Originally
posted by
martian:
(-1)*oldPV*(D/(1+oldIR))*(change in IR) = change in PV
well to be fair that does sort of work but it really I'm looking for.
Discount factor = PV/nominal value
if I have a duration D and an interest rate i then how does the discount rate change if I change i. I'm looking for a log formula which I can't remember:(.
The above sort of works because if we assume a nominal value of 1 then "PV" in that equation works and you get an answer. The reason why I want to avoid it is I'm looking at changes in i involving 100-200 bps so the above is less than ideal. The one I'm thinking of is closer.
Hmm can't say I remember anything with a log that would work with the variables you gave.. but the thing is, for large changes (100-200bps) duration doesn't work. Duration by definition assumes a linear relationship between IR and PV whereas like you implied it's not linear. Usually you solve this problem with the convexity adjustment but I can't see how you would calculate convexity with the variables you gave.
But more than likely I'm just forgetting something obvious. Sorry :P