I want to generate a bivariate time series data with given joint distribution. Let's say Yt = A + B*Yt-1 + Vt, where Yt = (Y1t,Y2t)'. The dependence between Y1t and Y2t is measured by a given copula function. I.e., C(F(Y1t), G(Y2t)) = H(Y1t,Y2t), where C is a copula function, F and G are the marginal functions of Y1t and Y2t and H is the joint distribution of Y1t and Y2t. How to generate Yt? Thanks for help. Xian _________________________________________________________________ ?????????????????????????????? MSN Hotmail?? http://www.hotmail.com