Does Corporate Policy Risk Affect Stock Liquidity? Panel Data Evidence from Listed Companies in a Major Emerging Market

Abstract

This study examines the impact of firms’ overall corporate policy risk on stock liquidity. This study constructs a novel overall corporate policies risk index (PRI) for firms by capturing risk embedded in managers’ different policy decisions, such as investment, financing, diversification, and cash management, by weighting each policy risk through the regression decomposition method. Using a large sample of 466 India-listed firms from the financial year 2003–2004 to 2022–2023, this study finds that there is a negative association between PRI and stock liquidity. The study further explores the information environment heterogeneity and finds that the adverse impact of a PRI is a more prominent firm that is hard to value or in a less transparent environment as compared to the transparent firms. Moreover, the adverse impact of PRI on stock liquidity is significantly more pronounced during financial crises, while its effect is less substantial during non-crisis periods. The robustness of these results is confirmed even after addressing endogeneity issues using various techniques, such as propensity score matching (PSM), two-stage least squares instrumental variable approach (2 SLS IV), and the system-generalized method of moments (System GMM).

1. Introduction

Corporate policy risk is defined as the risk embedded in firms’ strategic policy decisions, such as investment, financing, dividend, diversification, and cash management (Çolak & Korkeamäki, 2021). These decisions, shaped by top management, reflect an organization’s risk culture and influence the firm’s performance and market outcomes (Pan et al., 2017). Existing studies have advocated that high-risk corporate policies—such as aggressive investments or changes in capital structures—significantly affect firms’ equity returns and volatility (Cain & McKeon, 2016). For instance, aggressive investment strategies and high leverage amplify equity volatility (Cassell et al., 2012; Jiang & Feng, 2021). Further, excessive diversification increases organizational opacity, which in turn accentuates stock volatility (Amihud & Lev, 1981; Kang et al., 2017). Similarly, cash policies of the proper level of reserves promote financial stability and liquidity, whereas excessive reserves may limit the capacity to take risks or dampen market efficiency (Bates et al., 2009; Bolton et al., 2011; Opler et al., 1999). However, the cumulative risk embedded in a firm’s major corporate policies and its effect on stock liquidity remains underexplored.

This study fills that gap in the research by introducing a comprehensive framework that quantifies aggregate risk within corporate policies known as the policy riskiness index (PRI). The PRI does not hinge on one particular dimension but assimilates all total risk across investment, financial, diversification, and cash management policy decisions towards a holistic picture of a firm’s risk profile and its consequential impact on the liquidity of their stocks. The PRI is constructed using regression decomposition to connect current corporate policy choices to future realized risks, especially idiosyncratic stock volatility calculated through the Fama–French–Carhart model over 36 months. It provides a multidimensional measure of the contribution of each policy toward overall risk by weighing the contribution of each policy.

India’s economic growth and stock market evolution provide a compelling context for examining corporate policy risk and stock liquidity. As the fifth-largest economy, with a GDP growth of 8.2% in FY 2023–20241. Stock market capitalization has also increased from USD 1.2 trillion in 2014 to USD 5.66 trillion in 2024, with a market capitalization to GDP ratio of 140.1 in FY 2023–20242. The Indian stock market, characterized by its order-driven system, relies heavily on investor limit orders, differing from the quote-driven markets in developed countries like the US and UK, where market makers and central banks significantly influence liquidity (Bekaert et al., 2007). Indian firms tend to adopt conservative corporate strategies toward research and development (R&D) investment and leverage due to an unpredictable regulatory environment and evolving governance structures (Sasidharan et al., 2015). These firms often perceive R&D investments and high leverage as riskier, resulting in cautious financial strategies that impact stock liquidity. This contrasts with the aggressive strategies observed in developed markets, where stable regulatory frameworks bolster manager and investor confidence (Agnihotri & Bhattacharya, 2015; Dutta et al., 2019). The interaction between corporate policy risks and stock liquidity in India underscores the critical role of market microstructure, regulatory context, and managerial behavior (Sahu et al., 2024). Understanding these dynamics offers valuable insights into how corporate decisions influence liquidity in emerging markets, enhancing our comprehension of corporate strategy and financial outcomes in such contexts.

