10-StochDiffEq Page 280 Wednesday, February 4, 2004 12:51 PM
280 The Mathematics of Financial Modeling and Investment Management
t
xt()= Φ()t x () 0 + ∫Φ–^1 ()sas()sd , 0 ≤t < ∞
0
Let’s now go back to the stochastic equation
dXt = [At()Xt + at()]dt + σ()tdBt , 0 ≤t < ∞
X 0 = ξ
Using Itô’s formula, it can be demonstrated that the above linear sto-
chastic equation admits the following unique solution:
t t
Xt()= Φ()ξ t + ∫ Φ–^1 ()sas()sd + ∫Φ–^1 ()σs ()sdBs , 0 ≤t < ∞
0 0
This effectively demonstrates that the solution of the linear stochastic
equation is the solution of the associated deterministic equation plus the
cumulated stochastic term
t
∫Φ s ()dB
- 1
()σs s
0
To illustrate this, below we now specialize the above solutions in the
case of arithmetic Brownian motion, Ornstein-Uhlenbeck processes, and
geometric Brownian motion.
The Arithmetic Brownian Motion
The arithmetic Brownian motion in one dimension is defined by the fol-
lowing equation:
dXt = μdt + σdBt
In this case, A(t) = 0, a(t) = μ, σ(t) = σand the solution becomes
X = μt + σB
The Ornstein-Uhlenbeck Process
The Ornstein-Uhlenbeck process in one dimension is a mean-reverting
process defined by the following equation: