Plainx - You were actually right in the understanding as to what I was saying - you do receive the % of the notional upfront + 500 bps running. This is how the market is quoted so as to not have market quotes in the 1,000s bps and higher. Slym is right in the conversion of an upfront only/upfront and bps running and/or just bps running. ALbeit realize that you will (most like;y) not get the same executin Px as you would find through this calculation as you will be charged a liquidity premium. Risky duration is a combination of Discount factors, survival prob. and Day count. Ie. SUMDis. Factor(ti) x Expected Survival(t) x Day Count (t) = Risky Durationi assume you knwo how to calculate these components/?

Waiter do you know of any papers that cover Risky Duration? I've tried searching but haven't found much. When you say factor what is this please? Sorry if the question is very simple but I am new to CDOs.I have tried to emulate Gibson's results (2004) and get the following results (for 200,000 simulations - thus these are pretty stable):via Merton style model:0-3% 1590bps3-10% 320bps10-100% 7.5bpsvia Li Intensity model:0-3% 1556bps3-10% 319bps10-100% 6.2bpsThe results in Gibson's paper are:0-3% 1507bps3-10% 315bps10-100% 7bpsI am concerned about the difference between the equity tranches. Could these differences be explained by model differences, or should I be getting results much closer to Gibsons?

send me your email - I can send you some info. Can't post the formulas to the message board for some reason. Regards,Chris

Just to clarify this...On Equity tranche you get upfront% of notional + 500bps running on the tranche notional...1.Does the 500bps decrease with the tranche notional? 2.And why would the quoted upfront % decrease with losses? If the tranche is more likely to get hit, intuitively it should stay the same or increase if anything?Rgds,Ren.

renikus1.Does the 500bps decrease with the tranche notional? I believe the 500bps is fixed.2.And why would the quoted upfront % decrease with losses? If the tranche is more likely to get hit, intuitively it should stay the same or increase if anything?The quoted upfront should increase with losses. For example if the market spread is 40bps the equity might be 27% but when the market is 47bps the equity is 38%.

Double Post

The spread doesn't decrease for equity - it stays a constant 500 for the IG indexes. The equitys are quoted in percentage. Assuming a $1 billion portfolio, 38% upfront fee will result in the transfer of $11.4mm a few (usually three) days after the trade has settled.

Yes the tranche notional is 3% of the 1 billion and 38% of that is 11.4million.Does anyone have any comments upon the my results?via Merton style model:0-3% 1590bps3-10% 320bps10-100% 7.5bpsvia Li Intensity model:0-3% 1556bps3-10% 319bps10-100% 6.2bpsThe results in Gibson's paper are:0-3% 1507bps3-10% 315bps10-100% 7bps

All,one question still remains... we've established that the 500bps is constant, however, does it stop getting paid after the TOTAL tranche is wiped out? I am thinking from a pricing perspective... is it simply a case of taking the 500bps cash flow as given and then set the upfront % such that the tranche equals par? Is there anyone that would perhaps have the time to go over my Matlab code? It is very well annotated.Regards,Ren.

Yes the 500bps is stopped when the tranche is wiped out.The 500bps is only paid on the remaining principal of the tranche.

I think i have not been clear..I realise that the 500bps is a fixed premium throughout the contract, my question was on what NOTIONAL the 500bps is paid... from all of the below input, i think it is safe to say that the 500bps premium is fixed and paid on the REMAINING tranche notional, i.e. so the premium payment decreases with tranche loss as it is being paid on a lower notional?R.

In case no-one has answered this yet, the answer is A).Premium=[(Sum of all loss payments per each simulation) / (NSims)]/[(Sum of all premium leg payments per simulation) / (Nsims)]This can obviously be simplified, but I prefer to write it this way as it can be intuitively understood as "MC expectation" of each leg.