How is Earnings News Transmitted to Stock Prices?

• Published date: 01 March 2022
• Author: Vincent Grégoire,Charles Martineau
• Date: 01 March 2022
• DOI: http://doi.org/10.1111/1475-679X.12394

DOI: 10.1111/1475-679X.12394

Journal of Accounting Research

Vol. 60 No. 1 March 2022

Printed in U.S.A.

How is Earnings News Transmitted

to Stock Prices?

VINCENT GRÉGOIRE∗AND CHARLES MARTINEAU†

Received 19 June 2020; accepted 25 May 2021

ABSTRACT

We examine the speed and mechanism of the price discovery process follow-

ing earnings announcements in the after-hours market, a very illiquid trad-

ing environment. Prices reflect earnings surprises mostly through changes

in quotes rather than through trades. Following positive announcement sur-

prises, ask prices adjust quickly while bid prices are slower to adjust, and vice

∗HEC Montréal; †Rotman School of Management and UTSC Management, University of

Toronto

Accepted by Rodrigo Verdi. This paper is an extension of Martineau’s Ph.D. thesis Chap-

ter 1. We thank the anonymous associate editor and referee for valuable comments that have

greatly improved this manuscript. Pat Akey, Daniel Andrei, Philippe d’Astous, Markus Bal-

dauf, Oliver Boguth, Thomas Bourveau, Jonathan Brogaard, Mike Brolley, James Brugler,Mur-

ray Carlson, Tao Chen, Peter Christoffersen, Alex Edwards, Adlai Fisher, Mathieu Fournier,

Will Gornall, Dale Griffin, Terry Hendershott, Alexandre Jeanneret, Ali Lazrak, Nan Li, Ji-

asun Li, Kai Li, Kin Lo, Russell Lundholm, Chay Ornthanalai, Andreas Park, Neil Pear-

son, Ryan Riordan, Kumar Venkataraman, Sheng Jun Xu, Bart Yueshen, Terry Zhang, Zhuo

Zhong, Marius Zoican, and seminar participants at the UBC Finance and Accounting divi-

sions, HEC Montréal, University of Virginia, Temple University, University of Colorado Boul-

der, Nanyang Technology University, University of Melbourne, University of Toronto Finance

and Accounting divisions, Rice University, McGill University, Queen’s University, York Uni-

versity, Tsinghua University, Binghamton SUNY, the NFA, and the SFS Calvacade Asia Pacific

for their comments. We also thank Frank Hathaway for access to NASDAQ ITCH data, and

Compute Canada for high-performance computing support. We acknowledge financial sup-

port from the NASDAQ OMX Educational Foundation, Montreal Exchange, Bank of Mon-

treal Capital Group, Canadian Securities Institute Research Foundation, IVADO, and Social

Sciences and Humanities Research Council of Canada (SSHRC). An online appendix to this

paper can be downloaded at http://research.chicagobooth.edu/arc/journal-of-accounting-

research/online-supplements

261

©

© 2021 The Authors. Journal of Accounting Research published by Wiley Periodicals LLC on behalf of The

Chookaszian Accounting Research Center at the University of Chicago Booth School of Business.

This is an open access article under the terms of the Creative Commons

Attribution License, which permits use, distribution and reproduction in any medium,

provided the original work is properly cited.

[The copyright line for this article was changed on 22~~~ October 2022 after online publication].

[The copyright line year was corrected on 14 ~~~November 2022 after online publication].

262 v. grégoire and c. martineau

versa for negative surprises. Returns computed from trade prices underesti-

mate the speed and magnitude of price reactions following announcements

relative to returns computed from quotes. These findings emphasize the im-

portance of using quotes and not trade prices when examining intraday price

discovery. Because firm announcements such as earnings generally occur in

the after-hours market, using quotes is crucial as trading is sparse. We fur-

ther illustrate the importance of quotes when examining the price discovery

process around analyst recommendation revisions.

JEL codes: G10, G12, G14, M41

Keywords: after-hours market; analyst coverage; analyst recommendations;

asymmetric reaction; disclosure; earnings announcements; liquidity; price

discovery

1. Introduction

How do stock prices incorporate earnings news? Building on classical

frameworks of price formation (Glosten and Milgrom [1985], Kyle [1985]),

the model of Kim and Verrecchia [1994] predicts that stock prices reflect

earnings news through the arrival of trades because liquidity providers

are unsophisticated at processing such news. With the rise of automa-

tion and high-frequency traders becoming the dominant class of inter-

mediaries, recent theoretical models (e.g., Hoffmann [2014], Budish,

Cramton, and Shim [2015]) suggest that liquidity providers play a larger

role in the price discovery of news through changes in quotes without the

need for trades.

