The Effect of Corporate Governance on Earnings Quality and Market Reaction to Low Quality Earnings: Korean Evidence

Abstract

This study investigates whether corporate governance mechanisms are associated with earnings quality, especially accurate earnings reporting, and whether investors react differently to inaccurate earnings according to governance strength. Earnings accuracy is one of the key factors affecting a firm’s sustainability in the sense that reported earnings provide information about a firm’s long-term sustainability and further are directly associated with a firm’s cost of capital. In this paper, we employ the independence of the board of directors (BOD) and foreign ownership as governance mechanisms associated with the earnings gap between audited and unaudited earnings. Using 1976 non-financial firm-year observations listed on the Korea Stock Exchange from 2013 to 2016, we find that the gap between unaudited earnings and actual earnings is smaller for firms with independent BODs and foreign ownership, suggesting that earnings accuracy is higher for firms with effective corporate governance. This study also examines how investors react to the earnings gap. Stock returns to the earnings gap are less negative for firms with independent BODs and are more negative as foreign ownership increases, implying that each mechanism of corporate governance has different effects.

1. Introduction

This study examines the impact of corporate governance on earnings quality and investors’ reactions to low-quality earnings. One of the key factors to a firm’s continuing operation is effective corporate governance, which involves fostering of several elements of sustainability including finance [1]. As corporate governance outcomes, we focus on a firm’s earnings quality to comprehensively reflect corporate governance characteristics [2].

High-quality earnings are considered to be more sustainable because reported earnings provide information about a firm’s future earnings [3]. Additionally, given that capital market participants make their decisions regarding investments based on reported earnings, the quality of earnings is directly associated with a firm’s cost of capital. This implies that earnings quality is one of the key factors associated with corporate sustainability. Therefore, from the sustainability view, prior studies have investigated earnings quality and how market participants react differently to the quality of earnings [4,5].

There are many studies suggesting that independence of the BOD and foreign ownership influence firms’ earnings quality [6,7,8,9,10,11,12,13,14,15,16,17,18]. However, these studies have utilized audited earnings to measure earnings quality, which presents a limitation to verifying firms’ ability or tendency for high-quality earnings disclosure. Audited earnings do not fully show the incentive or the ability of management to produce high-quality earnings, because audited earnings are adjusted by external auditors. To retain auditing quality, external auditors tend to be more conservative to offset any existing management bias in the unaudited earnings [19]. However, the unaudited earnings, which are generally lower in quality than the audited earnings, reflect the management’s intention and/or ability to produce earnings information compared to the audited earnings.

To overcome this measurement limitation, this study utilizes the unaudited earnings that firms initially file with regulators in the natural experimental setting in Korea. In 2014, Korean regulators amended the Act on External Audit of Stock Companies (hereafter ‘External Audit Act’) to clarify responsibility for the preparation of financial statements. The act mandated that all listed firms must submit unaudited financial statements to the Korean Securities and Future Commission when they submit them to external auditors. Before this amendment, firms had been accused of relying on auditors when they prepared their financial statements, leading auditors to perform audits on financial statements prepared by themselves (the so-called ‘self-review threat’), thus damaging the auditors’ independence.

The current amendment provides a suitable setting for our studies. First, the amendment allows us to observe the quality of pre-audit earnings prepared by firms. Several studies utilizing this amendment confirm that the earnings gap between unaudited earnings and audited earnings increases after the amendment [20,21], implying that the unaudited earnings in the post-amendment period show the firm’s ability to produce earnings and do not contain the component adjusted by external auditors. Second, this amendment allows us to test which factors are associated with a firm’s earnings quality by showing that the earnings gap before and after the amendment is smaller for firms with a higher quality of human capital such as more capable managers or staff related to firm’s internal control system [21,22]. Therefore, the amendment provides a means to test whether a firm’s governance mechanisms affect its earnings quality. Third, given that the current amendment reveals firms’ actual quality of earnings, investors are able to adjust their perceptions of earnings credibility. Therefore, the market reaction to the earnings gap can capture the investors’ valuation of uncovered earnings quality.

In this setting, we explore the BOD and foreign investor ownership in two layers, ex-ante intervention in earnings generating process and ex-post valuation of erroneous earnings. To be specific, we analyze whether the independent BOD is effective for monitoring and whether investors value BOD independency even after disclosed errors in earnings. Second, we determine whether foreign investors are effective monitors and perform ‘exit’ strategies for firms performing undesirable behaviors.

As ex-ante governance mechanisms, both the BOD and foreign investor ownerships are expected to positively affect earnings quality; consequently, the average earnings gap is not greater after amendment.

However, the effects of valuation of erroneous earnings would not be the same. The BOD has only one strategy, the ‘voice’ and investors value the effectiveness of BOD. Investors consider the independence of the BOD when they assess earnings quality [23,24,25] and attribute the earnings gap to inevitable uncertainty, rather than from management’s opportunism of firms with independent BOD [26].

In contrast, foreign investors have a different strategy for discipline management, that of ‘exit’ [27,28] because they can significantly affect stock prices through trading [2,29]. Given that sophisticated investors tend to penalize firms when signals cause them to change their belief regarding earnings quality [30,31,32], a firm’s earnings quality revealed after amendment might affect foreign investors’ beliefs, resulting in sales of their shares. In particular, foreign investors penalize firms more severely for deteriorated earnings quality because they have a limited source of information and are more likely to rely on publicly disclosed information [18].

Thus, we predict that the effects of BOD and foreign ownership will act in opposite directions, producing less negative stock returns for BOD independence and more negative returns for higher foreign ownerships.

By assessing the earnings gap between unaudited earnings and audited earnings before and after the External Audit Act amendment, this study finds that firms in which an independent board and sophisticated investors monitor management’s behavior are more likely to report accurate earnings, supporting that effective corporate governance can improve firms’ ability to produce high-quality earnings.

We also find different market reactions to the earnings gap in the post-amendment period based on BOD independence and foreign investors’ ownership as expected. This implies that investors highly value an independent BOD, and foreign investors sell their shares when firms fail to meet their expectations.

