MLE Specification for a Binomial Density Function

[rubberbob]

Good day Gareth,

I am trying to specify the MLE for a binomial density function of the following form (please see attachment):

in :eviews6:

where the betas are the parameters of the explanatory variables and rho is the correlation between the asset returns for two borrowers which are both random and normally distributed.

Below is some of the program that I began to write already


logl AGG_DF
AGG_DF.append @logl lklhd
AGG_DF.append res=(df-c(1)-c(2)*ir(-3)-c(3)*gdp(-3)-c(4)*infl(-2))
AGG_DF.append rho = c(5)
AGG_DF.append lklhd = @sum log(@dbinom((((res-sqrt(rho)*latent)/@sqrt(1-rho))*(((1-(res-sqrt(rho)*latent)/@sqrt(1-rho)))),acc,ba))

equation adft.ls df c ir gdp infl latent roa

!m = @ncoef
for !k = 1 to !m
c(!k) = (adft.c(!k)
next
c(5) = (adft.@cor)

AGG_DF.ml(showopts)

Issues:

(1) How to specify the OLS (adft.ls) to get the correlation for just two variables (asset returns i.e. latent and roa) so that rho is accounted for in the likelihood function?
(2) What am I doing wrong in terms of the specified likelihood function for the binomial density function above?