September 8th, 2006, 2:33 am
I thought I should outline exactly what the model does since I made the comment that it replicates the results of Gibson's paper.It replicates the tranche spreads in Table 2 (p. 10). i.e. 1507 for Equity, 315 for Mez and 7 for Senior.It does this by monte-carlo simulation, rather than a one-factor model - which obviously gives the same result...It is all coded up in VBA and is therefore not terribly efficient. My Pentium M notebook does about 900 simulations in 1 minute, however my Centrino is about 3 times faster (2 years older than my Pentium M - go figure..). It does the default time sorting using the built in excel sort function. If this was replaced with an in-code quick sort it should make it much quicker, but when I was building it I didn't have time.I might code it up in C/C++ in the future if I have time.I currently don't have access to my email, but will send it to all who have asked when I do.If anyone who I send my model to has any Credit Deriv / Interest Rate Models (esp. HW1F/HW2F or BGM) or any other models, I'd be very appreciative if you could shoot them through to me.CheersMac