Estimation of Multi-Market Fix-Price Models: An Application of Pseudo Maximum Likelihood Methods
提出用蒙特卡洛模拟计算内生变量矩、再最大化伪似然函数来估计多市场固定价格模型，并以一个宏观经济模型为例验证该方法与全信息极大似然法结果高度一致，且成本更低。
The past decade has seen the econometric implementation of macroeconomic multi-market fix-price models for a number of European countries. The procedure in use, the full information maximum likelihood (FIML) method, unfortunately becomes very cumbersome and seems out of reach when additional features are incorporated in the model (disaggregation into micro markets, opinion surveys,...). One purpose of the present work is to prove the fruitfulness of the following estimation strategy: use Monte-Carlo simulations to compute the first and second order moments of the endogenous variables, and maximize a resulting pseudo likelihood function to estimate the parameters. We first describe the PML method in the context of the so-called canonical disequilibrium model. We then apply it to a small aggregated macroeconomic model previously studied under FIML by Artus, Avouyi-Dovi, Laroque (1985), where we allow for a disaggregation into micro markets. The results we obtain are strikingly similar to theirs. This both demonstrates in this example the robustness of the FIML procedure to the introduction of micro markets and stresses the usefulness of the PML method in obtaining reliable estimates at a smaller cost.