April 9th, 2006, 2:26 pm
QuoteOriginally posted by: ddrdoubledepend a lot also on the different underlyings, and especially what kind of instruments are available calibrate to...take a look at piterbargs paper about model with fx skew. But on many underlyings it is problem of liquid instruments, not only for volatility but even concerning forwards...in this way fx/ir is the easiest to price because of the liquidity of both markets, only difficult for vvery longterm structures..if you have commodities/ir it is getting pretty difficult to get liquid instruments, beside the futures and the main instruments like oil, copper, alu, (i put silver and gold under fx)if you need other informations send me a pmrgdsPartly agree with ddrdouble. In my experience most of the issues arise in hedging and pricing correlation between the underlyings. fx/ir, equities/credit and commodities/some equities is fairly straightforward, but mostly it is guesswork. Even when you get reasonable confidence in your correlations ATM, the correlations can look very different if spot levels change significantly during the lifetime of the product.Model-wise, most generally you would be looking at multifactor curve models, but in practice reduced factor models with curve spreads as the underliers are the norm.