Foreign Investors,External Monitoring,and Stock Price Crash Risk

Abstract

We investigate whether foreign investors help to reduce local firms’ future stock price crash risk through their external monitoring. We find that the entrance of foreign investors is associated with a significant reduction in local firms’ future crash risk. Further investigation reveals that foreign investors help to improve local firms’ financial reporting quality from the perspectives of accrual quality, conservatism, and annual report tone management. The evidence is consistent with our conjecture that foreign investors play an important external monitoring role, which reduces managerial bad-news hoarding and thereby lowers local firms’ future crash risk. We also find that the crash risk–reducing role of foreign investors is more pronounced when foreign investors are more familiar with the institutional background of the host country, when they have stronger incentives to monitor local firms, and when local firms have higher governance efficacy. A variety of robustness checks reveals that our results are unlikely to be driven by potential endogeneity.

Keywords

foreign investors stock price crash risk financial reporting quality institutional distance

Introduction

In this study, we investigate a hitherto under-researched question of whether and how foreign investors in the emerging market influence stock price crash risk of local firms. Prior research provides two opposing perspectives about how foreign investors would affect local firms’ information environment. One strand of research claims that foreign investors have a positive impact on the information environment, because they have more resources and expertise to process information about local firms and monitor them more effectively than domestic investors (e.g., Aggarwal, Erel, Ferreira, & Matos, 2011; Grinblatt & Keloharju, 2000). The other strand of research argues that foreign investors are less informed about local firms than local investors (e.g., Choe, Kho, & Stulz, 2005; Dvořák, 2005; Hau, 2001), and have to bear additional costs to overcome their informational disadvantage. Thus, foreign investors are less effective in monitoring local firms than domestic investors. Given that the impact of foreign investors on the information environment of local firms is inconclusive, it is ultimately an empirical question whether their presence reduces local firms’ future crash risk. Our study extends prior research on the efficacy of external monitoring by foreign investors. In so doing, our analysis focuses on two research questions:

Research Question 1: Whether foreign investors play an effective monitoring role and reduce local firms’ future crash risk.

Research Question 2: Whether and how the influence of foreign investors over local firms’ future crash risk varies across firms.

In examining the first research question, we focus on whether local firms’ crash risk reduces significantly after the entrance of foreign investors. If foreign investors do play an important role of external monitoring, it would be difficult for managers to withhold bad news. Thus, local firms’ future crash risk should be significantly reduced afterward, given that managerial bad-news hoarding is an important cause of crash risk (e.g., Jin & Myers, 2006; Piotroski, Wong, & Zhang, 2015). However, as aforementioned, there are also studies arguing that the effectiveness of foreign investors’ external monitoring is questionable, given that they are not familiar with local institutional background and have to bear additional costs associated with collecting and analyzing information. Thus, whether foreign investors help to reduce local firms’ crash risk is ultimately an empirical question.

Our second research question is whether foreign investors’ influence over local firms’ future crash risk varies across firms. We examine the variation from three perspectives: (a) the institutional distance between the host and home countries of foreign investors, (b) the strength of their monitoring incentives, and (c) the effectiveness of local firms’ governance system. We expect the influence of foreign investors on local firms’ future crash risk to vary conditional on these factors.

We perform empirical investigation in the A-share market of China for the following four reasons. First, over the past decades, the Chinese economy has attracted more and more investors worldwide. The Chinese government offers many preferential policies to foreign investors to boost foreign equity investments, including the introduction of Qualified Foreign Institutional Investors (QFII) and RMB Qualified Foreign Institutional Investors (RQFII) schemes.¹ Studying the role of foreign investors in monitoring local firms in China is therefore interesting and important in its own merit. Second, shareholder protection is still weak in the Chinese market. The few mechanisms to control the discretionary power of controlling shareholders make the external monitoring of foreign investors more important in the Chinese market than in other developed markets, where governance efficacy of listed firms is high and foreign investors’ influence on local firms might be limited. Third, as foreign investors in the Chinese stock market come from many different countries, there is a wide variation in their institutional backgrounds. This facilitates our investigation into the impact of institutional distance on the effectiveness of external monitoring. Finally, China experienced an astonishing market crash in June 2015, with the A-share market index plunged by nearly 30% and more than US $3 trillion of market value being wiped out. Studying potential factors that could help to reduce stock price crash risk in the Chinese market would have implications for market participants in China and in other emerging markets as well.

In examining foreign investors’ influence over local firms’ future crash risk, we first show that both the presence of foreign investors and their ownership are significantly negatively related to firms’ future crash risk. Moreover, we show that an increase in foreign ownership is followed by a significant reduction in local firms’ future crash risk, while a decrease in local firms’ crash risk is not followed by a significant change in foreign ownership. The evidence indicates that foreign investors help to reduce local firms’ crash risk, and that the findings are not likely to be driven by endogeneity, namely foreign investors’ tendency to invest in stocks with low crash risk in the first place.

We then employ the difference-in-differences (DiD) test design to examine the influence of foreign investors on local firms’ crash risk. We compare the change in the crash risk of local firms with foreign ownership (the treatment group) after the entrance of foreign investors with that of the local firms without foreign ownership (the control group) over the same period. The crash risk of the treatment group is comparable to that of the control group prior to the entrance of foreign investors. However, after the entrance of foreign investors, the treatment group experiences a significantly greater reduction in crash risk relative to the control group.

We further examine whether the reduction in crash risk is, at least partially, driven by the improved financial reporting quality after the entrance of foreign investors. We measure financial reporting quality based on accrual quality, conservatism, and annual report tone management. After the entrance of foreign investors, the improvement of financial reporting quality is significantly greater for the treatment group than for the control group over the same period. It supports our argument that foreign investors play an important external monitoring role and help to constrain managerial bad-news hoarding in the Chinese stock market.

We take one step further to examine the cross-sectional variation in such an influence conditional on three factors: the institutional distance between the home and host countries of foreign investors, the strength of foreign investors’ monitoring incentives, and the effectiveness of local firms’ governance system.

Kim, Li, Luo, and Wang (2015) argue that the institutional distance exacerbates the difficulty faced by foreign investors in monitoring local firms, which in turn decreases local firms’ financial reporting quality. They consider two types of institutional distance: formal and informal distances. Formal institutional distance focuses on a set of political, economic, and contractual rules and laws, while informal institutional distance originates from cultural differences and involves rules embedded in values, norms, and beliefs. We show that both types of institutional distance, which add to foreign investors’ unfamiliarity with local institutional background, constrain these investors’ ability to reduce local firms’ future crash risk.

Second, we examine the strength of foreign investors’ monitoring incentives from three perspectives: investor type, the length of investment period, and product market relation. There are mainly two types of foreign investors in our sample: financial institutions and nonfinancial corporations.² Foreign financial institutions usually are short-term investors and purchase shares through QFII to earn trading profits, while foreign nonfinancial corporations tend to be long-term investors and form strategic partnership with local firms. Consistently, we find that the reduction in local firms’ crash risk, after the entrance of foreign investors, is more evident for foreign nonfinancial corporations. Our results are in line with Li, Nguyen, Pham, and Wei (2011), which documents that foreign nonfinancial corporations are more committed to their investments than foreign financial institutions.

Furthermore, we classify foreign investors into long-term and short-term investors, and identify those with a product market relation with the local firm that they invest in. We expect long-term foreign investors to have stronger monitoring incentives, relative to short-term foreign investors, as it takes time for better external monitoring to produce a positive impact on firm value. Foreign investors’ external monitoring incentives, however, are expected to be weak if they maintain a product market relation with local firms that they invest in, as active intervention could potentially jeopardize such a relation. The empirical results are consistent with our expectations.

Third, we investigate the impact of local firms’ governance efficacy. While external monitoring may be more important when existing governance of local firms is weak (e.g., Aggarwal et al., 2011), there are other forces that may push the impact the other way around. For instance, Fang, Maffett, and Zhang (2015) contend that whether external monitors could bring about significant changes in local firms ultimately depends on the difficulty in overcoming existing institutional arrangements that insulate controlling shareholders from external disciplinary forces. Kho, Stulz, and Warnock (2009) conjecture that weak investor protection in emerging markets leads to a higher level of insider ownership in general, constraining the influence of foreign investors. Therefore, in emerging markets, it is possible that foreign investors are better able to exert their influence when the existing corporate governance system in local firms is relatively effective and thereby external oversight is not being largely hampered.

