The Information in Low Forecasts

Abstract

This article introduces an intuitive, earnings-based sentiment indicator that is useful for forecasting overall market direction. The indicator aggregates the difference between the mean and the low forecasts on individual stocks, with the rationale being that the indicator captures for the entire market what dispersion captures for a single stock. The basic results are that when the aggregate low forecast is considerably lower than the aggregate mean forecast, subsequent forecast revisions tend to be more negative and future index returns tend to be lower and more volatile. However, when the aggregate low forecast is extremely low compared with the aggregate mean forecast, investor and analyst sentiments tend to be overly pessimistic, and the market thus tends to perform strongly.

Key Findings

• ▪ When the aggregate low forecast (i.e., the sum of the low forecasts on each individual stock) is considerably lower than the aggregate mean forecast (i.e., the sum of the mean forecasts on each individual stock), subsequent forecast revisions tend to be more negative and future index returns tend to be lower and more volatile.

• ▪ The mechanism of the aggregate low forecast signal for a portfolio of stocks is comparable to the mechanism of the dispersion signal for individual stocks; they both signal that consensus earnings forecasts are too high.

• ▪ When the aggregate low forecast is extremely low compared with the aggregate mean forecast, it provides a strong signal that investor and analyst sentiments are overly pessimistic and that the market is likely to perform strongly.