Does the financial flexibility prevent stock price crash risk during COVID-19 crisis? Evidence from the Vietnamese stock market

Abstract

Stock price crash risk is of particular interest in developing countries as it poses a significant threat to investors and can have detrimental effects on the stability of emerging markets. This study investigates the role of financial flexibility in preventing stock price crash risk in the Vietnamese stock market, with a specific focus on the COVID-19 pandemic. Using the fixed-effect, system GMM, and quantile regression methods on a sample of 645 Vietnamese listed firms from 2011 to 2021, this study found that financial flexibility has a significant impact on preventing stock price crash risk. This effect was augmented during the COVID-19 crisis. Furthermore, this study found that financial flexibility mitigated the impact of the COVID-19 crisis on stock price crash risk. The findings provide important implications for firm regulators, shareholders, and investors to respond to similar future crises.

1. Introduction

Stock price crash risk (SPCR) refers to the likelihood or probability of a significant and sudden decline in the prices of stocks or securities in a financial market [1]. It represents the potential for a rapid and substantial decrease in the value of an individual stock, a specific sector, or even the overall market. It is for this reason that SPCR is a major concern for shareholders, investors, and regulators. Advancements in technology and the rise of algorithmic trading have led to increased concerns about SPCR [2,3]. The use of automated trading systems and complex financial instruments can amplify market movements and potentially contribute to abrupt price declines [4]. In developing countries, the financial system is often fragile, and investor sentiment is strongly affected by market fluctuations [[5], [6], [7], [8]]. Therefore, it is vital to study SPCR in the context of developing countries.

Several studies have attempted to analyze the factors that influence SPCR within different markets. Corporate governance studies have suggested that SPCR is a result of the management holding on to bad news and that SPCR therefore depends on the corporate governance structure of firms. For example, Dang and Nguyen [1] found that a strong board of directors (BOD) can increase SPCR, whereas a good external audit service can reduce such effect. Similarly, Wu et al. [9] found that a favorable corporate governance mechanism helps reduce SPCR in the Taiwanese stock market.

From an economic shock perspective, prior studies have provided evidence of the impact of macro-economic factors on SPCR. Callen and Fang [10] provided evidence that firms headquartered in countries with higher levels of religiosity exhibit lower levels of future SPCR. Xiao et al. [11] found that oil-price uncertainty increases SPCR in Chinese stock market. Similarly, Luo and Zhang [12] found a positive relationship between economic uncertainty and SPCR in the Chinese stock market. Nguyen and Dang [13] found that FinTech development increases SPCR in developing countries.

There are limited studies that focus on SPCR despite the excessive volatility and weak corporate governance within emerging equity markets in Vietnam. Although it is a separate market, Vietnam's stock market has similarities with other emerging markets. Specifically, Vietnamese stock market is developing at a fairly fast pace and tends to experience higher levels of market volatility compared to developed markets. This volatility can be influenced by investor factors such as economic instability, political uncertainties, and changes in global sentiment. Vietnamese stock market also has a relatively fast growth rate like other emerging markets. However, this is considered a market that is less sentient or well-established compared to developed markets. This can lead to concerns about transparency, corporate governance, and investor protection, which could lead to higher SPCRs compared to developed countries.

In the context of countries and firms heavily affected by COVID-19 [14], the stock markets in countries around the world fluctuated greatly during the COVID-19 pandemic. Harjoto et al. [15] found that COVID-19 caused a negative shock to the global stock markets, especially in emerging markets and for small firms. Contessi and De Pace [16] provide evidence that the collapses in the stock markets of 18 major countries during the first wave of the COVID-19 pandemic of 2020. Therefore, policymakers as well as shareholders in most countries, especially emerging countries, are trying to find solutions to limit SPCR in response to possible pandemics in the future. For those reasons, this study aims to extend the existing literature by focusing on the Vietnamese stock market and investigating the impact of a firm's financial flexibility on future stock price crash risk to address issues of interest to academia, policymakers and investors.

First, to the best of our knowledge, the relationship between financial flexibility and stock price crash risk remains underexplored, despite the acknowledged importance of financial flexibility in firms. This study found that financial flexibility can reduce SPCR, and such an effect becomes stronger during the COVID-19 crisis. This finding may provide important implications for both firm's management to control crash risk and for investors to manage their portfolios. Second, in the context of the increase in SPCR during the COVID-19 crisis, it becomes increasingly important to investigate effective ways to control SPCR and manage portfolios. To best knowledge, this is the first study to investigate the role of financial flexibility in preventing the negative effects of COVID-19 on SPCR. Furthermore, although some recent studies have investigated SPCR, few of these have been conducted in developing and emerging markets. The current paper attempts to provide new facts and insights into emerging markets in Asia, specifically in Vietnam.

The remainder of this article is structured as follows: In Section 2, we review the literature and develop the hypotheses; Section 3 introduces the research method; Section 4 examines the results of the hypotheses testing. Finally, the paper concludes in Section 5 with a review of the central insights and policy implications.

2. Literature review and hypothesis development

2.1. Financial flexibility and SPCR

Previous studies, conducted from multiple perspectives, have agreed that financial flexibility can make firms more stable. According to DeAngelo et al. [17], the primary aim of managerial financial strategies is to ensure that the company maintains its ability to acquire capital from the market, particularly in situations involving unanticipated cash flow shortages or investment prospects. It was discovered that companies frequently opt to issue temporary debt or equity, which may result in temporary deviations from their intended long-term capital structure. This approach is adopted to prepare for unforeseen external factors that may disrupt the company's cash flow requirements, as these factors are typically beyond the control of companies and difficult to predict. Similarly, Barclay et al. [18] support the utilization of proactive equity issuances as a valuable means to enhance financial flexibility. Their research reveals that significant seasoned equity offerings (SEO) often result in firms decreasing their leverage levels below the desired long-term target-debt ratio. Moreover, the study indicates that companies engaging in proactive equity issuances are typically in good financial health; they have low leverage, unused debt capacity, and substantial cash balances. These companies create confidence among investors, thereby avoiding large fluctuations in stock prices. Fundamentally, financial flexibility provides managers with the ability to effectively manage unpredictable cash flow requirements while also enabling firms to optimize their inherent growth opportunities. Given its significant impact on value creation, financial flexibility plays a crucial role in explaining why changes in a company's financial flexibility should be factored into stock market pricing. In addition, “unexpected” cash flow shortfalls may increase bad news hoarding and thus increase future crash risk [[19], [20], [21]].

