are very complex in nature
and require sophisticated mathematical tools for
analyzing them. The risk involved in these
products is large and needs to be taken care of.
To cover the risk, various types of financial
derivatives have been proposed. These
derivatives are used to hedge the risk and hence
are more complex in nature.

6
 Financial Market Jargon

• Asset - risky assets

• A riskless (or risk free) asset

• Portfolio

7
• Arbitrage

• Derivative - Contingent claim.

Examples - options, swaps, futures and forwards.

• Option is a right (and not an obligation) to sell (a Put)

or to buy (a Call) a product at (European) or before
(American) a predetermined time and at a pre-agreed
price.

• Attainable contingent claim with pay-off C

• Complete market model

8
Two Fundamental Theorems of Finance

• First Fundamental Theorem of Finance:

In a perfect market arbitrage does not exist.

• Second Fundamental Theorem of Finance:

For a market model to be complete, it is
necessary and sufficient that there exists one
(and only one) risk neutral probability
measure.

9
A risk- neutral probability measure P * on a
sample space Ω is such that the expected value
of an asset Ѕ at time t1 is given by

 
E P* St1  (1  r ) t1 t0
St0

where St is the value of the asset at time t0 ; r is

0

the risk-free rate of interest.

10
 Financial Derivatives and Stochastic
Processes
Portfolio vector    0 , 1 ,...,  n 
Price vector at time ‘t’
St   S0,t , S1,t ,..., S n ,t  ; t  0,1, 2,...N

The Value process- Vt ; t  0,1, 2,...N

Vt   St   0 S0,t   1S1,t  ...   n Sn,t

11
Portfolio allocation (portfolio strategy)
If  i   i denote the proportions of assets, then
   is a portfolio allocation
t ; t  0,1, 2,...N is a portfolio strategy
A portfolio strategy t ; t  0,1,2,...N is self-
financing if
t St   t1 S t1 ; t  1, 2,..., N  1

12
 Cox-Ross-Rubinstein Model (CRR) -
Binomial Model
A two assets only Portfolio    0 , 1 

Return on risk-free asset  0 = S0,t   0 1  r 

t

(1  u ) S1,t 1 if Rt  u
Return on risky asset  1 = S1,t  
(1  d ) S1,t 1 if Rt  d

(-1 < d < u)

13
Value of a self-financing portfolio at time ‘t’
Vt  V0    j  X j  X j 1 
t

j 1

Sj
where X j  j = 1,2,…,t.
1  r 
j

Discrete time stochastic integral of the portfolio

process  ; t  0,1, 2,...N  w.r.t. the random process
t

 X t ; t  0,1, 2,...N .

14
 Limiting case of a CRR model-
stochastic integral

  T x  rT 
 2T  
x2 / 2

lim C  lim f  S N   e  f  S0e

 rT 2
 e 2
dx
t  t  
  
where X N (0,1)
This is a stochastic integral.

A stochastic integral is the integral of a function

(mathematical or stochastic) w.r.t. a stochastic
process.
15
 Wiener process
A stochastic process Bt : t  R   such that
• B0  0 almost surely.
• The sample paths t Bt are continuous everywhere
with probability 1.
• For time sequence t0  t1  t2  ...  tn for all finite n, the
increments Bt1  Bt0 , Bt2  Bt1 ,..., Btn  Btn1are independent.
• For 0  s  t , Bt  Bs ~ N  0,(t  s) 2  .

Standard Brownian motion -  2  1

16
The total size of the movement in an interval
[0, T] is given by BT  Bs ~ N  0, T 

17
If the money invested in the asset with the price
St can be described as a function f (t ) of t, then
the value of the portfolio in 0,T  can be
described as T T

 f (t )dS
0
t    f (t )dBt
0
where
T

 f (t )dB   a  B 
n

t i ti  Bti1
0 i 1

ai in the interval ti 1 , ti  is independent of t.

18
 A square-integrable function f in the
space    2 


:  f  t  dt  
2 2

If f :  s. t. f 2  


0

  
 f (t )dB N  0,  f  t  dt 
2
Then t
0  0 
    
2

Further E   f  t  dBt    f  t  dt
2
  - Itô isometry
 0   0

Expectation of a stochastic integral is equal to a

deterministic integral.

19
 Itô stochastic integrals
A stochastic integral defined with respect to a
square- integrable adapted process

- Can be used to determine portfolio values as

generally portfolio processes are adapted to
random events occurring in the market

20
 Equation of an Itô process
For a Brownian motion, the Itô integral is
t t
1
f  Bt   f  B0    f '  Bs  dBs   f ''  Bs  ds
0
20

and hence, for square-integrable processes

t
u : t  R 
 and vt : t  R 
, if df  X t   vt dt  ut dBt
then the required equation is
t t
1
f  X t   f  X 0    vs ds   us dBs ; t  

0
20

21
 Itô formula for Itô processes-for a
continuous function of two variables
For a function f  t , x  continuous in R  ,
f f
t t
f  t , X t   f  0, X 0    vs  s, X s  ds   us  s, X s  dBs
0
x 0
x
f 
t t 2
1 f
   s, X s  ds +  us 2 
s, X s  ds
2

0
x 20 x
Equivalently
f f 1 2 2 f
df  t , X t    t , X t  dt   t , X t  dX t  ut 2 
t , X t  dt
t t 2 x
22
Major difference between this expression and
the deterministic calculus expression lies in the
inclusion of the second order term. The reason
for inclusion is the size of the increment in a
Brownian motion, i.e. dBt   dt which means
that  dBt   dt and hence second order terms
2

cannot be ignored.
So, if we integrate f t, Bt   Bt 2 , then we have
T
BT 2 T
0 Bs dBs  2  2
Note the additional second term.
23
Using Itô integral for a portfolio process
dSt   St dt   St dBt

 1 2  
we get St  S0 exp       t   Bt 
 2  
i.e. St : t  R  is a geometric Brownian


motion representing the price of the risky asset


at any time t  R .

