Covariance of correlated lognormal random variables.

40 Views Asked by At

Suppose $X_1,X_2,...,X_N$ are correlated lognormal random variables such that $$cov(\ln X_i,\ln X_j)=c_{ij}$$

and $\ln X_i\sim N(\mu_i,\sigma_i^2)$.

Is there a way to find $cov(X_i,X_j) $ with only the provided information?