Moreover, agency theory posits that managers have incentives to act in their own self-interest rather than maximizing shareholder wealth. However, firms with higher financial flexibility are less dependent on external financing and have greater freedom to make optimal investment decisions during crises, reducing agency conflicts that can negatively impact stock prices. Based on agency theory, previous studies have agreed that insiders always want to increase the private capture of the cash flow of their firm. Jin and Myers [22] argue that the lack of transparency and information asymmetry in firms allows insiders to decrease the flow of information to the market, thereby raising the risk of stock market crashes. Moreover, shareholders are always faced with the choice between preserving financial flexibility and preventing the waste of free cash. Companies that pursue a financial flexibility strategy often have shareholders accepting agency costs because these companies are considered to have low agency costs [23]. Therefore, companies pursuing a financial flexibility strategy can be a sign of low SPCR. Based on these discussions, we propose the following hypothesis.

H1

Financial flexibility reduces future SPCR.

2.2. Financial flexibility and the impact of the COVID-19 crisis on SPCR

The pecking order theory suggests that firms have a preference for internal financing over external financing [24]. This theory implies that companies with strong internal financial resources and a financial flexibility strategy are better positioned to navigate economic downturns and maintain their stock prices. Gamba and Triantis [25] found that firms adopting the financial flexibility strategy are able to avoid financial distress in the face of negative shocks. Similarly, Fahlenbrach et al. [26] investigated the role of financial flexibility of firms during the COVID-19 crisis and found that companies in a particular industry that possess strong financial flexibility encounter a 26 % decrease in their stock prices—equivalent to a reduction of 9.7% points—compared to companies with low financial flexibility.

Previous studies have found strong evidence of the importance of financial flexibility during economic crises. For example, Dittmar and Field [27] found that companies with higher cash holdings experienced lower stock price rundowns during the 2008 financial crisis. Similarly, Meier et al. [28] found that firms with higher financial flexibility had a lower likelihood of experiencing firm performance and making stock prices stable during economic downturns.

In this study, we argue that financial flexibility can mitigate the effect of the COVID-19 crisis on SPCR through several mechanisms. First, firms with higher financial flexibility are better positioned to absorb unexpected shocks and maintain their operations during crises. This resilience can instill confidence in investors and reduce the likelihood of panic selling, leading to lower SPCR. Second, financially flexible firms can take advantage of strategic opportunities that arise during a crisis, such as acquiring distressed assets or investing in innovative solutions [29]. These proactive actions can signal resilience to investors and positively impact stock prices. Finally, during the COVID-19 crisis, firms faced difficulties in accessing external financing [14]. However, financially flexible companies with ample cash reserves or available credit lines could navigate these constraints more effectively, mitigating the negative impact on their stock prices. Based on these discussions, we propose the following hypotheses.

H2

Financial flexibility reduced the impact of the COVID-19 crisis on SPCR.

3. Research methodology

3.1. Data

The data was sourced from various outlets. The financial information of companies and weekly stock price data was obtained from FiinPro, a professional third party specializing in providing the data of Vietnamese firms listed on the Ho Chi Minh City and Hanoi Stock Exchanges. Marco-economic variables were collected from World Bank, and the COVID-19 variable was collected from “Our World in Data” website. From the Vietnamese listed firms in the FiinPro database, we excluded all firms that did not have data after 2019 (i.e., after the beginning of the COVID-19 crisis) as well as all financial firms. After excluding outliers, our final data comprised 645 firms, across ten industries, with 4557 observations from 2011 to 2021. We only accounted for data from 2011 onwards in order to exclude the effect of the 2008 financial crisis.

3.2. Variable measures

3.2.1. Stock price crash risk measures

In this research, we employed two primary measures as proxies of SPCR, i.e., the negative skewness of firm-specific daily returns (NCSKW) and the down-to-up volatility of firm-specific daily returns (DUVO). These proxies have also been extensively utilized in previous studies [1,21,30,31].

To calculate NCSKW, we first computed the weekly returns specific to each firm (W) by taking the natural logarithm of 1 summed with the residual return obtained from the regression of the expanded market model for each firm and year:

$$r_{i,t}={\alpha }_j+{\gamma }_{1,i}{r}_{m,t-2}+{\gamma }_{2,i}{r}_{m,t-1}+{\gamma }_{3,i}{r}_{m,t}+{\gamma }_{4,i}{r}_{m,t+1}+{\gamma }_{5,i}{r}_{m,t+2}+{\epsilon }_{i,t}$$ (1)

where $r_{i/m,t}$ represents the returns of the market and the firm in week $t$, i corresponds to the firm, while m refers to the market, and ${\epsilon }_{i,t}$ denotes the component of a firm's stock return that cannot be attributed to the market factor, thereby representing the firm's individual weekly returns. Based on prior studies [1,31], we used ${\mathrm{w}}_{\mathrm{i},\mathrm{\tau }}$ = ln (1 + ${\mathrm{\epsilon }}_{\mathrm{j},\mathrm{t}}$) to calculate firm-specific weekly return for firm i in one week

Next, we used the following equation to compute NCSKW:

$$NSKW=-\left[n{\left(n-1\right)}^{\frac{3}{2}}\sum w_{j,\tau }^3\right]/\left[\left(n-1\right)\left(n-2\right){\left(\sum w_{j,\tau }^2\right)}^{3/2}\right]$$ (2)

where n represents the total count of trading weeks for a specific firm i in year t.

The skewness of firm-specific daily returns captures the “asymmetry of the return distribution.” Therefore, a negative NCSKW indicates that the data is skewed to the left, and vice-versa.

We applied the following equation to calculate the down-to-up volatility of firm-specific daily returns (DUVO):

$${DUVO}_{J,\tau }=\log \{\left(n_u-1\right)\sum _{Down}{w}_{j,\tau }^2/\left(n_d-1\right)\sum _{Up}{w}_{j,\tau }^2\}$$ (3)

For each firm i in year t, we separated the weekly returns of firms into “up” and “down” groups. This categorization of weeks as “down” or “up” is based on the comparison of each week's firm-specific weekly returns with the average value of firm-specific weekly returns for year t. If a particular firm's weekly return surpasses the mean value for year t, the week is classified as “up”. Conversely, if the firm's weekly return falls below the mean, the week is labeled as “down”. nd(nu) is the number of weeks that firm i's specific weekly returns are lower (or higher) than the average firm-specific weekly returns over year t. The standard deviation of weekly returns was computed separately for each of the two groups. A high value of DUVO indicates a correspondingly high value of SPCR.

