Introduction

Stock price crash risk (SPCR) is not rare in global marketplaces and can cause severe consequences (Cao et al. 2019; Jin et al. 2022), particularly for emerging economies with fragile capital markets and imperfect institutions. Taking several well-known crashes in China as an example, the stock price of Huishan Dairy, a Hong Kong listed company, plunged 91% due to fraudulent charges in 2017. Later, in 2018, the stock price of Changsheng Biologicals, a Chinese vaccine maker, decreased by 86% after being accused of fabricating data. Similarly, the stock price of Kangmei Pharmaceutical fell by more than 50% due to financial fraud in 2019. SPCR can undermine shareholder interests, erode investor confidence, and obstruct the progress of the capital market. The prosperity and stability of the stock market play a critical role in financial market development. Based on this, research on the SPCR in emerging markets deserves further investigation to prevent market risks and stabilize economic security.

The capital market developed unsteadily after the global financial crisis (Cao et al. 2022; Hayat et al. 2022), and scholars have focused on evaluating and predicting SPCR (Cao et al. 2019). SPCR could be driven by factors from macro, meso, and micro dimensions Ali et al., (2022). However, related studies have reached a consensus that this difference is fundamentally caused by differences in the information environment. For example, management manipulation and information asymmetry can both become significant drivers (Jin and Myers 2006; Hutton et al. 2009). Taken together, the breakdown of corporate governance can be credited with the occurrence of SPCR. As a key part of corporate governance, boards act to monitor management behaviors and alleviate agency conflicts. From this perspective, one potential mitigation mechanism for SPCR relies on board monitoring. In real-world business, a board comprises directors with heterogeneous characteristics and various backgrounds, which can lead to differences in board effectiveness. Therefore, board composition has increasingly attracted the attention of scholars, investors, and policymakers because of its vital role in alleviating agency concerns and protecting the interests of shareholders. However, the relationship between board composition and SPCR is largely unclear. With the exception of the perspectives of female directors (Qayyum et al. 2021) and informal board hierarchy (Jebran et al. 2019), there is scant evidence regarding the impact of board diversity (BD) on the SPCR (Jebran et al. 2020). To fill this gap, our aim is to explore the effect of BD on the SPCR that firms face.

As a growing area of corporate governance, BD has been studied from the dimensions of gender, age, educational background, professional background, personal experience, and so on. Prior studies have examined the link between BD and corporate behaviors, including innovative activities (Cumming and Leung 2021), social responsibility (Beji et al. 2021), risk-taking (Zhang and Luo 2021), and financial manipulation (Wahid 2019). However, the following aspects of this strand of literature still need further investigation. First, most studies narrow their focus to a single dimension of diversity while ignoring the fact that the effectiveness of board function depends on the joint influence of overall BD (Baranchuk and Dybvig 2009; Harjoto et al. 2018). Second, little evidence is available on the impact of BD on the capital market, particularly the SPCR in emerging markets, which raises increasing concerns from investors. Third, considering the dual nature of BD’s impact (Tasheva and Hillman 2019), there is scarce and conflicting evidence regarding the effect of BD on the efficacy of boards. Some scholars begin with agency theory and support the governance effect of BD (Aggarwal et al. 2019; Griffin et al. 2021; Fang et al. 2021), while others hold that BD aggravates conflicts and creates inefficiency (Adams and Ferreira 2009; Talavera et al. 2018). This inconclusive effect leads to different corporate outcomes, and the linkages between overall BD and SPCR remain an open question.

Our study answers these questions by constructing a multidimensional index system to capture BD and examining its influence on the SPCR based on Chinese listed companies from 2010 to 2020. Three reasons motivate us to focus on firms in China. First, understanding the factors that influence SPCR is vitally important, both from a scholarly perspective and for real-world applications, particularly in burgeoning economies such as China. In these environments, capital markets often suffer from inadequate investor safeguards and the shortcomings of nascent institutional frameworks, hindering their efficacy (Zhao et al. 2023; Allen et al. 2024; Deng et al. 2024). Second, boards of directors are pivotal in overseeing management activities and fostering company value creation, especially in emerging markets with ineffective external governance mechanisms (Li and He 2023). However, empirical research on the influence of BD on capital markets within developing economies is scarce (Su et al., 2024). Third, one of the key challenges of the diversity literature is that its definition differs from place to place, implying that diversity research in emerging economies may distinguish its definition and measure from that of developed economies (An et al., 2021). To measure the impact of BD on the SPCR in the Chinese context, BD must be captured from special attributes that make a difference for directors in China (Lu et al., 2022). After theoretical analysis and empirical tests, we verify that BD can significantly mitigate the SPCR. This conclusion holds robust after endogeneity and robustness tests. The results of the impact mechanism assessment demonstrate that agency costs, investment efficiency, disclosure and executive appointment are the four paths through which BD affects the SPCR.

The remainder of the paper is structured as follows: the second section reviews the relevant literature. The third section analyzes the impact of BD on the SPCR and the related mechanisms. The fourth section describes the methodology of the study, encompassing the selection of participants, the origins of the data, the quantification of variables, and the configuration of the analytical model. The fifth section presents the empirical results and analysis. Part VI presents the discussion and conclusions.

Literature review

Our study is related to two streams of literature. First, this paper builds upon the literature that explores BD and corporate risk. Boards consist of members from diverse backgrounds. Therefore, the effectiveness of BD depends on the combined effect of BD (Baranchuk and Dybvig 2009). When describing BD, scholars mostly consider the demographic characteristics of directors, such as gender, age, education, professional background, and experience in the field (Baker et al. 2020; Harjoto et al. 2018; Knyazeva et al. 2021). There is limited research on how the board as a whole affects its effectiveness (Ben Selma et al. (2022)), and focusing on only a single dimension of diversity may lead to biased findings on the governance effects of BD. In addition, a significant portion of the research concerning BD has predominantly concentrated on the advanced capital markets located in North America and Europe (Aggarwal et al. 2019), with a lack of attention given to emerging capital markets. An et al. (2021) constructed comprehensive indicators of BD based on six dimensions—educational background, demographic attributes, cultural attitudes, management traits, occupational background, and directors’ experience—which provides a reference for this paper in constructing indicators of BD in Chinese listed companies.

Related studies have examined BD’s influence on corporate risk-taking, risk disclosure, and capital market risk. Regarding corporate risk taking, one viewpoint posits that BD heightens the propensity for risk taking within corporations. (Poletti-Hughes and Briano-Turrent 2019) and promotes corporate innovation activities (Cumming and Leung 2021). At the same time, other scholars have reached the opposing conclusion that BD mitigates corporate risk (Zhang and Luo 2021), including credit risk (Kinateder et al. 2021) and financial risk (Li et al. 2022), while the specific skill and cultural diversity of the board can significantly reduce firm risk (Gallego‐Álvarez and Pucheta‐Martínez 2022). Concerning the disclosure of risk information, BD can increase the voluntary disclosure of risk information and hence promote firm performance (Reguera-Alvarado and Bravo-Urquiza 2020). It has also been found that BD can significantly contribute to risk information disclosure only if board gender diversity reaches a critical mass (Seebeck and Vetter 2022). Regarding capital market risk, related studies mostly focus on SPCR.

Second, our study intersects with existing scholarly work on the development of the SPCR. Extant work has linked SPCR with information disclosure, management incentives, capital markets, and formal and informal institutions (Habib et al. 2018). Despite its variety, most of the literature on this strand commonly attributes SPCR to management’s tendency to hoard bad news. Due to the uneven distribution of information between management and shareholders, managers often conceal negative information to expand their corporate dominance (Kothari et al. 2009), increase their option value (Kim et al. 2011), and retain their positions and salaries (Xu et al. 2014), thereby leading to improved SPCR (Jin and Myers 2006). Following this logic, Hutton et al. (2009) demonstrate that enhanced earnings management diminishes the clarity of information and intensifies the SPCR encountered by firms, which is also verified in the Chinese context (Chen et al. 2017). This suggests that managerial practices of withholding unfavorable information often stem from self-serving motives, frequently culminating in SPCR (Hutton et al. 2009). Therefore, the role of governance mechanisms, such as the board, in mitigating stock crash risk should be further highlighted. This is particularly the case in emerging economies such as China, where legalization and external governance mechanisms are relatively weak.

Closest to the current research, related studies about the link between board composition and SPCR usually begin with board structural characteristics, such as board size, the percentage of outside directors (Andreou et al. 2016), and internal characteristics concerning the background of directors. For example, Qayyum et al. (2021) and Cao et al. (2019) indicate that board gender diversity and experience diversity are both negatively associated with SPCR. Jebran et al. (2020) provide evidence regarding how BD alleviates the impact on the SPCR by capturing overall BD from the perspectives of gender, age, tenure, and education. Accordingly, grounded in emerging economies, evidence of the overall influence of BD on the SPCR is relatively inadequate, the measure of BD is still limited to insufficient perspectives, and its underlying mechanisms remain unknown. To bridge this research gap, we construct a multidimensional index system containing five attributes and eleven second-tier indicators to capture BD and explore its influence on the SPCR. To further open the black box, we identified four underlying mechanisms to explain the mitigating effect of BD on the SPCR.