We contribute to the literature in several ways. First, we extend the literature on the association with corporate governance and the earnings quality. Because corporate governance is fundamental to address firm finance sustainability, the role of corporate governance in producing earnings quality can show how corporate governance affects firm sustainability. Using unaudited earnings, we provide direct evidence that board independence and foreign ownership influence the management’s incentive or ability to disclose more accurate earnings. Second, the current amendment to the External Audit Act in Korea allows us to provide evidence that investors’ perception of earnings credibility, as well as any changes in this perception, affect their valuation of a firm. Moreover, foreign investors’ severe punishment of firms due to erroneous earnings in the post-amendment period implies that indicating higher earnings quality is important for retaining foreign investment in emerging markets Second, more importantly, this is different from prior studies on the monitoring effect of foreign investors in that we focus on an emerging issue of investor ‘exit’ threat. Prior studies show that foreign investors achieve high-quality earnings by inducing or motivating management as an effective monitor. However, because foreign investors exert a more profound influence on firm valuation, we suggest that foreign ownership is a powerful motivation for maintenance of firm earnings quality. Using the amendment in Korea, we can distinguish the exit due to dissatisfaction with a firm performance from exit resulting from other motives such as liquidity or portfolio rebalancing needs. Therefore, this paper shows foreign ownership as one of key factors for firm sustainability.

The remainder of this paper is organized as follows: Section 2 reviews the literature and develops the study’s hypotheses. Section 3 describes the study’s research design. Empirical results are reported in Section 4. Finally, Section 5 concludes the paper.

2. Literature Review and Hypotheses Development

2.1. The Impact of Corporate Governance on Earnings Quality

There are two streams of research on the association between corporate governance and a firm’s earnings quality. One posits that corporate governance is one of the determinants of improved earnings quality, and the other posits that investors anticipate corporate governance when they reflect earnings in the stock price.

There is well-established literature on corporate governance and earnings quality. Prior studies have shown that corporate governance is one of the most important determinants of earnings quality [33,34]. Strong monitoring of corporate governance curbs management’s opportunistic behavior and thus causes management to report high-quality earnings [6,7,11,34].

Many studies have focused on the role of the board of directors (BOD) as one governance mechanism [6,7]. They suggest that effective monitoring by the BOD can motivate management to disclose better quality of earnings to investors, leading to precise earnings forecasts or lower earnings management [7]. In addition, the BOD is involved in the earnings generation process as an internal governance mechanism and, thus, is generally associated with disclosed earnings quality [8,9]. Karamanou and Vafeas (2005) [7] suggest that an effective board contributes to more precise management earnings forecasts, as these governance mechanisms can motivate managers to disclose better-quality information to investors. These studies are consistent with the notion that an independent BOD improves earnings quality by monitoring management’s opportunistic behavior. Therefore, as the most influential organization within a firm, the BOD is closely involved in the earnings disclosure process and monitors management behavior; this implies that the effectiveness of a BOD depends on its independence [8,35].

In developing countries, foreign investors are perceived to be sophisticated investors who are better at collecting, processing, and analyzing financial statements [10,11,12,13]. Using their capabilities and resources, sophisticated investors can monitor managers’ financial misreporting behaviors, especially in emerging market [36]. Vintilă et al. (2014) [37] which investigate the association between foreign ownership and firm value in Romania show that foreign ownership performs better monitoring role to the certain point of ownership ratio. They suggest the nonlinear relationship between foreign ownership and firm value. According to their analyses results, the firm value increases as foreign ownership goes up to 62.33% and turn to decreases. However, the ratio of foreign ownership in our sample firms is 9.1% at average, which is much lower compared to their data (see

Table 1. Descriptive statistics.

	Full Sample (n = 1976)					REG = 1 (n = 1488)	REG = 0 (n = 488)
Variable	Mean	Std. Dev.	25%	50%	75%	Mean	Mean	T-Value	(p-Value)
DUM_DIFF	0.362	0.481	0.000	0.000	1.000	0.382	0.301	−3.218	(0.001) ***
AMT_DIFF	0.097	0.317	0.000	0.000	0.037	0.106	0.070	−2.122	(0.017) **
BOD	0.388	0.137	0.250	0.333	0.500	0.390	0.382	−1.062	(0.144)
FOR	0.091	0.122	0.010	0.037	0.127	0.091	0.091	−0.021	(0.492)
SIZE	20.222	1.668	19.042	19.988	21.161	20.235	20.183	−0.603	(0.273)
LEV	0.496	0.201	0.341	0.507	0.638	0.491	0.512	1.945	(0.026) **
ROA	0.009	0.075	−0.008	0.017	0.044	0.011	0.002	−2.296	(0.011) **
LOSS	0.289	0.454	0.000	0.000	1.000	0.273	0.340	2.850	(0.002) ***
INVREC	0.300	0.145	0.195	0.291	0.394	0.296	0.311	2.019	(0.022) **
SGRW	0.042	0.256	−0.067	0.012	0.095	0.044	0.035	−0.703	(0.241)
BIG4	0.678	0.467	0.000	1.000	1.000	0.663	0.717	2.131	(0.017) **
FIRST	0.143	0.350	0.000	0.000	0.000	0.151	0.119	−1.771	(0.038) **
CON	0.832	0.374	1.000	1.000	1.000	0.832	0.834	0.104	(0.459)
CRATIO	1.224	1.106	0.626	0.897	1.378	1.242	1.168	−1.287	(0.099) *
LEADDAY	3.775	0.282	3.584	3.784	4.007	3.758	3.826	4.615	(0.000) ***

Note: Please refer the variable definitions in research model 1. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

. Descriptive statistics). Therefore, we conjecture the positive impact of foreign investors as firms’ corporate governance. In addition, we conduct additional tests by dividing samples into subgroups according to the level of foreign ownership in this paper (see Section 4.3). Ajinkya et al. (2005) [6] find that management provides more accurate earnings forecasts when institutional investor ownership is higher because sophisticated investors monitor the management’s information disclosure more closely Chung and Zhang (2011) [38] also suggest that sophisticated institutional investors have an incentive and capabilities to promote accurate earnings reporting and discourage financial misreporting.

