if you value/hedge cliquets with lognormal model, you will get killed. you should reconsider looking into stochastic vol for these products

Hedging Cliquets:equivalent to series of fwd starting calls; lets say striking at t1 & maturing at t2.Pricing: I can find the fwd local vol surface if i know the vol surface for 0->t1 & 0->t2, Reason I am asking you to price it Local vol, because its a cap floor cliquet, something like a call spread having a forward skew exposure......Hedging exposures: 0-t11)delta is 02)vega : First of all y at all there should be a vega. as you increase the vol theres an opportunity of lower strike and higher spot at maturity, but if things turn out to be otherwise we are safe because of the global flooring at 0. Hence a positive vega, Now intuitively this vega is higher than the vega of an ATM already struck option. And we near the date t1 vega approaches the vega of a already struck ATM option. selling a 0-t2 straddle and buying a 0-t1 straddle in an appropriate ratio can ensure an approximate hedge.

I've seen a most general payoff function containing local and global cap, and local and global floor.but in reality, do these four guys always exist or only some appear?

QuoteOriginally posted by: PaulWell, you need a framework/model for the incompleteness and then some goal you are trying to achieve by hedging.numbersix will have lots to say on this!PPerhaps numbersix will comedown to the forum and answer the questions and help us. Assumption- Numbersix is around and kicking.yes, dr.Paul - How are you! After a long time.