Hypothesis development

Mitigating effect of BD on SPCR

Boards composed of directors with different backgrounds bring creativity, different cognitive styles, and a better understanding of marketplaces, which leads to high-quality decisions and promotes firm performance. Furthermore, BD offers additional assets, including expertise, abilities, information availability, and specialized industry knowledge. Accordingly, diversity enhances the board’s capacity to absorb and analyze information, leading to more effective problem solving and overall better performance (Carter et al. 2003; Harjoto et al. 2015). Specifically, a diverse board leads to lower agency costs and greater supervision efficiency and thus reflects the tougher monitoring of management (Carter et al. 2003; Adams et al. 2015). Specifically, Aguilera et al. (2008) start with opinion diversity and confirm its significance for the board’s monitoring function. Bernile et al. (2018) suggest that the diverse perspectives provided by board members from various backgrounds enhance the board’s ability to supervise effectively, ultimately improving its overall performance. Moreover, the diversity of professional backgrounds is found to have governance effects. Directors with financial and legal backgrounds help enhance the supervision of management and thus restrict earnings management (Wang et al. 2015). Concerning independent directors, a board with more independent directors performs stricter supervision, which helps improve the quality of financial reporting (Klein 2002). Centering on board gender diversity, prior work provides evidence of its positive role in bringing more board meetings, mitigating internal control weakness, facilitating within-board communication, enhancing information transparency, and ultimately enhancing board effectiveness in monitoring management (Adams and Ferreira 2009; Chen et al. 2016; Gul et al. 2011). Additionally, term diversity is confirmed to reduce the possibility of groupthink and foster the board’s supervising role (Li and Wahid 2018). Taken together, BD is expected to enhance board monitoring effectiveness and limit management’s self-serving behaviors, thereby mitigating SPCR. Drawing from prior analysis and current studies, this paper posits that BD influences SPCR through four paths.

The first is agency costs. In line with principal-agent theory, it is common for managers to prioritize their own self-interests, such as career advancement and building business empires, at the expense of shareholders, and boards of directors with a broad range of backgrounds can proficiently oversee managerial operations (Carter et al. 2003). Directors from different backgrounds are able to present diverse perspectives, thus reducing decision-making bias from a single perspective (An et al. 2021). Boards with a mix of backgrounds are better equipped to identify issues and opportunities, oversee management practices more efficiently, and ultimately curb any self-serving tactics by management (Wahid 2019). Directors from different backgrounds can more appropriately design management incentive plans to better serve the interests of shareholders. BD can reduce agency costs through effective monitoring and incentives (Liu et al. 2023). As monitoring efficiency increases, conflicts of interest between shareholders and management lessen, diminishing managers’ opportunities to conceal negative information. Consequently, the likelihood of management manipulating data decreases, and the chances of SPCR are also reduced (Wu et al. 2020).

The second factor is investment efficiency. Investment behavior is irreversible, and poor investment choices constitute a form of agency conduct that damages a company’s value and shareholder interests. These decisions frequently coincide with adverse outcomes such as project failures and declining business performance. Management is highly motivated to distort adverse information about a company to protect its own reputation and career prospects, which in turn intensifies the severity of the SPCR (Habib et al. 2018). The cumulative effect of poor investment decisions can increase gradually. Once the volume of negative information regarding these lackluster investments breaches a critical level, it may trigger the SPCR (Habib and Hasan 2017). Established studies confirm that board reform improves board performance, which can effectively enhance firms’ investment efficiency (Hu et al. 2020), including by curbing overinvestment and mitigating underinvestment (Yuan et al. 2024), thus mitigating the SPCR by weakening management’s incentives to manipulate information.

The third factor is information disclosure. The key to curbing the SPCR is to improve the quality of disclosure (Andreou et al. 2023). Revealing risks enhances the clarity of information, and evaluating these data enables managers to gauge the extent of danger precisely, thereby averting potential repercussions. Even though risky information is bad news, the SPCR is reduced when risky information is disclosed (Chen et al. 2017). Boards with a variety of perspectives are better attuned to the informational requirements of a myriad of stakeholders, which drives the disclosure of risk information (Radu and Smaili 2022). Risk disclosure can effectively explain market risk and increase stakeholders’ recognition of a firm’s prospective development hazards (Yang et al. 2024). Improving the corporate information environment and increasing information transparency (Chiu et al. 2018), bridging the information gap between internal and external companies can help mitigate the SPCR (Au et al. 2023).

Fourth, there are executive appointments. The varied composition of the board of directors mirrors the multifaceted culture of the organization. Under the influence of multiculturalism, the composition of the executive team will also show diversity (Gupta and Raman 2014). Similarity-attraction theory posits that individuals sharing common traits tend to gravitate toward one another (Zou et al. 2021). One of the core roles of directors is to appoint the management of the company, which means that BD is an important driver of executive diversity. Diversity in the executive team can bring different perspectives and experiences, which can increase the breadth and depth of information (Pham, Lo (2023)) and mitigate the accumulation of adverse information within the organization. When an executive team is composed of individuals hailing from diverse backgrounds, it is far more probable that they will scrutinize the company’s operations and financial situation from different perspectives, which can better process and share information, encourage the prompt release of unfavorable news, effectively improve the information environment (Wang 2023), and mitigate the SPCR.