Based on the above argument, foreign investors, as sophisticated investors, can positively affect firms’ earnings quality because they are equipped with more extensive resources to gather and analyze earnings information. Using their expertise and skills, foreign investors actively monitor firms by gathering information and adjusting pricing on earnings quality [14,39]. Because sophisticated investors closely monitor management’s information disclosure, firms are willing to provide high-quality earnings when sophisticated investors hold substantial ownership [6,15,40].

Especially, according to the home bias perspective, firms with foreign ownership have an incentive to provide more accurate earnings to investors to relieve foreign shareholder concern about the earnings quality of firms in emerging markets [16,17,41]. This is because foreign investors are at an information disadvantage relative to domestic investors, so they ask firms to disclose high-quality earnings to reduce the information risk. Since foreign investors have relatively limited source of information to domestic investors and are more likely to rely on the publicly disclosed information, the need to high-quality earnings is higher for foreign investors [18]. Specifically, Dou et al. (2018) [42] investigate the threat of ‘exit’ in regard to financial reporting quality. They find that reporting quality is greater when the exit threat increases, and the managers’ interests intensify the relationship between investor exit threat and financial reporting quality.

As discussed above, board independence and foreign ownership play roles of corporate governance, inducing firms to provide high quality earnings. Given that corporate governance positively affects a firm’s earnings reporting behavior, firms under such a governance mechanism would have reported accurate earnings in the pre-amendment period. The effect of corporate governance on earnings quality remained the same after the amendment because the amendment did not affect these firms’ earnings reporting process. Therefore, if corporate governance is associated with a firm’s higher earnings quality, the unaudited preliminary earnings would not differ from the audited actual earnings; thus, the earnings gaps of firms with good governance are not likely to increase after the amendment. This leads to the first hypothesis:

Hypothesis 1.

The increase in the earnings difference between unaudited earnings and actual earnings after the amendment is smaller for firms with effective corporate governance.

2.2. The Link between Market Reaction and Earnings Quality

The market reaction to the reported earnings partially contains the investors’ valuations of earnings reliability [43]. In other words, investors’ perceptions of high-quality earnings produce a positive effect on stock price.

When erroneous earnings are disclosed, investors reflect the error component of the earnings in the stock price because they cannot distinguish it from the true earnings. Upon a restatement of earnings, investors negatively react to the corrected earnings, and the magnitude of the reversal is much greater than that of stock returns for erroneous earnings [30,32]. Investors punish firms for their earnings corrections because they assume that the earnings were manipulated in favor of management and that the information content declined after the corrections [35].

However, negative market reactions to earnings gaps differ depending on the type of corporate governance, as the BOD is the internal firm organization that monitors the earnings reporting process and authorizes the disclosed earnings, whereas foreign investors are market participants trading the firm’s stocks.

Many studies find that investors value the independence of the BOD [24,25]. Vintilă et al. (2015) [44] provide analyses results of positive impact of independent board on firm value. Their results support that stock market participants consider the board independence into stock returns because they anticipate effective monitoring role of board of directors.

In terms of stock reaction to erroneous earnings reported under independent board of directors, Sohn et al. (2010) [26] also find that market reactions to earnings corrections are less negative for firms with independent BODs because investors are less likely to believe that the errors in earnings are derived from management’s opportunistic behavior. If investors believe that an earnings gap, even after the amendment, is not induced by management opportunism, a less severe negative reaction is expected.

Given that effective governance is an important factor in earnings quality, investors consider the independence of the BOD when they assess earnings quality [23,24,25]. Thus, while there is generally a negative reaction to earnings gaps, investors are less negative toward earnings gaps for firms with independent boards, because investors consider the earnings gap to come from inevitable uncertainty, rather than from management opportunism [26]. Apart from the market reaction to board independence, foreign investors, as entities who trade shares on the stock market, can significantly affect stock prices through trading [2,29]. Foreign investors have two alternatives to discipline managers, ‘voice’ and ‘exit’ [27]. Although many studies have focused on the role of ‘voice’ of informed traders by providing their monitoring role on earnings quality [45,46], a few studies recently have shown that investors sell their shares, decreasing the stock price and thus punishing the manager ex-post. Edmans (2014) [28] shows that sophisticated investors exert their influence by selling shares, i.e., exit. McCahery et al. (2016) [47] provide survey results that substantial investors have exited when they are dissatisfied with the governance or firm performance, suggesting investors consider ‘exit’ as an effective discipline strategy complementary to the ‘voice’ strategy.

This stream of research implies that investors can play a role of monitors by exiting, i.e., selling their shares, if managers take undesirable actions. If earnings gaps are reported after the amendment, foreign investors can identify the actual earnings quality of firms. In this case, the threat of exit does not work as an ex-ante strategy for management. Rather, foreign investors decide to take an ex-post strategy of ‘exit’, selling their shares to punish management, resulting in decreasing share prices.

If foreign investors perceive that the cost of exit is lower than the benefits, they are more likely to re-balance their portfolios by eliminating dissatisfying firms.

To influence stock prices significantly, the ownership should be large. Therefore, firms with higher foreign ownership are more likely to suffer negative stock returns.

Therefore, foreign investors discipline management by selling their shares when the earnings quality does not meet their expectations. Jeon and Sohn (2005) [18] show that only foreign investors react negatively when earnings are revealed to be overstated. Bardos et al. (2011) [30] find that negative reactions to misstated earnings are observed around the restatement date and may persist for several months, or even up to two years. If an earnings gap that is reported after the amendment suggests earnings credibility to investors, foreign investors are more likely to penalize firms that reported an earnings gap after the amendment; this is because the amendment allows them to see that firms perceived as having high-quality earnings, in fact, have low-quality earnings. Therefore, foreign investors are likely to penalize firms disclosing an earnings gap after the amendment.

