A new estimator of the Box-Cox transformation model using moment conditions

The maximum likelihood estimator (MLE) under the normality assumption of error terms is widely used to estimate the Box-Cox transformation model. However, since the error terms cannot be normally distributed, it is not a proper estimator. In other words, the estimator is inconsistent. In this paper, I propose a new estimator of the Box-Cox transformation model that modifies the MLE in key ways. I demonstrate that the estimator is consistent, and that an asymptotic distribution is obtained. The results of Monte Carlo experiments are also presented.