3.2.2. Financial flexibility measures

Financial flexibility refers to the extent to which a company possesses unused borrowing potential. In the study conducted by De Jong et al. [32], financial flexibility is quantified by calculating the disparity between the maximum debt ratio that a firm can maintain without encountering immediate financial distress and its actual debt ratio. To determine this measure, it is essential to have comprehensive data on a company's debt rating. However, since we lack such information, we adopted the approach proposed by Yung et al. [33] and Marchica and Mura [34], which employs a regression model to assess financial flexibility. This regression model is implemented by the following equation:

$${LEV}_{it}=\alpha {LEV}_{it-1}+\beta 1{INLEV}_{it-1}+\beta 2{MTB}_{it-1}+\beta 3{Size}_{it-1}+\beta 4{Tangibility}_{it-1}+\beta 5{Profibility}_{it-1}+\beta 6{Inflation}_{t-1}+u_{it}$$ (4)

To address potential issues of endogeneity between the dependent and independent variables, we incorporated lagged independent variables into our model, following the recommendation of Yung et al. [33] and Marchica and Mura [34]. The variables used in this analysis are defined in Table 1. According to our model, companies that exhibit a negative deviation between their actual leverage and the predicted leverage are identified as having untapped borrowing potential. Consistent with the approach of Yung et al. [33] and Marchica and Mura [34], we establish a criterion that requires a firm to demonstrate a minimum of three consecutive years of unused debt capacity before being categorized as financially flexible. In this study, we refer to the financial flexibility calculated from equation (4) as FF1, which serves as our primary measure of financial flexibility. FF1 is a binary variable that assumes a value of 1 when a firm exhibits a minimum of three consecutive years of unused debt capacity, and 0 otherwise.

Table 1.

Variable definition and source.

Variables	Definitions	Source
SPCR
NCSKW	“Asymmetry of return distribution,” measured using equation (2) and based on equation (1)	Authors' calculation
DUVO	“Down-to-up volatility,” measured using equation (3)	Authors' calculation
Financial flexibility variables
FF1	A dummy variable that takes a value of 1 if the firm demonstrates at least three consecutive years of unused debt capacity, and 0 otherwise (Based on equation (4))	Authors' calculation
FF2	A dummy variable that takes a value of 1 if the firm demonstrates at least three consecutive years of unused debt capacity, and 0 otherwise (Based on equation (5))	Authors' calculation
COVID-19 crisis
COVID	Natural logarithm of stringency index	Our World in Data” website
Control variables
BSIZE	The number of directors on the BOD	Annual report
BIND	The proportion of independent directors on the BOD in relation to the total number of BOD members	Annual report
IOW	The percentage of institutional investor ownership	Annual report
ERNM	Earnings management, as evaluated by Dechow and Dichev [35], quantified using the absolute value of abnormal accruals	Authors' calculation
ROA	The proportion of firm earning after tax on total asset	FiinPro
MB	The ratio of market-to-book value	FiinPro
SIGMA	Standard deviation of weekly stock returns in year	FiinPro
SIZE	Natural logarithm of total assets	FiinPro
LEV	Firm leverage: Ratio of long-term debt on total assets	FiinPro
Other variables to calculate financial flexibility
INLEV	Industry leverage: Total debt to total assets ratio, averaged by industry	Authors' calculation
TANG	Fixed assets to total assets	FiinPro
PROF	Profit after tax to total assets	FiinPro
INF	Annual inflation rate	World Bank
CASH	Cash and equivalents to total assets	FiinPro
MATU	The ratio of debt repayable after one year to total debt	FiinPro
DIV	Dividend payout ratio calculated as total cash dividends divided by sales	FiinPro
NDTS	Depreciation to total assets	FiinPro

We also use an extended model of Yung et al. [33] and Marchica and Mura [34] to compute an alternative measure of financial flexibility. This augmented model is as follows:

$${LEV}_{it}=\alpha {LEV}_{it-1}+\beta 1{INLEV}_{it-1}+\beta 2{MTB}_{it-1}+\beta 3{SIZE}_{it-1}+\beta 4{TANG}_{it-1}+\beta 5{PROF}_{it-1}+\beta 6{\mathit{INF}}_{t-1}+\beta 7{CASH}_{it-1}+\beta 8{MATU}_{it-1}+\beta 9{\mathit{DIV}}_{it-1}+\beta 10{TAX}_{it-1}+{\beta 11{NDTS}_{it-1}+u}_{it}$$ (5)

This alternative model is abbreviated as FF2. In equations (4), (5), conventional indicators of financing constraints, such as dividends and firm size, are incorporated as control variables. As a result, the leverage models outlined in equations (4), (5) have the ability to differentiate companies that genuinely possess financial flexibility from those that have low debt levels due to inherent financial limitations or constraints.

3.2.3. COVID-19 crisis measures

We used the stringency index (SINDEX) to measure the severity of the COVID-19 crisis. SINDEX is calculated using the Oxford Coronavirus Government Response Tracker project. It is a combined measure of nine response metrics related to the coronavirus. These metrics include the closure of schools and workplaces, cancellation of public events, limitations on public gatherings, shutdown of public transportation, stay-at-home orders, public information campaigns, restrictions on internal movements, and controls on international travel. The index is determined by taking the average score of these nine metrics, with each metric assigned a value between 0 and 100. We applied the natural logarithm to scale this variable.

3.2.4. Other control variables

First, we used board independence (BIN) and board size (BSI) to control internal corporate governance, because BOD is considered the “apex” of internal corporate governance [[36], [37], [38]]. Second, we controlled for institutional investor ownership (IOW), as suggested by Xiang et al. [39]. Third, based on Dechow and Dichev [35], we controlled for earnings management (ERM), since ERM is considered to increase SPCR [1]. Fourth, we controlled for the firm's growth opportunities and firm profitability by using return-on-assets ratio (ROA) and market-to-book value (MB), because growth opportunities and high profitability can increase future SPCR [1]. Fifth, we included the standard deviation of weekly stock returns in year t (SIGMA) to control for the stock return volatility. Finally, we controlled for other firms' characteristics that can affect SPCR as found in the literature [21,[40], [41], [42]], including firm leverage (LEV) and firm size (SIZE), where LEV is the long-term debt-to-total asset ratio and SIZE is the natural logarithm of total assets. All variables have been summarized and explained in Table 1.

3.3. Models and estimation methods

3.3.1. Models

First, to test the hypothesis of relationship between financial flexibility and future stock price crash risk, we estimated the following model:

$${SPCR}_{it}={\alpha }_0+{\alpha }_1{FF}_{it-1}+{\alpha }_2{COVID}_{it}+{\alpha }_j\sum _{j=3}^n{CONT}_{jt-1}+{\epsilon }_{it}$$ (6)

Second, to test the role of financial flexibility in preventing crash risk during the COVID-19 crisis period, we estimated the following model:

$${SPCR}_{it}={\alpha }_0+{\alpha }_1{FF}_{it-1}+{\alpha }_2{FF}_{it-1}*{COVID}_{it}+{\alpha }_j\sum _{j=3}^n{CONT}_{jt-1}+{\epsilon }_{it}$$ (7)

where SPCR is the stock price crash risk measure by NCSKW and DUVO, FF is financial flexibility, whose measures are discussed above, COVID is the COVID-19 crisis variable, CONT is the vector of control variables, and ε is the error term. All variables are defined in Table 1. The selected variables are based on previous studies [1,21,30,40].