We capture the effectiveness of local firms’ governance system using analyst coverage, state ownership, and regional institutional development. Yu (2008) asserts that analysts play a role in enhancing firms’ internal corporate governance. State ownership, which has been widely documented as a determinant of firms’ internal governance effectiveness, is expected to be associated with low governance quality (e.g., Borisova, Brockman, Salas, & Zagorchev, 2012). Regional institutional development, which is associated with strong monitoring effect of institutional investors as well as investor protection, is expected to promote governance efficacy (e.g., Li & Qian, 2013; Yu, Zhang, & Zheng, 2015).

Subsample tests show that foreign investors’ negative influence over local firms’ future crash risk is stronger for firms with analyst coverage, with lower state ownership, or located in more institutionally developed regions. The evidence confirms that foreign investors are better able to play their external monitoring role and reduce crash risk when existing governance system is more effective.

Some might concern that our results are driven by endogeneity. As aforementioned, we find that the change in firms’ foreign ownership is negatively related with the change in future crash risk, while the change in firms’ crash risk does not significantly affect the change in future foreign ownership. Moreover, results of DiD tests reveal that prior to the entrance of foreign investors, the average crash risk of treatment firms is comparable to, if not higher than, that of control firms. Thus, the results that local firms’ crash risk decreases significantly after the entrance of foreign investors could not be attributed to their tendency to select and invest in local firms with lower crash risk in the first place. Furthermore, in robustness checks, we control firm fixed effects, and all results remain unaltered.

Our study contributes to the literature on the consequences of foreign investments. Existing studies have shown that foreign investors affect firm value and performance (Ferreira & Matos, 2008), improve governance (Aggarwal et al., 2011), and discipline financial reporting (Beuselinck, Blanco, & García Lara, 2017; Fang et al., 2015). To our knowledge, our study is one of the few, if not the first, to establish foreign investors as an economically important determinant of local firms’ crash risk in an emerging stock market like China, and show that their influence over local firms’ crash risk could be affected by forces including the institutional distance between the home and host countries of foreign investors, the strength of their monitoring incentives, and the effectiveness of local firms’ governance system.

Our study echoes Hu, Ke, and Yu (2018) and Trabelsi (2018). They argue that transient institutions analyze and process negative firm information precisely. If the trading of transient institutions reveals negative signals sufficiently, it should help to reduce crash risk. Our results, however, show that foreign financial institutional investors, which focus on short-term trading profits and belong to transient institutions, actually have weaker impacts than foreign nonfinancial corporations on local firms’ future crash risk. The discrepancy might result from the fact that we focus on foreign investors while Hu et al. (2018) and Trabelsi (2018) consider all institutional investors. The foreign financial institutional investors in our sample mainly invest through QFII and have quite low average ownership in local firms (around 1.17%). Thus, they do not have strong incentives to monitor local firms and are not likely to impose considerable impacts on prices and reveal their information accordingly, resulting in their limited influence over firms’ crash risk.

The article proceeds as follows. The “Data and Method” section introduces sample, data, and variable construction. The “Foreign Investors, External Monitoring, and Local Firms’ Crash Risk” section examines whether foreign investors help to reduce local firms’ crash risk and the acting channel. The “Cross-Sectional Analyses” section investigates the cross-sectional variation. The “Robustness Checks” section performs robustness checks. The final section concludes.

Data and Method

Sample and Data Source

Our sample covers all nonfinancial firms on China’s A-share market over the period from 1999 to 2013.³ We require sample firms to have foreign investors among their top 10 largest shareholders, as shareholder’s identity is only disclosed for those who are on the top 10 shareholder list. Firms with missing financial information or negative book equity values are also deleted.

We hand-collect information on foreign investors’ nationality, ownership in local firms, and business with local firms from firms’ annual reports and prospectus, or through the Internet. Foreign investors examined in this study include financial institutions, corporations, and individuals, but not branches of Chinese firms.⁴ Our final sample consists of 702 unique firms and 2,057 firm-year observations.

Panel A of Table 1 presents statistics for firm-year observations with foreign investors as well as for the A-share market full sample (with and without foreign ownership). There are no noticeable differences between the two groups, except that the sample with foreign investors tends to have lower state ownership, be larger in size, have a higher book-to-market ratio, and be more profitable.

Table 1.

Summary Statistics.

Panel A: Firm characteristics of firm-year observations with foreign investors and all A-share firms
	Firm-year observations with foreign investors			All A-share firm-year observations
	(1)	(2)	(3)	(4)	(5)	(6)
	M	Median	SD	M	Median	SD
NCSKEW_t	–0.221	–0.181	0.640	–0.233	–0.211	–0.221
RET_t	–0.001	–0.001	0.001	–0.001	–0.001	–0.001
SIGMA_t	0.045	0.043	0.016	0.044	0.042	0.045
STATEOWN_t	15.970	0.000	22.480	18.880	0.000	24.520
MV_t	8.526	4.237	12.530	5.504	2.734	8.526
BM_t	0.489	0.417	0.315	0.435	0.372	0.489
LEV_t	0.191	0.179	0.150	0.197	0.186	0.191
ROA_t	0.041	0.039	0.054	0.031	0.032	0.041
Observation numbers	2,057			23,921

Panel B: Distribution of firm-year observations with foreign investors by year
Year	Firm-year observation numbers	% of firm-year observations in the market (observation numbers)	% of firm-year observations in the market (market capitalization)	Ownership by lead foreign investors (OWN_F)
1999	35	4.63	6.05	12.40
2000	41	4.90	6.31	14.00
2001	42	4.40	5.46	13.97
2002	43	4.18	4.38	12.75
2003	58	5.30	8.20	10.51
2004	75	6.61	11.38	8.91
2005	144	11.75	19.72	6.65
2006	196	16.21	26.71	4.81
2007	177	14.14	15.03	7.09
2008	169	12.04	32.01	8.74
2009	222	15.48	15.03	6.63
2010	212	13.57	15.11	7.50
2011	204	10.60	15.54	10.60
2012	228	10.44	15.83	10.89
2013	211	9.05	14.47	9.44
Total	2,057	—	—	—

Panel C: Distribution of firm-year observations with foreign investors by foreign investor type
Foreign investor type	No. of firm-year observations	% of shares held by lead foreign investors (OWN_F)
Financial institutions	1,180	1.17
Nonfinancial corporations	867	18.48
Individuals	10	35.71
Total	2,057	—

Panel D: Institutional distance between foreign investors’ home countries/regions and China
Countries/regions	Observations	ID	CD	OWN_F
Australia	4	1.018	2.439	1.080
Austria	6	1.012	2.476	2.060
Belgium	78	0.899	2.461	3.581
Canada	10	1.040	2.435	8.054
Denmark	5	1.135	2.408	22.476
France	123	0.863	2.250	7.111
Germany	74	0.976	2.283	4.964
Hong Kong	513	0.862	0.542	9.084
India	13	0.397	1.303	26.058
Japan	208	0.787	1.730	11.496
Korea	19	0.601	1.550	13.191
Netherlands	45	1.081	2.373	1.662
New Zealand	11	1.098	2.451	8.493
Norway	25	1.067	2.370	0.587
Singapore	114	0.956	1.516	8.207
Switzerland	125	1.075	2.201	4.522
Taiwan	142	0.671	1.169	27.058
Thailand	18	0.336	1.585	16.708
United Kingdom	223	1.006	2.487	4.000
United States	297	0.923	2.436	5.088
Vietnam	4	0.112	0.942	37.500

Note. This table reports summary statistics of our sample. Panel A reports firm characteristics. Panels B and C report the distribution of firm-year observations with foreign investors by year and by foreign investor type, respectively. Panel D reports the institutional distances between home countries/regions of foreign investors and China. All variables are defined in Appendix Table A1.

Panel B shows that the number of Chinese listed firms with foreign investors being one of the top 10 shareholders increases greatly from 35 in 1999 to 211 in 2013. Columns 2 and 3 show that the percentage of the whole market represented by firm-year observations with foreign investors has increased from 4.63% in 1999 to 9.05% in 2013 in terms of observation numbers, and from 6.05% in 1999 to 14.47% in 2013 in terms of market capitalization. Column 4 shows that the foreign equity ownership has a time-series average of 9.66% in our sample.

Panel C shows the distribution of observations across different foreign investor types. Foreign investors are mainly financial institutions and nonfinancial corporations, which account for 57% and 42% of the whole sample, and the majority of financial institutions are QFII (over 80% of the group).

Construction of Key Research Variables

Foreign investors

If a local firm has more than one foreign investor among its top 10 largest shareholders, we identify the one with the greatest foreign ownership as the “lead” foreign investor.⁵ We rely on the lead foreign investor when calculating institutional distances, identifying foreign investor types, and examining foreign investors’ product market relation with local firms.

