I explain the reason for my question below, but for those not interested in the story I simply ask:Why would the following two payoffs produce a different option value using Black76:Payoff 1: [$]max[Strike - (Forward + spread),0][$]Payoff 2: [$]max[(Strike - spread) - Forward,0][$]where [$]spread[$] is a flat spread on the forward rate.If you reshuffle the insides, the payoffs are the same, but the option value from B76 is different.Now, by simply looking at the parameters of B76, the first thing you can say is that [$]\ln(\frac{Forward + spread}{Strike})[$] is not the same as [$]\ln(\frac{Forward}{Strike - spread})[$], which explains mathematically why, but I would like to understand ("economically") what is going on here.Or am I missing something silly!?Story background:As part of a different project I recently validated our Front Office trading system's Interest Rate Cap pricer; stock standard Black-76 at its best.All went well until I added "a constant spread to the resultant forward rate" when all of a sudden I could not match the value anymore.After liaising with the developers they informed me that the spread is actually subtracted from the strike instead of added to the forward. Accounting for this change I match the system again. So the original question remains, why are the option values different when the payoffs are similar?

max has two arguments and, with that, it's clear they are different

I'd have said (assuming you simply forgot max( ... ; 0 ) in the payoff description) if you put forward+spread and strike into B76 you get a lognormal distribution for forward+spread (aka shifted lognormal distribution for the forward with shift = spread), while with forward and strike-spread you have a lognormal distribution for the forward. Which is quite different of course, e.g. the possible values for the forward in the former case are (-spread,\infty) while in the latter case they are (0,\infty).

Alan...have fixed the little typo, thanks.@pcaspers, thanks very much, I think that is it!