This paper by J. M. C. Santos-Silva and Silvana Tenreyro extends the simulation results from their earlier work (Santos-Silva and Tenreyro, 2006) to consider data generated as a finite mixture of gamma variates. This type of data can naturally include a large proportion of zeros, which is relevant for models like the gravity equation. The authors confirm that the Poisson pseudo-maximum likelihood (PPML) estimator performs well in a wide range of scenarios, even when the conditional variance is not proportional to the conditional mean. The study also evaluates the performance of other estimators, such as the gamma pseudo-maximum likelihood (GPML) and various log-linearized models, finding that the PPML and GPML are generally more robust to departures from the implicit heteroskedasticity assumptions. The results support the use of the PPML estimator for estimating constant elasticity models, particularly in the context of gravity equations.This paper by J. M. C. Santos-Silva and Silvana Tenreyro extends the simulation results from their earlier work (Santos-Silva and Tenreyro, 2006) to consider data generated as a finite mixture of gamma variates. This type of data can naturally include a large proportion of zeros, which is relevant for models like the gravity equation. The authors confirm that the Poisson pseudo-maximum likelihood (PPML) estimator performs well in a wide range of scenarios, even when the conditional variance is not proportional to the conditional mean. The study also evaluates the performance of other estimators, such as the gamma pseudo-maximum likelihood (GPML) and various log-linearized models, finding that the PPML and GPML are generally more robust to departures from the implicit heteroskedasticity assumptions. The results support the use of the PPML estimator for estimating constant elasticity models, particularly in the context of gravity equations.