# Thread: Geometric random walk with Normal steps

1. ## Geometric random walk with Normal steps

Let X1, X2,....be a geometric random walk with normal steps. More formally, we assumet that, for any positive integer k,

Xk = X0exp(r1+.....rk)

where X0 is a fixed positive constant and r1, r2, .....are i.i.d N(μ,σ^2)

1. Determine the expectation and the variance of the random variable Yk = log(Xk/X0) in terms of k, μ and σ^2.

2. Let μk = kμ + log(X0) and σ^2 k = kσ^2. Then, show that Xk is Lognormal (μk, σ^2 k)

3. Determine the c.f.d and the p.d.f. of Xk

4. Show that the expectation of Xk is exp(μk + (σ^2 k)/2 ) = exp(kμ + log(X0) + k(σ^2/2)

Hint: Start with the definition of the expectation and use the change of variable y=log(x)-μk. You will also need the expression of the p.d.f. of the normal distribution with arbitrary parameters.

5. Show that E(X^2 k ) = {E(Xk)}^2 exp(kσ^2), and deduce the variance of Xk

Could you please walk me through the these questions.

2. Originally Posted by matthew2040
Let X1, X2,....be a geometric random walk with normal steps. More formally, we assumet that, for any positive integer k,

Xk = X0exp(r1+.....rk)

where X0 is a fixed positive constant and r1, r2, .....are i.i.d N(μ,σ^2)

1. Determine the expectation and the variance of the random variable Yk = log(Xk/X0) in terms of k, μ and σ^2.
$Y_k=\log(X_0)+(r_1+r_2+..r_k)-\log(X_0) =r_1+r_2+..r_k$

the sum of $k$ iid RV $\sim N(\mu.\sigma^2)$

CB