there is a certain structure with a certain payoff i.e a floorthat knocks out when a certain cumulative level is reached.i.e pay max{8%- 3mLibor(set in arrears) , 0}and the whole structure is knocked out when the cumulativepayout exceeds 12% of principalCould someone discuss the motivations and pricing ramificationsof having 3mLibor set in arrears as opposed to setting in advance?Is it simply a convexity issue? i.e if reset in advance the payoff would like {Notional*L(t,t+3/12)} / {1+daycount*L(t,t+3/12)}where as if it reset in arrears the same period payoff would be{Notional*L(t+3/12,t+6/12)} / {1+daycount*L(t,t+3/12)}Am i thinking of this correctly?Are there other issues?Why would someone prefer this structure as opposed to one resetting in advance(pay max{8%- 3mLibor(set in advance) , 0}thank you all for your insights

you could have it setting in advance or arrears - it's not a big deal. the current shape of the fwd curve is positive so you expect libor in arrears to set higher than libor in advance.the implication for this trade is that setting in arrears enables you to offer a bigger "headline" coupon (8%) thansetting libor in arrears (in the current positive curve). given that your target investor market is indifferent between adavnce/arrears (they just see "libor") setting in arrears gives you a better headline without revealing the actual cost. try it on a simple 3yr receive irs. say my client is receiving 3.50% fixed (ie 15 under mid) on $100m. if i switch to arrears rate setting and keep the rate constant, i pick up $770869 (27 swap points)so rather than a convexity issue, i'd say it's more about marketing and P/L.

thanks kurve