Excess Volatility in Professional Stock Return Forecasts

Consensus professional forecasts of stock returns are three times more volatile than those of non-professionals and econometricians. This "excess" volatility in professional forecasts is not due to noise. Rather, professional forecasts respond immediately, strongly, and countercyclically to macro shocks. Business cycle, technology, and oil supply shocks each account for 20% to 40% of the total variation in the difference between professional and other forecasts. We also replicate prior work’s finding that professional forecasts of GDP growth are no more responsive to macro shocks than those of non-professionals and econometricians. We argue that professionals’ forecasts of stock returns respond differently from (i) other forecasts of stock returns and (ii) professional forecasts of macro variables, because they are better able to identify discount rate shocks. To this end, we run a two-stage test using standard discount rate proxies. In the first stage, we identify "shock-sensitive" proxies as those for which macro shocks are strong instruments. In the second stage, the shock-sensitive proxies, and particular the part instrumented using macro shocks, strongly predict professional forecasts. In fact, we show that this predictive relation is both qualitatively and quantitatively consistent with ex post realized returns. We conclude that professionals are unique in correctly identifying discount rate fluctuations from macro shocks. In contrast to us, a large literature in macro-finance finds that consensus forecasts of professionals are no better than other forecasts