[Limdep Nlogit List] Simultaneous probit model

[Richard Harris]

Dear all

Maddala sets out in his 1983 book (P246-247 Model 6. y1 and y2 both binary)
the 2-equation simultaneous probit model whereby endogenous dichotomous
variables are replaced by their predicted values (from reduced-form
probits). He also shows how to obtain corrected-standard errors in this
two-variable dichotomous model. Some earlier LIMDEP manuals have the code
needed for this model.

Is anyone aware of how to proceed with a system of three endogenous
dichotomous variables? How can the variance-covariance matrix be corrected
in this instance? 

Any help much appreciated.

Thanks

Richard Harris