An analysis of private investors’ stock market return forecasts

Abstract

The study analyses data on stock index forecasts made by private investors. The implied returns calculated from these forecasts exhibit negative skewness and excess kurtosis. Past returns have a positive impact on the implied returns, consistent with investors expecting positive momentum. Females are less optimistic than males, but their forecasts have higher standard deviation. Consistent with the weekend effect, implied returns from estimates entered on weekends are significantly lower than those entered on weekdays. Implied returns are not consistently related to the weather conditions on the day the forecast was made.

Acknowledgements

I thank comdirect bank AG for providing the data. I further thank Mathias Binswanger, Markus Nöth, Dirk Schiereck, Heinz Zimmermann and participants of the 7th conference of the Swiss Society for Financial Market Research and the 2004 annual meeting of the European Financial Management Association for valuable comments.

Notes

¹ Besides, there is a quite substantial literature analysing expectations implicit in survey data. See, for example, Frankel and Froot (1987) and Ito (1990) for research on exchange rate expectations and Lovell (1986) for a survey of other related research. Other papers analyse sentiment measures that are not derived from survey data. For an example see Wang (2003) who analyses a sentiment indicator derived from positions in the futures market.

² The record indicated that the estimate was entered in the year 2010.

³ The test result is insensitive to the way the data are grouped into bins. The table shows the result obtained when making the number of observations (both female and male) equal across bins. Using equidistant classes instead also leads to a clear rejection of the null hypothesis.

⁴ Huberman (2001) argues that this may be one cause for the home bias in equity investments.

⁵ Krämer and Runde (1997) replicate Saunders' analysis using data from the German stock market. They conclude that no systematic relationship between stock returns and the local weather in Frankfurt exists.

⁶ The cities are Berlin, Dresden, Düsseldorf, Frankfurt, Hamburg, Karlsruhe, Munich and Stuttgart.

⁷ They suggest that market makers or news providers might cause the weather effect.