Stock return predictability of the cumulative abnormal returns around the earnings announcement date: Evidence from China

• Published date: 01 March 2023
• Author: Ping‐Wen Sun,Zipeng Wen
• Date: 01 March 2023
• DOI: http://doi.org/10.1111/irfi.12380

ORIGINAL ARTICLE

Stock return predictability of the cumulative

abnormal returns around the earnings

announcement date: Evidence from China

Ping-Wen Sun | Zipeng Wen

Minjiang University, Newhuadu Business

School, Fuzhou, China

Correspondence

Ping-Wen Sun, Newhuadu Business School,

Minjiang University, Fuzhou, China.

Email: sunpingwen@gmail.com

Funding information

Minjiang University, Grant/Award Number:

MYS20011

Abstract

We find that cumulative abnormal returns adjusted by size,

book-to-market, and momentum around the earnings

announcement date (DGTW_CAR3 hereafter) significantly

and positively predict stock returns in the 6-month period

from May 2005 to October 2020 in the China's A-shares

market. The monthly equally-weighted DGTW_CAR3 pre-

miums are 0.47% and 0.67% after risk adjustment. Although

stock price delay fails to fully account for the DGTW_CAR3

premium, we find that the DGTW_CAR3 premium is more

significant for illiquid stocks and during periods with high

investor sentiment. This result suggests that market ineffi-

ciency explains the DGTW_CAR3 premium. Further analysis

shows that, in addition to earnings information, the opti-

mism reflected in the management discussion and analysis

section of the annual or half-year report also contributes to

the DGTW_CAR3 premium. This finding implies that

DGTW_CAR3 may contain new fundamental information

that correlates significantly and positively with future stock

performance. Finally, we find that the institutional owner-

ship change of a stock associated with DGTW_CAR3 also

significantly and positively predicts the stock's return,

suggesting that institutional investors adjust their holdings

according to DGTW_CAR3 and consequently influence the

demand for the stock in the China's A-shares market.

Received: 9 May 2021 Revised: 21 January 2022 Accepted: 6 March 2022

DOI: 10.1111/irfi.12380

© 2022 International Review of Finance Ltd.

58 International Review of Finance. 2023;23:58–86.

wileyonlinelibrary.com/journal/irfi

KEYWORDS

cumulative abnormal returns, earnings announcement, earnings

surprise, institutional investors, price delay

JEL CLASSIFICATION

G11, G12, G14, G23

1|INTRODUCTION

Several studies have documented the post-earnings announcement drift (PEAD) phenomenon in the stock mar-

ket and shown that the magnitude of the earnings surprise on the earnings announcement date is positively cor-

related with the PEAD.

1

Although many studies attribute the PEAD to a delayed price response, we conjecture

that new information other than earnings news released on the earnings announcement date also contributes to

the PEAD.

2

In this study, we directly use the cumulative abnormal returns adjusted b y size, book-to-market

(BM), and momentum proposed by Daniel et al. (1997) around the earni ngs announcement date (DGTW_CAR3

hereafter) as a proxy for the new information released on the earnings announcemen t date and examine

whether DGTW_CAR3 is also responsible for the stock return after the earnings announcement date in the

China's A-shares market. We find that DGTW_CAR3 significantly and positively predicts cross-sectional stock

returns in the following 6 months. Our further analysis suggests that, in addition to the earnings news,

DGTW_CAR3 also contains new information such as the optimism reflected in the management discussion and

analysis section of the annual or half-year report, which significantly and positively correlates with stock perfor-

mance in the near future.

Many studies have shown that stock returns around the earnings announ cement date contain useful infor-

mation about a firm's fundamentals. For example, Easton and Zmijewski (1989) contend that if earnings

announcements release information that revises investors' expectations about future dividends, stock prices on

the earnings announcement date will react positively to the magnitude of the revisions. Ch apman (2018)shows

that investor reactions around the earnings announcement date are mitigated if a firm provides earnings notifi-

cations, showing that investors pay special attention to earnings-associated news. Engelberg et al. (2018)find

that returns of stock anomalies are higher on earnings announcement days and argue that investors' biased

expectations responsible for the anomaly returns are corrected upon news arrival. Abarbanell and Kim (2010)

find that the realized returns of earnings announcements can predict firms' fut ure earnings because there is

more informed trading for firms with more analyst coverage and for big firms on earnings announcement days.

Beaver et al. (2020) document an increase in investor response to earnings announcements and argue that, in

addition to earnings news, management guidance, analyst forecast, andfinancial statement information bundled

with the earnings announcement are responsible for the increased response. In addition to stock returns around

the earnings announcement date, scholars document that the information released on the e arnings announce-

ment date can predict the PEAD. For example, Truong (2011) finds that in the China's A-shares market, thetop

earnings surprise quintile stocks outperform the bottom earnings surprise quintile stocks in the year after the

earnings announcement date. Truong (2011) contends that the PEAD is positively related to a firm's future

financial performance.

Although these studies show that firm fundamental news released on the earnings announcementdate is associ-

ated with abnormal returns around the earnings announcement date and PEAD, none of these studies directly inves-

tigates whether abnormal returns around the earnings announcement date significantly predict the PEAD. The

answer to this question is important because it can help us better understand why the PEAD exists. On the one

hand, if we find that DGTW_CAR3 significantly predicts the PEAD because of market inefficiency, we may observe

that stock price delay subsumes the return predictability of DGTW_CAR3, and that the DGTW_CAR3 premium is

SUN AND WEN 59