Star Analysts, Overreaction, and Synchronicity: Evidence from China and the United States

• Published date: 01 September 2017
• Author: Jing Lin,Mingshan Zhou,Yunbi An
• Date: 01 September 2017
• DOI: http://doi.org/10.1111/fima.12164

Star Analysts, Overreaction,

and Synchronicity: Evidence from China

and the United States

Mingshan Zhou, Jing Lin, and Yunbi An∗

This paper examines star analyst coverage, investor overreaction, and stock price synchronicity

in the Chinese and US markets. In China, we find that star analyst coveragecan induce investor

overreaction, such that it is negatively correlated with price synchronicity. This overreaction

effect is particularly pronounced for stocks with primarily individual investors. In contrast, in

the United States, we find that star analyst coverage is positively related to synchronicity and

is not significantly associated with investor overreaction. Our overall findings imply that the

heterogeneousnature of investors in a market drives the association among star analyst coverage,

overreaction, and stock price synchronicity.

Studies on stock price synchronicity (R2from a market model of stock returns), such as Morck,

Yeung, and Yu (2000), generally relate synchronicity to a firm’s price informativeness. A low

(high) R2means that the stock price of the firm is informative (uninformative). The logic is that

withalowR2, stock price variations carry more information about the f irm itself as compared to

a market-tracking portfolio or vice versa. The primary role of analysts is information production.

If analysts are indeed able to provide firm-specific, rather than market-wide, information about a

firm, then that f irm’s R2should be low, or vice versa. Chan and Hameed (2006), in a cross-country

study of synchronicity, suggest that R2in emerging markets is generally positively correlated

with analyst coverage. The authors posit that analysts in opaque emerging markets produce less

firm-specif ic information. Xu et al. (2013), in a study of Chinese firms, consider the hetero-

geneous nature of analyst ability and find that star analyst coverage is negatively related to R2,

while nonstar analyst coverage is still positively related to R2. This branch of the synchronicity

literature assumes that the role of analysts is confined to information production.

An emerging body of literature, however, questions the assertion that R2primarily relates to

a firm’s price informativeness (Ashbaugh-Skaife, Gassen, and Lafond, 2006; Teoh, Yang, and

Zhang, 2009; Chang and Luo, 2010). Specifically, Hou, Peng, and Xiong (2013) find that the

extent of investoroverreaction to news is negatively correlated with R2. They argue that according

to the behavioral finance literature, investors overreact to news. If investors overreact to analyst

recommendations and forecasts for a stock, investors update their beliefs about the value of that

stock to a degree that is much larger or smaller in relation to its fundamental value. Thus, the

overreaction causes the stock price’s variability to deviate significantly from the stock’s market

trend. Ceteris paribus, the R2of the firm will be lower than the case without such overreaction. A

We thank Marc Lipson (Editor), an anonymous referee, Bing Han, and participants at the 6th China Finance Review

International Conference in 2013 and the 10th China Annual Finance Conference in 2013 for helpful comments. An

acknowledges the support fromthe Odette School of Business at the University of Windsor. The usual caveats apply.

∗Mingshan Zhou is a Professor in the School of Finance at Southwestern University of Finance and Economics in

Chengdu, Sichuan, China. Jing Lin is a Ph.D. student in the Schoolof Economics at Peking University in Beijing, China.

Yunbi An is a Professorin the Odette School of Business at the University of Windsor in Windsor,Ontario, Canada.

Financial Management •Fall 2017 •pages 797 – 832

798 Financial Management rFall 2017

low R2could be due to investoroverreaction induced by star analyst coverage rather than superior

information produced by star analysts.

The objective of this study is to examine whether the information production or investor

sentiment role of star analysts dominates in explaining R2in the Chinese and US markets. We

focus on the Chinese market, as it is the largest emerging market in the world and its trading

volume is generated primarily by individual investors. Several studies indicate that individual

investors are prone to behavioral bias, such that investors overreact to news in China.1We

contend that market overreaction to star analyst coverageprevails in China, and that the sentiment

role of star analysts is more important than their information production role in explaining the

negative correlation between star analyst coverage and R2. Unlike the Chinese market, the US

market is more mature and is one in which trading is dominated by large institutional investors.

As such, overreaction is ameliorated. We replicate the analysis with US data to contrast with the

findings in the Chinese market, and further clarify the different role of analysts in the relation

between stock price synchronicity and analyst coverage in different markets.

Using the Chinese data, similar to Xu et al. (2013) and Zhou, Lin, and Xu (2016), we find

that star analyst coverage is negatively correlated with R2, while nonstar analyst coverage is

positively correlated with R2in the full sample. In addition, we determine that star analyst

coverage and recommendations are positively related to short-term price momentum, long-term

price reversal, and abnormal trading volume. This suggests that star analyst coverage is related

to market overreaction. We find that this relation is more pronounced for stocks with primarily

individual investorsthan those with mostly institutional investors. Moreover,we confir m that even

if both star and nonstar analysts have similar earnings forecasting abilities, star analyst coverage

continues to be negatively related to synchronicity, while nonstar analyst coverage proves to

be positively related to synchronicity. Thus, the information production role of star analysts

in emerging markets appears to be limited. Additionally, the average R2of firms followed by

star analysts with the most accurate earnings forecasts is not statistically different from those

followed by star analysts with the least accurate earnings forecasts suggesting that it is not the

ability of star analysts that most relates to R2. Rather, star analysts move firm R2via inducing

market overreaction. In contrast, our findings based on the US data indicate that both star and

nonstar analyst coverage are positively related to synchronicity, and star analyst coverage is not

associated with investor overreaction. Thus, in the United States, the role of star analysts in

inducing investor sentiment is limited. Our overall findings imply that the heterogeneous nature

of investors in a market drives the association among star analyst coverage, overreaction, and

stock price synchronicity.

