Scribd
Upload
10 views

Fin1819 Sample

Uploaded by

ziruigu2000
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
10 views

Fin1819 Sample

Uploaded by

ziruigu2000
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Sample Questions MA5248

Question 1. (SDE and Itô formula)


Let Wt be a one-dimensional standard Brownian Motion.
(a). Consider the stochastic differential equation:
1/3 2/3
dXt = 3Xt dt + 3Xt dWt

with initial condition X0 = 0. Show that both Xt = Wt3 and X ≡ 0 (i.e. Xt = 0


for t ≥ 0) are solutions. Do the coefficients of the SDE satisfy the Lipschitz
condition? Note that the coefficients satisfy the linear growth condition.
(b). Let dXt = at dt + σt dWt . We want to prove that if X ≡ 0, then a ≡ 0,
σ ≡ 0.
(i). Apply Itô formula to Yt = exp(−Xt2 ).
(ii). Deduce the desired result.

Question 2. (Integration by parts). Let W1 (t) and W2 (t) be two indepen-


dent one-dimensional Brownian motions. Let Yt = tX1 (t)X2 (t) with

dX1 (t) = f (t)dt + σ1 (t)dW1 (t)


p
dX2 (t) = σ2 (t)(ρdW1 (t) + 1 − ρ2 dW2 (t)), ρ ∈ [−1, 1]

where f (t), σ1 (t) and σ2 (t) are deterministic functions such that the integral
and stochastic integrals are well defined. Compute dYt .

Question 3. (Properties of Brownian motion)


Z t Let Wt be a one-dimensional standard Brownian Motion. The process Yt =
Wu du is then a Gaussian process. Compute its expectation and covariance
0
(i.e. cov(Yt , Ys ) for 0 ≤ s ≤ t).
We assume that the order of expectation and integration can be inter-
changed.

Question 4. (Delta-Gamma neutral portfolio)


Suppose a portfolio is delta neutral and has a gamma of -1800. The delta
and gamma of a particular call option are 0.5 and 1.80, respectively.
How to make the portfolio delta and gamma neutral with the underlying
asset and the call option?

1
Question 5. (Change of probability and Barrier options)
Let Wt be a standard Brownian motion under P. Let µ be a constant.
Introduce Zt := exp(µWt −µ2 t/2). For T > 0, denote by P0 the new probability
defined by
P0 (A) := E[Z(T )1A ], A ∈ FT .
(a). By Bayes formula, prove that if M 0 (t) is a martingale under the new
probability P0 , then M (t) := Z(t)M 0 (t) is a martingale under the original
probability P, i.e.

E[M (t) |Fs ] = M (s), 0 ≤ s ≤ t ≤ T.

(b). Introduce W̃t = Wt − µt. Without doing any integration, for x ≥ y and
y < 0, compute
 n o   
E exp 2W̃T − 2T · I W̃T ≥ x, min W̃t ≤ y · ZT .
0≤t≤T

By using twice the change of probability argument, this expectation can be


reduced to a probability.
We admit the result that when x ≥ y and y < 0, we have
   
2µy/σ 2 −x + 2y + µt
P Wµ,σ (t) ≥ x, min Wµ,σ (s) ≤ y = e Φ √
0≤s≤t σ t
where Wµ,σ (t) := µt + σWt is a Brownian motion with drift µ and volatility σ.

Question 6. (Self-financing condition and numeraire)


Consider a market with two assets S (1) and S (2) where S (1) is positive. A
(1) (2)
portfolio Vt = πt1 St + πt2 St is self-financing if
(1) (2)
dVt = πt1 dSt + πt2 dSt .
(1) (2)
We choose the first asset S (1) as numeraire and denote Ṽt = Vt /St , S̃t =
(2) (1)
St /St , where
(1) (2)
Ṽt = Vt /St = πt1 + πt2 S̃t .
The objective of this question is to prove that the self-financing condition
does not depend on the choice of numeraire, i.e.
(1) (2) (2)
dVt = πt1 dSt + πt2 dSt implies dṼt = πt2 dS̃t .

(i). Preliminary question. Let X and Y be two continuous processes (we don’t
give their dynamics which are useless for the following questions). Write down
Itô formula for the product XY (i.e. d(Xt Yt )). Deduce that, for a positive
process X (i.e. Xt > 0 for t ≥ 0),
   
1 1 1
Xt d + dXt + dXt · d = 0.
Xt Xt Xt

2
   
1 (1) 1 (2)
(ii). Write down dVt · d (1) as a function of dSt ·d (1) and dSt ·
St St
 
1
d (1) .
St
    
(2) 1 1 (2) (2) 1
(iii). Prove that dṼt = πt2 St d (1) + (1) dSt + dSt ·d (1) .
St St St
 
(2)
St (2)
(iv). Prove that dṼt = πt2 d (1) = πt2 dS̃t .
St

You might also like