FOR is an indicator that equals 1 for firms with foreign investors and 0 otherwise. OWN_F is the logarithm of one plus the ownership of lead foreign investors. POST is an indicator that equals 1 for firm-year observations after the first-time entrance of foreign investors and 0 otherwise. Corporation is an indicator that equals 1 if the lead foreign investor is a nonfinancial corporation and 0 otherwise. LONGTERM_F is an indicator for long-term lead foreign investors, who are those with an investment period exceeding the sample median. Product is an indicator variable that equals 1 if a firm’s lead foreign investor establishes a product market relationship with the firm and 0 otherwise.

Crash risk

We employ two measures of stock price crash risk: (a) the negative coefficient of skewness of firm-specific weekly returns (NCSKEW) and (b) the down-to-up volatility of firm-specific weekly returns (DUVOL). We require at least 30 weekly return observations for the calculation, and estimate firm-specific weekly returns for each firm in each year using the augmented market model of Piotroski et al. (2015):

\begin{matrix} r_{j, τ} = α + β_{1} Marke t_{China, j, τ} + β_{2} Marke t_{China, j, τ - 1} + β_{3} Marke t_{China, j, τ - 2} + β_{4} Marke t_{US, j, τ} \\ + β_{5} Marke t_{US, j, τ - 1} + β_{6} Marke t_{US, j, τ - 2} + ε_{j, τ}, \end{matrix}

(1)

where r_j,τ is stock j’s return on week τ, and Market_China,j,τ and Market_US,j,τ are value-weighted returns for China’s A-share market and the U.S. market on week τ, obtained from China Stock Market and Accounting Research (CSMAR) and Center for Research in Security Prices (CRSP) respectively. NCSKEW is computed as follows:

NCSKE W_{j, t} = \frac{- (n {(n - 1)}^{\frac{3}{2}} \sum R_{j, τ}^{3})}{((n - 1) (n - 2) {(\sum R_{j, τ}^{2})}^{\frac{2}{3}})},

(2)

where R_j,τ is the natural log of one plus the residual return from Equation 1, and n is the number of R_j,τ during year t. A higher NCSKEW indicates greater crash risk.

In constructing DUVOL, we first label weeks with R_j,τ above (below) the mean of the year as “up” (“down”) weeks. Then, for each stock j over year t, we divide the standard deviation of R_j,τ during down weeks by that during up weeks:

DUVO L_{j, t} = \log {\frac{(n_{u} - 1) \sum_{DOWN} R_{j, τ}^{2}}{(n_{d} - 1) \sum_{UP} R_{j, τ}^{2}}},

(3)

where n_u and n_d are the number of up and down weeks over year t, respectively. Similar to NCSKEW, a higher value of DUVOL suggests higher crash risk.

Financial reporting quality

We measure financial reporting quality using (a) accrual quality, (b) conservatism, and (c) annual report tone management. Our accrual quality measure, or AQ, is constructed following Dechow and Dichev (2002):

\frac{Δ W C_{it}}{T A_{it - 1}} = l_{0} + l_{1} (\frac{CF O_{it - 1}}{T A_{it - 2}}) + l_{2} (\frac{CF O_{it}}{T A_{it - 1}}) + l_{3} (\frac{CF O_{it + 1}}{T A_{it}}) + ε_{it},

(4)

where ΔWC is change in working capital, TA is total assets, and CFO is cash flows from operations.⁶AQ is the absolute value of the residual from Equation 4. A higher AQ indicates lower accrual quality and therefore lower financial reporting quality.

We follow Chen, Chen, Lobo, and Wang (2010) and measure conservatism in Chinese listed firms using Ball and Shivakumar’s (2005) specification:

AC C_{i, t} = α_{0} + α_{1} CF O_{i, t} + α_{2} DCF O_{i, t} + α_{3} CF O_{i, t} \times DCF O_{i, t} + (error),

(5)

where ACC_i,t and CFO_i,t are total accruals and cash flow from operations, respectively, both of which are scaled by total assets in year t– 1. DCFO_i,t is an indicator for negative CFO_i,t. A higher α ₃ is indicative of greater conservatism.

We follow Huang, Teoh, and Zhang (2013) and capture annual report tone management using two measures. The first one is firm’s tone (Tone_i,t) in the Management’s Discussion and Analysis (MD&A) section of the annual report. We use the word list of National Taiwan University NTUSD Vocabulary for Sentiment Analysis to classify the frequency of optimistic versus pessimistic words, and define Tone_i,t as the frequency difference between positive and negative words scaled by the total number of nonnumerical words in MD&A. The second measure is the discretionary component of tone (AbTone_i,t) estimated using the residual of the following model⁷:

\begin{matrix} Ton e_{i, t} = β_{0} + β_{1} \times RO A_{i, t} + β_{2} \times RE T_{i, t} + β_{3} \times Siz e_{i, t} + β_{4} \times B M_{i, t} + β_{5} \times STD_RE T_{i, t} \\ + β_{6} \times STD_RO A_{i, t} + β_{7} \times AG E_{i, t} + β_{8} \times LOS S_{i, t} + β_{9} \times Δ RO A_{i, t} + (error), \end{matrix}

(6)

where ROA_i,t is return-on-asset, RET_i,t is annual stock return, Size_i,t is the logarithm of market capitalization, BM_i,t is the book-to-market ratio, STD_RET_i,t and STD_ROA_i,t are the standard deviation of monthly stock returns and quarterly ROA_i,t over the year, AGE_i,t is the logarithm of one plus the number of years since firm i went listed, LOSS_i,t is an indicator of negative ROA_i,t, and ΔROA_i,t is the change in ROA_i,t. A higher Tone_i,t or AbTone_i,t indicates greater annual report tone management.

Formal and informal institutional distances

Formal institutional distance is measured in reference to the Worldwide Governance Indicators (WGI) issued by World Bank, including a country’s voice and accountability, governance effectiveness, regulatory quality, rule of law, and control of corruption. Informal institutional distance is quantified in reference to Hofstede’s (1980, 2001) cultural dimensions including (a) small versus large power distance, (b) high versus low uncertainty avoidance, (c) individualism versus collectivism, (d) masculinity versus femininity, and (e) short-term versus long-term orientation.

The institutional distance measures are constructed following the methodology that Nahata, Hazarika, and Tandon (2014) adopt in building their cultural distance index. For each lead foreign investor, the formal institutional distance (ID) between its home country and China is defined as

ID = \frac{{(\sum_{k = 1}^{n} {(I_{kj} - I_{kc})}^{2})}^{\frac{1}{2}}}{n},

(7)

where $I_{kj}$ represents the $k th$ formal institution attribute score of foreign country j, $I_{kc}$ is the $k th$ formal institution attribute score of China, and n is the number of formal institutional attributes (n = 5).

The informal institutional distance (CD) is defined similarly:

CD = \frac{{(\sum_{k = 1}^{m} {(I_{sj} - I_{sc})}^{2})}^{\frac{1}{2}}}{m},

(8)

where $I_{sj}$ is the $s th$ informal institution attribute score of foreign country j, $I_{sc}$ is the $s th$ informal institution attribute score of China, and m is the number of informal institutional attributes (m = 5).

Panel D of Table 1 presents both ID and CD between each of the 21 foreign countries/regions in our sample and China. Among the 2,057 firm-year observations, about 25% have investors from Hong Kong. This is as expected given the close relationship between mainland China and Hong Kong. And it is the least common to find foreign investors from Australia and Vietnam among the top 10 largest shareholders in Chinese listed firms. And investors from Vietnam have the highest foreign ownership (OWN_F), with an average of 37.5% in a single local firm.

Control Variables

Following the literature (e.g., Chen, Hong, & Stein, 2001; Kim, Li, & Zhang, 2011), we control variables that are known to affect stock price crash risk, including the change in monthly share turnover (DTURN_t) defined as the average monthly share turnover over fiscal year t minus that in year t– 1; lagged negative coefficient of skewness for firm-specific weekly returns (NCSKEW_t); lagged standard deviation and average of firm-specific weekly returns over the year, or SIGMA_t and RET_t; state ownership (STATEOWN_t); book-to-market ratio (BM_t); leverage ratio (LEV_t); return-on-asset (ROA_t); and the logarithm of firm size (SIZE_t). We obtain the data from CSMAR and winsorize continuous variables at the top and bottom 1% levels. Detailed description is provided in Appendix Table A1.