Thus, when earnings are restated or corrected, investors react negatively because the firm’s earnings credibility has been damaged. However, the negative reaction should be lessened when investors believe that an independent BOD has authorized the earnings and that the misstated earnings are not the product of management’s opportunistic behavior. Meanwhile, foreign investors, known to require high-quality earnings, are more likely to sell their shares when they find that firms have reported inaccurate earnings under the amended act. These observations lead to the second hypothesis:

Hypothesis 2.

Negative stock returns due to earnings corrections after the amendment differ according to the effectiveness of corporate governance.

Hypothesis 2-1.

Negative stock returns due to earnings corrections after the amendment are smaller for firms with an independent BOD.

Hypothesis 2-2.

Negative stock returns due to earnings corrections after the amendment are larger for firms with higher foreign investor ownership.

3. Research Design

3.1. Research Model Specification

This study verifies H1 by using research model 1. Earnings differences are measured in two ways: one is a dummy variable indicating the difference in reporting (DUM_DIFF); the other is the absolute amount of the earnings difference (AMT_DIFF). The variable of interest is corporate governance (GOV). This study uses two corporate governance variables, board independence (BOD) and foreign ownership (FOR). BOD is the ratio of outside directors in the BOD, and FOR is the ratio of foreign investors’ shareholding. Another independent variable, REG, is an indicator variable equal to 1 if the year is in the post-amendment period, or 0 if the year is in the pre-amendment period. If the interaction of GOV × REG has a negative coefficient, the first hypothesis is supported. Several control variables drawn from prior studies are included [20,21,22,26]. We control for firm size (SIZE) and leverage ratio (LEV). As a firm’s profitability can affect the earnings gap, return on assets (ROA), net loss (LOSS), and firm growth (SGRW) are included. To control for firm complexity on accuracy of earnings, the amount of inventory and account receivables (INVREC), a dummy variable for consolidated earnings disclosure (CON), and current ratio (CRATIO) are included in the model. Because this paper measures DIFF as the difference between unaudited and audited earnings, the characteristics of the external auditor could be an important factor. Therefore, this paper includes factors related to the external auditor, a dummy variable indicating a Big 4 firm (BIG4), and the first year of audit (FIRST). Finally, the number of days from the fiscal year end is included because it is related to the accuracy of earnings. All variables (except the dummy variables) are winsorized at the top and bottom 1%, and firm-clustered standard errors are used. Thus,

Diff_i,t = β₀ + β₁GOV_i,t + β₂REG_i,t + β₃GOV_i,t × REG_i,t + β₄SIZE_i,t + β₅LEV_i,t + β₆ROA_i,t
+ β₇LOSS_i,t + β₈INVREC_i,t + β₉SGRW_i,t + β₁₀BIG4_i,t + β₁₁FIRST_i,t + β₁₂CON_i,t
+ β₁₃CRATIO_i,t + β₁₄LEADDAY_i,t + year dummies

(1)

where

• DIFF: differences between actual net income and unaudited net income
  • DUM_DIFF: dummy variable equal to 1 if differences exist and 0 otherwise
  • AMT_DIFF: |(actual net income − unaudited net income)/actual net income|
• GOV: governance variables
  • BOD: outside board ratio
  • FOR: foreign shares ratio
• REG: dummy variable equal to 1 if the period is after the amendment and 0 otherwise
• SIZE: ln (total asset)
• LEV: leverage ratio
• ROA: return on assets
• LOSS: equal to 1 if net income is less than 0 and 0 otherwise
• INVREC: (inventory + receivable)/total asset
• SGRW: sales growth
• BIG4: dummy variable equal to 1 if the auditor is a Big 4 auditor and 0 otherwise
• FIRST: dummy variable equal to 1 if auditor tenure is 1 and 0 otherwise
• CON: dummy variable equal to 1 if the firm reports consolidated financial statements and 0 otherwise
• CRATIO: current ratio (= current assets/current liabilities)
• LEADDAY: ln (calendar days from fiscal year end to earnings reporting date)

To examine H2, this paper uses model 2. CAR denotes the market response as the market-adjusted one-year cumulative abnormal returns from April at year t to March at year t + 1 [26]. The research model is run by a separate group classified using the median value of BOD or FOR. If ABS_DIFF × REG are not the same, H2 is supported. For H2-1, ABS_DIFF × REG should have a negative coefficient for less independent BOD sample. For H2-2 to be supported, the interaction term of ABS_DIFF × REG should have a negative coefficient for the higher foreign ownership sample. Thus,

CAR_i,t = β₀ + β₁ABSDIFF_i,t + β₂REG_i,t + β₃ABSDIFF_i,t × REG_i,t + β₄SIZE_i,t + β₅LEV_i,t + β₆ROA_i,t
+ β₇LOSS_i,t + β₈INVREC_i,t + β₉SGRW_i,t + β₁₀BIG4_i,t + β₁₁FIRST_i,t + β₁₂CON_i,t  
+ β₁₃CRATIO_i,t + β₁₄LEADDAY_i,t + year dummies            

(2)

where,

• CAR: cumulative abnormal returns from April at t to Mar at t + 1. We measure monthly cumulative abnormal returns (CAR) as

  $${\mathrm{AR}}_{\mathrm{it}}={\mathrm{R}}_{\mathrm{it}}-{\mathrm{R}}_{\mathrm{mt}}$$

  (3)

  $${\mathrm{CAR}}_{\mathrm{it}}=\sum {\mathrm{AR}}_{\mathrm{it}}$$

  (4)

  where, ${\mathrm{R}}_{\mathrm{it}}$ = monthly stock returns for firm i, ${\mathrm{R}}_{\mathrm{mt}}$ = monthly stock market returns

Other variables are as defined in research model 1.

3.2. Sample Selection

This study uses Korean firms listed in the Korea Stock Price Index from 2013 to 2016. The sample firms met the following conditions:

(1)

Listed in the KOSPI market from 2013 to 2016

(2)

Non-financial

(3)

Fiscal year ends in December

(4)

Disclosed unaudited earnings through the KRX disclosure system (KINDS)

(5)

Have data available on KIS-VALUE

The research period covers the years 2013 to 2016. Unaudited earnings dates reported through KINDS are hand-collected. Other data are acquired using KIS-VALUE. The final sample comprises 1976 firm-year observations. As the amendment came into force in 2014, the post-amendment period sample comprises 1499 firm-year observations, and the pre-amendment period sample comprises 488 firm-year observations.

