December 27th, 2005, 6:29 pm
Hi,Most traders on funds options mark the volatility using benchmarks (true only if the fund can be benchmarked of course !!) :As, Return Fund = Beta . Retun Index + Error, we deduce for SE and thus historical volatility,Vol Fund²=Sqrt(Beta².Vol Index² + Vol Error²)But this relation works for historical not for implied volatilities !There, I see 3 ways for estimating volatility1/ getting values of options on other baskets of funds issued on the market and then deduce a level of volatility according to a geographical area and type of fund)This is the way the market price impled correlation for CDOs but the funds market semmes even more illiquid2/ using the previous relation and apply it to the implied volatility of the benchmark for every strike and maturityIt works pretty well around the moneyness and for maturities between 6m to 2Y but after you can seriously misprice the skew or the term structure3/ implying the volatility using the previous relation and transposing it to a Market Distribution Model or a DupireOnce again, the skew and term structure can be very far from market reality !!Any ideas on the subjet ? What would you take for volatility input when pricing a Call ? I do not like Historical vol.Thanks in advance,