Does High Stock Return Synchronicity Indicate High or Low Price Informativeness? Evidence from a Regulatory Experiment

• Date: 01 December 2018
• DOI: http://doi.org/10.1111/irfi.12157
• Published date: 01 December 2018

Does High Stock Return

Synchronicity Indicate High or Low

Price Informativeness? Evidence

from a Regulatory Experiment*

SHUO KAN

†

AND STEPHEN GONG

‡

†

School of Banking and Finance, University of International Business and Economics,

Beijing, China and

‡

Graduate School of Management, University of Auckland Business School, Auckland,

New Zealand

ABSTRACT

We investigate the link between stock return synchronicity and price infor-

mativeness by exploiting the Regulation SHO pilot program, which removed

short-selling price tests for randomly selected stocks (“pilot stocks”) in May

2005. A difference-in-differences analysis reveals that relative to non-pilot

stocks, pilot stocks saw a significantly larger increase in both price informa-

tiveness and return synchronicity when the pilot program started, but such

difference disappeared when Regulation SHO removed the short-selling price

tests for all stocks in July 2007. The results suggest that high return synchro-

nicity reflects high, rather than low price informativeness.

JEL Codes: G14; G1

Accepted: 29 August 2017

I. INTRODUCTION

Stock return synchronicity, usually measured as transformed R

2

from an asset

pricing model, reflects the extent of co-movement between an individual

stock’s return and the market return. Stock return synchronicity is now a widely

used proxy for stock price informativeness (Bae et al. 2013; Li et al. 2014). Yet it

remains hotly contested whether synchronicity reflects informativeness or

noise (Teoh et al. 2009; Chan and Chan 2014; Gassen et al. 2016).

1

* Comments and suggestions from Henk Berkman, Paul Griffin, Yongxiang Wang, Songtao Tan,

Jinyu Liu and participants at the New Zealand Finance Colloquium 2017 and Renmin University

Research Workshop are gratefully acknowledged. The authors are particularly indebted to the anony-

mous referees and the Editor (Doug Foster) for their valuable comments and suggestions. Shuo Kan

gratefully acknowledges the financial support received from the China Scholarship Council. All

remaining errors are their own.

1 Henceforth we refer to stock return synchronicity as “synchronicity”and stock price informa-

tiveness as “informativeness”for short.

© 2017 International Review of Finance Ltd. 2017

International Review of Finance, 18:4, 2018: pp. 523–546

DOI: 10.1111/irfi.12157

Since Roll’s (1988) seminal work documenting a low market-model R

2

statis-

tic (i.e., low synchronicity) for US stocks, a growing literature has examined the

interpretation and causes of low synchronicity.

2

One strand of the literature

(e.g., Morck et al. 2000; Durnev et al. 2003, 2004) has suggested that low syn-

chronicity reflects high informativeness because individual stock returns should

exhibit lower co-movement with market returns when stock prices reflect more

firm-specific information, due to better information (e.g., disclosure) and insti-

tutional (e.g., investor protection) environments. On the other hand, Dasgupta

et al. (2010) show, both theoretically and empirically, that a more informative

stock price today should be associated with less firm-specific variation in stock

prices, or higher synchronicity, in the future. Hou et al. (2013) also propose a

theoretical model that investigates the relation between informativeness and

synchronicity. They show that, in a rational setting, the average return syn-

chronicity over time is independent of price informativeness. However, when

there is idiosyncratic mispricing driven by investor sentiment, greater synchro-

nicity indicates less idiosyncratic noise and higher informativeness. It seems fair

to say that there is support for each of these vastly different viewpoints.

3

Given the important theoretical and empirical implications of price informa-

tiveness and the popularity of synchronicity as a measure of informativeness, it is

imperative to understand the empirical relation between synchronicity and infor-

mativeness. The widely used methodology in prior studies on this issue is to test

the statistical relation between synchronicity and other proxies for informative-

ness, typically in a cross-sectional setting. For instance, if a proxy “X”(e.g., firm

size, analyst coverage) is expected to be positively associated with informativeness,

the statistical relation between “X”and synchronicity is tested. If it is found that

stocks with high “X”also have relatively high (or low) synchronicity, one then

draws the conclusion that high synchronicity reflects high (or low) informative-

ness. Examples applying this approach include Durnev et al. (2003), Hutton

et al. (2009), Teoh et al. (2009), and Chan and Chan (2014), among others.

The empirical methodology in prior literature is plagued by serious endo-

geneity concerns. One concern is omitted variables. When we examine the sta-

tistical relation between synchronicity and other proxies for informativeness

(like “X”mentioned above), it is almost impossible to rule out the possibility

that some important (perhaps unobserved) factors which are related with both

synchronicity and “X”are left out, causing the statistical relation between syn-

chronicity and “X”to be unreliable. Hutton et al. (2009), for instance, use earn-

ings quality as a measure of stock information efficiency and show that poorer

earnings quality is associated with higher synchronicity. However, both

2 After controlling for exposure to market and industry factors and the occurrence of identifia-

ble public information, Roll (1988) continues to find a low R

2

. He concludes that the majority

of returns is explained either by private information or a “frenzy”unrelated to firm-specific

information.

3 For more discussions of the extensive but contradictory literature on synchronicity and infor-

mativeness, see Alves et al. (2010), Morck et al. (2013), Kelly (2014), Chan and Chan (2014)

and Bramante et al. (2015).

© 2017 International Review of Finance Ltd. 2017524

International Review of Finance