4. Empirical Results

4.1. Descriptive Statistics and Correlation Analysis

Table 1 presents the descriptive statistics for the variables and a mean comparison by amendment period. For the full sample, the mean value of DMU_DIFF is 0.362, implying that 36.2% of observations report an earnings gap. In the sub-groups, the mean DUM_DIFF in the post-amendment period is 0.382, in contrast, the mean DUM_DIFF in the pre-amendment period is 0.301, indicating that the more firms report an earnings gap after the amendment, in line with results of prior studies [20,21].

The mean value of AMT_DIFF is 0.097 for the full sample. The mean value of AMT_DIFF is 0.106 in the post-amendment period and 0.070 in the pre-amendment period. Thus, the earnings gap increases 3.6% after the amendment.

The mean value of BOD is 0.388, which indicates that firms have 38.8% outside directors in their BODs. The mane value of FOR is 0.091, representing foreign investors own on average 9.1% of firm shares. The mean values of BOD and FOR pre- and post-amendment are not statistically different.

Table 2. Correlations.

	DUM DIFF	AMT DIFF	BOD	FOR	SIZE	LEV	ROA	LOSS	INVREC	SGRW	BIG4	FIRST	CON	CRATIO
AMT_DIFF	0.40	1.00
	(0.00)
BOD	−0.09	−0.06	1.00
	(0.00)	(0.01)
FOR	−0.16	−0.09	0.33	1.00
	(0.00)	(0.00)	(0.00)
SIZE	−0.12	−0.03	0.60	0.51	1.00
	(0.00)	0.16	(0.00)	(0.00)
LEV	0.14	0.14	0.18	−0.10	0.27	1.00
	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)
ROA	−0.16	−0.09	0.03	0.24	0.16	−0.30	1.00
	(0.00)	(0.00)	(0.22)	(0.00)	(0.00)	(0.00)
LOSS	0.13	0.12	−0.03	−0.19	−0.13	0.30	−0.67	1.00
	(0.00)	(0.00)	(0.13)	(0.00)	(0.00)	(0.00)	(0.00)
INVREC	−0.01	0.01	−0.16	−0.16	−0.27	0.17	−0.02	0.02	1.00
	(0.62)	(0.61)	(0.00)	(0.00)	(0.00)	(0.00)	(0.29)	(0.32)
SGRW	0.03	−0.00	−0.05	0.03	−0.01	−0.05	0.18	−0.16	−0.02	1.00
	(0.25)	(0.98)	(0.05)	(0.25)	(0.53)	(0.03)	(0.00)	(0.00)	(0.40)
BIG4	−0.02	0.01	0.28	0.25	0.47	0.10	0.09	−0.10	−0.09	0.00	1.00
	(0.29)	(0.77)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.99)
FIRST	0.10	0.05	−0.01	−0.07	−0.03	0.11	−0.05	0.08	−0.01	0.00	−0.07	1.000
	(0.00)	(0.05)	(0.76)	(0.00)	(0.14)	(0.00)	(0.02)	(0.00)	(0.72)	(0.99)	(0.00)
CON	0.04	0.04	0.13	0.18	0.30	0.08	0.05	−0.01	−0.04	0.04	0.10	−0.03	1.00
	(0.05)	(0.08)	(0.00)	(0.00)	(0.00)	(0.00)	(0.04)	(0.52)	(0.06)	(0.09)	(0.00)	(0.28)
CRATIO	−0.07	−0.08	−0.15	0.04	−0.30	−0.69	0.15	−0.14	0.02	−0.00	−0.14	−0.04	−0.06	1.00
	(0.00)	(0.00)	(0.00)	(0.07)	(0.00)	(0.00)	(0.00)	(0.00)	(0.35)	(0.93)	(0.00)	(0.06)	(0.00)
LEADDAY	0.03	0.04	−0.26	−0.27	−0.31	0.08	−0.15	0.18	0.07	−0.01	−0.15	0.05	0.15	0.02
	(0.21)	(0.07)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.57)	(0.00)	(0.04)	(0.00)	(0.34)

Note: Please refer the variable definitions in research model 1. p-values are reported in the parentheses.

shows the correlations among the variables used in this study. Both BOD and FOR are negatively correlated with DUM_DIFF and AMT_DIFF, implying that effective corporate governance is related to high-quality earnings in terms of accuracy.

4.2. Multivariate Analyses Results

Table 3. Test results of H1: Logit analysis using the dependent variable of DUM_DIFF.

	GOV = BOD		GOV = FOR
	Model 1	Model 2	Model 1	Model 2
Intercept	6.666	6.165	6.265	6.081
	(4.129) ***	(3.737) ***	(3.854) ***	(3.717) ***
GOV	−0.996	0.467	−1.755	−0.329
	(−1.752) *	(0.532)	(−2.484) **	(−0.367)
REG		1.268		0.703
		(3.619) ***		(4.494) ***
GOV × REG		−1.835		−1.996
		(−2.159) **		(−1.856) *
SIZE	−0.317	−0.321	−0.301	−0.298
	(−5.223) ***	(−5.313) ***	(−4.971) ***	(−4.893) ***
LEV	2.486	2.507	2.276	2.287
	(5.539) ***	(5.567) ***	(5.067) ***	(5.076) ***
ROA	−3.042	−3.092	−2.755	−2.791
	(−2.962) ***	(−3.016) ***	(−2.718) ***	(−2.745) ***
LOSS	0.029	0.014	0.023	0.014
	(0.187)	(0.089)	(0.149)	(0.09)
INVREC	−1.492	−1.493	−1.452	−1.456
	(−3.094) ***	(−3.101) ***	(−3.021) ***	(−3.028) ***
SGRW	0.385	0.398	0.408	0.413
	(1.944) *	(2.001) **	(2.034) **	(2.051) **
BIG4	0.34	0.35	0.337	0.334
	(2.390) **	(2.453) **	(2.397) **	(2.369) **
FIRST	0.391	0.384	0.375	0.365
	(2.949) ***	(2.893) ***	(2.819) ***	(2.746) ***
CON	0.671	0.669	0.704	0.712
	(3.808) ***	(3.789) ***	(4.062) ***	(4.103) ***
CRATIO	0.085	0.086	0.088	0.09
	(1.161)	(1.175)	(1.204)	(1.222)
LEADDAY	−0.674	−0.668	−0.695	−0.696
	(−2.810) ***	(−2.778) ***	(−2.918) ***	(−2.919) ***
Year dummies	Included	Included	Included	included
Chi square	123.18	131.2	124.34	126.99
Pseudo R2	0.0749	0.0765	0.0772	0.0786
n	1976	1976	1976	1976