Taken together, BD may bring different opinions, enhance board independence, and provide adequate resources, thus leading to improved board monitoring and mitigating SPCR. We anticipate that BD will alleviate the SPCR by diminishing the prevalence of agency conflicts and curbing imprudent investments. Moreover, it is projected to enhance the transparency of disclosures and broaden the diversity within executive ranks. Therefore, we propose the following hypothesis.

H1a: The higher the BD is, the lower the SPCR, and BD mitigates SPCR.

H1a-1: BD mitigates SPCR by reducing agency costs.

H1a-2: BD mitigates SPCR by reducing inefficient investment.

H1a-3: BD mitigates SPCR by increasing disclosure.

H1a-4: BD mitigates the SPCR by increasing executive team diversity.

Exacerbating effect of BD on SPCR

In addition to the mitigating effects discussed above, prior work also contends that BD has negative consequences, supporting the exacerbating effect of BD on the SPCR. First, BD leads to higher decision costs and slower decision-making. Giannetti and Zhao (2019) argue that BD results in inefficiencies in decision-making and creates conflicts within the board, which damages mutual trust and cooperation among the directors, making it even harder to reach an agreement. Therefore, BD can present challenges in consensus decision-making, which ultimately leads to poor monitoring of management (Harjoto et al. 2015). In addition, building on social identity theory, differences in gender, education, and experience lead to the formation of within-board subgroups. The literature on group behavior also shows that group heterogeneity makes it difficult for individuals with different backgrounds to understand each other, creating more conflicts and communication barriers among them. Following this logic, BD may increase in-group and out-group diversity and hence harm group cohesion, leading to a failure to effectively monitor management and allowing for more self-serving behaviors (Salloum et al. 2019; Van Peteghem et al. 2018). Poor governance ultimately exacerbates the SPCR.

In summary, BD lengthens the decision-making process, increases decision costs, fosters the formation of subgroups, creates more conflicts within the board, and damages group cohesion, which leads to board inefficiency in monitoring management and thus exacerbates SPCR. Based on the negative governance perspective of BD, this paper can infer that BD exacerbates the SPCR by escalating agency costs, catalyzing inefficient investments and reducing disclosure. From the executive hiring perspective, we believe that BD significantly increases executive team diversity. However, executive team diversity can undermine team cohesion, trigger relationship conflicts within the team, and reduce trust and respect among team members. Noncooperation between executives will lead to communication barriers, and the quality and accuracy of information will be greatly diminished, which in part exacerbates the SPCR. Therefore, we propose the following hypothesis.

H1b: The greater the BD is, the greater the SPCR, and BD exacerbates SPCR.

H1b-1: BD exacerbates SPCR by increasing agency costs.

H1b-2: BD exacerbates the SPCR by reducing investment efficiency.

H1b-3: BD exacerbates SPCR by reducing disclosure.

H1b-4: BD exacerbates the SPCR by increasing executive team diversity.

Research design

Data source and sample selection

We compiled a research sample of all companies listed on the Chinese A-shares market between 2010 and 2020, sourced from the CSMAR,Footnote 1 resulting in an initial dataset of 28,462 firm-year observations. Then, we exclude firms (1) in the financial and insurance industry sectors; (2) that are ST, *ST,Footnote 2 and delisted during the research period; (3) listed for less than one year; (4) with missing financial data; and (5) with fewer than 30 annual cumulative trading weeks. Finally, our dataset includes 17412 firm-year observations. To mitigate the impact of outliers, all continuous variables were winsorized at the 1st and 99th percentiles. To overcome concerns about endogeneity, we set the research period to 2010–2019 for BD and control variables and 2011–2020 for SPCR.

Definition of variables

  1. (1)

    Our dependent variable was the SPCR. We measure SPCR as negative conditional skewness (NCSKEW) and down-to-up volatility (DUVOL) following prior work (Kim et al. 2011; Kim et al. 2016)), which are widely used by researchers. Equations (1–3) show our calculations. Specifically, Eq. (1) is used to eliminate the impact of market variables on the returns of individual stocks, where Ri,t represents the return on stock i in week t, which is the weekly market return in week t. The residual εi,t represents the segment for which market elements fail to account. \({W}_{i,t}=ln(1+{\varepsilon }_{i,t})\) is the firm-specific weekly return of stock i in week t.

    $$\begin{array}{l}{R}_{i,t}={\beta }_{0}+{\beta }_{1}{R}_{m,t-2}+{\beta }_{2}{R}_{m,t-1}+{\beta }_{3}{R}_{m,t}\\\qquad+{\beta }_{4}{R}_{m,t+1}+{\beta }_{5}{R}_{m,t+2}+{\varepsilon }_{i.t}\end{array}$$
    (1)

