TESTING A CLASS OF SEMI- OR NONPARAMETRIC CONDITIONAL MOMENT RESTRICTION MODELS USING SERIES METHODS

This paper proposes a new test for class of conditional moment restrictions (CMRs) whose parameterization involves unknown, unrestricted expectation functions. Motivating examples such CMRs arise from models discrete choice under uncertainty including certain static games incomplete information. The proposed may be viewed as semi-/nonparametric extension the Bierens (1982, Journal Econometrics 20, 105–134) goodness-of-fit parametric model mean. Estimating expectations using series methods and employing Gaussian multiplier bootstrap to obtain critical values, is shown asymptotically correctly sized consistent. Simulation studies indicate good finite-sample properties. In an empirical application, used study validity game-theoretical discount store market entry, treating equilibrium beliefs nonparametric expectations. indicates that Walmart Kmart entry decisions do not result game information with linearly specified profits.