A METHOD OF SIMULATED MOMENTS FOR ESTIMATION OF DISCRETE RESPONSE MODELS WITHOUT NUMERICAL INTEGRATION

摘要: This paper proposes a simple modification of conventional method moments estimator for discrete response model, replacing probabilities that require numerical integration with estimators obtained by Monte Carlo simulation. simulated (MSM) does not precise estimates these consistency and asymptotic normality, relying instead on the law large numbers operating across observations to control simulation error, hence can use simulations practical size. The is useful models such as high-dimensional multinomial probit (MNP), where computation has restricted applications.