Building on this comprehensive framework, the study seeks to answer the question: How does corporate policy risk influence stock liquidity in India? By isolating firm-level policy risks from external factors, the research aims to provide a clearer understanding of how risk is involved in all major corporate strategies, such as investment, financial, diversification, and cash management decisions, impacting stock liquidity.

Using a comprehensive sample of 466 non-financial firms listed on the National Stock Exchange of India (NSE), this study employs a fixed-effects regression model to estimate the relationship between corporate policy risk and stock liquidity. The findings reveal a robust and significant negative relationship between the policy riskiness index (PRI) and stock liquidity. Further, we find that our baseline results are robust after addressing endogeneity issues by employing serval endogeneity tests, including propensity score matching (PSM), a two-stage least squares instrumental approach (2 SLS IV), and two-step system generalized method of moments (System GMM). The analysis shows that increased policy-induced risk, often due to managerial biases or decisions that do not align with market expectations, further increases market uncertainty and discourages long-term investors. Thus, it may reduce stock turnover and increase stock illiquidity. In addition, this chapter also reveals that the negative effect of corporate policy risk on stock liquidity is particularly significant for firms with less transparent information environments, especially those that are small in size, highly volatile, and provide more complex financial reports. Overall, the study argues that there is an imperative necessity for balanced and transparent corporate strategies to minimize risks, protect stock liquidity, and improve market efficiency.

This study makes significant contributions to the fields of corporate finance and market microstructure. First, this study bridges the literature gap by introducing a firm-specific determinant of stock liquidity through a novel, comprehensive measure of corporate policy risk that reflects the multidimensional nature of firm-level policy decisions (Bernile et al., 2018; Cain & McKeon, 2016; Cassell et al., 2012). Second, it explains the vital link between managerial decision-making and stock market behavior, especially stock liquidity, an essential component of market efficiency. Therefore, this study contributes to the existing literature on managerial risk attitudes and market microstructure by providing novel comprehensive corporate risk culture determinants of stock liquidity (Coles et al., 2006; Dang et al., 2022; Ferris et al., 2017; Korkeamäki et al., 2017; Stereńczak & Kubiak, 2022). In practical terms, our findings have implications for managers, investors, and policymakers, as balanced corporate strategies offer a means for mitigating policy-induced risks, stabilizing markets, and building confidence among investors.

The subsequent sections of this study are structured as follows. Section 2 reviews the literature and outlines the formulation of testable hypotheses. Section 3 delineates the criteria for sample selection, the methodology employed in constructing the PRI and stock liquidity measures, and the model specifications. Section 4 reports the findings and discussion of the baseline regression alongside the implications of the information environment and financial crisis. Section 5 presents the findings from various endogeneity testing. The conclusion of the paper is presented in Section 6.

2. Literature Review and Hypotheses Development

2.1. Theoretical Framework

Corporate policy risk is a concept that includes a variety of strategic decisions made by firms, especially by top managers, such as investment, financial, diversification, and cash management policies. Each of these policies is critical to a corporation’s risk profile and its behavior in the stock market, with important implications for stock liquidity.

Several theoretical frameworks and the empirical evidence support the interplay between corporate policy risk (driven by top manager’s decisions) and stock market behavior, especially stock liquidity. Market microstructure theory posits that stock returns inversely relate to liquidity, where lower liquidity necessitates higher returns to compensate investors for trading costs and risks (Amihud, 2002). Corporate policy risk, arising from managerial decisions directly affects liquidity by amplifying uncertainty and informational asymmetry (Bradley et al., 2016; Cain & McKeon, 2016; Serfling, 2014). Furthermore, a recent strand of literature has observed a sudden evaporation of liquidity in the financial markets, especially during economic uncertainty, which is coupled with persistent increases in corporate policy risk (Cui et al., 2021; Wen et al., 2021). Thus, one can intuitively argue that corporate policy risk could be a potential determinant of stock liquidity.