    After obtaining the firm-specific weekly returns of individual stocks, we further use Eqs. (2) and (3) to capture the SPCR, where n represents the number of weeks that stock i is traded in the market in year t. In Eq. (3), ni,t,up (ni,t,down) indicates how often the weekly return of stock i specific to the company exceeds (or falls below) the average annual return in year t.

    $$\begin{array}{l}NCSKE{W}_{i,t}=-\left[n{(n-1)}^{3/2}\sum {W}_{i,t}^{3}\right]/\\\qquad\qquad\qquad\left[(n-1)(n-2){\left(\sum {W}_{i,t}^{2}\right)}^{3/2}\right]\end{array}$$
    (2)
    $$\begin{array}{l}DUVO{L}_{i,t}=\,{\rm{Log}}\left\{\left[({n}_{i,t,up}-1)\mathop{\sum}\limits _{down}{W}_{i,t}^{2}\right]/\right.\\\left.\qquad\qquad\qquad\left[({n}_{i,t,down}-1)\sum _{up}{W}_{i,t}^{2}\right]\right\}\end{array}$$
    (3)
  2. (2)

    Our independent variable is BD. Following An et al. (2021), we develop a multidimensional index system designed to gauge BD based on various director traits. Considering prior work on diversity, data availability, and the special context of China, we incorporate demographic characteristics, educational background, professional background, personal experience, and managerial traits of directors into one system. The data for measuring BD in this paper all come from CSMAR, a database that discloses the basic characteristics, background characteristics, and concurrent positions of board members of all Chinese listed companies, which facilitates this paper’s measurement of BD. Table 1 shows our index system for capturing BD. Specifically, we quantify the diversity of board size through the total number of board members. An increase in the number of board members represents greater board size diversity. We measure the diversity of age and outside directors using the coefficient of variation. All other indicators are gauged using the Herfindahl–Hirschman Index (HHI). Equation (4) presents the calculation of these indicators.

    Table 1 BD indicator system and its calculation.
    Full size table
    $$Diversity=1-\mathop{\sum }\limits_{i=1}^{n}{p}_{i}^{2}$$
    (4)

    where pi is the proportion of board members with certain background characteristics. n is the number of background characteristics. Taking educational diversity as an example, we categorize the educational qualifications of directors into six distinct groups: those with a technical high school diploma or less, associate degree holders, bachelors, masters, doctoral degree recipients, and a category for all other educational backgrounds. p1 represents the proportion of board directors who have completed their education at a technical secondary school or at a lower level, and p2 represents the proportion of board directors holding a junior college degree. At this time, n is 6. For gender diversity, n is coded 2, where p1 and p2 are the ratios of male to female directors on the board, respectively. Other indicators are measured in a similar way. For diverse professional backgrounds, we follow Anderson et al. (2011) and categorize board members into two distinct groups depending on their expertise in finance or law. Furthermore, we measure position diversity based on whether the board members work on the executive team at the same time. In particular, considering the special Chinese context, we incorporate the critical personal experience of “the Great Chinese famine” from 1959 to 1961 into the overall BDFootnote 3. This indicator is captured by whether the board member was born at or before 1961, which equals 1 if so and 0 otherwise.

    We then rank each second-tier indicator by year, segment our dataset into ten distinct clusters based on the valuation of each metric, and assign each group a value of 1–10. The group with the lowest value equals 1, and the group with the highest value equals 10. We then sum the assigned value of each second-tier indicator to capture the first-tier indicator. Finally, we obtain the overall BD by summing all first-tier indicators, which are further divided by 100 to match the range of dependent variables.

  3. (3)

    Control variables. We follow prior work (Kim et al. 2011; Kim et al. 2016) and control for the mean of the firm-specific return (RET), volatility of the firm-specific return (SIGMA), turnover rate (DTURN), firm size (SIZE), gearing ratio (LEV), profitability (ROA), and book-to-market ratio (BM). In addition, this paper measures corporate transparency (ACCM) using the sum of the absolute value of manipulable accrued profits over the past three years (Hutton et al. 2009). Additionally, this document incorporates fixed effects for both year and industry within the regression analysis. Table 2 shows the definitions of the key variables.