We make two contributions to the literature. First, in recognizing the heterogeneous nature of

analyst coverage, we relate star analyst vis-`

a-vis nonstar analyst coverage to R2in the presence

of investor sentiment rather than focusing on the relation between investor sentiment and R2,as

in Hou et al. (2013). Hou et al. (2013) study the relation between stock price momentum and R2,

but do not relate analyst coverage to R2. In addition, unlike the cross-country study of Hou et al.

(2013), our one country study allows us to control for the impact of a country’s legal, cultural,

and socioeconomic factors on the conclusions. Our findings offer support to the overreaction

1The short-term contrarian profits are due to overreaction to firm-specif ic information (Kang, Liu, and Ni, 2002);

compared to US investors, Chinese investors are more overconfident (Chen, Kim, and Nofsinger, 2007). Investors

overreact to star analyst forecasts and recommendations, particularly bull recommendations (Zhou et al., 2016); Ng and

Wu (2007) and Wu (2011) also find that the pure contrarian strategy can, in the Chinese stock market, produce positive

excess return, and a strategy combining mean reversion and momentum can outperform a buy-and-hold strategy, which

collectively supports the overreaction hypothesis.

Zhou, Lin, & An rStar Analysts, Overreaction, and Synchronicity 799

explanation of R2in Hou et al. (2013). Wef ind that after we classify analysts into star or nonstar,

Hou et al.’s (2013) explanation of R2and investor sentiment still holds.

In addition, we explain why star analysts can induce market overreaction in emerging markets

resulting in a low R2of star analyst covered firms. We find that the investor sentiment role

of star analysts dominates their information production role in China, while this is not true

in the United States where trading activity is dominated by institutional investors and investor

overreaction to analyst reports is ameliorated. We demonstrate that the trading behavior of

individual investors playsa critical role in explaining market overreaction induced by star analyst

coverage in China, therebydriving the association among star analyst coverage, overreaction, and

stock price synchronicity.Our results complement those of Xu et al. (2013) and Zhou et al. (2016)

who primarily find a negative relation between star analyst coverage and R2in China. Xu et al.

(2013) attribute low R2to the superior information production role of star analysts, while Zhou

et al. (2016) relate low R2to investoroverreaction and star analyst coverage without documenting

the causes for these relations. In contrast, we examine the role of star analysts in inducing investor

sentiment and in explaining stock price synchronicity in the Chinese market compared with their

role in the US market and draw inferences as to what constitutes the differences.

The remainder of this paper is organized as follows. Section I reviews the literature related to

our study and proposes hypotheses to be tested. Section II describes the data used in this study

and provides definitions of the variables. Section III analyzes the empirical results. Section IV

reports the empirical results based on US data, while Section V provides our conclusions.

I. Literature Review and Hypothesis Development

A. Literature Review

There are two strands of literature related to our study.The f irst discusses the relation between

synchronicity (R2) and stock price informativeness. Roll (1988) is the first to suggest using R2

to gauge stock price informativeness. Morck et al. (2000) argue that emerging markets generally

have a higher R2due to their weak protection of investor property rights as compared with

developed markets. A high R2indicates that a large percentage of a stock’s variation can be

explained by general stock market movement. Thus, for a firm with high R2, its stock price does

not provide much information about the firm itself. Durnev et al. (2003) and Jin and Myers (2006)

find a negative correlation between a market’s opaqueness and a firm’s R2.

In contrast, Ashbaugh-Skaife et al. (2006), in a cross-country study, suggest that R2does not

necessarily reflect the price informativeness of a firm. Instead, when a firm’sstock price contains

noise, R2can be low. Hou et al. (2013) develop a model that demonstrates that a low R2can

be due to investor overreaction. That is, investors’ rational reaction to information makes the

stock return unpredictable and, more importantly, R2does not necessarily relate only to the price

informativeness of the stock, but also depends upon investor sentiment. In their cross-country

study, Hou et al. (2013) find that the medium-term price momentum and long-term price reversal

of stock returns are negatively related to R2suggesting that investor overreaction is negatively

correlated with R2. In sum, the determinants of R2are likely due to two factors: 1) the amount

of firm-specif ic information in the stock price and 2) the extent of investor sentiment. However,

it is unclear whether information production or investor sentiment dominates the explanation of

the R2variation.

The second strand of the literature examines the relation between analyst coverage and R2.

Chan and Hameed (2006) argue that due to capital market impediments, analysts provide less