Whether liquidity providers or liquidity takers are now generally respon-

sible for price discovery is key in identifying the appropriate methodology

to measure the intraday price formation following news events. Recently,

a growing number of studies in accounting use intraday data to examine

the impact of corporate news on stock prices. Such recent studies include

the analysis of high-frequency price formation around analyst conference

calls (Matsumoto, Pronk, and Roelofsen [2011]), SEC insider trading filing

releases (Rogers, Skinner, and Zechman [2016, 2017]), and analyst recom-

mendations (Altınkılıç and Hansen [2009], Li et al. [2015]). As argued by

Li et al. [2015], disentangling the effects of various news on prices using

daily data is challenging when multiple events occur on the same day, thus

the importance of using intraday data.

In this paper, we study how stock prices incorporate earnings news us-

ing high-frequency data in the after-hours market (4 p.m. to 9:30 a.m.)

from 2011–2015 for S&P 1500 firms. Our objectives are to understand how

quickly and through what mechanism, that is, through trading or changes

in quotes, prices mostly incorporate earnings news. Further, we shed light

on the price formation process in the after-hours market because 99.1%

of earnings are announced in this environment, which is characterized by

high trading costs, a high degree of information asymmetry, and low trad-

ing volume (Barclay and Hendershott [2003, 2004]).

how is earnings news transmitted to stock prices? 263

If today’s price discovery generally occurs through changes in quotes and

trading is sparse, using trade prices to compute returns can bias measure-

ments of price discovery in multiple ways. First and foremost, this will cause

trade returns to lag quote adjustments. Trade returns will then underes-

timate the speed and magnitude of price adjustments and more so when

markets are illiquid. After-hours trading is indeed sparse; the median num-

ber of trades following earnings announcements in the after-hours market

is only 16. Second, the use of trade returns can lead to a selection bias

because it forces the exclusion of events with no trades even if price dis-

covery takes place. For instance, we observe no trades following earnings

announcements during the after-hours market for approximately 20% of

our stock-earnings announcement observations.

We first examine the dynamics in cumulative returns computed from

midquote and trade prices following earnings announcements in the after-

hours market. We find midquote price adjustments to news occur quickly

but are followed by price drifts for stocks with large surprises. We further

show returns computed from trade prices significantly lag the adjustment

in midquotes for S&P MidCap 400 and S&P SmallCap 600 stocks, while

there is no significant difference for S&P 500 stocks. The fact trade prices

lag midquote prices provides a first indication that price discovery is most

likely attributable to liquidity providers updating prices through changes in

quotes without necessarily relying on trades.

Midquote price drifts following announcements might indicate slow

price discovery and profitable trading opportunities. Still, the ask and bid

prices are what indicate to liquidity takers whether trading on the earnings

surprise is profitable. We show bid–ask spreads are wide before announce-

ments, in most cases already encompassing the post-announcement closing

price at 4 p.m. of the following trading day. These large pre-announcement

spreads indicate liquidity providers face adverse selection from other fast

traders who can rapidly process the news.1Following announcements, the

bid and ask prices gradually tighten around the post-announcement price,

leaving no profits on average for informed liquidity takers that trade in the

direction of the surprise.

Another key characteristics of spreads following announcements is the

asymmetry in the speed of adjustment: following a positive surprise, liquid-

ity providers instantly adjust the best ask price to the earnings surprise but

only slowly adjust the best bid price, and vice versa for negative surprises.

It is the slow adjustment in the bid price following positive news and in

the ask price following negative news that generates midquote price drifts.2

1So and Wang [2014] show market makers demand higher expected returns prior to earn-

ings announcements because of increased inventory risks. Wide spreads before news are con-

sistent with the models of Hoffmann [2014], and Budish, Cramton, and Shim [2015].

2Li [2016] develops a trading strategy following earnings announcements in the after-hours

market and also documents a slow price adjustment in midquotes. We show why midquotes

slowly adjust to earnings news.