3.3.2. Estimation method

In this study, we implemented firm fixed-effect regressions to estimate equations (6), (7), as suggested by prior studies [12,31,43,[44], [45]]. Furthermore, we performed the Hausman test and the result showed that fixed-effect method is more appropriate than random effect for our models. Based on previous studies relating to stock price crash risk, cross-section independence and homogenous slope assumption are acceptable for panel data with firms in a country. As a robustness test, we applied the system generalized method of moments (SGMM) and the quantile regression method as robustness tests to treat the potential endogeneity problem and heterogeneity problem, as suggested in previous studies [1,[46], [47], [48], [49]]. To assess the validity of the model's specifications, we employed the Hansen J test to examine the over-identification of restrictions. This test evaluates the absence of correlation between the instruments and the error term. Additionally, the AR2 statistics are used to gauge the presence of second-degree serial correlations. We also applied quantile regression method as another robustness test to treat the heterogeneous effect of independent variables.

4. Research results

4.1. Descriptive statistic and correlation matrix

The information provided in Table 2 presents the statistical characteristics of all variables. Our dataset comprised a total of 4557 observations of firm-year data before applying a one-year lag. The average values for the variables NCSKW and DUVO are 0.168 and 0.053, respectively. These values are similar to those found in comparable studies performed in Vietnam, such as Dang and Nguyen [1] and Nguyen and Dang [13]. However, they are higher than the values observed in the cross-country sample by Ben-Nasr and Ghouma [31], suggesting that the SPCR in developed countries might be lower than developing countries such as Vietnam. Furthermore, the variables FF1 and FF2 have relatively low means, of 0.412 and 0.408, respectively, indicating that not many listed firms in Vietnam adopt the financial flexibility strategy.

Table 2.

Descriptive statistics.

Variables	No of Obs	Mean	Std.Dev	Min	Max
NCSKW	4557	0.168	1.211	−2.983	3.685
DUVO	4557	0.053	1.553	−3.271	8.312
FF1	4557	0.412	1.325	0.000	1.000
FF2	4557	0.408	0.925	0.000	1.000
COVID	4557	0.609	3.312	0.000	4.062
BSIZE	4557	9.758	3.251	3.000	21.000
BIND	4557	0.411	1.357	0.000	1.000
IOW	4557	0.218	1.028	0.000	0.628
ERNM	4557	0.051	0.280	0.003	0.172
ROA	4557	0.013	0.203	−0.466	0.135
MTB	4557	0.769	0.581	0.221	1.731
SIGMA	4557	2.123	1.118	0.257	4.258
SIZE	4557	7.758	1.358	3.589	11.528
LEV	4557	0.483	0.825	0.001	0.954

Note: The information in this table presents the statistical characteristics of all the variables. The “No of Obs” column indicates the number of observations in the complete data before applying a one-year lag. The “Mean” column represents the average value of each variable. “Std.Dev” refers to the standard deviation, while “Min” and “Max” represent the minimum and maximum values, respectively. For detailed definitions of the variables, please refer to Table 1.

Table 3 presents a correlation matrix for all variables. The correlation coefficients between FF1 and FF2 are negative and statistically significant with both NCSKW and DUVO. On the other hand, COVID shows a significant positive correlation with DUVO. These findings suggest that there is a relationship between the COVID-19 crisis, financial flexibility, and SPCR. However, it is important to note that pair-wise correlation measures alone may not be reliable indicators of the complex relationships among multiple variables that influence SPCR. Therefore, we conducted further tests using a multiple regression framework to examine our hypotheses. Additionally, the highest correlation value in Table 3, which is 0.582 between NCSKW and DUVO, indicates that our models are unlikely to encounter the issue of multicollinearity.

Table 3.

Correlation matrix.

	NCSKW	DUVO	FF1	FF2	COVID	BSIZE	BIND	IOW	ERNM	ROA	MTB	SIGMA	SIZE	LEV
NCSKW	1.000
DUVO	0.582***	1.000
FF1	−0.347***	−0.251***	1.000
FF2	−0.227**	−0.243*	0.218	1.000
COVID	0.152**	0.178***	−0.285**	0.135	1.000
BSIZE	−0.123	0.164	0.084***	0.313**	0.121	1.000
BIND	0.057**	−0.103***	0.161*	0.223***	0.148***	0.228*	1.000
IOW	−0.159	0.258	−0.115	−0.219*	−0.228**	0.272	0.129	1.000
ERNM	0.289***	0.267***	0.019***	0.023	0.291*	0.352	−0.132***	0.351***	1.000
ROA	0.124	0.229	−0.019	−0.263***	−0.031	−0.215***	0.162**	−0.097***	0.231**	1.000
MTB	0.171	−0.069**	0.241***	0.158	−0.241***	−0.243**	−0.218	0.391	0.159	0.226	1.000
SIGMA	0.317***	0.229**	0.145*	0.334*	0.228	0.162*	0.319	−0.263***	0.126	−0.291**	0.231	1.000
SIZE	−0.113	0.087***	−0.210	0.029***	0.048***	0.229	0.085***	−0.337	0.192***	0.253	−0.238*	0.179	1.000
LEV	0.216	0.472	−0.129***	−0.142***	−0.227	0.253	0.187	0.143	−0.269	0.065*	−0.384	0.159**	0.128	1.000

Note: This table reports the Pearson's correlation matrix. Statistical significance is indicated as ***, **, and * at the 1 %, 5 %, and 10 % levels, respectively.

4.2. Main results

4.2.1. Financial flexibility and SPCR

In Table 4, we present the results for equation (6) estimation by applying the fixed-effect estimation method. Firstly, the results show that both coefficients on FF1 and FF2 are negative and statistically significant with SPCR variables in all regressions, indicating that the financial flexibility strategy can significantly reduce future SPCR. This result strongly supports hypothesis H1, as per our expectations and indicates that financial flexibility can help firms reduce the agency problem and prevent managers from withholding bad news. Thus, this result is consistent with agency theory and empirical evidence from previous studies. Furthermore, this supports previous studies which state that reducing “unexpected” cash flow shortfalls can reduce bad news hoarding and thus reduce future crash risk [[19], [20], [21]]. Our research also provides an important implication for Vietnamese listed firms that do not prioritize pursuing a financial flexibility strategy, indicating that more attention should be paid to pursuing a financial flexibility strategy. In addition, domestic and foreign investors can consider the financial flexibility of Vietnamese listed companies as a sign to assess the possibility of a stock crash and then adjust their investment portfolio appropriately.

