Uncertainty of Macroeconomic Forecasters and the Prediction of Stock Market Bubbles

Abstract

We assess the contribution of macroeconomic uncertainty - approximated by the dispersion of the real GDP survey forecasts - to the ex post and ex ante prediction of stock price bubbles. For a panel of six OECD economies covering 24 years, two alternative binary chronologies of bubble periods are determined and subjected to panel logit regressions conditioning on macroeconomic indicators and expectation uncertainty. Measures of macroeconomic uncertainty improve the ex ante signalling of stock price booms and bubbles.