Previously, **Stock price simulation **, we saw that stock prices can be described by geometric brownian motion (GBM).

In order to solve a partial differential equation GBM, we need to appreciate Ito’s lemma.

Let us consider G is a function of two variables, x and t.

The change in G for a small changes in variables can be solved appliying partial derivatives by differentiating with respect to one varible at a time, keeping the other variable constant:

Taylor series are used in the formula above to get a more accurate estimate.

Suppose we have a variable x that follows the Ito process: Δx = aΔt + b*e*sqrt(Δt), where “e” is a random number drawn from a normal distribution with mean = 0 and standard deviation = 1.

Now, we take this Ito process to the power 2 and see that power(Δx,2) = power(b*e,2)*Δt, where all Δt with power higher than 1 are ignored because the value is too small.

Also, the expected value of a random number is too small and can be ignored as well.

Thus, power(Δx,2) = power(b,2)*Δt. In the Taylor series we take approximate value of G:

Plugging into the equation for x in ito’s lemma, we obtain:

or it can be rewritten as:

This is called Ito’s lemma.

Now, consider the stock price to be lognormally distributed, G=ln(S), x=S, a= μS and b=σS(according to GBM formula in **Stock price simulation **) and taking into account that:

We obtain:

That can be simply rewritten:

And it gives a stock movement formula:

Additional iformation is **here.**