Note: Please refer the variable definitions in research model 1. T-value are presented in the parentheses. *, **, and *** represent significance level at the 10%, 5%, and 1%, respectively.

and

Table 4. Test results of H1: Regression analysis using the dependent variable of AMT_DIFF.

	GOV = BOD		GOV = FOR
	Model 1	Model 2	Model 1	Model 2
Intercept	0.298	0.224	0.301	0.277
	(1.495)	(1.085)	(1.474)	(1.348)
GOV	−0.155	0.047	−0.133	0.015
	(−2.215) **	(0.373)	(−2.807) ***	(0.162)
REG		0.154		0.072
		(3.045) ***		(3.254) ***
GOV × REG		−0.259		−0.212
		(−2.065) **		(−1.916) *
SIZE	−0.013	−0.014	−0.016	−0.015
	(−1.988) **	(−2.077) **	(−2.408) **	(−2.297) **
LEV	0.24	0.243	0.218	0.218
	(3.398) ***	(3.436) ***	(3.048) ***	(3.047) ***
ROA	−0.017	−0.022	0.019	0.014
	(−0.128)	(−0.169)	(0.143)	(0.104)
LOSS	0.061	0.059	0.061	0.06
	(1.906) *	(1.865) *	(1.907) *	(1.876) *
INVREC	−0.074	−0.074	−0.069	−0.07
	(−1.475)	(−1.481)	(−1.378)	(−1.379)
SGRW	0.02	0.022	0.023	0.024
	(0.906)	(0.99)	(1.052)	(1.06)
BIG4	0.037	0.039	0.037	0.037
	(2.000) **	(2.069) **	(2.005) **	(1.968) **
FIRST	0.009	0.007	0.007	0.006
	(0.409)	(0.341)	(0.336)	(0.265)
CON	0.037	0.037	0.041	0.042
	(2.100) **	(2.087) **	(2.342) **	(2.386) **
CRATIO	0.004	0.005	0.004	0.004
	(0.512)	(0.526)	(0.455)	(0.463)
LEADDAY	−0.021	−0.019	−0.02	−0.02
	(−0.674)	(−0.622)	(−0.656)	(−0.662)
Year dummies	Included	Included	Included	included
F-value	5.666	5.612	5.699	5.557
adj R2	0.033	0.035	0.032	0.033
n	1976	1976	1976	1976

Note: Please refer the variable definitions in research model 1. T-value are presented in the parentheses. *, **, and *** represent significance level at the 10%, 5%, and 1%, respectively.

shows the analysis results for the H1 test. Though the hypotheses concern the effect of corporate governance on the earnings gap after the amendment, the analyses are conducted using model 1, without the REG variables and the interaction term GOV × REG, to confirm that corporate governance is positively associated with earnings quality.

Table 3 shows the logistics analysis results using the dummy variable for reporting an earnings gap. In model 1, both BOD and FOR have negative coefficients, implying that earnings gaps are less likely to be reported as outside directors fill more seats on BODs and the shareholding ratio of foreign investors increases. This result is consistent with the notion that earnings quality is higher when corporate governance is effective. In model 2, including REG and GOV × REG, the interaction variable GOV × REG has statistically negative coefficients for both BOD and FOR, supporting H1. Therefore, firms monitored by independent BODs and foreign investors are less likely to report an earnings gap after the amendment.

Table 4 shows the results using the AMT_DIFF. In model 1, both BOD and FOR have significant negative coefficients, which is not qualitatively different from the results with DUM_DIFF. Thus, an independent board and greater ownership of foreign investors play roles in enhancing earnings quality as effective monitors. In model 2, REG has a positive coefficient, indicating that the frequency of reporting an earnings gap increases after the amendment. This is consistent with prior studies [20,21,22], indicating that the current amendment shows firms’ actual earnings quality. The variable of interest, GOV × REG also has a negative coefficient for both BOD and FOR. These results support our hypothesis 1 that firms with effective corporate governance can produce high of earnings after the amendment.

In conclusion, firms monitored by more outside directors and higher foreign ownership are more likely to report accurate unaudited earnings, and their earnings gap is smaller. Specifically, the results after the amendment show more clearly that firms with effective corporate governance produce high-quality earnings.

The coefficients of most of the control variables do not differ from those of prior studies [20,21,22]. Larger firms are less likely to report accurate earnings. Financially constrained firms are more likely to report an earnings gap (which is larger), while more profitable and higher-growth firms are not likely to do so. BIG4 has positive coefficients, suggesting that Big4 auditors are more likely to detect and report errors. As auditors make more of an effort in the first year of an audit, FIRST has a positive coefficient, but the size of gap is not significant.

Table 5. Test results for H2.

