Page 1 of 1

Pricing Basket Options

Posted: July 20th, 2006, 12:31 pm
by JohnLaw
Is there anyone in charge to price options with underlying baskets of homogeneous or heterogeneous indices?Any help will be appreciate a lot!Thanks in advance,P.

Pricing Basket Options

Posted: July 20th, 2006, 2:45 pm
by DavidF
I worked a little bit on this. Can you formulate your question more precisely ?

Pricing Basket Options

Posted: July 24th, 2006, 1:32 pm
by JohnLaw
i am in charge to price exotic option with underlying a basket of indices which is computed as the weighted average. How you behave when you have to price this kind of basket when the index are not necesarrely correlated....how can you simulate the underlying to compute the payoffs ?thanksP.

Pricing Basket Options

Posted: July 24th, 2006, 1:40 pm
by DavidF
One way, if you are able to calibrate a local volatility model for each of the components of the basket, is to have a multi-dimensionnal Monte-Carlo simulation of all of them, with the correlations being estimated for example form historical data.

Pricing Basket Options

Posted: August 10th, 2006, 4:10 pm
by mrblue1978
There is also moment-matching approximation (ie sum of lognormals is lognormal + calibration of first moments). You can google 'moment-matching basket options' or check this page : http://www.fabiomercurio.it/papers.html

Pricing Basket Options

Posted: August 11th, 2006, 5:28 am
by WillK
QuoteOriginally posted by: DavidFwith the correlations being estimated for example form historical data.Which is the usual way of estimating correlation for pricing exotic option on baskets : is it a pure historical correlation and if yes, which observation window and which frequency ? Or is it some kind of percentile of historical correlation or more complex models ?

Pricing Basket Options

Posted: August 15th, 2006, 12:08 pm
by jokeoh
basket options typically priced as single-factor.use sensible correlations: e.g. for a 2Y ATM call on a basket on NKY, SPX, SX5E, use 2 years worth of data and 180-day weekly sampling.