Foreign Investors, External Monitoring, and Local Firms’ Crash Risk

Ordinary Least Squares (OLS) Specification Based on All Firms Listed in the A-Share Market

To investigate the influence of foreign investors over local firms’ future crash risk, we first regress future crash risk, measured by NCSKEW_t₊₁ or DUVOL_t₊₁, on the dummy variable indicating the presence of foreign investors (FOR_t) or foreign ownership (OWN_F_t). A set of control variables as well as year- and industry-fixed effects are included in the regressions. Panel A of Table 2 reports the results.

Table 2.

The Influence of Foreign Investors on Local Firms’ Future Crash Risk Based on the Overall Sample.

Panel A: The influence of foreign investors on local firms’ future crash risk
	(1)	(2)	(3)	(4)
	NCSKEW_t ₊₁	DUVOL_t ₊₁	NCSKEW_t ₊₁	DUVOL_t ₊₁
FOR_t	–0.082***	–0.045**
	(–3.19)	(–2.46)
OWN_F_t			–0.086***	–0.046**
			(–2.79)	(–2.12)
DTURN_t	–0.009	–0.014	–0.009	–0.014
	(–0.65)	(–1.48)	(–0.64)	(–1.47)
NCSKEW_t	0.145***	0.095***	0.145***	0.095***
	(20.18)	(19.63)	(20.20)	(19.65)
RET_t	85.618***	57.107***	85.712***	57.138***
	(3.30)	(3.08)	(3.31)	(3.08)
SIGMA_t	6.531***	4.328***	6.542***	4.333***
	(4.71)	(4.35)	(4.71)	(4.35)
SIZE_t	0.006	–0.001	0.006	–0.002
	(1.10)	(–0.35)	(1.06)	(–0.38)
BM_t	–0.142***	–0.102***	–0.142***	–0.102***
	(–6.99)	(–6.89)	(–6.99)	(–6.90)
LEV_t	0.061*	0.062**	0.061*	0.062**
	(1.79)	(2.54)	(1.79)	(2.54)
ROA_t	–0.011	–0.045	–0.010	–0.044
	(–0.12)	(–0.68)	(–0.10)	(–0.67)
STATEOWN_t	–0.001***	–0.000**	–0.001***	–0.000**
	(–2.72)	(–2.33)	(–2.71)	(–2.31)
Intercept	0.128	0.004	0.129	0.005
	(0.42)	(0.02)	(0.42)	(0.02)
Year- and industry-fixed effects	YES	YES	YES	YES
Observation numbers	23,921	23,921	23,921	23,921
Adjusted R²	.095	.100	.095	.100

Panel B: Changes tests: Regress changes in future crash risk on changes in foreign ownership
	(1)	(2)
	ΔNCSKEW_t₊₁	ΔDUVOL_t₊₁
ΔOWN_F_t	–0.098**	–0.082**
	(–2.00)	(–2.29)
Changes in the control variables of Panel A	YES	YES
Intercept, year- and industry-fixed effects	YES	YES
Observation numbers	18,007	18,007
Adjusted R²	.284	.226

Panel C: Reverse changes tests: regress future foreign ownership changes on crash risk changes
	(1)	(2)
	ΔOWN_F_t₊₁	ΔOWN_F_t₊₁
ΔNCSKEW_t	0.001
	(0.59)
ΔDUVOL_t		0.000
		(0.14)
Changes in the control variables of Panel A	YES	YES
Intercept, year- and industry-fixed effects	YES	YES
Observation numbers	19,385	19,385
Adjusted R²	.022	.022

Note. Panel A examines the influence of foreign investors on local firms’ future crash risk. Panels B and C perform changes and reverse changes tests. All variables are defined in Appendix Table A1. The t statistics, computed with robust standard errors clustered at the firm level, are reported in parentheses.

Significance at 10% level. **Significance at 5% level. ***Significance at 1% level.

The coefficients on both FOR_t (t = −3.19 in column 1 and t = −2.46 in column 2) and OWN_F_t (t = −2.79 in column 3 and t = −2.12 in column 4) are significantly negative. The findings that the presence of foreign investors and their ownership in local firms is both significantly negatively associated with future crash risk are supportive of our arguments.

To address potential endogeneity, we follow Aggarwal et al. (2011) and Beuselinck et al. (2017) to perform changes and reverse changes tests. In the changes regressions, the change in future crash risk is regressed on the change in foreign ownership. In the reverse changes tests, the change in future foreign ownership is regressed on the change in local firms’ crash risk.

Panel B of Table 2 reports the results. Following Aggarwal et al. (2011), all control variables are expressed in terms of changes. The coefficient on ΔOWN_F_t is significantly negative at the 5% level in both columns 1 and 2. It confirms that an increase in foreign ownership is followed by a decrease in local firms’ future crash risk. More importantly, in Panel C, where we conduct the reverse changes regression analysis, the coefficients on both ΔNCSKEW_t and ΔDUVOL_t are insignificant. It suggests that foreign investors, in selecting their investment targets, are not particularly attracted by stocks with lower past crash risk. The evidence renders strong support to our argument that the negative relation between foreign presence/ownership and local firms’ future crash risk is not driven by foreign investors’ tendency to invest in firms with low crash risk in the first place.

Baseline DiD Regressions

To further mitigate concerns over potential endogeneity, we employ a DiD regression design. First, in each year, we identify treatment firms for which foreign investors initiated their investments. Then, for each of the treatment firms, we find a control firm without foreign investors and (a) is in the same industry in the same year as the treatment firm and (b) has the closest odds of having foreign investors as the treatment firm, conditional on variables including DTURN_i,t-1, NCSKEW_i,t-1, SIGMA_i,t-1, RET_i,t-1, SIZE_i,t-1, STATEOWN_i,t-1, BM_i,t-1, LEV_i,t-1, and ROA_i,t-1. These variables are chosen to include firm and risk characteristics associated with both the entrance of foreign investors (Dahlquist & Robertsson, 2001) and future crash risk (e.g., Chen et al., 2001; Jin & Myers, 2006; Piotroski et al., 2015).

To examine the influence brought by foreign investors, we keep only firm-year observations 5 years before and 5 years after the entrance of foreign investors in the DiD analysis, and we require sample firms to have at least 1 year of financial data in each of the 5-year period.⁸

Following Armstrong, Jagolinzer, and Larcker (2010), we examine the covariate balance between the treatment and control samples to ensure that the observable dimensions of the matched pairs are similar aside from the presence of foreign investors in Panel A of Table 3. As shown in column 6, the absence of statistically significant differences across all the nine variables, which are selected based on Equation 9, indicates that differences in these observed variables between the matched pairs are not likely to confound our estimates of the treatment effect.

Table 3.

The Influence of Foreign Investors on Local Firms’ Future Crash Risk: Difference-in-Difference Analysis.

Panel A: Covariate balance between treatment firms and control firms
	Mean before propensity score matching(PSM)			Mean after propensity score matching(PSM)
	(1)	(2)	(3) = (1) – (2)	(4)	(5)	(6) = (4) – (5)
	Treatment	Control	t statistics	Treatment	Control	t statistics
DTURN_t _-1	–0.060	0.011	–2.89***	–0.058	–0.066	0.33
NCSKEW_t _-1	–0.242	–0.225	–0.63	–0.241	–0.243	0.03
RET_t _-1	–0.001	–0.001	–0.08	–0.001	–0.001	0.29
SIGMA_t _-1	0.044	0.044	0.52	0.044	0.045	–0.35
SIZE_t _-1	15.135	14.704	11.96***	15.124	15.089	0.58
BM_t _-1	0.526	0.408	11.00***	0.523	0.552	–1.61
LEV_t _-1	0.189	0.200	–1.72*	0.189	0.195	–0.76
ROA_t _-1	0.043	0.030	5.76***	0.042	0.039	1.27
STATEOWN_t _-1	25.035	20.415	4.38***	24.844	26.407	–0.99