Table 4.

Estimation results for financial flexibility and SPCR.

	NCSKW				DUVO
	Co-eff	t-stat	Co-eff	t-stat	Co-eff	t-stat	Co-eff	t-stat
	(1)		(2)		(3)		(4)
FF1	−0.05***	−2.39			−0.02***	−2.93
FF2			−0.03***	−3.25			−0.01*	−1.87
COVID	0.19***	3.91	0.09****	2.94	0.01***	3.11	0.00**	2.19
BSIZE	0.12	1.66	0.07	1.28	0.02*	1.76	−0.01	−1.56
BIND	−0.02***	−2.31	−0.03**	−1.98	−0.11	−1.55	−0.00*	−1.71
IOW	0.13	1.51	0.02	1.01	−0.04*	−1.88	−0.00*	−1.79
ERNM	0.04	1.33	0.07	1.51	0.02	1.28	0.02**	1.92
ROA	−0.01***	−2.62	−0.01**	−1.89	−0.01*	−1.82	−0.01	−1.14
MTB	0.03	1.12	0.04	1.23	−0.16	−0.49	0.19	0.63
SIGMA	0.05	1.61	0.07*	1.75	0.17***	2.48	0.04*	1.75
LEV	0.12	1.35	−0.01	−0.08	0.12	0.89	0.08	1.02
SIZE	−0.14	−1.57	−0.29	−1.13	0.08**	1.87	0.43***	3.24
Const	1.62***	2.87	1.21***	2.43	1.53**	1.98	1.59***	2.93
Year dummy	yes		yes		yes		yes
Industry dummy	yes		yes		yes		yes
R-squared	0.146		0.139		0.189		0.191
Prob F-statistics	0.00		0.00		0.00		0.00
Obs	3918		3918		3918		3918

Note: The table showcases the estimation outcomes of equation (6) through the utilization of the fixed effect estimation technique. Regressions 1–2 and 3–4 display the estimation results of Eq. (6) using NCSKW and DUVO as dependent variables, respectively. The variables in the models have a one-year lag. The definitions of the variables can be found in Table 1. Statistical significance is indicated as***, **, and * at the 1 %, 5 %, and 10 % levels, respectively.

Second, the coefficients on COVID are positive and statistically significant with both NCSKW and DUVO in all regressions, indicating that COVID-19 crisis significantly increases SPCR. This result extends the literature on the negative impact of the COVID-19 crisis on the stock market and the economic environment [2,14,50]. This result is also consistent with the fact that Vietnam's stock market was heavily affected by the COVID-19 pandemic during 2019–21. Thus, these results confirm the negative impact of economic shocks that stock markets in emerging countries like Vietnam need to prepare for. Our results are consistent with some studies performed in other markets. For example, Huang et al. [51] found that COVID-19 crisis increase SPCR of energy firms in Chinese stock market. Similarly, Shu et al. [52] found that COVID-19 is the reason for 2020 US stock market crash. Overall, previous studies show that the COVID-19 crisis increases SPCR in both developing and developed countries. Our research result provides an implication that policymakers and investors need to prepare for future epidemic crises.

Table 4 also provides important results regarding control variables. First, we found that an independent board can reduce future SPCR, because the coefficients on BIND are negative and statistically significant in most regressions. This result is consistent with the literature which indicates that a good corporate governance can reduce future SPCR [1]. Second, we found that stock return volatility positively affects future SPCR. Therefore, stock return volatility can be seen as a signal for investors to assess the level of SPCR. This result is consistent with the results in previous studies [53]. Finally, we provided evidence that firm performance is negatively associated with future SPCR. The coefficients on ROA are negative and statistically significant with both dependent variables in most regressions. Previous studies have agreed that low firm performance can create uncertainty about a company's future prospects and increase perceived risk [12,53].

4.2.2. COVID-19 crisis and the relationship between financial flexibility and SPCR

Table 5 presents the estimation results for Eq. (7) by applying the fixed effect method. First, we found that the coefficients on FF1 and FF2 are negative and statistically significant with all SPCR variables in all regressions, implying that hypothesis H1 is strongly supported. Similarly, the coefficients on COVID are positive and significant with all SPCR variables in all regressions. These results are consistent with the results in Table 4. Second, the coefficients on FF1*COVID and FF2*COVID are negative and significant with both NCSKW and DUVO in all regressions, indicating that financial flexibility strategy can help firms reduce the impact of the COVID-19 crisis on SPCR. Thus, this result strongly supports hypothesis H2 and is consistent with the pecking order theory that companies with strong internal financial resources and a financial flexibility strategy are better positioned to navigate economic downturns and maintain stock prices.

Table 5.

Financial flexibility and COVID-19 crisis-SPCR relationship.

	NCSKW				DUVO
	Co-eff	t-stat	Co-eff	t-stat	Co-eff	t-stat	Co-eff	t-stat
	(1)		(2)		(3)		(4)
FF1	−0.03**	−2.13			−0.00***	−2.87
FF2			−0.07***	−2.41			−0.01**	−2.24
FF1*COVID	−0.01**	−1.93			−0.01***	−3.12
FF2*COVID			−0.02*	−1.87			−0.00***	−2.68
COVID	0.03***	3.15	0.04**	2.11	0.02*	1.78	0.01**	1.93
BSIZE	0.04	1.61	0.15	1.43	−0.00	−0.69	0.01*	1.77
BIND	−0.05*	−1.82	−0.08	−1.57	−0.01**	−2.31	−0.02	−1.05
IOW	−0.09	−1.08	−0.15**	−1.91	0.02	0.25	0.01	1.38
ERNM	0.03*	1.82	−0.02	−1.05	0.01*	1.68	0.00*	1.84
ROA	−0.15*	−1.75	−0.08**	−1.99	−0.07	−0.85	−0.05	−0.69
MTB	0.03	0.39	0.05	0.87	0.00	1.21	0.01	0.98
SIGMA	0.15*	1.81	0.11**	1.97	0.02*	1.71	0.01**	1.93
LEV	0.06	1.53	−0.02	−0.42	0.00	1.15	0.02	1.31
SIZE	0.08	0.91	0.05	1.24	0.13	1.08	0.12	1.05
Const	2.23***	3.91	2.42***	2.71	1.19***	2.63	1.62**	2.03
Year dummy	yes		yes		yes		yes
Industry dummy	yes		yes		yes		yes
R-squared	0.185		0.178		0.237		0.214
Prob F-statistics	0.00		0.00		0.00		0.00
Obs	3918		3918		3918		3918

Note: The information in this table displays the results of estimating equation (7). Two measures of financial flexibility were utilized, and the fixed effect estimation method was employed. The results of estimating equation (7) are shown separately for regressions 1–2 and 3–4 using NCSKW and DUVO as dependent variables, respectively. All control variables have a lag of one year. See Table 1 for definitions of the variables. Statistical significance is indicated as ***, **, and * at the 1 %, 5 %, and 10 % levels, respectively.

