OkTwo underlyings, Two Currencies. Non-financial payoff, exercise into the underlying. Anyone have any ideas on how to value such a product?

OkTwo underlyings, Two Currencies. Non-financial payoff, exercise into the underlying. Anyone have any ideas on how to value such a product? >>Is this a question or a Nike commercial?I think you're asking how to value the option to swap one million shares of MSFT for an office building in London with an expiry of one year. Is that it?

Close but no cigar!!!I had a chat with one of our quants and this is what he says..Here is the payoff: MAX(U1 * NOK/EUR - U2 - X,0) The U2 and Strike are in Euros. My quant says to just sum the vols like so sig spread = (sig1 + sig(n/e))^2 + sig2^2 - 2 rho sig2 *(sig1 + sig(n/e)). But I do not think this works properly. There is obviously an additional correlation risk due to the FX, but how can I adjust a spread option approx model (HAUG) to account for this? Other than Monte Carlo is there another way to price/manage this option? I guess I could price it up ignoring FX, but a 1000bps swing in the FX does some change to the spread.I am positive the rate could swing by 5000 either way over the life of the option....Can you help Dr. Aaron?

Sorry to keep answering questions with questions, but I'm still not sure I understand. What currency is U1 measured in? Looking at the formula, EUR is the natural bet, because then you're converting the price to NOK. But then you're subtracting EUR from NOK, which usually is not meaningful. If U1 and U2 are special assets with strong currency correlations this might make some sense, but then your quant's answer doesn't make sense.If you could tell me what U1 and U2 are, that might help me understand.If you change sig1 to sig1^2 in your quant's formula, you have the formula for the variance of (U1 + NOK/EUR - U2) under the assumption that NOK/EUR is uncorrelated with U1 and has the same correlation as U1 with U2. If U1 is always near 1, and has small standard deviation relative to NOK/EUR, and you estimate rho by the correlation coefficient between NOK/EUR and U2, this formula might be reasonably accurate. But it doesn't smell right to me, even as a quick approximation.

This is exactly what I am looking for.. Any answers!Thanks,Ram

ok,the payoff is: MAX(U1 * NOK/EUR - U2 - X,0) The U2 and Strike are in Euros. 1. this implies you have U1 in NOK, U2 in EUR, and so the strike AND the notional must be in EUR. This means also your payoff formula is wrong because your currency exchange rate should be EUR/NOK as below (this reads EUR per NOK also noted NOK-EUR which is the invert of the standard currency market quote EUR-NOK). payoff should be : MAX(U1 * EUR/NOK - U2 - X,0) = MAX(U1 * EUR-NOK^(-1) - U2 - X,0).2. from there it's all easy, this is a simple spread option. - Start by calculing the correlation of U1 with EUR-NOK and calculate the cross vol for that asset (v1). Be careful with the signs !.- Calculate the correlation (corr) of the cross asset (A1 = U1 / EUR-NOK) with U2. - Evaluate the spread option between A1 (vol =v1) and U2 (vol=v2), using the correlation corr as above.Other possibilities3. your payoff was right, meaning your client wants you to "quanto" U1 into a scalar which happens to be stochastic. It means nothing economically, but maybe your client has a good reason for it. Just ask the client if the payoff formula is right though.4. your payoff was right, BUT U1 is in EUR, as U2, Strike and Notional are in NOK and not in EUR as mentioned. Follow the same path as 2. but this time you will get a price in NOK per spread.Hope it helpsJerome