[R] Bootstrapping Standard errors of nonlinear GMM obtained from BFGS (Optimx)

[Olu Ola]

Hello,
I am running a 2 equation system of nonlinear GMM using BFGS in optimx. Using the conventional way of calculating the standard error of the estimates gives NAN's for some of the standard errors. As a result, I want to bootstrap the standard error. 

A way forward on how to bootsrap the standard errors will be greatly appreciated.