    Table 2 The definition of variables.
    Full size table

Model specification

To test the above hypotheses, we construct regression model (5) and mainly focus on the coefficient of BD, which supports H1a if it is negative and H1b if it is positive.

$$\begin{array}{l}NCSKE{W}_{i,t+1}/DUVO{L}_{i,t+1}={\alpha }_{0}+{\alpha }_{1}B{D}_{i,t}+Control{s}_{i,t}\\\qquad\qquad\qquad\qquad\qquad\qquad\quad+\,\sum YEAR+\sum INDUSTRY+{\varepsilon }_{i,t}\end{array}$$
(5)

Empirical results and analysis

Descriptive statistics

Table 3 shows the results of the descriptive statistics for the key variables. The mean, minimum, maximum, and standard deviation of the negative conditional skewness (NCSKEWt+1) are −0.273, −2.332, 1.444, and 0.670, and −0.180, −1.296, 0.949, and 0.461, respectively, for down-to-up volatility (DUVOLt+1). This variation underscores the extensive diversity among Chinese publicly traded companies with respect to their SPCRs. Regarding our independent variable, the standard deviation of BD stands at 0.107, exceeding the 0.060 value reported by An et al. (2021) in their study of U.S. listed companies, demonstrating that BD varies more dramatically in Chinese capital markets than in the United States.

Table 3 Descriptive statistics.
Full size table

Univariate analysis

We carry out a univariate analysis on the primary variables before the baseline regression. Specifically, we segmented our sample into two cohorts by using the median BD as the dividing line and then examined the variations in the primary variables between these groups. As shown in Table 4, the means of SPCR in the high BD group were −0.299 and −0.198, respectively, which are notably less than the −0.247 and −0.162 recorded for the group with low BD. The results show comparable trends when analyzed using the median z test, indicating that companies with varied board membership experience have reduced levels of SPCR. In addition, except for information transparency (ACCM), the means and medians of the control variables reveal notable disparities between the two groups at the 1% significance level, underscoring the imperative to control these factors in the subsequent empirical analysis.

Table 4 Univariate analysis.
Full size table

Baseline regression

Table 5 displays the findings from our regression analysis. We take negative conditional skewness (NCSKEWt+1) as the dependent variable in Columns (1)–(2) and down-to-up volatility (DUVOLt+1) in Columns (3)–(4). Controlling for year and industry fixed effects, we gradually add control variables to the regression. Our analysis shows that the coefficients relating BD to SPCR are consistently negative and statistically significant at the 1% level, irrespective of the inclusion of control variables. This finding implies that improved BD leads to lower SPCR, which supports H1a and primarily reflects the governance effect of BD. Specifically, the coefficients of BD on NCSKEWt+1 and DUVOLt+1 are −0.188 and −0.128, respectively, controlling for other factors and fixed effects. Therefore, we can evaluate the economic significance by capturing the impact of a one standard deviation change in BD on SPCR, which is 7.368% for NCSKEWt+1 and 7.609% for DUVOLt+1.Footnote 4 The R2 of all regressions ranged between 0.04 and 0.06, which is generally consistent with the literature on the impact of directors’ influence on the SPCR (Fang et al. 2021; Jebran et al. 2019; Hu et al. 2020).

Table 5 Regression results of BD on SPCR.
Full size table

Endogeneity tests and robustness tests

The results are presented in the Appendix.

Mechanism test

Prior empirical evidence suggests a correlation between elevated instances of BD and a diminished occurrence of SPCR. This paper argues that the dampening effect of BD on the SPCR works through four channels: reducing agency costs, reducing inefficient investment, improving disclosure, and appointing diverse executives.

To provide empirical evidence, we apply mediating effect models (6) and (7) to test the above four channels. Model (6) serves to examine the influence of BD on the mediating variables, and model (7) is constructed by incorporating the mediating variables into model (5). We mainly focus on the significance of the coefficients β1, γ1, and γ2, and the mediating effect holds if both β1 and γ2 are significant.

$${M}_{i,t}={\beta }_{0}+{\beta }_{1}B{D}_{i,t}+Control{s}_{i,t}+\sum YEAR+\sum INDUSTRY+{\varepsilon }_{i,t}$$
(6)
$$NCSKE{W}_{i,t+1}/DUVO{L}_{i,t+1}={\gamma }_{0}+{\gamma }_{1}B{D}_{i,t}+{\gamma }_{2}{M}_{i,t}+Control{s}_{i,t}+\sum YEAR+\sum INDUSTRY+{\varepsilon }_{i,t}$$
(7)

where M is the mediating variableFootnote 5. The other variables are consistent with those in the previous section.

  1. (1)

    Reducing agency costs. According to Carter et al. (2003), a diverse board is effective at overseeing management and thus can diminish the agency conflicts between shareholders and management, curbing the self-interested actions of the latter. Therefore, management is less likely to manipulate information, thus alleviating SPCR. To test empirically, we take agency costs (M1) as the mediating variable and measure it as the quotient of combined administrative and sales outlays over the principal business’s operating income. Table 6, Panel A, displays the results of the regression analysis. We find that the coefficient of BD in Column (1) is significantly negative, indicating that improved BD leads to lower agency costs. In Columns (2) and (3), the coefficients of agency cost (M1) on SPCR are both positive and pass the significance test, which suggests that agency problems can exacerbate SPCR. The coefficients of BD on SPCR are −0.183 and −0.123, respectively, which are highly significant, with a certainty level of 99%. This implies that BD alleviates the agency problem, ultimately mitigating SPCR, which supports the mechanism for reducing agency costs (H1a-1).