Panel B: Difference-in-difference analysis
	(1)	(2)	(3)	(4)
	NCSKEW_t ₊₁	DUVOL_t ₊₁	NCSKEW_t ₊₁	DUVOL_t ₊₁
FOR_t	0.053	0.031
	(1.59)	(1.39)
POST_t	0.038	0.020	0.010	–0.001
	(1.36)	(0.97)	(0.42)	(–0.07)
FOR_t×POST_t	–0.164***	–0.110***
	(–4.22)	(–3.96)
OWN_F_t	0.025	0.014	0.087***	0.049**
	(0.83)	(0.70)	(2.95)	(2.51)
OWN_F_t×POST_t			–0.157***	–0.097***
			(–4.70)	(–4.08)
DTURN_t	–0.038	–0.022	–0.038	–0.022
	(–0.90)	(–0.77)	(–0.90)	(–0.77)
NCSKEW_t	0.066***	0.043***	0.067***	0.043***
	(3.54)	(3.26)	(3.58)	(3.30)
RET_t	108.558*	67.589	115.875*	72.102*
	(1.80)	(1.60)	(1.91)	(1.70)
SIGMA_t	9.223***	5.664**	9.624***	5.917**
	(2.65)	(2.41)	(2.76)	(2.51)
SIZE_t	0.020	0.006	0.021	0.006
	(1.49)	(0.62)	(1.59)	(0.69)
BM_t	–0.052	–0.040	–.049	–0.038
	(–0.97)	(–1.14)	(–0.92)	(–1.07)
LEV_t	0.222**	0.134**	0.222**	0.134**
	(2.48)	(2.30)	(2.48)	(2.30)
ROA_t	0.721**	0.277	0.710**	0.270
	(2.56)	(1.45)	(2.54)	(1.42)
STATEOWN_t	–0.000	0.000	–0.000	0.000
	(–0.14)	(0.13)	(–0.17)	(0.09)
Intercept	–1.156***	–0.607***	–1.182***	–0.622***
Intercept	(–4.22)	(–3.32)	(–4.27)	(–3.37)
Year- and industry-fixed effects	YES	YES	YES	YES
Observation numbers	4,297	4,297	4,297	4,297
Adjusted R²	.088	.103	.088	.102

Note. Panel A presents results of the covariate balance tests between treatment and control firms. Panel B presents results of the difference-in-difference tests. All variables are defined in Appendix Table A1. The t statistics, computed with robust standard errors clustered at the firm level, are reported in parentheses.

Significance at 10% level. **Significance at 5% level. ***Significance at 1% level.

We then estimate the following DiD regression that links the economic consequences with our test variables, that is, FOR, POST, and FOR×POST:

D V_{i, t} = α_{0} + α_{1} FO R_{i, t} + α_{2} POS T_{i, t} + α_{3} FO R_{i, t} \times POS T_{i, t} + (Controls) + (error),

(9)

where DV refers to firm i’s future crash risk, FOR is an indicator that equals 1 for treatment firms and 0 for control firms, and POST is an indicator that equals 1 for the post-period and 0 otherwise. A set of control variables as well as year- and industry-fixed effects are included. The results are reported in Panel B of Table 3.

The coefficient on FOR×POST is highly significantly negative in columns 1 and 2, suggesting that the decrease in crash risk for firms with foreign investors from the 5-year period before entrance to the 5-year period after entrance is significantly greater than that for firms without foreign investors over the same period. In columns 3 and 4, we replace FOR with OWN_F. The coefficient on OWN_F×POST is still significantly negative at the 1% level in both columns. It is as expected as foreign investors with a greater stake in local firms have stronger incentives to monitor these firms, resulting in lower future stock price crash risk. It is worth noting that the coefficients on FOR and OWN_F are either insignificant or significantly positive. It suggests that prior to the entrance of foreign investors, the average crash risk for the treatment group is comparable to, if not higher than, that of the control group. The evidence alleviates concerns over the possibility that foreign investors tend to invest in local firms with lower crash risk in the first place.

Overall, results in Table 3 reveal that the reduction in crash risk for treatment firms with foreign investors after the entrance of these investors is significantly greater than that for control firms during the same period. The results confirm our findings in Table 2 that foreign investors help to reduce local firms’ crash risk.

The Impact of Foreign Investors’ Monitoring on Financial Reporting Quality

We further investigate whether improving financial reporting quality is a mechanism through which the entrance of foreign investors reduces local firms’ future crash risk. We estimate DiD regressions to examine the changes in reporting quality for the treatment group after the entrance of foreign investors, relative to that of the control group over the same period.

Results in columns 1, 3, 5, and 7 of Table 4 are all supportive of our predictions. The significantly negative coefficient on FOR_i,t×POST_i,t (t = −2.90) in column 1 suggests that the increase in accrual quality is significantly greater for the treatment group after the entrance of foreign investors than for the control group over the same period. The significantly positive coefficient on FOR_i,t×POST_i,t×CFO_i,t×DCFO_i,t (t = 2.39) in column 3 implies that, from the pre- to the post-period, the increase in reporting conservatism is significantly greater for the treatment group than for the control group. In columns 5 and 7, the coefficient on FOR_i,t×POST_i,t is significantly negative at the 1% level in both columns, confirming that the reduction in tone-management behavior is greater in local firms with foreign investors than in the control group from the pre- to the post-period.

Table 4.

The Influence of Foreign Investors on Local Firms’ Financial Reporting Quality.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	AQ_t	AQ_t	ACC_t	ACC_t	Tone_t	Tone_t	AbTone_t	AbTone_t
FOR_t	0.006***		–0.003		–0.043		–0.003
	(2.74)		(–0.56)		(–0.77)		(–0.06)
POST_t	0.004**	0.002	–0.005	–0.007*	0.075***	0.070***	0.090***	0.081***
	(2.26)	(1.51)	(–1.21)	(–1.86)	(2.61)	(2.72)	(3.35)	(3.41)
FOR_t×POST_t	–0.007***		0.002		–0.115***		–0.099***
	(–2.90)		(0.32)		(–3.08)		(–2.75)
FOR_t×POST_t×CFO_t×DCFO_t			0.514**
			(2.39)
OWN_F_t	–0.002	0.003	0.009**	0.005	0.028	0.025	0.021	0.042
	(–1.43)	(1.30)	(2.32)	(0.89)	(0.43)	(0.63)	(0.35)	(1.06)
OWN_F_t×POST_t		–0.006**		0.011*		–0.134***		–0.110**
		(–2.25)		(1.88)		(–2.76)		(–2.56)
OWN_F_t×POST_t×CFO_t×DCFO_t				0.536***
				(3.64)
Control of CFO_t, DCFO_t, CFO_t× DCFO_t, and related interaction termsControl of SIZE, BM, SGR, LEV, ROA, and STATEOWN	NOYES	NOYES	YESYES	YESYES	NOYES	NOYES	NOYES	NOYES
Intercept, year- and industry-fixed effects	YES	YES	YES	YES	YES	YES	YES	YES
Observation numbers	4,047	4,047	4,134	4,134	4,044	4,044	3,805	3,805
Adjusted R²	.051	.049	.810	.809	.134	.133	.015	.015

Note. This table examines the influence of foreign investors on local firms’ financial reporting quality. Columns 1 and 2 examine the accrual quality, columns 3 and 4 examine reporting conservatism, and columns 5 to 8 examine annual report tone management. All variables are defined in Appendix Table A1. The t statistics, computed with robust standard errors clustered at the firm level, are reported in parentheses.

Significance at 10% level. **Significance at 5% level. ***Significance at 1% level.

We further replace FOR_i,t with OWN_F_i,t and report the results in columns 2, 4, 6, and 8 in Table 4. Again, all pieces of evidence are supportive of our argument. Table 4 shows that foreign investors are effective external monitors and help to improve local firms’ financial reporting quality. It is consistent with findings of Beuselinck et al. (2017). Given that managerial bad-news hoarding is an important cause of crash risk, these results help us to better understand why foreign investors help to reduce local firms’ crash risk, as documented in Tables 2 and 3.

Cross-Sectional Analyses

We argue that foreign investors’ influence over local firms’ crash risk might not be constant across firms. It could vary conditional on (a) the institutional distance between foreign investors’ home and host countries, (b) the strength of their monitoring incentives, and (c) the effectiveness of local firms’ governance system. Accordingly, we examine the cross-sectional variation in this section.

Institutional Distance and Local Firms’ Crash Risk

We expect foreign investors’ ability to reduce local firms’ crash risk to be stronger (weaker) when their home countries are less (more) institutionally distant from China. We examine the influence of institutional distance using Equation 10:

D V_{i, t} = α_{0} + α_{1} I D_{i, t} + α_{2} C D_{i, t} + α_{3} POS T_{i, t} + α_{4} I D_{i, t} \times POS T_{i, t} + α_{5} C D_{i, t} \times POS T_{i, t} + (Controls) + (error),

(10)

where ID_i,t and CD_i,t refer to formal and informal institutional distance, respectively; POST equals 1 for observations after the entrance of lead foreign investors and 0 otherwise; and the dependent variable, DV_i,t+1, refers to firm i’s crash risk.

