Maximum likelihood in the frequency domain: the importance of time-to-plan

We illustrate the use of various frequency-domain tools for estimating and testing dynamic, stochastic general equilibrium models. We show how to exploit the well-known fact that the log, Gaussian density function has a linear decomposition in the frequency domain. We also propose a new resolution to the problem that the phase angle between two variables is not uniquely determined. These methods are applied to the analysis of business cycles. Our substantive findings confirm existing results in the literature, which suggest that time-to-plan in the investment technology has a potentially useful role to play in business cycle analysis. (C) 2003 Published by Elsevier Science B.V.