Our findings provide strong implications for firms that financial flexibility plays an important role in controlling SPCR in emerging markets. It can not only reduce future SPCR directly but also reduce the impact of economic shock on SPCR. Investors can evaluate the SPRC of a company's stock by analyzing that company's financial flexibility strategy. Shareholders and managers must pay more attention to the company's financial flexibility strategy to avoid large stock price fluctuations. Our findings extend the literature about the role of financial flexibility, such as reducing agency cost [23] and reducing corporate liquidity [54], thus providing evidence that financial flexibility can reduce future SPCR.

Regarding control variables, the sign of BIND, ROA, and SIGMA in Table 5 remain unchanged with Table 4. However, in Table 5, we found that the coefficients on ERNM are positive and statistically significant in regressions 1, 3, and 4, indicating that earnings management can increase future SPCR. This result supports previous studies which state that earnings management is an indication of the manager holding on to bad news [55,56]. Overall, the results in Table 5 strongly support hypotheses H1 and H2.

4.3. Robustness test

In this study, we applied some robustness tests in this study. First, we used the SGMM method for equations (6), (7) to treat the potential dynamic endogeneity problem. Second, to test the heterogeneity relationship between financial flexibility, the COVID-19 crisis, and SPCR, we used the quantile regression method for equation (7). By using quantile regression, we were able to examine the heterogeneity impact of financial flexibility on future SPCR as well as the heterogeneity impact of the COVID-19 crisis on SPCR. This testing is important because the role of financial flexibility may be different among Vietnamese listed firms (firms with low and high level of SPCR).

Table 6 presents the estimation results for equation (6) by applying the SGMM method. In this table, the signs of FF1 and FF2 coefficients are negative and significant. Therefore, hypothesis H1 is strongly supported. Similarly, the coefficients on COVID are similar to the results in Table 4, indicating a positive impact of the COVID-19 crisis on SPCR. The other coefficients do not differ considerably from the results in Table 4. The effectiveness of the GMM estimator in Table 6 is confirmed through the assessment of instrument validity using the Hansen J test as well as the examination of the second-order auto-correlation of the error terms using the AR (2) test developed by Arellano and Bond [57]. The p-value of the AR (2) and Hansen J test are higher than 10 % in all regressions.

Table 6.

SGMM estimation results for financial flexibility and SPCR.

	NCSKW				DUVO
	Co-eff	t-stat	Co-eff	t-stat	Co-eff	t-stat	Co-eff	t-stat
	(1)		(2)		(3)		(4)
FF1	−0.12***	−3.31			−0.01**	−2.16
FF2			−0.09***	−3.61			−0.00***	−2.81
COVID	0.19**	2.25	0.11*	1.74	0.01***	2.65	0.01***	3.12
BSIZE	0.02*	1.86	0.06	1.68	0.00	1.55	0.01*	1.75
BIND	−0.05*	−1.81	−0.08*	−1.82	−0.03***	−2.55	−0.00	−1.22
IOW	0.17	1.28	0.12*	1.71	0.01**	−1.97	0.00*	1.82
ERNM	0.03	1.68	0.05	1.26	−0.00	−0.83	−0.01	−1.52
ROA	−0.03**	−1.92	−0.04	−1.69	−0.01*	−1.82	−0.00*	−1.74
MTB	0.02	0.82	−0.04	−1.29	0.00	1.15	0.08	0.93
SIGMA	0.05*	1.71	0.07*	1.74	0.05	1.43	0.02	1.36
LEV	0.11**	1.95	−0.08*	−1.78	−0.02**	−1.89	0.05*	1.72
SIZE	−0.04	−1.67	−0.29**	−1.83	−0.03	−1.15	0.06*	1.74
Const	2.61***	3.89	2.06***	3.41	1.91***	2.95	1.27**	2.08
Hansen J (p-value)	0.19		0.26		0.21		0.28
AR (2) (p-value)	0.28		0.31		0.19		0.22
No of instruments	103		98		102		101
Obs	3918		3918		3918		3918

Note: The information presented in this table showcases the results of estimating equation (6) using the SGMM estimation method. The estimation results are separately displayed for regressions 1–2 and 3–4, with NCSKW and DUVO as the dependent variables, respectively. All control variables have a lag of one year. See Table 1 for definitions of the variables. Statistical significance is indicated as ***, **, and * at the 1 %, 5 %, and 10 % levels, respectively.

Table 7 shows the estimation results for equation (7) using the SGMM method. First, we found that the coefficients on FF1 and FF2 are negative and significant with SPCR variables in all regressions, indicating that this result supports hypothesis H1. Second, the coefficients on the sign of COVID remained negative and statistically significant in all regressions, and the coefficients on FF1*COVID and on FF2*COVID remained negative and significant with both dependent variables in all regressions. This result supports the idea that financial flexibility can reduce the positive impact of the COVID-19 crisis on SPCR. In other words, the results in Table 7 are consistent with the results in Table 5, and hypothesis H2 is supported. The p-value of AR (2) and the Hansen J test are higher than 10 %, indicating that the instruments are valid and the estimation results are reliable. Finally, the reliability of the Hansen J statistics is enhanced since the number of instruments employed in the models in Table 6, Table 7 is lower than the panel.

Table 7.

SGMM estimation results for financial flexibility and COVID-19 crisis-SPCR relationship.

	NCSKW				DUVO
	Co-eff	t-stat	Co-eff	t-stat	Co-eff	t-stat	Co-eff	t-stat
	(1)		(2)		(3)		(4)
FF1	−0.02***	−3.14			−0.02**	−2.16
FF2			−0.04**	−2.05			−0.00*	−1.80
FF1*COVID	−0.03*	−1.81			−0.00*	−1.82
FF2*COVID			−0.02**	−1.99			−0.01***	−2.47
COVID	0.04**	2.11	0.06***	2.53	0.01**	1.91	0.01*	1.83
BSIZE	0.15	1.50	0.23	1.48	0.01	1.49	0.02	1.64
BIND	−0.08	−1.26	0.13	0.59	−0.00***	−2.63	−0.02*	−1.76
IOW	−0.04*	−1.78	−0.09*	−1.76	0.01	1.22	−0.00**	−1.88
ERNM	0.06***	2.84	0.02	1.35	0.00**	1.93	0.01	1.64
ROA	−0.09	−1.06	−0.13*	−1.79	−0.04	−1.15	−0.05*	−1.76
MTB	0.08	1.08	−0.03	−1.47	0.00	0.75	0.01	0.98
SIGMA	0.19**	1.93	0.15**	1.92	0.06	1.29	0.00**	1.97
LEV	0.05	1.42	0.05	1.42	0.02	1.24	0.00	1.14
SIZE	0.07	1.01	−0.05	−0.84	0.07*	1.78	0.19	1.55
Const	2.33**	2.17	2.87*	1.71	1.29**	2.18	1.78*	1.73
Hansen J (p-value)	0.22		0.31		0.25		0.25
AR (2) (p-value)	0.42		0.22		0.16		0.27
No of instruments	94		105		97		102
Obs	3918		3918		3918		3918

