ESTIMATION AND INFERENCE IN THE LINEAR-QUADRATIC INVENTORY MODEL

We summarize some recent work of ours on estimation and hypothesis tes ting on the parameters of the linear-quadratic inventory model. For so me data-generating processes calibrated to estimates from some existin g studies, this work uses (1) asymptotic theory to compare alternative estimators on the basis of the asymptotic efficiency of parameter est imates, (2) asymptotic theory and simulations to consider how likely o ne will be to get sharp estimates of the parameters of the model, and (3) simulations to see how accurately sized are hypothesis tests about the parameters of the model.