Moment matching versus Bayesian estimation : backward-looking behaviour in the new-Keynesian three-equations model

The paper considers an elementary New-Keynesian three-equations model and contrasts its Bayesian estimation with the results from the method of moments (MM), which seeks to match the model-generated second moments of inflation, output and the interest rate to their empirical counterparts. Special emphasis is placed on the degree of backward-looking behaviour in the Phillips curve. While, in line with much of the literature, it only plays a marginal role in the Bayesian estimations, MM yields values of the price indexation parameter close to or even at its maximal value of one. These results are worth noticing since the matching thus achieved is entirely satisfactory. The matching of some special (and even better) versions of the model is econometrically evaluated by a model comparison test. -- Inflation persistence ; autocovariance profiles ; goodness-of-fit ; model comparison