We introduce reprojection as a general purpose technique for characterizing the dynamic response of a partially observed nonlinear system to its observable history. Reprojection is the third step of a procedure wherein first data are summarized by projection onto a Hermite series representation of the unconstrained transition density for observables; second, system parameters are estimated by minimum chi-squared, where the chi-squared criterion is a quadratic form in the expected score of the projection; and third, the constraints on dynamics implied by the nonlinear system are imposed by projecting a long simulation of the estimated system onto a Hermite series representation of the constrained transition density for observables. The constrained transition density can be used to study the response of the system to its observable history. We utilize the technique to assess the dynamics of several diffusion models for the short-term interest rate that have been proposed and to compare them to a new model that has feedback from the interest rate into both the drift and diffusion coefficients of a volatility equation.
We derive a seminonparametric utility function containing the constant relative risk aversion (CRRA) function as a special case, and we estimate the associated Euler equations with U.S. consumption data. There is strong evidence that the CRRA function is misspecified. The correctly specified function includes lagged effects of durable goods and perhaps nondurable goods, is bounded as required by Arrow's Utility Boundedness Theorem, and has a positive rate of time preference. Constraining sample periods and separability structure to be consistent with the generalized axiom of revealed preference affects estimation results substantially. Using Divisia aggregates instead of the NIPA aggregates also affects results.
We consider the quasi-maximum likelihood estimator (qmle) obtained by replacing each transition density in the correct likelihood for a non-Markovian, stationary process by a transition density with a fixed number of lags. This estimator is of interest because it is asymptotically equivalent to the efficient method of moments estimator as typically implemented in dynamic macro and finance applications. We show that the standard regularity conditions of quasi-maximum likelihood imply that a score vector defined over the infinite past exists. We verify that the existence of a score on the infinite past implies that the asymptotic variance of the finite lag qmle tends to the asymptotic variance of the maximum likelihood estimator as the number of lags tends to infinity.