Say, I am pricing vanilla call on Basket option. C = Max(Sum(w(i)*F(i)) - K), i = 1 to NThe underlying is Futures contract with different expiration. I can use different models to priceoption such as Gentle, Log-Normal(Levy), Reciprocal Gamma(Milevsky), Monte Carlo ....In all of this models, one of the input parameters will be Sigma(i) which is volatility of F(i)My question would be:Let's say, I am pricing option on Calendar 2010, so T1 = Jan10 - today, ...T(12) = Dec10 - today.Topt = Dec09 - today. I have Sigma(1) .... Sigma(12) for each Futures contract.Should I adjust volatiles Sigma(i) of each F(i) to account that Jan10, Feb10 ... Dec10 later then expiration of the option (Topt = 12/31/2008) ?Thank you in advance !

Hi,Yes, I would say you should ... because typically the vol of longer-dated futures contracts tends to be lower than shorter-dated futures contracts (by vol, I mean both ATM option vols as well as historic vol of futures contracts). I have seen this for natural gas futures (and nat gas futures options) and similarly for crude oil. I say "typically" because for a base metal (or agricultural product - can't quite remember) I have seen different behaviour.This behaviour is called the Samuelson Effect.

Hi Ahew,thank you for the message, i googled Samuelson effect but cannot find any good article about how to damp volatility. can you suggest any report/article ? thank you in advance

Hi,If you can get hold of it, "Understanding Sam" by Sang Baum Kang and Mark Klein is quite good - it's at http://www.eprm.com/public/showPage.htm ... on=2005-09 - you'll need to create a trial login.Also look at http://www2.warwick.ac.uk/fac/soc/wbs/r ... 05-243.pdf.

Thank you very much for those links !