Columns 1 and 2 of Table 5 report the results. The coefficient on POST is significantly negative across both columns, suggesting that stock price crash risk declines in general from the pre- to the post-period, consistent with our previous findings. The significantly positive coefficients on ID×POST and CD×POST across columns 1 and 2 confirm that foreign investors’ ability to reduce local firms’ crash risk, as captured by the negative coefficient on POST, is significantly attenuated if they come from foreign countries that are institutionally more distant from China.

Table 5.

Cross-Section Variation: The Type of Foreign Investors.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	NCSKEW_t ₊₁	DUVOL_t ₊₁	NCSKEW_t ₊₁	DUVOL_t ₊₁	NCSKEW_t ₊₁	DUVOL_t ₊₁	NCSKEW_t ₊₁	DUVOL_t ₊₁
ID_t	–0.400	–0.243
	(–1.44)	(–1.53)
CD_t	–0.046	–0.015
	(–0.89)	(–0.42)
POST_t	–0.806***	–0.549***	0.038	0.020	0.038	0.019	0.037	0.019
	(–2.99)	(–2.98)	(1.35)	(0.95)	(1.34)	(0.95)	(1.34)	(0.95)
ID_t×POST_t	0.541*	0.354*
	(1.86)	(1.78)
CD_t×POST_t	0.095**	0.065**
	(2.56)	(2.19)
Corporation_t			0.197*	0.196**
			(1.74)	(2.53)
Corporation_t×POST_t			–0.393**	–0.249***
			(–2.47)	(–2.73)
LONGTERM_F_t					0.003	–0.005
					(0.02)	(–0.06)
LONGTERM_F_t×POST_t					–0.340**	–0.194*
					(–2.29)	(–1.89)
Product_t							–0.374***	–0.255***
							(–4.42)	(–4.81)
Product_t×POST_t							0.185**	0.232***
							(2.49)	(4.37)
Control of FOR_t, FOR_t×POST_t, and OWN_F_t	NO	NO	YES	YES	YES	YES	YES	YES
Control variables as in Table 3	YES	YES	YES	YES	YES	YES	YES	YES
Intercept, year- and industry-fixed effects	YES	YES	YES	YES	YES	YES	YES	YES
Observation numbers	2,195	2,195	4,297	4,297	4,297	4,297	4,297	4,297
Adjusted R²	.100	.104	.089	.104	.089	.103	.089	.103

Note. This table examines the cross-sectional variation in foreign investors’ influence over local firms’ crash risk, conditional on institutional distance and the strength of their monitoring incentives. All variables are defined in Appendix Table A1. The t statistics, computed with robust standard errors clustered at the firm level, are reported in parentheses.

Significance at 10% level. **Significance at 5% level. ***Significance at 1% level.

The Strength of Foreign Investors’ Monitoring Incentives

If the negative relation between the presence of foreign investors/foreign ownership and local firms’ future crash risk results from foreign investors’ external monitoring effect, such a negative relation should be more (less) evident when foreign investors’ incentives to monitor are stronger (weaker). We expect foreign investors’ monitoring incentives to be stronger (weaker) if they are nonfinancial corporate (financial institutional) investors, long-term (short-term) investors, and have no (have) product market relation with local firms. We replace ID (CD) in Equation 10 with an indicator of foreign corporations Corporation, of long-term foreign investors LONGTERM_F, or of foreign investors with product market relation with local firms Product. The results are shown in columns 3 to 8 of Table 5.

The coefficient on Corporation_i,t×POST_i,t is significantly negative in columns 3 and 4, supporting our argument that foreign nonfinancial corporations have stronger incentives to monitor local firm than foreign financial institutions, as the former group usually has great and long-term stake in local firms and form strategic partnership with them while the latter group usually makes small and short-term investments to earn trading profits.

In columns 5 to 8, the coefficients on LONGTERM_F_i,t×POST_i,t and Product_i,t×POST_i,t are significantly negative and positive, respectively. It is consistent with the findings of Fang et al. (2015) that foreign investors with long-term stakes have stronger incentives to monitor local firms to ensure their long-term interests in the firm. It is also consistent with our expectation that foreign investors having product market relationship with local firms are less motivated to monitor these firms as active intervention may jeopardize their relationship. Overall, the results confirm that foreign investors’ influence on local firms’ crash risk is stronger when these investors’ external monitoring incentives are stronger.

The Effectiveness of Local Firms’ Governance System

We further examine whether the effectiveness of local firms’ governance systems moderates the influence of foreign investors on local firms’ crash risk. As discussed in the “Introduction” section, we expect governance efficacy to be higher among local firms with greater analyst coverage, with lower state ownership, and located in regions with greater institutional development.⁹ Accordingly, we partition our sample based on these three governance efficacy determinants and estimate the baseline DiD model specified in Equation 9 in each subsample. The results are reported in Table 6.

Table 6.

Cross-Section Variation: The Effectiveness of Local Firms’ Governance System.

Panel A: Analysts coverage
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	With analyst coverage				Without analyst coverage
	NCSKEW_t ₊₁	DUVOL_t ₊₁	NCSKEW_t ₊₁	DUVOL_t ₊₁	NCSKEW_t ₊₁	DUVOL_t ₊₁	NCSKEW_t ₊₁	DUVOL_t ₊₁
FOR_t	0.091**	0.062**			0.003	–0.016
	(2.33)	(2.27)			(0.05)	(–0.41)
POST_t	0.066**	0.026	0.030	–0.007	–0.002	0.022	–0.006	0.024
	(2.02)	(1.07)	(1.06)	(–0.32)	(–0.03)	(0.50)	(–0.13)	(0.64)
FOR_t×POST_t	–0.212***	–0.141***			–0.039	–0.033
	(–4.73)	(–4.33)			(–0.49)	(–0.57)
OWN_F_t	0.031	0.016	0.136***	0.077***	0.003	0.010	0.008	0.005
	(0.84)	(0.70)	(3.54)	(3.07)	(0.05)	(0.28)	(0.19)	(0.14)
OWN_F_t×POST_t			–0.211***	–0.118***			–0.041	–0.050
OWN_F_t×POST_t			(–5.06)	(–3.84)			(–0.62)	(–1.11)
Control variables as in Table 3	YES	YES	YES	YES	YES	YES	YES	YES
Intercept, year- and industry-fixed effects	YES	YES	YES	YES	YES	YES	YES	YES
Observation numbers	2,805	2,805	2,805	2,805	1,492	1,492	1,492	1,492
Adjusted R²	.089	.093	.090	.092	.098	.122	.098	.123

Panel B: State ownership
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	Low state ownership				High state ownership
	NCSKEW_t ₊₁	DUVOL_t ₊₁	NCSKEW_t ₊₁	DUVOL_t ₊₁	NCSKEW_t ₊₁	DUVOL_t ₊₁	NCSKEW_t ₊₁	DUVOL_t ₊₁
FOR_t	0.068*	0.011			0.042	0.050
	(1.66)	(0.39)			(0.78)	(1.46)
POST_t	0.069**	0.032	0.025	0.008	0.005	0.004	0.018	0.006
	(1.98)	(1.28)	(0.82)	(0.36)	(0.10)	(0.11)	(0.43)	(0.18)
FOR_t×POST_t	–0.237***	–0.143***			–0.015	–0.017
	(–4.85)	(–4.13)			(–0.23)	(–0.33)
OWN_F_t	0.017	0.018	0.099***	0.046*	0.042	0.019	0.092*	0.064**
	(0.51)	(0.71)	(2.92)	(1.82)	(0.68)	(0.55)	(1.95)	(2.18)
OWN_F_t×POST_t			–0.214***	–0.129***			–0.078	–0.045
			(–4.93)	(–4.19)			(–1.41)	(–1.23)
Control variables as in Table 3	YES	YES	YES	YES	YES	YES	YES	YES
Intercept, year- and industry-fixed effects	YES	YES	YES	YES	YES	YES	YES	YES
Observation numbers	2,431	2,431	2,431	2,431	1,866	1,866	1,866	1,866
Adjusted R²	.098	.108	.097	.106	.083	.109	.084	.108
Panel C: Regional institutional development index
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
		High regional institutional development index				Low regional institutional development index
	NCSKEW_t ₊₁	DUVOL_t ₊₁	NCSKEW_t ₊₁	DUVOL_t ₊₁	NCSKEW_t ₊₁	DUVOL_t ₊₁	NCSKEW_t ₊₁	DUVOL_t ₊₁
FOR_t	0.080*	0.034			0.048	0.045
	(1.90)	(1.20)			(0.87)	(1.17)
POST_t	0.087**	0.054**	0.058*	0.033	–0.033	–0.037	–0.033	–0.034
	(2.56)	(2.29)	(1.90)	(1.54)	(–0.60)	(–0.86)	(–0.73)	(–0.97)
FOR_t×POST_t	–0.209***	–0.140***			–0.052	–0.017
	(–4.42)	(–4.21)			(–0.72)	(–0.29)
OWN_F_t	0.025	0.024	0.146***	0.082**	0.019	–0.001	0.063	0.037
	(0.73)	(0.94)	(3.10)	(2.55)	(0.32)	(–0.02)	(1.52)	(1.38)
OWN_F_t×POST_t			–0.231***	–0.142***			–0.083	–0.041
OWN_F_t×POST_t			(–4.36)	(–3.74)			(–1.29)	(–0.84)
Control variables as in Table 3	YES	YES	YES	YES	YES	YES	YES	YES
Intercept, year- and industry-fixed effects	YES	YES	YES	YES	YES	YES	YES	YES
Observation numbers	2,484	2,484	2,484	2,484	1,813	1,813	1,813	1,813
Adjusted R²	.080	.086	.080	.085	.077	.097	.077	.097

