This work is based on Turnbull und Wakeman (1991). We describe their quick algorithm for pricing a European Average Rate Call Option. The main challenge in this, is to determine the distribution of a sum of lognormal random variates. First, we get to know the lognormal distribution. Afterwards we learn how to derive cumulants of a distribution from its moments. We consider the representation of a probability density function through another probability density function, using the generalized Edgworth Series Expansion from Boyle (1977), that uses the cumulants. After we got to know the assumptions and implications of the Black Scholes Model, we approximate the desired distribution by another lognormal random variable, referring to Mitchell (1968). We calculate the moments of the exact distribution by a recursion, which takes advantage of a multiplicative decompositon of the lognormal distribution. We set the first two moments of the approximation and the exact distribution equal. The derived approximation algorithm is tested against Monte Carlo Estimates and found to be faster, while sufficiently accurate.
«
This work is based on Turnbull und Wakeman (1991). We describe their quick algorithm for pricing a European Average Rate Call Option. The main challenge in this, is to determine the distribution of a sum of lognormal random variates. First, we get to know the lognormal distribution. Afterwards we learn how to derive cumulants of a distribution from its moments. We consider the representation of a probability density function through another probability density function, using the generalized Edg...
»