Does Past Volatility Affect Investors' Price Forecasts and Confidence Judgments?

Abstract

This study investigates the influence of past volatility on individual investors' forecasting behavior. We conducted two experiments in which we used real stock prices to construct low and high volatility time series, and asked participants to make both point estimates and interval forecasts of future values. We focus on two main aspects of investors' future expectations: (1) forecasts of future price, and (2) subjective confidence of future stock price, which includes 50%, 70% and 90% confidence intervals. Our results show that past volatility has a weak effect on future forecasts that are sensitive to minor changes in the characteristics of price series. We found strong evidence that low past volatility increases participants' confidence and improves forecast accuracy. The calibration of the confidence intervals was not affected by the stocks volatility. However, most confidence intervals were skewed suggesting that participants use asymmetric confidence intervals to hedge their price forecasts.

Keywords: Calibration, Judgmental forecasting, Overconfidence, Investment, Volatility

JEL Classification: D81, D84, D91, G10, G30, M40