Academic journal article Quarterly Journal of Business and Economics

Analyst Forecast Dispersion and Future Stock Return Volatility

Academic journal article Quarterly Journal of Business and Economics

Analyst Forecast Dispersion and Future Stock Return Volatility

Article excerpt

Introduction

Analysts' forecast dispersion refers to the disagreernent among analysts with regard to the expected earnings per share (EPS) of a given firm. It is a forward-looking variable that embeds analysts' expectations about the firm's future profitability. It is also used as a proxy variable for differences in investor-opinion for a given firm. Previous research has shown the usefulness and importance of forecast dispersion in forming profitable trading strategies. Ackert and Athanassakos (1997) report that a strategy of buying (selling) low (high) dispersion stocks at the beginning of the year produces positive abnormal returns. Dische (2002) shows that positive abnormal returns can be achieved by applying earnings momentum strategies to stocks with low analysts' forecast dispersion. Ang and Cicone (2002) find that buying low dispersion stocks and selling high dispersion stocks on June 30 of each year and holding the portfolio over the next twelve months generates significantly positive returns that are not related to size or book-to-market. Diether, Malloy, and Scherbina (2002) find that high dispersion stocks earn relatively lower future returns.

In this paper, we examine the relationship between analysts' forecast dispersion and future stock return volatility, using monthly data for a cross section of 160 U.S. firms over the period 1981-1996. Such a relationship, if it exists, would be of great importance for active portfolio management, option pricing, and arbitrage trading strategies. For example, mutual fund managers can use volatility forecasts based on analyst dispersions to perform efficient volatility-timing strategies. Similarly, option investors can buy (sell) straddles whenever volatility forecasts based on analysts' forecast dispersion generate straddle prices that are higher (lower) than corresponding market prices. While much of previous work on analysts' forecast dispersion and stock return volatility has studied periods of earnings announcements, we look at such a relationship on an on-going basis, without reference to formal accounting events and disclosures.

Earlier research has also examined the relationship between stock return volatility and forecast dispersion. Abarbanell, Lanen, and Verrechia (1995) present a rational expectations model, which predicts that analysts' forecast dispersion and the ex-ante variance of price changes should be positively related during earnings announcement periods. They indicate that high forecast dispersion implies low forecast precision, which in turn suggests reduced public information and greater uncertainty in the market at the time of the earnings announcement. Early empirical studies by Ajinkya and Girl (1985) and Daley, Senkow, and Vigeland (1988) find that the ex-ante variability of stock returns around earnings announcements (obtained as implied volatility from option prices) is positively related to analysts' forecast dispersion. Daley et al. (1988), in particular, find a positive relationship between dispersions and average implied volatilities of options maturing after the earnings announcement dates. More recently, Lobo and Tung (2002) also find a strong and positive relationship between future stock price volatility and analyst forecast dispersion in periods surrounding quarterly earnings announcements for the sample period 1987-1990. They find that firms with high forecast dispersion experience high price variability over a longer time window surrounding earnings announcements compared to firms with low dispersion. Their findings support the theoretical predictions of Abarbanell et al. (1995).

At the same time, a large body of literature has examined the properties of financial analysts' EPS forecasts and the analysts' incentives to issue optimistic forecasts (Ali, Klein, and Rosenfeld, 1992; De Bondt and Thaler, 1990; and Diether et al., 2002). Ackert and Athanassakos (1997, 2003) show that analyst optimism and uncertainty are positively related. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.