Search

In this paper, we propose a Bayesian estimation and prediction procedure for noncausal autoregressive (AR) models. Specifically, we derive the joint posterior density of the past and future errors and the parameters, which gives posterior predictive densities as a byproduct. We show that the posterior model probability provides a convenient model selection criterion and yields information on the probabilities of the alternative causal and noncausal specifications. This is particularly useful in assessing economic theories that imply either causal or purely noncausal dynamics. As an empirical application, we consider U.S. inflation dynamics. A purely noncausal AR model gets the strongest support, but there is also substantial evidence in favor of other noncausal AR models allowing for dependence on past inflation. Thus, although U.S. inflation dynamics seem to be dominated by expectations, the backward-looking component is not completely missing. Finally, the noncausal specifications seem to yield inflation forecasts which are superior to those from alternative models especially at longer forecast horizons.

We propose an estimation method of the new Keynesian Phillips curve (NKPC) based on a univariate noncausal autoregressive model for the inflation rate. By construction, our approach avoids a number of problems related to the GMM estimation of the NKPC. We estimate the hybrid NKPC with quarterly U.S. data (1955:1–2010:3), and both expected future inflation and lagged inflation are found important in determining the inflation rate, with the former clearly dominating. Moreover, inflation persistence turns out to be intrinsic rather than inherited from a persistent driving process.

We propose a noncausal autoregressive model with time-varying parameters, and apply it to U.S. postwar inflation. The model .fits the data well, and the results suggest that inflation persistence follows from future expectations. Persistence has declined in the early 1980.s and slightly increased again in the late 1990.s. Estimates of the new Keynesian Phillips curve indicate that current inflation also depends on past inflation although future expectations dominate. The implied trend inflation estimate evolves smoothly and is well aligned with survey expectations. There is evidence in favor of the variation of trend inflation following from the underlying marginal cost that drives inflation.

In this paper, we propose a simulation-based method for computing point and density forecasts for univariate noncausal and non-Gaussian autoregressive processes. Numerical methods are needed for forecasting such time series because the prediction problem is generally nonlinear and therefore no analytic solution is available. According to a limited simulation experiment, the use of a correct noncausal model can lead to substantial gains in forecast accuracy over the corresponding causal model. An empirical application to US inflation demonstrates the importance of allowing for noncausality in improving point and density forecasts.