Note: The estimation results of equation (7) are presented in this table, utilizing two proxies of financial flexibility and applying the SGMM estimation method. Regressions 1–2 and 3–4 display the estimation results of equation (7), with NCSKW and DUVO serving as the dependent variables, respectively. All control variables have a lag of one year. See Table 1 for definitions of the variables. Statistical significance is indicated as ***, **, and * at the 1 %, 5 %, and 10 % levels, respectively.

In general, the findings from the SGMM estimation presented in Table 6, Table 7 support the notion that, even when accounting for potential unobserved heterogeneity, simultaneity, and dynamic endogeneity, there exists a relationship between financial flexibility and SPCR. This relationship aligns with the anticipated expectations.

As another robustness test, we applied the quantile regression method for equation (7), and the results are reported in Table 8. First, the coefficients on FF1 are negative and statistically significant with NCSKW in regressions 1–4 and with DUVO in regressions 10–12. Similarly, the coefficients on FF2 are negative and significant with the NCSKW variable in regressions 6–8 and with DUVO in regressions 14–16. The signs of these coefficients remain unchanged through all quantiles, indicating that the negative impact of financial flexibility on future SPCR remains unchanged from low to high level of SPCR. This continues to support hypothesis H1. However, the value of these coefficients increases from low to high quantiles, indicating that the impact of financial flexibility on future SPCR is strengthened in firms with a high level of crash risk. In other words, financial flexibility plays a more important role in preventing crash risk in firms with a high level of crash risk. This may be because firms with a high level of crash risk to be more vulnerable to negative shocks. Greater financial flexibility becomes crucial for such firms to effectively manage and mitigate the potential impact of adverse events [25]. This can include accessing additional funding sources, adjusting capital structure, or implementing risk management strategies. Thus, the impact of financial flexibility on reducing crash risk becomes more pronounced in high-risk firms. Although there is evidence of the heterogeneity impact of financial flexibility on future SPCR, hypothesis H1 is well supported.

Table 8.

Quantile regression results for financial flexibility and SPCR during the COVID-19 crisis.

	NCSKW								DUVO
	Q25	Q50	Q75	Q90	Q25	Q50	Q75	Q90	Q25	Q50	Q75	Q90	Q25	Q50	Q75	Q90
	Co-eff (t-stat)	Co-eff (t-stat)	Co-eff (t-stat)	Co-eff (t-stat)	Co-eff (t-stat)	Co-eff (t-stat)	Co-eff (t-stat)	Co-eff (t-stat)	Co-eff (t-stat)	Co-eff (t-stat)	Co-eff (t-stat)	Co-eff (t-stat)	Co-eff (t-stat)	Co-eff (t-stat)	Co-eff (t-stat)	Co-eff (t-stat)
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)	(16)
FF1	−0.03*	−0.06**	−0.08**	−0.18***					−0.01	−0.01*	−0.02**	−0.03***
	(-1.73)	(-1.85)	(-1.88)	(-2.43)					(-1.49)	(-1.76)	(-1.79)	(-2.28)
FF2					−0.03	−0.07*	−0.09**	−0.16**					−0.01	−0.01*	−0.03**	−0.05***
					(-1.56)	(-1.75)	(-1.93)	(-2.11)					(-1.05)	(-1.71)	(-1.87)	(-2.29)
FF1*COVID	−0.02**	−0.01*	−0.02*	−0.03					−0.00**	−0.01*	−0.00	−0.00
	(-1.90)	(-1.82)	(-1.71)	(-155)					(-1.97)	(-1.78)	(-1.65)	(-1.69)
FF2*COVID					−0.04	−0.03*	−0.03*	−0.04**					−0.00	−0.01*	−0.01*	−0.01**
					(−1.61)	(-1.74)	(-1.79)	(-1.86)					(-1.61)	(-1.74)	(-1.79)	(-1.96)
COVID	0.07***	0.06**	0.06**	0.07*	0.05**	0.06**	0.06*	0.07	0.00***	0.01***	0.01**	0.00*	0.01***	0.00***	0.01**	0.01*
	(2.67)	(2.06)	(1.98)	(1.84)	(1.97)	(1.93)	(1.82)	(1.61)	(3.46)	(2.98)	(2.17)	(1.88)	(2.79)	(2.54)	(2.01)	(1.79)
BSIZE	−0.21	−0.12	−0.06	0.01*	−0.00	−0.00	−0.01	0.00	−0.17	0.05	0.07	0.15	0.00	0.01	0.02	−0.00
	(-0.66)	(-1.18)	(-0.56)	(1.71)	(-1.12)	(-1.21)	(-1.52)	(0.98)	(-0.13)	(0.89)	(1.23)	(1.31)	(1.41)	(1.02)	(1.05)	(-1.32)
BIND	−0.21*	−0.15**	−0.12**	−0.08***	−0.01***	−0.00*	−0.02	0.03	−0.33	−0.13*	−0.08**	−0.06**	0.11	−0.02	−0.01*	−0.00***
	(-1.71)	(-1.85)	(-1.91)	(-2.41)	(-2.22)	(-1.74)	(-1.23)	(0.87)	(-1.03)	(-1.77)	(-1.88)	(-1.91)	(0.99)	(-0.87)	(-1.71)	(-3.14)
IOW	0.12	0.09	0.11	0.14	−0.12	0.16	0.11	0.21*	0.01	0.01	0.02	−0.00	−0.01	−0.00*	0.00	0.01
	(0.89)	(1.03)	(1.21)	(1.13)	(-0.67)	(0.91)	(1.35)	(1.72)	(0.55)	(0.89)	(1.28)	(-0.86)	(-1.15)	(-1.77)	(0.65)	(0.66)
ERNM	0.04*	0.06	0.05	0.08	0.11**	0.18*	0.15	0.13	0.02**	0.03*	0.01*	0.01	0.02***	0.02**	0.01**	0.00
	(1.70)	(1.61)	(1.42)	(1.18)	(2.05)	(1.84)	(1.51)	(1.48)	(2.15)	(1.83)	(1.77)	(1.68)	(3.48)	(2.12)	(1.92)	(1.53)
ROA	−0.11***	−0.93	−0.17	0.18	−0.18**	−0.21	0.22	0.53	−0.01	−0.03	−0.02*	0.08	−0.02**	−0.03*	−0.01	−0.02
	(-3.18)	(-1.52)	(-1.49)	(0.95)	(-2.12)	(-1.54)	(1.23)	(1.11)	(-1.23)	(-1.38)	(-1.72)	(1.14)	(-2.23)	(-1.87)	(-1.33)	(-1.29)
MTB	0.19	0.16	0.18*	0.17*	0.12	0.15	0.11*	0.16*	0.03	0.07	−0.08	−0.04	−0.01	−0.01	0.05	0.04
	(1.16)	(1.21)	(1.74)	(1.79)	(1.15)	(1.21)	(1.73)	(1.76)	(0.94)	(1.19)	(-0.92)	(-1.13)	(-1.29)	(-1.33)	(1.08)	(1.16)
SIGMA	0.08	0.22	0.15	0.14	−0.18	−0.23	0.12	0.18	−0.05	−0.05	0.02	0.04	0.06	0.02	0.02	0.03
	(1.42)	(1.55)	(1.46)	(1.04)	(-1.23)	(-1.38)	(0.97)	(1.12)	(-1.19)	(-1.37)	(0.85)	(0.72)	(1.49)	(1.52)	(1.61)	(1.58)
LEV	0.11*	0.05	0.08	0.05	−0.09	−0.09	0.16*	0.18*	0.05***	0.07*	0.01**	0.04	0.01***	0.02*	0.02	−0.01
	(1.74)	(1.62)	(1.60)	(1.55)	(-1.21)	(-0.77)	(1.83)	(1.85)	(2.43)	(1.81)	(1.97)	(1.59)	(2.33)	(1.85)	(1.54)	(−1.53)
SIZE	0.09*	0.05**	0.07	0.05	0.09	0.17	0.12	0.15	0.09***	0.05***	0.06**	0.07	0.03*	0.02	0.02	0.03
	(1.77)	(1.94)	(1.34)	(1.09)	(1.22)	(1.18)	(1.37)	(1.23)	(3.16)	(2.92)	(2.17)	(1.67)	(1.81)	(1.66)	(1.63)	(1.51)
Obs	3918	3918	3918	3918	3918	3918	3918	3918	3918	3918	3918	3918	3918	3918	3918	3918

