Mapping of bespokes CDO using Base Correlation - the Square Root Moneyness Approach

Mida · March 23rd, 2007, 10:40 am

b = bespoke pooli = index poolrho = base correlationL = k = detachment point of the bespoke poolr = scaling factorThe above formulae explains the Moneyness Approach in the mapping of bespoke tranches. Here, r is set to 1.However we find when r = 0.5, we find that the bespoke CDO price somehow gives prices closer to market.1. Would you agree with the above?2. Would anyone know why when r=0.5, we get better results? Anything to do with dispersion of the pool you think?Thanksmida

smile · March 26th, 2007, 2:03 am

It is the simple EL mapping with a scaling factor for 'sticky prices'... I think grandularity of the portfolio could also be a factor.

Money · March 26th, 2007, 12:07 pm

I saw the presenatino slides from Gregory.Don't quite understand what does it mean by sticky ? what's the advantage of having power to the "r" ?

sadsack2 · March 27th, 2007, 1:13 am

sadsack2 · March 27th, 2007, 1:16 am

sorry for the blank post ---- i'd say that the word 'sticky' just comes from the fact that for r<1 you don't quite scale your correlations linearly, i.e., you stick closer to the index correlation. That is, if the bespoke pool has 2x the EL of the index, then you don't double your detachment point but you multiply it by 1.414 (say if r=0.5). Sounds fine, tho I haven't checked if it 'matches prices better', it probably depends on the dealer and if they use this approach. I haven't seen the presentation --- can someone post it?

Sanskript · March 29th, 2007, 3:17 pm

Can anyone please post the presentation?Was it by Jon Gregory?

smile · March 30th, 2007, 3:12 am

I think my version was by Jeff Meli.

Sanskript · April 2nd, 2007, 8:07 am

Can you post the whole presentation please?