Estimation and Inference by Stochastic Optimization: Three Examples

This paper illustrates two algorithms designed in Forneron and Ng (2020): the resampled Newton-Raphson (rNR) quasi-Newton (rQN) algorithms, which speed up estimation bootstrap inference for structural models. An empirical application to BLP shows that computation time decreases from nearly five hours with standard just over one hour rNR only 15 minutes using rQN. A first Monte Carlo exercise accuracy of method a probit IV regression. second additionally statistical efficiency gains relative simulation-based dynamic panel regression example.