    Table 6 Mechanism test.
    Full size table
  2. (2)

    Alleviating inefficient investment. The well monitoring enabled by BD could make a difference by enhancing the firm’s investment efficiency, reducing the management’s incentive to manipulate information, and hence mitigating SPCR. To test this channel, we measure inefficient investment (M2) following Richardson (2006) and take it as the mediating variable. It should be noted that inefficient investment is a negative indicator, and an increase in this indicator represents fewer efficient investments. Panel B of Table 6 shows that the coefficient of BD on inefficient investment (M2) is −0.024 and demonstrates compelling statistical relevance at the 99% confidence interval, suggesting that BD alleviates a firm’s inefficient investment. In Columns (5) and (6), the coefficients of inefficient investment (M2) on SPCR (NCSKEWt+1, DUVOLt+1) are both significantly positive, while the coefficients of BD on SPCR (NCSKEWt+1, DUVOLt+1) remain significantly negative, which validates that BD alleviates inefficient investments and hence mitigates SPCR (H1a-2).

  3. (3)

    Improving information disclosure. High-level BD allows a firm to improve information disclosure and disseminate more information (Abad et al. 2017; Gul et al. 2011). To test for this channel, we capture the amount of risk information disclosed by listed companies as a proxy for information disclosure (M3). An increased amount of risk information disclosed by a firm represents greater information disclosure and lower information asymmetry. Similarly, we take information disclosure as the mediator and present the results in Panel C of Table 6. Column (1) shows that the coefficient of BD on information disclosure (M3) is significantly positive, indicating that BD helps enhance information disclosure. After incorporating the mediator into the regression model, Columns (2) and (3) show that the coefficients of BD on SPCR (NCSKEWt+1, DUVOLt+1) remain significantly negative, and the coefficients of information disclosure (M3) on SPCR (NCSKEWt+1, DUVOLt+1) are negative and pass the significance test at the 10% level. This result verifies the channel of improving information disclosure, suggesting that BD mitigates SPCR by improving information disclosure (H1a-3).

  4. (4)

    Appointing diverse executives. Building on social categorization theory and similar attraction theory, board members tend to select executive members with similar background characteristics. Therefore, we propose that improved BD leads to a diverse executive team, which shapes effective internal oversight within the TMT. Mutual supervision among executives curbs the self-interested actions of management. To test for this channel, we capture executive diversity (M4)Footnote 6 through a multidimensional index system consistent with BD and take it as the mediator. Table 6, illustrated in Panel D, shows that the coefficient of BD on executive diversity (M4) is 0.249, with statistical significance at the 1% level according to the results of Column (4). Columns (5) and (6) show that the coefficients of BD on SPCR (NCSKEWt+1, DUVOLt+1) are still significantly negative. Similarly, the coefficients of executive diversity (M4) on SPCR (NCSKEWt+1, DUVOLt+1) are both negative. However, one possible concern is the collinearity issue caused by the similar measures of BD and executive diversity. Given this, we conduct a correlation test between them, and the correlation is less than 0.5, suggesting that the collinearity issue is not severe. This finding validates that BD leads to diverse executive teams, thereby effectively mitigating SPCR (H1a-4).

Discussion and conclusion

Conclusion

Stock price crashes are destructive, contagious, and concealed and can shake investors’ confidence, harm the firm’s interests, and thus hamper the stability of the financial market. Even worse, it would trigger systemic financial risks and lead to an economic decline (Liu et al. 2022). Therefore, mitigating the SPCR has drawn increasing attention from scholars and policymakers. A vast amount of literature regards corporate governance failure as the fundamental cause of SPCR (Andreou et al. 2016). Building on this, one potential mitigation mechanism of SPCR is board monitoring, which is largely influenced by board composition.

We construct a composite index of BD utilizing information from publicly traded Chinese corporations and explore its impact and mechanism on the SPCR. The findings indicate that BD effectively curtails the SPCR, and H1a was tested. This conclusion holds robust after passing endogeneity and robustness tests. We expanded the research perspective of BD from single dimensions such as gender and age (Abou-El-Sood 2021; Cumming and Leung 2021; Jebran et al. 2020) to a multidimensional perspective. Our study showed that BD can effectively mitigate SPCR. Our findings are consistent with the BD literature focusing on a single dimension (Qayyum et al. 2021; Cao et al. 2019; Jin et al. 2022; Jebran et al. 2020), validating the mitigating effect of BD from both a general and single perspective.

The results of the impact mechanism test show that BD can mitigate the SPCR by reducing agency costs and inefficient investment and improving disclosure and executive team diversity, and H1a-1, H1a-2, H1a-3, and H1a-4 are tested. BD has been demonstrated to reduce SPCR (Jebran et al. 2020), but the mechanism by which BD affects SPCR remains unknown. Our study reveals the “black box” of BD affecting SPCR.

