When to use continuous compounding

[kenndanielso]

Hi David,

In the exam,
(1) Should I use the linear approximation of computing for forward rates ((R2T2 - R1T1)/(T2-T1)) or the more accurate way of (((1+R2)^T2)/((1+R1)^T1))^(1/(T2-T1)?
(2) When computing for the forward exchange rate, should I use continuous compounding (e^(rd-rf)) or non-continuous compounding ((1+rd)/(1+rf))?

Thanks!

 

[David Harper CFA FRM]

Hi kenndanielso,

I would love to be able to tell you which, I have consistently requested to GARP that they standardize on compound frequencies, for at least four years. The best we've got it a promise for that the question will be explicit, see the source email here from GARP's FRM Program Direct (Bill) @ http://forum.bionicturtle.com/threads/continuous-vs-non-continuous-and-rounding-issues.4743/

wrt (1), I did not realize continuous is an approximation of the more accurate? Isn't one merely continuous and the other merely annual discrete, i am not sure i understand why you say the latter is "accurate" except insofar as the underlying rates are expressed annually? Thanks,