Tell me more ×
Cross Validated is a question and answer site for statisticians, data analysts, data miners and data visualization experts. It's 100% free, no registration required.
up vote 0 down vote favorite

Can someone please help me with this? I need to do a regression in order to estimate the alpha's in this following equation, but I can't seem to get this equation to be in a format that i could use regression with.Ii was just looking to simplify this equation so that I can get something in the form of Y = constant + alpha0 * something + alpha1 * something .. etc so that the regression coefficients will be my alphas.

Any advice would be greatly appreciated, thanks a lot! D(T) stands for the balance at time T, and R(T) stands for the interest rate at time T.

D(T) = D(0) * exp[alpha0 * T + alpha 1 * T^2 / 2 + Alpha 3 * (R(T) - R(0))] * ((1 / df(T)) ^ Alpha3

where df(T) = exp(summation from i to N of - R(TI) * dTI) (so df(T) is basically a constant number)

share|improve this question
Actually I think the ending part of the equation ((1+df(t))^alpha3 is not part of the exp() part of the equation. Does that make sense? Would there then be a way of simplifying for regression? Thanks a lot! – alex Dec 22 '11 at 0:07
alex, I've merged your two accounts, but you we'll need to register once for all to get system-wide notifications, be able to comment, update your question, etc. – chl♦ Dec 22 '11 at 0:13

1 Answer

up vote 1 down vote

I am assuming that the "Alpha 3" inside the exp[...] part of your formula should be an "Alpha 2", as otherwise you are missing an "Alpha 2" and have two "Alpha 3"s.

Just take the logs of both sides; simplest to divide both sides by D(0) first:

$\log (D_t/D_0) = \alpha_0T + \alpha_1 T^2/2 + \alpha_2 (R_t-R_0) + \alpha_3\log(1/df_t)$

share|improve this answer
thanks so much! – alex Dec 21 '11 at 23:44

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.