Need some help with estimation a CCAPM model

[terance]

Well, I have 10 assets, and though I have 10 equations
sample eq:

c(i1)*(r1*((cons(1)/const)^c(2i))=1

where c(i1) and c(i2) are the parameters, which are to be obtained using General Method of Moments.

The data is provided by this site: http://www.econ.kuleuven.be/gme/ (pricing.wf1)

When I start to estimate this parameters I have the same error: matrix is near to singular. How can I avoid this problem?May be I have to choose another start point to calucalate this parametrs using gmm?

[startz]

Yes.

Try starting the parameters at reasonable (and non-zero) values and see what happens.

[terance]

nothing changed...I've tried 20 diiferent start-points, but none helped me. May be any analytical ideas should help with this problem?

[startz]

As a wild guess only, is there some chance the denominator should be cons rather than const? In other words, do you have consumption growth specified correctly?

[terance]

may be i should use something another, to calculate theese parametres?

[terance]

As a wild guess only, is there some chance the denominator should be cons rather than const? In other words, do you have consumption growth specified correctly?

it is okay with consumption=)

[trubador]

First of all, I think the consumption data (CONS) is already defined in growth terms, so you do not have to transform it again. Second, please keep in mind that for the GMM estimator to be identified, there must be at least as many instruments as there are parameters in the model.