24
 Black –Scholes’ Partial Differential
Equation
For a self-financing portfolio strategy such that
the value of the portfolio is of the form Vt  g t, St 
where g .,. is a continuous function of t and St
on space R   R. Then g t, St  satisfies the
Black- Scholes’ PDE
g g 1 2 2 2 g
r.g  t , St    t , St   r.St t , St    St 2 t , St  ; t  R 

t St 2 St
and   g ; t  R 
St
t

25
If the portfolio is hedging an option with pay-off
C  f  ST  , then this is the price of the option
which satisfies the boundary conditions
g T , ST   f  ST 

i.e. C  g T , ST   T AT   T ST

26
 Black-Scholes option pricing formula
The solution of this equation is the famous
Nobel Prize winning Black-Scholes formula
g  t, St   St   d2   Ker (T t )  d1 
where x
1
  x  e
t 2 / 2
dt ; x  R
2 

 t
S   2

log     r   T  t 
K  2 
d2  ; d1  d2   T  t
 T t
27
The price of a European call option is given by

ST  K ; ST  K
C
0; ST  K

and the price of a European put option is

 K  ST ; K  ST
P
0; K  ST
Put-call parity
P  t , St   C  t , St   St  Ke r (T t )

28
 Jump processes
Count Process

Nt = 1Tk , (t ); t  R 

k 1

1 if t  Tk
1Tk , (t )   is the jump size
0 if 0  t  Tk

Poisson process – independent, homogeneous

and regular count process- a jump process

29
A Poisson process is given by

 t
e ( t )k
P  Nt  k   ; k  0,1, 2...;   0
k!

 is called the intensity of the process

30
Compensated Poisson process-  N  t : t  R 
t


- a martingale under its own filtration

Compound Poisson process- Variable jump size

 Nt

Yt =  Z k ; t  R  where Zk : k  1, 2,...
 k 1 

are i.i.d. random variables with probability

distribution Z (.) on R

31
A stochastic integral with respect to a
compound Poisson process
Stochastic process t
 : t  R 

Compound Poisson process Yt : t  R  
T NT
Stochastic integral   dY
t t  
k 1
Tk Zk
0

32
This representation finds a natural
interpretation in financial markets, explaining
the value of a portfolio at time T when the
portfolio has a (fractional) quantity t of a risky
asset whose price evolves according to random
return Z k at random times Tk

33
Stochastic integral of a jump-diffusion
process with respect to a compound
Poisson process
t
 : t  R 
 - a standard Brownian motion
 t t

 X t   us ds   s ds  Yt : t  R 


 0 0 
- a jump-diffusion process
The stochastic integral is
T T T NT

  dY
t t   t ut dt   tt dt   Tk Z k ; T  0
0 0 0 k 1

34
 Itô formula For a general jump
process (a Lèvy process)
t t t

 
f  X t   f  X 0    us f ' X s 
1
 
ds   f ''  X s  us ds   s f ' X s dYs
20
2

0 0

  
t

 
  f  X s   f X s  X s f ' X s d  N s  s 
0

  
t

 
  f  X s   f X s  X s f ' X s ds; t  R
0

A Poisson process, a compound Poisson process

and a jump-diffusion process, all are examples of a
Lèvy process.

35
 Solutions of SDEs of a jump process
(i) dSt  St  St   St ; t  0  St  S0 1    N ; t  0
 
t

(ii) dSt  t St dt  t St  dNt   dt 

 
t t

 s ds  s ds
 1   ;
Nt
 St  S 0 e 0 0
Tk t  R
k 1

(iii) dSt  t St dt  t St  dYt   E  Z  dt    t St d t

 

t t t t
1
 s ds E Z   
 s ds   s ds   s ds Nt 
2

 
1  Tk Z k ; t  R 
2
 St  S 0 e 0 0 0 0

k 1
36
Girsanov Theorem for Jump Processes
- allows for a change in probability measure and
hence pricing of a contract.

So, if an asset price is modeled by the equation

dSt   St dt   St dt  St  dYt

Then it’s price is 1 N

t t   2t t
St  S0 e 2
 1  Zk ;
k 1
t  R

37
These were some of the techniques which can
be used to price a portfolio of riskless assets,
risky assets and (vanilla option) derivatives.

These can be used to price other derivatives

such as bonds, swaps, futures, forwards and
interest-rate derivatives, exotic options as well

38
Some other techniques are change in probability
measures and change of numèraire , Reflexion
principle and martingales.

39
 Bibliography
1. Stochastic Finance- An Introduction with
Market Examples (2014): Privault N. - Chapman
& Hall/CRC Financial Mathematics Series.
2. Investment Science: David G. Luenberger (2015)
- Oxford University Press(South Asian edition).
3. Statistics of Financial Markets- An Introduction
(2011): Franke, J., Hardle, W.K. And Hafner, C.M. –
3rd Edition, Springer Publications.
4. A Course on Statistics for Finance (2012)-
Stanley L. S. - Chapman and Hall/CRC.

40
41