Hello all,I've implemented Monte Carlo pricing of European and American Basket Options, but there's something that keeps bothering me that I want to make sure is not an indication of a bug/mistake. The American options are priced based on the Longstaff-Schwartz algorithm, and Europeans are priced pretty much using a multi-variate (correlated) Monte Carlo approach (discounted average of payoffs at maturity).With vanilla stock options, if the underlying stock pays no dividends, one expects the American and European options to have the same price given that the optimal excercise date for the American is at option maturity. Now, I'm not finding that relationship with basket options, and I was wondering whether this is an indication of an error in my part or if this is expected behavior. I understand that the combination of log-normally distributed basket instruments will not produce a log-normally distributed basket, and therefore Black-Scholes is not quite applicable. Could this be the possible explanation?Thanks for any help/insight,LB