The spillover effects of a bright-line regulation: evidence from China

• Pages: 22-44
• DOI: https://doi.org/10.1108/IJAIM-05-2019-0058
• Date: 08 January 2020
• Published date: 08 January 2020
• Author: Yunling Song,Shihong Li,Ling Zhou
• Subject Matter: Accounting methods/systems,Accounting & Finance

The spillover ef‌fects of a bright-line

regulation: evidence from China

Yunling Song

School of Economics and Management, Inner Mongolia University,

Hohhot, China, and

Shihong Li and Ling Zhou

Department of Accounting, Anderson School of Management,

University of New Mexico, Albuquerque, New Mexico, USA

Abstract

Purpose –The purpose of this paper is to investigate the spillover effects of a bright-line disclosure

regulation that requiredChinese listed f‌irms to provide earnings forecasts if they anticipatedspecif‌ied, large

earningschanges.

Design/methodology/approach –The paper examines the discontinuity of the earnings change

distribution of f‌irmslisted on the Shenzhen Stock Market between 2010 and 2014. The paper f‌inds that f‌irms

no longer subject to the bright-line test still exhibited discontinuity in earnings change distribution. The

discontinuitylasted for at least three years with magnitude comparable to that of the f‌irms still subject to the

bright-line test. In addition,newly listed f‌irms that had never experienced the bright-line test showed similar

tendency to avoidthe same threshold. There is some evidence that thesef‌irms’avoidance of the 50 per cent

changeswas partly becauseof market pressure.

Research limitations/implications –Research on bright-line tests has to date focused on their

immediate and direct effects on f‌irms currently subject to such tests. This study f‌inds that a bright-line

disclosureregulation’sinf‌luence is not limited to the f‌irms directlygoverned by the regulation. It could lead to

widespread andlong lasting distortions in f‌inancial reportingbehaviors of f‌irms not currently subject to such

tests.

Practical implications –The paper has implications for regulators who study the economic

consequencesof bright-line regulations in general and analysts of the Chinesecapital market in particular.

Originality/value –This is the f‌irst empirical report that bright-line disclosure regulations affected the

f‌inancialreporting behavior of f‌irms that were not directly subject to the bright-line tests.

Keywords China, Earnings management, Bright-line regulation,

Discontinuities in earnings distributions, Earnings benchmark, Prospect theory

Paper type Research paper

1. Introduction

Regulations containingbright-line tests can induce managerial attempts to avoid the bright-

line thresholds to reduce compliance cost (Jiang and Wang, 2008;Gao et al.,2009).

Presumably, the avoidance behaviorwould cease with the removal of the bright-line tests if

compliance cost were the manager’s major concern. However, we have not seen any

empirical evidence. This paper investigates into this assumption by exploring a unique

regulatory setting, wherein a bright-line test that was applicable to all was removed from

some f‌irms but not the others. The answer to this question could shed light on the

JEL classif‌ication –M41, G18, G41

IJAIM

28,1

22

Received31 May 2019

Revised2 August 2019

Accepted29 August 2019

InternationalJournal of

Accounting& Information

Management

Vol.28 No. 1, 2020

pp. 22-44

© Emerald Publishing Limited

1834-7649

DOI 10.1108/IJAIM-05-2019-0058

The current issue and full text archive of this journal is available on Emerald Insight at:

https://www.emerald.com/insight/1834-7649.htm

behavioral complexity of managerial decisions and have signif‌icant implications to

policymakers and investors.

Before September 2011, f‌irms on the Chinese stock market were required to issue

earnings forecasts only if theyanticipated a loss, or a prof‌it after reporting a loss in the same

period of the previous year, or an earnings increase or decrease by 50 per cent or more.

Otherwise, they did not have to issue earnings forecasts. Thus, f‌irms were subject to three

bright-line thresholds: zero earnings and earnings increases or decreases by 50 per cent.

Prior literature documents unusuallylow (high) frequencies of earnings changes just below

(above) 50 per cent after but not before the implementation of these thresholds (Fan et al.,

2015;Huang et al., 2018). Researchers view the kink at the 50 per cent as a result of

managerial manipulation to circumvent the 50 per cent benchmark, i.e. f‌irms avoid

reporting earnings just below 50 per cent to evade the costs of mandatory management

earnings forecasts. No discontinuity at the þ50 per cent earnings change threshold was

detected probably because the costs of managing earnings down outweigh the costs of

issuing forecasts[1][2].

Effective in the last quarter of 2011, these thresholds of material changes were removed

for one type of f‌irms, those trading on the ChiNextmarket (SZSE, 2011a)[3]; they must issue

earnings forecasts regardless of the level of expected earnings. In contrast, other f‌irms

continued to follow the original rule. In this paper, we investigate whether ChiNext f‌irms

cease to avoid the 50 per cent earnings change threshold afterwards. We do not examine

f‌irms’behaviors aroundþ50 per cent earnings changes because the original bright-linetests

did not induce discontinuity at this threshold (Fan et al.,2015;Huang et al.,2018). After the

removal of the bright-line test, avoiding the 50 per cent earnings change would no longer

be able to save ChiNext f‌irms from making earnings forecasts. If evading mandatory

earnings forecasts is the main incentive for this avoiding behavior, it is plausible to expect

that the distribution kink at the 50 per cent earnings change will disappear for ChiNext

f‌irms. However, we f‌ind the same discontinuityin these f‌irms’earnings change distribution

at magnitude comparableto that of the f‌irms still subject to the test. Bright-line tests seem to

affect behaviors of f‌irms not currentlysubject to such tests, i.e. they have spillover effects. In

addition, the spillover effects are long lastingas the avoidance behavior persists for another

three years at least. This pattern suggests that althoughthe forecast related bright-line test

gave rise to the discontinuity at 50 per cent earnings change, forecasting costs might not

be the only reason behindChiNext managers’avoidance of this point.

We explore two plausible explanations for this phenomenon. First, ChiNext f‌irms may

avoid the 50 per cent earnings change to compete for resources on the stock market

against f‌irms that are still subject to bright-line tests. This is a rational strategy if, all else

equal, investors prefer stocks with slightly above 50 per cent earnings changes to those

with slightly below 50 per cent earnings changes or investors overreact to earnings

changes that cross the 50 per cent threshold because they use it as a heuristic cutoff in

valuation. We investigate this possibility by examining market responses to earnings

changes in the neighborhood of 50 per cent. We f‌ind that the average abnormalreturn for

earnings changes in therange (55%, 50%] was signif‌icantly lower than that for earnings

changes in the range (50%, 45%], consistent with the notion that ChiNext f‌irms avoid

the 50 per cent earnings changeto cater to investors.

Second, ChiNext f‌irms might exhibit the avoiding behavior simply out of inertia.

Organizational imprintingtheory proposes that organizational behaviors are affectedby the

initial conditions set at the foundation of the organization. An organization’s original

structures and practices designed to f‌it the initial conditions can persist because of inertia

and institutionalization (Stinchcombe, 1965;Marquis and Tilcsik, 2013). This explanation

Ef‌fects of a

bright-line

regulation

23