Does Limited Attention Constrain Investors’ Acquisition of Firm‐specific Information?

• Published date: 01 November 2014
• Author: Yi Dong,Chenkai Ni
• DOI: http://doi.org/10.1111/jbfa.12098
• Date: 01 November 2014

Journal of Business Finance & Accounting

Journal of Business Finance & Accounting, 41(9) & (10), 1361–1392, November/December 2014, 0306-686X

doi: 10.1111/jbfa.12098

Does Limited Attention Constrain

Investors’ Acquisition of Firm-specific

Information?

YIDONG AND CHENKAI NI*

Abstract: According to the framework outlined in Peng and Xiong (2006), attention-

constrained investors tend to process more market- and sector-level information. We empirically

test this theory. We find that firms with higher media coverage have lower contemporaneous

stock return synchronicity. Such an effect is robust to analyses within size deciles, inclusion of

firm fixed effects, estimation using a matched sample, and a two-stage least squares approach.

The effect becomes less pronounced during the financial crisis period when both the quantity

and quality of firm-specific information decrease. Further, the attention from media coverage

has a spillover effect on the firm’s industry peers without media coverage. Finally, investors of

firms with higher media coverage are more efficient in incorporating future firm performance

into current stock prices. Collectively, our findings support the theory in Peng and Xiong (2006)

that investors increase their acquisition of firm-specific information when a firm captures their

attention.

Keywords: mass media, limited attention, stock return synchronicity, information acquisition

1. INTRODUCTION

Attention is a scarce cognitive resource, and investors are subject to the limited

attention constraint (Kahneman, 1973; Merton, 1987). It has been suggested that such

a constraint will generate undesirable economic consequences such as a higher equity

premium and lower informational efficiency around analyst recommendations and

earnings announcements (Peress, 2008; Fang and Peress, 2009; Loh, 2010).

Firm information affects stock prices only when investors pay attention to it. Prior

studies build upon the theory of limited attention and examine how it affects the

dynamics of asset prices (Merton, 1987; Sims, 2003; Peng, 2005; Peng and Xiong,

∗The first author is from School of Banking and Finance, University of International Business and

Economics, Beijing, China. The second author is from Department of Accounting, School of Business,

Renmin University of China, Beijing, China. The authors thank Oliver Zhen Li, Yupeng Lin, and seminar

participants at the University of International Business and Economics for their helpful comments. Yi Dong

acknowledges financial support from the National Science Foundation for Distinguished Young Scholars of

China (Grant No. 71402026) and the Fundamental Research Funds for the Central Universities in UIBE

(CXTD5-03).

Address for correspondence: Chenkai Ni, Department of Accounting, School of Business, Renmin Univer-

sity of China, 59 Zhongguancun Street, Haidian District, Beijing, China, 100872.

e-mail: nichenkai@rbs.org.cn

C

2014 John Wiley & Sons Ltd 1361

1362 DONG AND NI

2006). Specifically, Peng and Xiong (2006) construct a discrete-time model with

an infinite number of periods, multiple risky assets, and one risk-free asset in the

economy. Assuming that investors have a learning process regarding information

processing, they show that limited attention can lead to a category-learning behavior.

In particular, investors who are subject to the limited attention constraint tend to

allocate more attention to market- and sector-level factors, and less to firm-specific

factors. The intuition behind this assertion is as follows. As there are a large number

of sectors in one market and a large number of firms in one sector, it is rational for

an attention-constrained investor to concentrate more on processing common factors

(e.g., market factors and sector factors) which tend to generate more uncertainties in

her portfolio.

We empirically test the theoretical link proposed in Peng and Xiong (2006).

Specifically, we analyze the association between a firm’s media coverage and its

contemporaneous stock return synchronicity, the latter of which serves as our proxy

for the proportion of firm-specific information contained in stock prices (Roll, 1988;

Morck et al., 2000; Hutton et al., 2009). Our sample consists of S&P1500 firms covering

the period from 2006 to 2011. We hand-collect news articles covering these firms

from the LexisNexis database, published in The Wall Street Journal,The New York Times,

The Washington Post and USA Today. We construct media coverage as the number of

news articles covering a firm during a specific year. Stock return synchronicity is

defined based on the R2from a firm–year regression of firms’ weekly stock returns

on market returns and industry returns, with both lead- and lag-terms included. We

further apply a log transformation to the R2measure, which is constrained to lie

between 0 and 1. Our results suggest that a higher media coverage is associated with

a lower level of stock return synchronicity (i.e., a lower R2and more firm-specific

information). To put this result into economic perspective, for firms without media

coverage, one additional media article is associated with a 1.88% reduction in the R2.

