Hi, given I would like to price this:3-month XAU-call / USD-put, strike 1,300I would need to plug-in the corresponding implied volatility to calculate the option price (premium). But how do I know the respective delta of this strike 1,300 and then look-up the correct vol to use?I'm a newbie for this...thanks in advance.

You mean: you have a delta-vol mapping and you'd like to know the implied vol for a strike of 1.30? There isan algorithm for that in Wystup's book.

Actually I have the volatility surface of XAU/USD in Reuters but I don't know which vol to use in order to price the option...(omitted the notional size in my original thread)3-month XAU-call / USD-put, XAU 1,000 oz, strike 1,300

how is your surface defined ? Is it by strike or by delta ?

dupe

The surface is defined by tenor and delta...that one shown in Reuters as "XAUVOLSURF"

1. Convert delta to strike (see Leonidas' post).2. Interpolate.M.

Sorry would you pls recommend the book and the specific method to covert delta into strike? thanks

delta = exp(-q*T)*N(d) for spot (change this appropriate for forward)your take the inverse of the standard normal (NORMSINV in Excel)d = N^-1(delta * exp(qT))now d = (log(F/K) + .5*vol*vol*T)/(vol*sqr(T))solve for K which you can write directly in terms of F, vol, T and dhth

Almost all there is to be known on FX smile can be found in this article. Very detailed regarding the different conventions and how it affects the smile construction.

Actually there are 2 ways the FX smile is quoted:- one is the delta-vol version as used by reuters (note that this version is usually the output of an atm, strangle, rr version)- the at-the-money, risk-reversal, strangle version as discussed in the paper referenced by PierreGYou obviously use reuters and see the delta-vol version. Yeah, you can transform each delta-vol pair as for example shown bydaveangel. And then interpolate in strike space. People do also interpolate in delta space and use an algorithm to get thestrike-vol version. However, you need to find out which delta they quote. Good luck with that! Don't trust anyone

I would suggest A. Castagna's book on FX options, where you have even more detailed explanation of issues related to various ways of quoting vols and construction of volatility surface.

QuoteOriginally posted by: daveangeldelta = exp(-q*T)*N(d) for spot (change this appropriate for forward)your take the inverse of the standard normal (NORMSINV in Excel)d = N^-1(delta * exp(qT))now d = (log(F/K) + .5*vol*vol*T)/(vol*sqr(T))solve for K which you can write directly in terms of F, vol, T and dhthHi, I've a doubt about your calculation; should the formula for d bed = (log(F/K) + (rate - div + .5*vol*vol)*T)/(vol*sqr(T))?Why did you ignore rate and dividend yield? Is there some convention that I ignore?Talking about fx option, rate and div should be domestic rate and foreign rate respectevely (e.g. for an option on EURUSD, dividend yiled=EUR rate) do you agree?thx

Quote Why did you ignore rate and dividend yield? Is there some convention that I ignore? because they are in the forward price, F

Yes, of course