On the empirical separability of news shocks and sunspots

Abstract

In this note we discuss the possibility of empirically evaluating the relative importance of different drivers of forecast errors in linear rational expectations frameworks, using the predictions generated by the theory. By means of a few simple examples, we show that, when accounting for indeterminate equilibria, empirical difficulties are likely to arise in distinguishing between determinate models driven by news shocks or rather by indeterminate ones under non-fundamental – or arbitrarily related to fundamentals – sunspot noise.