Stock price reaction to public and private information

https://doi.org/10.1016/j.jfineco.2005.07.011Get rights and content

Abstract

I use Easley and O’Hara's [1992, Journal of Finance 47, 577–604] private information-based trading variable, PIN, together with a comprehensive public news database to empirically measure the effect of private and public information on the post-announcement drift. I show that stocks associated with high PIN, consensus public news surprises, and low media coverage experience low or insignificant drift. Thus not all information acquisition variables have the same effect on the market's efficiency. Whether information is public or private is irrelevant; what matters is whether information is associated with the arrival rate of informed or uninformed traders.

Introduction

In an efficient market, security prices at any given time fully reflect all available information. A priori, there is good reason to believe that stock markets are efficient because such markets are paradigmatic examples of competition. Yet, rather than adjusting immediately to news surprises, stock prices tend to drift over time in the same direction as the initial surprise. This phenomenon is referred to as post-earnings announcement drift (PEAD) or earnings momentum. Previous research suggests three explanations for its existence and persistence.1 First, because announcements with unexpectedly high (low) earnings makes investing in these firms more (less) risky, the drift may be explained as a risk premium. Second, the persistence of this anomaly may be due to high transaction costs (limits of arbitrage). Third, the drift may be a function of the type of information agents receive. This paper focuses on the third explanation; prior literature shows that risk cannot account for the drift, and Korajczyk and Sadka (2003) examine the transaction costs explanation empirically and conclude that transaction costs do not fully explain price momentum.

The third explanation comprises two competing theories: the “behavioral” theory (Daniel et al., 1998) and the “rational structural uncertainty” theory (Brav and Heaton, 2002). The behavioral theory emphasizes the distinction between public and private information. In particular, Daniel, Hirshleifer, and Subrahmanyam predict that investors underreact to public information and overreact to private information. The rational structural uncertainty theory, on the other hand, argues that it is the distribution of information in the economy that matters. This latter theory predicts that a high (low) arrival rate of informed traders is associated with low (high) structural uncertainty and hence low (high) drifts.

In this paper I try to determine which of these two theories best describes the data by using a combination of variables that distinguish between public/private information and the arrival rate of informed/uninformed traders: PIN, the probability of private information-based trading, SUR, consensus public news surprises, and MEDIA, the number of days a particular firm is mentioned in the news.

I find that stocks associated with high PIN experience low or insignificant drift while stocks associated with high MEDIA experience significant drift. This evidence alone is not sufficient to determine which theory best describes the data because PIN can proxy for private information as well as informed trading while MEDIA can proxy for public information as well as uninformed trading. Rather, the key variable is SUR: it reflects public information, yet the results show that it is associated with informed trading and insignificant drift. That is, the results for SUR reveal that what matters is whether information is concentrated (informed) or diffuse (uninformed), not whether information is private or public.

In addition to exploring the role of information in explaining the post-earnings announcement drift, this paper also contributes to the understanding of what drives abnormal order flow (i.e., excess buying or selling pressure). I show that PIN, a function of abnormal order flow, is contemporaneously positively correlated with SUR. This finding goes against the standard assumption in the microstructure literature that public information, rather than affecting order flow, is directly incorporated into prices. However, it is possible that the private signals agents receive are triggered by public information that is not easily interpreted (see, e.g., Kim and Verrecchia, 1994, Kim and Verrecchia, 1997). In other words, PIN is not exclusively an insider trading measure as it also captures informed trading by investors who are particularly skillful in analyzing public news. This second interpretation further motivates recent studies (Brandt and Kavajecz, 2004; Green, 2004; Pasquariello and Vega, 2005; Evans and Lyons, 2002) that analyze the role of order flow as a proxy for private information-based trading in bond and foreign exchange markets. In these markets insider trading, by definition, does not exist, hence informed trading must refer to sophisticated agents who, given public news, are able to better predict future macroeconomic activity.

This paper is most closely related to two recent studies, Chan (2003) and Young (2004), who test Daniel's et al. (1998) model. Chan (2003) documents that firms covered by the media experience larger drift (underreaction) and firms that experience stock price jumps without observable public news (private information) experience returns reversal. He interprets his results to be consistent with Daniel's et al. (1998) model. Using the same news database as Chan (2003) employs, I also document that firms covered by the media experience larger drift. However, this paper goes a step further in linking media coverage to the arrival rate of uninformed traders and asks whether the distinction between private and public information that the behavioral theory emphasizes is relevant. In contrast to Chan (2003), I find that sometimes investors underreact to public announcements and sometimes public announcements make prices more informative. The key finding in this paper is that public announcements that generate underreaction are associated with the arrival rate of noise traders, while public announcements that make markets more efficient are associated with the arrival rate of informed traders.

