Statistical estimation and moment evaluation of a stochastic growth model with asset market restrictions

Abstract

This paper estimates the parameters of a stochastic growth model with asset market and contrasts the model’s moments with moments of the actual data. We solve the model through log-linearization along the line of Campbell (1994) [Journal of Monetary Economics 33(3), 463] and estimate the model without and with asset pricing restrictions. As asset pricing restrictions we employ the riskfree interest rate and the Sharpe-ratio. To estimate the parameters we employ, as in Semmler and Gong (1996a) [Journal of Economics Behavior and Organization 30, 301], a ML estimation. The estimation is conducted through the simulated annealing. We introduce a diagnostic procedure which is closely related to Watson (1993) [Journal of Political Economy 101(6), 1011] and Diebold et al. (1995) [Technical Working Paper No. 174, National Burea of Economic Research] to test whether the second moments of the actual macroeconomic time series data are matched by the model’s time series. Several models are explored. The overall results are that sensible parameter estimates may be obtained when the actual and computed riskfree rate is included in the moments to be matched. The attempt, however, to include the Sharpe-ratio as restriction in the estimation does not produce sensible estimates. The paper thus shows, by employing statistical estimation techniques, that the baseline real business cycle (RBC) model is not likely to give correct predictions on asset market pricing when parameters are estimated from actual time series data.