RE: [Quantlib-dev] decreasing discount factors

[Luigi B. <lui...@fa...>]

At 03:09 PM 11/8/02 +0100, Ferdinando Ametrano wrote:
 > Negative forward rates would be an arbitrage opportunity.

Not if the two spot rates from which you bootstrap the curve are both on 
the bid side or both on the ask side.

>Which leads me back to my original argument. Marco, what kind of benefit 
>are you getting out of the "relaxed" constraints? What are you using the 
>discount curve for?

Why, to represent market data, of course :)
A customer using our libraries stumbled upon a legitimate case such as the 
one Andre' presented. They were including overnight rates, and the 
impedence mismatch between the very-short and the short rate region caused 
a hump on the discounts---a very real one, but not an arbitrage 
opportunity, mind you, because they were on the same side. Combining a bid 
and an ask side would not give increasing discounts, of course.

Bye,
         Luigi