Note. The estimation results of equation (7) are reported in this table using the quantile regression method. Regressions 1–8 and 9–16 display the estimation results of equation (6), with NCSKW and DUVO as the dependent variables, respectively. All control variables have a lag of one year. See Table 1 for definitions of the variables. Q25-Q90 indicate the quantile 25–90, respectively. Statistical significance is indicated as ***, **, and * at the 1 %, 5 %, and 10 % levels, respectively.

Furthermore, the coefficients on FF1*COVID and FF2*COVID are negative in all regressions and significant in most regressions. The values of these coefficients do not showcase much difference among quantiles, indicating that there is no evidence of the heterogeneity impact of financial flexibility on the relationship between the COVID-19 crisis and SPCR. Although not all coefficients of FF1*COVD and FF2*COVID are statistically significant, the results in Table 8 still support hypothesis H2, i.e., that financial flexibility can reduce the positive effect of the COVID-19 crisis on SPCR. Finally, the coefficients on COVID are negative and statistically significant in all regressions, but these values do not differ much among quantiles. Therefore, the effect of the COVID-19 crisis on SPCR may not be different among Vietnamese listed firms. Overall, after examining the heterogeneity impact of financial flexibility on future SPCR as well as on the relationship between the COVID-19 crisis and SPCR, the results continue to support hypotheses H1 and H2.

5. Conclusion

We conducted an analysis using a dataset comprising 645 non-financial listed firms in the Vietnamese stock market. The data covers the period from 2011 to 2021, and our analysis focused on examining the influence of financial flexibility on future SPCR as well as the role of financial flexibility in mitigating the impact of the COVID-19 crisis on SPCR. Our findings revealed a significant negative association between financial flexibility and SPCR in the Vietnamese stock market. Firms with higher levels of financial flexibility experience lower probabilities of stock price crashes, indicating the crucial role of financial flexibility in safeguarding firm value and investor interests. Furthermore, our results demonstrated that financial flexibility diminishes the positive effect of COVID-19 on SPCR, suggesting that firms with greater financial flexibility are better positioned to withstand the challenges imposed by the pandemic. Overall, this research contributes to the literature by providing empirical evidence on the role of financial flexibility in reducing stock price crash risk and moderating the impact of COVID-19 on stock price crash risk in the context of Vietnam. It offers valuable insights into the mechanisms through which firms can enhance their resilience and protect shareholder value during periods of heightened uncertainty and crisis.

The implications of this study are manifold. First, it highlights the importance of financial flexibility as a risk management tool, emphasizing the need for firms to maintain a flexible financial structure to mitigate SPCR. Second, it underscores the significance of financial preparedness in times of crisis, as firms with higher financial flexibility are more resilient to exogenous shocks such as the COVID-19 pandemic. Policymakers, investors, and market participants can benefit from these insights by integrating financial flexibility considerations into the risk-assessment and decision-making processes.

Our study is limited in that it was only conducted within one emerging country due to data limitations. In addition, due to the lack of data after the COVID-19 pandemic, the role of financial flexibility in SPCR during the recovery period has not been analyzed. Therefore, further studies can approach using multinational data or evaluate the role of financial flexibility in developed countries. In addition, future studies can evaluate the role of financial flexibility in the post-pandemic recovery period to have a more general view.

Data availability statement

Data will be made available on request.

Funding

This research is funded by the University of Economics Ho Chi Minh City (UEH).

CRediT authorship contribution statement

Quang Khai Nguyen: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Conceptualization. Van Cuong Dang: Data curation.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Distribution of data by industries

No	Industry name	No of firms	%	No of Observations	%
1	Science and Technology	28	4.34 %	172	3.77 %
2	Mining and Petroleum	15	2.33 %	420	9.22 %
3	Industry	236	36.59 %	1887	41.41 %
4	Agriculture forestry seafood	11	1.71 %	58	1.27 %
5	Trading, service	122	18.91 %	615	13.50 %
6	Construction	94	14.57 %	626	13.74 %
7	Health	17	2.64 %	127	2.79 %
8	Warehousing and transportation	58	8.99 %	321	7.04 %
9	Real estate	36	5.58 %	186	4.08 %
10	Information, communication	28	4.34 %	145	3.18 %
	Total	645	100.00 %	4557	100.00 %

Note: The industry is classified by “Industry Classification Benchmark”.