Note. This table examines the cross-sectional variation in foreign investors’ influence over local firms’ future crash risk, conditional on the effectiveness of local firms’ governance system. The effectiveness of local firms’ governance system is captured by analyst coverage in Panel A, state ownership in Panel B, and regional institutional development index in Panel C. The low and high groups are partitioned based on the median of the conditioning variable being examined. All variables are defined in Appendix Table A1. The t statistics, computed with robust standard errors clustered at the firm level, are reported in parentheses.

Significance at 10% level. **Significance at 5% level. ***Significance at 1% level.

Panel A presents results of subsample tests based on analyst coverage partition. In columns 1 to 4, where firms with analysts coverage are examined, the coefficients on the interaction terms FOR_i,t×POST_i,t and OWN_F_i,t×POST_i,t are both strongly significantly negative at the 1% level. In columns 5 to 8, where firms without analysts coverage are examined, the coefficients on FOR_i,t×POST_i,t and OWN_F_i,t×POST_i,t turn to be indistinguishable from 0.

Panels B and C show results based on state ownership and regional institutional development index partition, respectively. The coefficients on both FOR_i,t×POST_i,t and OWN_F_i,t×POST_i,t are significantly negative at the 1% level when firms with low state ownership or located in regions with a high institutional development index are examined. The coefficients of these interaction terms, however, turn to be insignificant when firms with high state ownership or located in regions with a low institutional development index are examined.

In sum, the above evidence suggests that local firms’ governance system does moderate the relation between the presence of foreign investors/foreign ownership and local firms’ future crash risk. When local firms’ corporate governance is more effective, foreign investors could better play their external monitoring role and reduce local firms’ future crash risk to a greater extent.

Robustness Checks

Firm Fixed Effect

To address endogeneity concerns, we redo previous tests with firm fixed effects controlled in Table 7. Panels A and B reexamine the influence of foreign investors on local firms’ crash risk as well as financial reporting quality. Columns 1 to 4 of Panel A show that the coefficients on both FOR×POST and OWN_F×POST remain significantly negative at the 1% level after controlling for firm fixed effects. Results in Panel B remain qualitatively similar to those in Table 4.

Table 7.

Robustness Checks.

Panel A: The influence of foreign investors on local firms’ future crash risk
	(1)	(2)	(3)	(4)
	NCSKEW_t ₊₁	DUVOL_t ₊₁	NCSKEW_t ₊₁	DUVOL_t ₊₁
FOR_t×POST_t	–0.196***	–0.139***
	(–4.29)	(–4.14)
OWN_F_t×POST_t			–0.177***	–0.115***
			(–4.33)	(–3.67)
Intercept and control variables as in Table 3	YES	YES	YES	YES
Firm fixed effects	YES	YES	YES	YES
Observation numbers	4,297	4,297	4,297	4,297
Adjusted R²	.140	.136	.138	.134

Panel B: The influence of foreign investors on local firms’ financial reporting quality
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	AQ_t	AQ_t	ACC_t	ACC_t	Tone_t	Tone_t	AbTone_t	AbTone_t
FOR_t×POST_t	–0.007***		0.003		–0.067*		–0.093**
	(–2.81)		(0.56)		(–1.69)		(–2.48)
OWN_F_t×POST_t		–0.006*		0.014***		–0.072*		–0.067
		(–1.75)		(2.59)		(–1.76)		(–1.60)
FOR_t×POST_t×CFO_t×DCFO_t			0.280**
			(2.31)
OWN_F_t×POST_t×CFO_t×DCFO_t				0.367**
				(2.57)
Intercept and control variables as in Table 4	YES	YES	YES	YES	YES	YES	YES	YES
Control of CFO_t, DCFO_t, CFO_t×DCFO_t, and related interaction terms	NO	NO	YES	YES	NO	NO	NO	NO
Firm fixed effects	YES	YES	YES	YES	YES	YES	YES	YES
Observation numbers	4,047	4,047	4,134	4,134	4,044	4,044	3,805	3,805
Adjusted R²	0.235	0.234	0.872	0.873	0.488	0.488	0.438	0.437

Note. Panels A and B examine the influence of foreign investors on local firms’ future crash risk and financial reporting quality, respectively, with the control of firm fixed effects. Coefficients on control variables are suppressed for brevity. All variables are defined in Appendix Table A1. The t statistics, computed with robust standard errors clustered at the firm level, are reported in parentheses.

Significance at 10% level. **Significance at 5% level. ***Significance at 1% level.

In unreported tests, we repeat cross-sectional variation tests in Tables 5 and 6 with the control of firm fixed effects, and the results are qualitatively similar. These results confirm that our previous findings are robust even after controlling time-invariant omitted firm-level factors and alleviate concerns over endogeneity.

Alternative Crash Risks Measure

In unreported tests, we also check whether our main findings are robust to an alternative crash risk measure, which is constructed following Jin and Myers (2006) and Callen and Fang (2015) based on the downside frequencies minus the upside frequencies. For each stock in each week, if its weekly return is more than 3.2 standard deviations below (above) its average weekly return over the year, it is defined to have a large negative (positive) stock price drop (jump). We reestimate Equation 9 with the alternative crash risk measure, and the results are very similar to those in Panel B of Table 3.

Firms With Foreign Investors From Certain Regions/Countries Excluded

Some sample firms simultaneously issue both A-shares to domestic investors and foreign shares such as B- or H-shares to foreign investors and may face different regulatory and informational environment compared with other sample firms (e.g., Gul, Kim, & Qiu, 2010). We thus conduct subsample tests where we exclude cross-listed firms and reestimate Equation 9. And since Hong Kong has a unique relationship with China, we also repeat the main analysis after excluding firms with lead foreign investors from Hong Kong. The results for these subsample tests remain robust and strong. These results are not reported for brevity but are available upon request.

Conclusion

In this study, we find that the entrance of foreign investors is followed by a significant reduction in local firms’ stock price crash risk. We further show that foreign investors contribute to improving local firms’ financial reporting quality, which could be one possible channel through which they help to reduce crash risk as their monitoring constrains managerial bad-news hoarding.

We further examine the cross-sectional variation in foreign investors’ influence over local firms’ future crash risk from three perspectives: the institutional distance between foreign investors’ home and host countries, the strength of their monitoring incentives, and the governance efficacy of local firms. We find that the role of foreign investors in reducing crash risk is more evident when foreign investors are coming from countries that are closer to China in terms of the institutional distance, when foreign investors have stronger monitoring incentives, and/or when local firms’ governance system is more effective. In these cases, foreign investors are better able to play their external monitoring role and thus reduce local firms’ future crash risk.

We perform a series of robustness tests to address concerns over potential endogeneity, including changes and reverse changes regressions, DiD tests, and regressions with the control of firm fixed effects. All these tests produce consistent results, suggesting that the findings are unlikely to be driven by endogeneity.

Footnotes

Appendix

Table A1.

Variable Definitions.

