Can anyone provide a formula for a polynomial interpolation of the corrected lognormal distribution used to model returns traditionally resulting from the wrong Brownian motion generated model? Obviously I am only looking for a sufficiently good approximation. Thanks.