Young (2004) focuses on the plausibility of the two behavioral assumptions in Daniel's et al. (1998) model, which are associated with the overreaction of investors to their private signals: overconfidence and biased self-attribution. The empirical evidence in his paper supports the overconfidence assumption, but it does not support the biased self-attribution assumption. In contrast, this paper is concerned with investors’ reactions to all types of information, not only with investors’ reactions to private information signals.

The remainder of the paper proceeds as follows. In Section 2, I describe the data and the different measures of public and private information. In Section 3, I present the empirical results, and I conclude in Section 4.

Section snippets

Data description and definition of variables

The empirical tests of this paper require data from six different sources: Center for Research in Securities Prices (CRSP), Compustat, Institute for the Study of Security Markets (ISSM), Trade and Quote (TAQ), I/B/E/S and Dow Jones Interactive. CRSP and COMPUSTAT data are used to compute abnormal returns, ISSM and TAQ data are used to estimate the private information variable, PIN, and I/B/E/S and Dow Jones Interactive are used to calculate the public news variables. First I explain how the

Empirical results

In this section, I characterize the effect of public and private information acquisition variables on earnings momentum. I start by computing earnings momentum returns for portfolios of stocks that are characterized by different degrees of public and private information at the time of formation. Then I estimate the relation between information acquisition variables and the arrival rate of informed traders, noise traders, uncertainty, and a proxy for dispersion of beliefs. Finally, I estimate

Concluding remarks

The goal of this paper is to deepen our understanding on how private and public information received by agents prior to earnings announcements affects the post-earnings announcement drift. The details of the linkage are intriguing, as not all information acquisition variables have the same effect on drift. The differences in information acquisition are not as simple as private information having a different effect on stock prices than public information, as Daniel et al. (1998) conjecture.

References (54)

  • E. Odders-White

    On the occurrence and consequences of inaccurate trade classification

    Journal of Financial Markets

    (2000)
  • A.R. Admati et al.

    A theory of intraday patterns: volume and price variability

    Review of Financial Studies

    (1988)
  • A. Atkins et al.

    Market structure and reported trading volume: Nasdaq versus the NYSE

    Journal of Financial Research

    (1997)
  • R. Ball et al.

    An empirical evaluation of accounting income numbers

    Journal of Accounting Research

    (1968)
  • V.L. Bernard et al.

    Post-earnings announcement drift: delayed price response or risk premium?

    Journal of Accounting Research

    (1989)
  • M. Brandt et al.

    Price discovery in the U.S

    Treasury market: the impact of order flow and liquidity on the yield curve. Journal of Finance

    (2004)
  • A. Brav et al.

    Competing theories of financial anomalies

    Review of Financial Studies

    (2002)
  • L.K. Chan et al.

    Momentum strategies

    Journal of Finance

    (1996)
  • Chordia, T., Shivakumar, L., 2005. Earnings and price momentum. Journal of Financial Economics,...
  • K.D. Daniel et al.

    Investor psychology and security market under- and over-reactions

    Journal of Finance

    (1998)
  • K.D. Daniel et al.

    Evidence on the characteristics of cross-sectional variation in common stock returns

    Journal of Finance

    (1997)
  • J.B. De Long et al.

    Noise trader risk in financial markets

    Journal of Political Economy

    (1990)
  • D.W. Diamond et al.

    Liquidity and the cost of capital

    Journal of Finance

    (1991)
  • K.B. Diether et al.

    Differences of opinion and the cross section of stock returns

    Journal of Finance

    (2002)
  • E. Dyl et al.

    Market structure and trading volume

    (2003)
  • D. Easley et al.

    Time varying arrival rates of informed and uninformed trades

    (2003)
  • D. Easley et al.

    Is information risk a determinant of asset returns?

    Journal of Finance

    (2002)
  • Cited by (255)

    • Rumors in the sky: Corporate rumors and stock price synchronicity

      2023, International Review of Financial Analysis
    • Progress and prospects of data-driven stock price forecasting research

      2023, International Journal of Cognitive Computing in Engineering
    View all citing articles on Scopus

    This article is an extension of the first chapter of my doctoral dissertation at the University of Pennsylvania. I am indebted to my dissertation committee Francis X. Diebold, Simon Gervais, and Frank Schorfheide. I greatly appreciate Wesley Chan's generosity in giving me access to his public news database. I thank Sun Le and Laura X. L. Liu for their research assistance. I’m particularly grateful to the anonymous referee for his insightful comments. I appreciate discussions with Rui Albuquerque, Mike Barclay, Alon Brav, Liz Demers, Soeren Hvidkjaer, Wei Jiang, Yuichi Kitamura, John Long, Bill Schwert, Vesna Strasser, Michela Verardo, and Ivo Welch as well as participants at the University of Pennsylvania econometrics seminar and the University of Rochester finance seminar.

    View full text