Serving the Quantitative Finance Community

 
User avatar
khamsing
Topic Author
Posts: 0
Joined: May 12th, 2005, 7:13 am

Path dependent option pricing with MC

March 27th, 2006, 10:51 am

Hi,I am facing a problem trying to implement model for the pricing of path dependent basket option (Lookback type, Anapurna, Emerald...) with payoff depending on the performance of several (but not all) assets in in initial basket with caps and floors applied to the performance of each assets. The result I obtain vary significantly with the number of step I choose (daily, monthly or yearly).Is there a "right" number of steps to use? How do we pick it? Many thanks for your help!!!!
 
User avatar
figaro
Posts: 7
Joined: October 3rd, 2005, 5:49 pm

Path dependent option pricing with MC

March 27th, 2006, 1:02 pm

You don't mean caps and floors - you mean barriers. Look up "brownian bridges".
 
User avatar
ppauper
Posts: 11729
Joined: November 15th, 2001, 1:29 pm

Path dependent option pricing with MC

March 27th, 2006, 1:21 pm

QuoteOriginally posted by: figaroYou don't mean caps and floors - you mean barriers. Look up "brownian bridges".not the original poster but he may have meant caps and floors.If for example the option pays the average price of the option, but for the purposes of the average any daily price over $100 is treated as $100, I'd call that a cap not a barrier.
 
User avatar
khamsing
Topic Author
Posts: 0
Joined: May 12th, 2005, 7:13 am

Path dependent option pricing with MC

March 28th, 2006, 6:23 am

Perhaps it will be more understandable with an example. Let say I have basket option with the following features: - Maturity 10 years - Underlying: 20 shares - Payoff = 200% of the initial investment if at each yearly observation date (10 observations) none of the shares has been observed below 50% of its initial strike price (discrete barrier) - Payoff = performance of the basket at maturity equals the average performance of the individual shares included in the basket. Individual performance capped at 10%; i.e: if at maturity the share has a performance of 50%; we will only take the performance up to 10%.With or without barriers, my problem is to set a consistent timestep.Many thanks.
 
User avatar
figaro
Posts: 7
Joined: October 3rd, 2005, 5:49 pm

Path dependent option pricing with MC

March 28th, 2006, 6:43 am

performance = max( S_final - S_initial, 0) ? Is this a collection of call spreads with a discrete down-and-out cross knock out?Unless I seriously misunderstand what you are saying, all you need is to make sure your time grid includes the knock-out dates, and sample the curvature of the implied vol surface.
 
User avatar
Paolos
Posts: 1
Joined: November 12th, 2004, 2:15 pm

Path dependent option pricing with MC

March 28th, 2006, 6:46 am

khamsing, in this case you need only a number of timestep equal to the number of observations. Setting more timesteps doesn't have any impact on the result, it's just a waste of timeCheersP.