Dear all,

I completeley agree with Ben that Treasuries are the most liquid and reliable 
benchmarks. I also think that FMAs have their own advantages and 
disadvantages.

However, it looks like it would be very difficult for us to avoid using LIBOR 
as benchmark - at least to price anything related to Capital Markets (for 
example, credit default swaps) . Our prices will always be different from 
that on Capital Markets if we use benchmarks different from the Market.

As a half-way solution we can have different benchmarks - for example, use 
anything, that we decide is closest to risk-free, to value products 
significantly different from that priced on Capital markets,and LIBOR for 
everything priced on Capital Markets .

 We can adjust by manipulating other coefficients (event, liquidity, 
recovery) for our purposes, but it looks like benchmark should be LIBOR.

To avoid negative default probabilites (for highly rated reference entities) 
we can add cost fo repo and haircuts (as all bankers do).

Let us continue our discussion on this.

Slava





Benjamin Parsons
13/03/2000 20:07
To: Bryan Seyfried/LON/ECT@ECT, Martin McDermott/LON/ECT@ECT, Viacheslav 
Danilov/LON/ECT@ECT, Stuart Ffoulkes/LON/ECT@ECT, Eklavya Sareen/LON/ECT@ECT, 
Tomas Valnek/LON/ECT@ECT, John Metzler/LON/ECT@ECT, Simon Brooks/LON/ECT@ECT, 
Konstantin Malinovski/LON/ECT@ECT
cc: Vasant Shanbhogue/HOU/ECT@ECT, Vince J Kaminski/HOU/ECT@ECT, Steven 
Leppard/LON/ECT@ECT, Jitendra Patel/LON/ECT@ECT, William S 
Bradford/HOU/ECT@ECT, Harry Arora/HOU/ECT@ECT 

Subject: Choosing the Riskfree Rate

Dear all,

Following on from the discussions had with Vince et al a few weeks ago, 
regarding the problems inherent in using Treasuries as our benchmark riskfree 
curve, I've written the attached short document. My preliminary conclusions 
are as follows:

Recent turmoil in the Treasury market mostly hit 10 and 30-year bonds.
Treasuries are still the most liquid curve, and that used by banks for 
benchmarking risky debt.
There is a considerable a liquidity/scarcity premium inherent within the 
corporate-Treasury spread, which we are not taking into account. More 
accurate measurement of this premium should allow us to calculate bankruptcy 
prices based on Treasuries, which are consistent with the market.

All further comments are welcome of course.

Regards

Ben