Aggregate Dispersion in Economists' Opinion on Macroeconomic Forecasts

I construct a novel measure of aggregate dispersion in economists' opinion on macroeconomic forecasts. The measure aggregates dispersion across relevant economic releases over one week and is used to predict stock returns in the subsequent week. I provide evidence that higher aggregate dispersion predicts lower future stock market returns by 27 basis points in the subsequent week. The impact is both statistically significant and economically large. My evidence does not support that aggregate dispersion is a risk measure and supports the theoretical explanations that heterogeneous beliefs lead to lower future stock returns