Theoretical implications

Our research provides valuable insights into the existing body of literature on BD and SPCR.

For the literature related to BD, we investigate the influence of BD on capital market risk from a general perspective in the emerging economy. Board composition is extremely important for board effectiveness and strategic decision-making and has attracted growing attention in the field of corporate governance (Adams et al. 2015). We respond to the call to estimate the overall influence of BD and extend the literature on this topic. Previous research has predominantly focused on the impact of BD through a singular lens, such as gender and age (Abou-El-Sood 2021; Cumming and Leung 2021), while few scholars have studied the joint effect of BD from multiple dimensions in emerging economies. Although a few studies on this topic have attempted to capture BD from a larger scope in the context of emerging economies (Jebran et al. 2020), their efforts are still limited, and unique factors embedded in the setting are commonly overlooked. Therefore, we construct a multidimensional index system containing five attributes to measure BD in the Chinese context and empirically examine how BD affects the SPCR. BD can effectively alleviate the SPCR, which provides fresh insights into the impact of widespread BD on the mechanisms of governance in emerging economies.

For the research on the SPCR, we contribute to the existing scholarly discussion by exploring the underlying mechanisms that animate the relationship between BD and the SPCR, thereby bringing clarity to this opaque area of study. The antecedents of SPCR primarily comprise factors related to information disclosure, management incentives, capital markets, and institutions, which have been well established in previous work (Habib et al. 2018). SPCR is confirmed to be essentially caused by ineffective corporate governance (Andreou et al. 2016). Research within this field has delved into the factors influencing SPCR from the viewpoint of the board (Jebran et al. 2020; Qayyum et al. 2021; Cao et al. 2019; Jin et al. 2022). Limited evidence in prior work shows that BD is associated with lower SPCR (Qayyum et al. 2021; Cao et al. 2019; Jin et al. 2022; Jebran et al. 2020), but the mechanism underlying the relationship between BD and SPCR remains unknown. Therefore, our study opens the black box of BD and the SPCR by validating four underlying mechanisms for reducing agency costs, decreasing inefficient investments, improving information disclosure, and appointing diverse executive teams.

Managerial implications

Building on our findings, we propose the following managerial implications.

First, our findings have important implications for board composition for listed companies, particularly in emerging markets. We confirmed the effectiveness of BD in mitigating SPCR. To thrive within the competitive landscape of the capital market, firms ought to recruit board members from varied backgrounds, thereby assembling a multifaceted board of directors. At the same time, adequate information flow and communication channels within the board should be ensured, which will help bring different opinions and improve board monitoring effectiveness. However, despite the positive consequences of BD, previous work has shown that excessive BD can lead to more conflicts, slower decision-making, and damage to group cohesion. Therefore, firms should consider both the cost and benefit of BD and try to determine the optimal board composition.

Second, our research provides significant insights for regulators and policymakers, underscoring the importance of BD. Stock price crashes are unpredictable, concealed, and destructive and can seriously hinder the development of firms and capital markets in the long run. Our findings show that a diverse board reduces agency costs, alleviates inefficient investments, improves information disclosure, and appoints a diverse executive team, which mitigates the SPCR. Therefore, regulators and policymakers could pay more attention to BD and guide firms to enhance their BD through external supervision and regulations.

Third, we confirm that BD is an effective mitigation mechanism for SPCR, which could provide key guidance for investors. Faced with information asymmetry, investors are vulnerable to changes in stock prices and may suffer from severe damage in the SPCR, particularly in emerging markets with weak investor protection. Our study provides empirical evidence that BD is associated with fewer agency conflicts, less inefficient investments, less information asymmetry, and diverse executive teams and ultimately alleviates the SPCR. Therefore, BD signals the quality of corporate governance and predicts SPCR. Investors should factor this in while deciding where to allocate their investments and try to evaluate a firm’s BD from multiple dimensions.

Limitations and future research

This study has the following limitations. First, we confirm that agency costs, investment efficiency, disclosure, and executive tenure are the mechanisms by which BD mitigates the SPCR, but we do not consider the interactions between the different mechanisms. In future research, we can further explore the path through which BD affects the SPCR by introducing a chain mediation model. Second, due to limited data availability, we measure BD based on directors’ personal attribute characteristics. In subsequent studies, we can consider factors such as cognitive diversity and cultural diversity to explore more accurate and comprehensive measures of BD. Third, this paper delves into the effects of BD on SPCR within the context of emerging capital markets. There are differences in institutions, cultures, market environments, and laws and regulations in different countries and regions, which may lead to the impact of BD on the SPCR showing different characteristics in different environments, and this is one of the directions that should be researched in the future.