	Hypothesis 2-1		Hypothesis 2-2
	BOD > Median	BOD < Median	FOR > Median	FOR < Median
Intercept	0.399	1.785	0.754	1.481
	(1.240)	(5.107) ***	(2.648) ***	(4.000) ***
ABS_DIFF	−0.028	0.116	0.078	0.001
	(−0.618)	(1.296)	(1.184)	(0.020)
REG	−0.099	−0.197	−0.199	−0.108
	(−2.909) ***	(−5.616) ***	(−6.775) ***	(−2.675) ***
ABS_DIFF × REG	−0.076	−0.231	−0.196	−0.106
	(−0.937)	(−2.496) **	(−2.582) **	(−1.312)
SIZE	−0.041	−0.080	−0.034	−0.088
	(−4.365) ***	(−5.230) ***	(−3.948) ***	(−6.122) ***
LEV	0.079	0.335	0.202	0.237
	(0.854)	(3.260) ***	(2.314) **	(2.246) **
ROA	1.455	0.869	1.283	1.027
	(6.131) ***	(2.785) ***	(4.446) ***	(3.950) ***
LOSS	−0.024	−0.098	−0.031	−0.100
	(−0.679)	(−2.442) **	(−0.827)	(−2.614) ***
INVREC	0.204	−0.128	0.019	0.036
	(1.856) *	(−1.285)	(0.222)	(0.348)
SGRW	0.137	0.115	0.152	0.096
	(2.548) **	(1.555)	(2.684) ***	(1.183)
BIG4	−0.031	−0.038	−0.070	−0.007
	(−0.852)	(−1.487)	(−2.444) **	(−0.258)
FIRST	0.063	−0.042	0.038	−0.030
	(1.758) *	(−1.152)	(1.191)	(−0.777)
CON	−0.023	0.053	0.014	0.026
	(−0.589)	(1.652) *	(0.410)	(0.739)
CRATIO	−0.015	0.005	−0.004	0.003
	(−1.244)	(0.365)	(−0.425)	(0.155)
LEADDAY	0.135	−0.052	−0.001	0.064
	(2.643) ***	(−1.024)	(−0.019)	(1.152)
F-value	8.697	11.106	8.381	11.014
adj R2	0.152	0.136	0.136	0.139
n	913	1063	989	987

Note: Please refer the variable definitions in research model 1. T-value are presented in the parentheses. *, **, and *** represent significance level at the 10%, 5%, and 1%, respectively.

shows the results for H2. We conjecture that market reactions differ according to independence of BOD and foreign ownership. If investors perceive that the BOD effectively monitors management and its earnings reporting procedure both pre- and post-amendment, negative reactions would be smaller than in firms monitored by less independent BODs.

The analysis results for board independence are as follows. For an independent board sample (BOD > median), the coefficient of ABS_DIFF and ABS × REG are negative but statistically not significant. For the non-independent group (BOD < median), the coefficient of ABS_DIFF is positive and marginally significant, and ABS_DIFF × REG has a significantly negative coefficient. This result indicates that investors differentiate firms’ earnings gaps according to board independence and react to such gaps more severely when the earnings are produced under less independent boards since amendment is effective, supporting H2-1.

Our results are in line with those of Sohn et al. (2010) [26] in that investors attribute incorrect earnings to inevitable uncertainty rather than manager opportunism derived from less effective board monitoring. Therefore, negative market reaction is observed only for a less independent board.

Regarding foreign investor ownership, we split the sample based on the median value of foreign ownership. The results are as follows. ABS_DIFF is not significant for either subsample. The coefficient of ABS_DIFF × REG is negatively significant for the group with higher ownership but insignificant for the group with less foreign ownership, suggesting H2-2.

This result can be interpreted that foreign investors more severely punish firms reporting an earnings gap after the amendment, supporting that foreign investors adjust their valuation of earnings for less credible firms. Also, these results are evidence of the ‘exit’ strategy of foreign shareholders. As prior studies [2,28,47] suggest, the threat of exit of foreign investors is realized if they observe low-quality earnings. These results distinguish foreign shareholders from BOD in that foreign shareholders are actually trading shares of firms and have the additional strategy of ‘exit’.

Taken together, the results for the market reaction to earnings gap emphasize that investors recognize the earnings gap after the amendment as a signal of firm actual earnings quality. However, investors differently react to earnings correction based on their perception of board independence and their previous belief or attitude on firm earnings quality.

4.3. Additional Test

We conjecture that foreign investors punish firms by selling shares when they are dissatisfied with earnings quality. Even though we provide results supporting our hypothesis, such an exit strategy might not be prevalent for all firms owned by foreign investors because investors have a preference on the strategies which are voices or exit. McCahery et al. (2016) [47] suggest that investor ownership size is a determinant to actual exit; however, it has opposing effects on use of the exit strategy. Foreign investors with large ownership are more likely to sell their shares when dissatisfied; however, it may difficult to do because of the large price impact. Thus, there is an optimal ownership size that is an effective threat to managers.

According to this suggestion, we divide our sample into four groups based on foreign ownership amount and conduct analyses. The results presented in

Table 6. Additional test: analyses by subgroup according to foreign ownership.

	Group 1 (>12.6%)	Group 2 (>3.7%)	Group 3 (>1.3%)	Group 4 (≥0%)
	b/t	b/t	b/t	b/t
Intercept	0.558	1.255	1.229	1.714
	(1.632)	(2.833) ***	(2.461) **	(3.163) ***
ABS_DIFF	−0.002	0.125	−0.063	0.020
	(−0.021)	(2.383) **	(−0.590)	−0.206
REG	−0.16	−0.232	−0.147	−0.069
	(−3.893) ***	(−5.264) ***	(−2.842) ***	(−1.080)
ABS_DIFF × REG	−0.071	−0.255	−0.024	−0.145
	(−0.558)	(−4.049) ***	(−0.199)	(−1.399)
SIZE	−0.026	−0.063	−0.081	−0.097
	(−2.589) **	(−3.963) ***	(−4.156) ***	(−4.591) ***
LEV	0.193	0.307	0.255	0.291
	(1.917) *	(2.118) **	(2.002) **	(1.804) *
ROA	1.704	0.660	1.44	0.726
	(4.949) ***	(1.344)	(4.491) ***	(1.876) *
LOSS	0.018	−0.089	−0.06	−0.128
	(0.401)	(−1.393)	(−1.146)	(−2.199) **
INVREC	−0.021	0.016	0.059	0.016
	(−0.210)	(0.120)	(0.408)	(0.107)
SGRW	0.117	0.205	0.240	−0.010
	(1.609)	(2.346) **	(1.878) *	(−0.106)
BIG4	−0.027	−0.070	0.022	−0.041
	(−0.628)	(−1.760) *	(0.592)	(−0.943)
FIRST	0.031	0.060	−0.023	−0.035
	(0.749)	(1.207)	(−0.410)	(−0.658)
CON	−0.011	0.029	0.050	0.006
	(−0.212)	(0.692)	(1.095)	(0.123)
CRATIO	−0.009	0.009	0.03	−0.013
	(−0.891)	(0.505)	(0.972)	(−0.545)
LEADDAY	0.004	0.010	0.071	0.061
	(0.083)	(0.157)	(0.862)	(0.805)
F-value	4.831	7.563	7.784	4.842
adj R2	0.109	0.158	0.174	0.107
n	495	494	493	494

