November 9th, 2005, 8:43 am
one thought about cliquet hedging - imagine we are talking about a vanilla cliquet, uncapped. let's imagine a one year monthly cliquet of calls, so one spot start call and 11 forward start calls.in a way this has some of the properties of a variance swap. a variance swap always has a vega, no matter where the spot goes. you could in principle trade a one year variance swap with monthly settings, and this would probably act very much the same as a cliquet. in fact the payoff function of a variance swap can be viewed as a "power cliquet". as each setting looks like [ ( S(ti) - S(ti-1) ) / S(ti-1) ]^2 you can see the analogy. it is a cliquet of straddles with a "squared" payoff!so maybe the best vol hedge for a cliquet of this type is a variance swap? (or a vol swap?)for a cliquet of call spreads of course you have got a sort of supercharged skew position, like a skew swap would have if you could deal it.for a globally floored, locally capped cliquet ... hmmm ...