Media Coverage and the Cross‐Section of Stock Returns: The Chinese Evidence

• Published date: 01 December 2019
• Author: Yishun Wang,Kien Dinh Cao,Liping Zou
• DOI: http://doi.org/10.1111/irfi.12191
• Date: 01 December 2019

Media Coverage and the Cross-

Section of Stock Returns: The

Chinese Evidence

LIPING ZOU

†

,KIEN DINH CAO

‡

AND YISHUN WANG

†

†

School of Economics and Finance, Massey University, Auckland, New Zealand and

‡

Faculty of Business Administration, Foreign Trade University, Hanoi, Vietnam

ABSTRACT

Using hand-collected news headlines for a large sample of listed firms in

China over a period of 2000–2015, we investigate the cross-sectional relation

between media coverage and stock returns. Our results document that no-

media coverage stocks earn 55 basis points a month higher than stocks that

are featured in the media. This result is robust after controlling for common

risk factors and is not driven by short-run return reversals. Further analysis

provides evidence to support the investor recognition hypothesis, suggesting

that mass media may play an incremental role in providing a supplement to

traditional channels of information dissemination. Therefore, results in this

paper are of interests to both investors and regulators on drivers of stock

returns.

JEL Codes: G12; G14; G15

Accepted: 16 March 2018

I. INTRODUCTION

The effect of media on various aspects of stock markets has recently received

great attention from researchers and practitioners. Media is found to contribute

to the efficiency of the stock market due to its dissemination of information to

the public (Peress 2014), predict stock returns (Tetlock 2007; Fang and Peress

2009), increase trading activities (Engelberg and Parsons 2011), enhance the

quality of corporate governance (Liu and McConnell 2013), and reduce the

information asymmetry among investors (Tetlock 2010; Dai et al. 2015; Rogers

et al. 2016). Extant literature also find that media affects investors’behavior in

a biased way (Gurun and Butler 2012; Hillert et al. 2014; Ahern and

Sosyura 2015).

However, almost all of studies on the media effect are conducted for the US

markets and very little is known for non-US markets. The purpose of this paper

is to fill this gap and provide evidence on the relation between media coverage

and the cross-section of stock returns for a large sample of Chinese stocks. We

hand-collect news headlines for 1500 Chinese stocks listed in Shanghai and

© 2018 International Review of Finance Ltd. 2018

International Review of Finance, 19:4, 2019: pp. 707–729

DOI: 10.1111/irfi.12191

Shenzhen stock exchanges during the 2000–2015 period. We document that

media coverage is significantly related to a number of firm characteristics. Spe-

cifically, we find that large stocks and stocks with high idiosyncratic volatility

attract relatively high media attention. In addition, stocks covered by analysts

are also more likely to be featured in the media. However, firms with greater

institutional holdings and state ownerships exhibit lower media coverage. We

also find that past month absolute returns are negatively related to media cover-

age. In a cross-sectional setting regarding the relation between the media cover-

age and the stock return, we find strong evidence of return premiums for stocks

with no news compared to those that are featured in the news, even after con-

trolling for common risk factors. This no-coverage premium does not exhibit a

reversal when we extend the holding period of up to 12 months. Further analy-

sis also reveals that the no-media return premium is due to investor recognition

hypothesis proposed by Merton (1987). However, our evidence on illiquidity

hypothesis is weak, as opposed to Fang and Peress (2009).

This study contributes to the increasingly attractive literature on the media

effect for a number of reasons. Firstly, it is among few papers that examine the

media coverage—stock return relationship for non-US markets. Ferguson

et al. (2015) investigate the effects of both tone and volume of firm-specific

news media on the stock returns in the UK. They show the significance of both

news effects with the volume effect being more pronounced. Aman

et al. (2016) find that television media in Japan is associated with higher trad-

ing volume and greater price change. Griffin et al. (2011) examine the informa-

tion effect of media for 56 countries around the world and find that the effect

of public news on a firm’s stock price on news days is significant for developed

markets. They do not find significant news effects for emerging markets, which

is mainly explained by the existence of insider trading in these markets.

Second, the news data for this study is hand collected from a media outlet

that is quite popular to all types of investors, big or small, individual or institu-

tional. These media headlines are in local language, meaning that local inves-

tors, who are dominant in Chinese stock markets, are able to read and assess

the news about the firms. This media sample is unique and in contrast to other

studies that their media sample is in English. Our sample of 1500 firms is much

larger than the 235 Chinese firms included in Griffin et al. (2011) and covers a

longer time period (2000–2015 versus 2003–2009). In addition, this study

examines the effect of media coverage on stock return premium, which is differ-

ent from examining the relevance of news by Griffin et al. (2011).

Third, the examination of media and stock returns in China is of interests to

both investors and regulators. China is characterized as one of the countries

with heavy control of media with its freedom of press being consistently ranked

at the low end in the world by the Reporters without Borders Organization.

1

This heavy control of media by the government could induce a strong bias on

the tone of the media since the government is more likely to see media report

1 https://en.wikipedia.org/wiki/Press_Freedom_Index

© 2018 International Review of Finance Ltd. 2018708

International Review of Finance