Note: Please refer the variable definitions in research model 1. T-value are presented in the parentheses. *, **, and *** represent significance level at the 10%, 5%, and 1%, respectively.

are as follows. We found no significant market reactions to the earnings gap after the amendment in Group 3 or Group 4, which are smaller ownership groups. This is consistent with the results in Table 5.

For high ownership groups, the coefficients of ABS_DIFF × REG are different across sub-groups. Group 1, a subsample with foreign ownership greater than 12.6% has a non-significant coefficient. Group 2, a subsample with foreign ownership greater than 3.7% and less than 12.6% has a significantly negative coefficient on ABS_DIFF × REG. These results indicate that finding in Table 5 is driven by Group 2.

These results show that when foreign ownership is too large, they cannot easily exit firms. Specifically, the group of foreign ownership over 3.7% and less than 12.6% is more likely to take an action of exit upon the signal of the lower quality of earnings. Our results are in line with those of McCahery et al. (2016) [47], which show that investors believe that at least 2% ownership is needed to perform exit as an effective strategy and exit threat of investors with larges ownership is less credible. Their survey results show that institutional investors rarely own more than 10%, and the respondents who answer that ownership above 10% is effective to exit threat is small. Even though McCahery et al. elaborate about the ownership of individual institution shareholder, this result can be interpreted to mean that an exit strategy is effective when foreign investors own from 2% to 10% of shares. This is because that the foreign investors’ shareholding ratio is likely to be high when individual foreign investors own large ownership of firms.

5. Conclusions

This study investigates whether board independence and foreign ownership are associated with the disclosure of high earnings quality and whether market reactions to corrected earnings differ according to board independence and foreign ownership. Even though numerous prior studies investigate the role of board independence, they generally measure the effects on the auditor’s effort or auditor change, rather than measuring its effects directly. Contrary to previous studies, we use unaudited earnings, free of external auditors’ intervention, and focus on the management’s incentive or ability to produce earnings quality. Using the amendment of Korea’s External Audit Act in 2014, this paper finds that increases in the earnings gaps between unaudited earnings and actual earnings after the amendment are smaller for firms with independent BODs and higher foreign investor ownership. This study provides direct evidence that an independent board and ownership by foreign investors are effective corporate governance mechanisms to enhance and/or sustain high-quality earnings.

Moreover, negative stock returns due to corrections after the amendment are lower for firms with independent BODs, implying that investors can justify corrections made by firms with good governance. As foreign investor ownership increases, however, the negative market returns grow more intense because foreign investors, as active market participants, raise questions about the credibility of the firm’s earnings and its ability to produce them at a high quality after the amendment. Our results show that investors consider independence of the BOD when they value erroneous earnings. However, foreign investors exit when dissatisfied with firm earnings quality.

In sum, we show an ex-ante effect of corporate governance on earnings quality and an ex-post market reaction to errors in earnings.

This study contributes to the literature in several ways. First, by using unaudited earnings data, this paper attempts to provide direct evidence concerning firms’ earnings quality before auditors become involved. Using the earnings gap between unaudited and actual earnings, this study shows that effective corporate governance can improve firms’ ability to produce high-quality earnings. This finding differs from those of prior studies, most of which have focused on appointments of, or switching to, high-quality external auditors to improve a firm’s earnings quality.

Second, our findings concerning market reactions according to corporate governance type show how differently various investors react to low-quality earnings. Investors can justify inaccurate earnings reporting when they believe that the firm is monitored by an independent board. Foreign investors, however, penalize firms providing low-quality earnings more severely when they question the firm’s past earnings credibility. Therefore, firms with higher foreign ownership care more about shareholders’ need for higher earnings quality than do firms with lower foreign ownership. Overall, this study shows that effective corporate governance improves a firm’s ability to produce high-quality earnings, and these high-quality earnings benefit investors.

In this paper, we underline that effective corporate governance is a key determinant to high-quality earnings, and each governance mechanism has a unique strategy inducing managers to provide high-quality earnings. Especially, we revisit independence of the BOD and foreign ownership, whose roles are in question in an emerging market. Given that corporate governance is closely related to firm sustainability, and that earnings quality is the financial outcome of corporate governance, this paper sheds light on the role of corporate governance to strengthen firm sustainability in an emerging market.

This paper has a political implication. The amendment of the external act in Korea was conducted to improve firms’ ability to generate higher quality of earnings and revealed the actual quality of earnings of firms. After the amendment, managements do not rely on external auditors when they prepare financial statements, leading to the improvement of earnings quality. Our results imply that firms’ sustainable development is induced by regulations of enhancing board independence and lowering information asymmetry between firms and investors. Therefore, our paper gives implications to regulators who are interested in the improvement of earnings quality and encouraging foreign investment.

This paper has several limitations. First, because we use only firms in Korea, the results may not be generalized to other countries. Second, we focus on the short-term effects of the amendment. Even though changes in investor perception upon amendment are instant, it is worth investigating the long-term effects. Third, even though there are studies supporting that unaudited earnings after the amendment is free of auditors’ intervention, there still exists the possibility of external auditors’ intervention in generating unaudited earnings.