We find similar results when we conduct analyses within size-deciles, incorporate firm-

fixed effects, estimate the regression using a matched sample, and address potential

endogeneity using a two-stage least squares (2SLS) approach. We further decompose

the R2into one component related to market returns and another component (semi-

partial R2) related to industry returns (Haggard et al., 2008). Results suggest that the

explanatory power of both market returns and industry returns decrease as media

coverage increases. Evidence here is consistent with the argument that media coverage

captures investors’ attention and increases their incentive to acquire more firm-specific

information.

Next, we investigate whether the effect of media coverage on stock return syn-

chronicity varies along the business cycle. Two lines of arguments suggest a mitigated

effect of media coverage during an economic recession. First, both the quantity and

quality of firm-specific information are expected to decrease during economic reces-

sions (Veldkamp, 2005; Nieuwerburgh and Veldkamp,2006). In contrast, information

production becomes a positive externality of aggregate economic activities during ex-

pansions, as investments endogenously generate information signals. Second, investors

tend to shift their attention to resolve macroeconomic uncertainty when the economy

begins to decline because the market factor carries a large weight in their portfolios

(Peng and Xiong, 2006; Peng et al., 2007). Consistent with these arguments, we find

that the association between media coverage and stock return synchronicity is less

pronounced during the recent financial crisis.

C

2014 John Wiley & Sons Ltd

LIMITED ATTENTION AND INFORMATION ACQUISITION 1363

We then incorporate the possibility of attention spillover from firms with media

coverage to their industry peers without media coverage. As firms compete with

their industry peers in the same product market, serve as each other’s benchmarks

in relative performance evaluation, and are exposed to similar technology shocks,

investors are likely to pay attention to a firm because of its peers’ media coverage.

Based on a subsample including only firms that are not covered by the mass media,

we examine whether their industry peers’ media coverage has an impact on their

stock return synchronicity. To isolate the effect of attention spillover from the effect

of information spillover, we split the sample based on the number of firms in the

industry.1Consistently, we find that uncovered firms who have industry peers with

higher media coverage enjoy the benefit of attention spillover and, in turn, have

a lower level of stock return synchronicity. Such an effect is more pronounced in

industries with a larger number of firms where attention spillover is more likely to

be detected.2

Finally, we examine how media coverage affects stock prices’ efficiency in incor-

porating information on future earnings. If stock prices contain more firm-level

information, we expect that the current period stock return reflects a firm’s future

earnings to a larger extent. Building upon the framework outlined in Kothari and

Sloan (1992), we find empirical evidence suggesting that higher media coverage is

associated with a more timely incorporation of future earnings into current stock

prices.

Our study contributes to the literature in several important dimensions. First, we

investigate how investors’ attention affects the informational efficiency of stock prices.

Theory suggests that limited attention constrains investors’ incorporation of firm-

specific information into stock prices (Peng and Xiong, 2006). Empirical studies

have touched upon whether media coverage spurs trading around specific dates of

information releases and whether earnings announcements accompanied by media

coverage generate a more efficient response and less delay (Peress, 2008; Engelberg

and Parsons, 2011). However, these studies predominantly focus on a short window

and do not specifically address the allocation of attention to firm-specific information

versus market/industry information. Our study differs from prior studies in that we

examine long-term informational efficiency measured by stock return synchronicity.

In so doing, we directly capture the proportion of firm-specific information in stock

prices.

Second, we add to the stream of literature on the impact of media coverage on in-

vestors’ behaviors and security prices. Fang and Peress (2009) find that media coverage

explains the cross-section of stock returns, and Engelberg and Parsons (2011) suggest

that local media coverage has a direct impact on local investors’ trading behaviors.

We add evidence that media coverage increases investors’ acquisition of firm-specific

1 Information spillover suggests that a firm’s information speaks to the performance of other firms, e.g.,

firms in the same industry generally have correlated quarterly earnings. In contrast, attention spillover

conveys the idea that, when an investor is paying attention to a firm, she will also be directed to pay attention

to the firm’s peers.

2 For industries with few firms, information spillover is more likely to arise from media coverage. For

example, one leading firm’s information affects other firms to a larger extent in an oligopolistic industry

compared with another industry with perfect competition. Different from attention spillover, information

spillover drives industry peers’ stock prices to move together and results in a higher stock return

synchronicity. Attention spillover thus becomes more detectable in industries with large numbers of firms

when we utilize stock return synchronicity to capture investors’ information acquisition activities.

C

2014 John Wiley & Sons Ltd