Firm-level crash risk measure
NCSKEW	The negative coefficient of skewness of firm-specific weekly return. It is defined as the negative of third moment of firm-specific weekly return scaled by cubed standard deviation of firm-specific weekly return in the year.
DUVOL	The log of the ratio of the standard deviation of firm-specific weekly returns for the “down-day” sample to standard deviation of firm-specific weekly returns for the “up-day” sample over the fiscal year.
Foreign institutions holding characteristics
OWN_F	The natural logarithm of one plus ownership held by the lead foreign investors. A lead foreign investor is the one with the greatest foreign ownership. In a few cases where two foreign investors have similar ownership in the local firm, we identify the one that enters the firm earlier as the lead foreign investor.
FOR	An indicator variable that equals 1 for treatment firms and 0 for control firms.
POST	An indicator that equals 1 for firm years after the first-time entrance of lead foreign investors and 0 otherwise.
Corporation	An indicator that equals 1 if the lead foreign investor is a nonfinancial corporation, and 0 otherwise.
LONGTERM_F	An indicator for long-term lead foreign investors, who are those with an investment period exceeding the sample median.
Product	An indicator variable that equals 1 if a firm’s lead foreign investors establish product market relationships with local firms, and 0 otherwise.
Institutional distance
ID	Formal institutional distance between the home country of the firm’s lead foreign investors and China, measured by Cartesian distance between the two nations’ five Worldwide Governance Indicators released by the World Bank annually.
CD	Informal institutional distance between the home country of the firm’s lead foreign investors and China, measured by Cartesian distance between the two nations’ scores on Hofstede’s five cultural dimensions.
Firm characteristics
AQ	Accrual quality, the absolute value of discretionary accruals based on Dechow and Dichev (2002) model.
ACC	Total accruals scaled by total assets in year t– 1.
CFO	Cash flow from operations scaled by total assets in year t– 1.
DCFO	An indicator variable that equals 1 if CFO is negative, and 0 otherwise.
Tone	(#positive words – #negative words)/total nonnumerical words in Management’s Discussion and Analysis section of the annual report. Source: National Taiwan University NTUSD Vocabulary for Sentiment Analysis.
AbTone	Discretionary tone, calculated based on Equation 6 following Huang, Teoh, and Zhang (2013).
DTURN	The average monthly turnover in fiscal year t minus the average monthly turnover in fiscal year t– 1.
RET	The average firm-specific weekly returns over the fiscal year.
SIGMA	The standard deviation of firm-specific weekly returns over the fiscal year.
STATEOWN	State ownership.
SIZE	The natural logarithm of market value of equity, measured at the fiscal year end.
SGR	Sales growth in fiscal year t.
MV	Market value of equity (billion RMB), measured at the fiscal year end.
BM	The book-to-market ratio measured at the fiscal year end.
LEV	Total debt divided by total assets, measured at the fiscal year end.
ROA	Net income divided by total assets, measured at the fiscal year end.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Dr. Li acknowledges the financial support from National Natural Science Foundation of China (NSFC Project Number 71802094). Dr. Luo acknowledges the financial support from National Natural Science Foundation of China (NSFC Project Number 71772049) and HKU-Fudan IMBA Joint Research Fund (No. JRF1718_0602). Professor Wang acknowledges the financial support from National Natural Science Foundation of China (NSFC Project Numbers 71272072, 71572042).

ORCID iD

Yan Luo

Notes

References

Aggarwal

Erel

Ferreira

Matos

(2011). Does governance travel around the world? Evidence from institutional investors. Journal of Financial Economics, 100, 154-181.

Armstrong

C. S.

Jagolinzer

A. D.

Larcker

D. F.

(2010). Chief executive officer equity incentives and accounting irregularities. Journal of Accounting Research, 48, 225-271.

Ball

Shivakumar

(2005). Earnings quality in UK private firms: Comparative loss recognition timeliness. Journal of Accounting & Economics, 39, 83-128.

Beuselinck

Blanco

García Lara

J. M.

(2017). The role of foreign shareholders in disciplining financial reporting. Journal of Business Finance & Accounting, 44, 558-592.

Borisova

Brockman

Salas

J. M.

Zagorchev

(2012). Government ownership and corporate governance: Evidence from the EU. Journal of Banking & Finance, 36, 2917-2934.

Callen

J. L.

Fang

(2015). Religion and stock price crash risk. Journal of Financial and Quantitative Analysis, 50, 169-195.

Chen

J. Z.

Lobo

G. J.

Wang

(2010). Association between borrower and lender state ownership and accounting conservatism. Journal of Accounting Research, 5, 973-1014.

Chen

Hong

Stein

J. C.

(2001). Forecasting crashes: Trading volume, past returns, and conditional skewness in stock prices. Journal of Financial Economics, 61, 345-381.

Choe

Kho

B.-C.

Stulz

R. M.

(2005). Do domestic investors have an edge? The trading experience of foreign investors in Korea. Review of Financial Studies, 18, 795-829.

10.

Dahlquist

Robertsson

(2001). Direct foreign ownership, institutional investors, and firm characteristics. Journal of Financial Economics, 59, 413-440.

11.

Dechow

Dichev

I. D.

(2002). The quality of accruals and earnings: The role of accrual estimation errors. The Accounting Review, 77(Suppl.), 35-59.

12.

Dvořák

(2005). Do domestic investors have an information advantage? Evidence from Indonesia. The Journal of Finance, 60, 817-839.

13.

Fan

Wang

X. L.

(2011). NERI index of marketization of China’s provinces. Beijing, China: Economics Science Press. (In Chinese)

14.

Fang

V. W.

Maffett

Zhang

(2015). Foreign institutional ownership and the global convergence of financial reporting practices. Journal of Accounting Research, 53, 593-631.

15.

Ferreira

M. A.

Matos

(2008). The colors of investors’ money: The role of institutional investors around the world. Journal of Financial Economics, 88, 499-533.

16.

Grinblatt

Keloharju

(2000). The investment behavior and performance of various investor types: A study of Finland’s unique data set. Journal of Financial Economics, 55, 43-67.

17.

Gul

F. A.

Kim

J.-B.

Qiu

A. A.

(2010). Ownership concentration, foreign shareholding, audit quality, and stock price synchronicity: Evidence from China. Journal of Financial Economics, 95, 425-442.

18.

Hau

(2001). Location matters: An examination of trading profits. The Journal of Finance, 56, 1959-1983.

19.

Hofstede

(1980). Culture’s consequences: International differences in work-related values. Newbury Park, CA: SAGE.

20.

Hofstede

(2001). Culture’s consequences: Comparing values, behaviors, institutions and organizations across nations. Thousand Oaks, CA: SAGE.

21.

(2018). Can transient institutions correctly interpret small negative earnings surprises in the absence of access to management’s private information? Journal of Accounting, Auditing & Finance, 33, 3-33.

22.

Huang

Teoh

S. H.

Zhang

(2013). Tone management. The Accounting Review, 89, 1083-1113.

23.

Jin

Myers

S. C.

(2006). R2 around the world: New theory and new tests. Journal of Financial Economics, 79, 257-292.

24.

Kho

B.-C.

Stulz

R. M.

Warnock

F. E.

(2009). Financial globalization, governance, and the evolution of the home bias. Journal of Accounting Research, 47, 597-635.

25.

Kim

J.-B.

Luo

Wang

(2015). Institutional distance and the monitoring effect of foreign investors (Fudan University Working Paper).

26.

Kim

J.-B.

Zhang

(2011). CFOs versus CEOs: Equity incentives and crashes. Journal of Financial Economics, 101, 713-730.

27.

Kim

Sung

Wei

S.-J.

(2014, May). The diffusion of corporate governance to emerging markets: Evaluating two dimensions of investor heterogeneity (ECGI–Finance Working Paper No. 421/2014). Retrieved from https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2439306

28.

Nguyen

Q. N.

Pham

P. K.

Wei

S. X.

(2011). Large foreign ownership and firm-level stock return volatility in emerging markets. Journal of Financial and Quantitative Analysis, 46, 1127-1155.

29.

Qian

(2013). Principal-principal conflicts under weak institutions: A study of corporate takeovers in China. Strategic Management Journal, 34, 498-508.

30.

Nahata

Hazarika

Tandon

(2014). Success in global venture capital investing: Do institutional and cultural differences matter? Journal of Financial and Quantitative Analysis, 49, 1039-1070.

31.

Piotroski

J. D.

Wong

T. J.

Zhang

(2015). Political incentives to suppress negative information: Evidence from Chinese listed firms. Journal of Accounting Research, 53, 405-459.

32.

Trabelsi

(2018). Are transient institutions sophisticated enough to interpret small negative earnings surprises? Journal of Accounting, Auditing & Finance, 33, 34-39.

33.

F. F.

(2008). Analyst coverage and earnings management. Journal of Financial Economics, 88, 245-271.

34.

Zhang

Zheng

(2015). Corporate governance, political connections, and intra-industry effects: Evidence from corporate scandals in China. Financial Management, 44, 49-80.