Problem with estimation

[ElineP]

There might be a very simple explanation for the problem we are having but we can't find the answer.

We are trying to estimate (by OLS) the following model:

Mt = K(y) Yt + K(Tau) T - K(i) It + Sommation from s=-3 to 3 ( (K(y)^s)*d(y(t+s)) + (K(i)^s)*d(i(t+s)) ) + Et

where Et=residuals, K(y), K(Tau) and K(i) are the parameters to be estimated.

The problem is getting this formula into Eviews (Version 7). So far we have:

M = C(1)*Yt - C(2)*It + C(3)*T + C(1)^(-3)*d(Yt(-3)) + C(1)^(-2)*d(Yt(-2)) + C(1)^(-1)*d(Yt(-1)) + d(Yt) + C(1)*d(Yt(1)) + C(1)^(2)*d(Yt(2)) + C(1)^(3)*d(Yt(3)) + C(2)^(-3)*d(It(-3)) + C(2)^(-2)*d(It(-2)) + C(2)^(-1)*d(It(-1)) + d(It) + C(2)*d(It(1)) + C(2)^(2)*d(It(2)) + C(2)^(3)*d(It(3))

where C(1) = K(y), C(2) = K(i) and C(3)=K(Tau).

The results are clearly wrong and we were wondering whether or not this is because of our 'translation' into eviews is wrong....
We don't know how else to make the coefficients Ky and Ki reappear in the sommation. Suggestions?

Any help would be more than welcome since we have a deadline!

Thank you!

[trubador]

Your specification seems allright, try initializing coefficients (the "c" vector in your workfile) with different values.

[ElineP]

That's the problem, I'm not sure we need starting values. We need estimates for the parameters that stay the same no matter how many times we press specify equation.
Should I try to do it in some other way or is the equation above correct?
Thx