According to the upper echelons theory (Hambrick & Mason, 1984), this is the foundational understanding of how manager characteristics affect corporate strategies and embed risks into policies. The framework assumes that individual traits, for example, age, experience, education, and risk preferences, considerably impact strategic decisions. For example, Cain and McKeon (2016) show that sensation-seeking CEOs, as indicated by personal hobbies like piloting, are likelier to adopt high-risk investment policies, such as mergers and acquisitions (M&A). This tendency increases equity return volatility and may negatively affect stock liquidity. Younger CEOs tend to be more willing to take risks than older CEOs, who may be more conservative because of their experience (Serfling, 2014). In addition, the social capital of CEOs (their network and relational capital) can increase their access to resources and information, influencing their risk-taking behavior (Ferris et al., 2017).

The relationship between corporate policy risk and stock liquidity is further elucidated by the agency theory framework of Jensen and Meckling (1979). The critical point of this theory is that managers may undertake risky projects to align their interests with those of shareholders. However, compensation structures, social capital, and managerial power may also be sources of risk-taking behaviors. Coles et al. (2006) and Gormley et al. (2013) find that performance-linked incentives induce risk-seeking behaviors, which may lead to stock market behaviors. Ferris et al. (2017) also document that CEOs with highly developed social networks and high social capital tend to increase corporate risk through aggressive policies, increasing corporate operational earnings and stock volatility.

Behavioral finance theory takes things a step further by introducing overconfidence and hubris into the analysis. Malmendier and Tate (2005) find that overconfidence drives speculative investments that exacerbate liquidity problems by raising volatility. Similarly, Adam et al. (2015) documented that managerial overconfidence leads to corporate risk management, as managers increase speculative derivatives activities following gains but not following losses and thus exhibit selective self-attribution. Such behavior affects corporate decisions by matching risk management practices to stock market anomalies and speculative dynamics. Li and Tang (2010) extend this argument by demonstrating that hubristic CEOs overestimate their abilities, making destabilizing, high-risk decisions.

2.2. Empirical Literature

Corporate Policies Risk and Stock Liquidity

Investment policies, which are defined as allocating resources to achieve the firm’s growth opportunities, are crucially determinant of corporate risk and stock liquidity. Eberhart et al. (2004) demonstrate that increased research and development (R&D) expenditures are associated with large positive abnormal stock returns and improved long-term operating performance benefiting shareholders. Therefore, strategic investment in R&D disclosure transparency and investor trust can enhance stock liquidity. Nevertheless, high-risk investments (e.g., R&D or mergers and acquisitions) may raise stock volatility and reduce stock liquidity (Chan et al., 2001; Fang et al., 2014). Furthermore, Xiang et al. (2020) demonstrate that R&D expense uncertainty causes higher stock volatility and lower returns due to disruptive costs and earnings management. Empirical evidence underscores this duality: Cassell et al. (2012) show that firms with aggressive investment policies and high leverage amplify financial and investment risks, increasing stock return volatility and eroding liquidity.

Moreover, corporate leaders’ risk preferences and characteristics play a big role in investment decisions. For instance, Ferris et al. (2017) demonstrate how CEOs with high social capital take on riskier investment projects and increase stock volatility but perhaps secure the benefits of stronger investor confidence from their networking advantages. Coles et al. (2006) demonstrate that higher CEO wealth sensitivity to stock volatility, measured as vega—the change in an executive’s wealth for a 1% change in stock-return volatility—drives riskier corporate policies. These include increased R&D investment, higher leverage, and a shift in compensation structures toward greater vega and reduced pay-performance sensitivity (delta), ultimately amplifying stock-return volatility.

Leverage and debt management decisions constitute the financial policy of a corporation, which significantly affects a corporation’s overall risk and market behaviors. High leverage amplifies financial risk, reducing stock return volatility and deterring liquidity providers because uncertainty is increased (Cassell et al., 2012; Serfling, 2014). For instance, firms led by CEOs with private pilot licenses (a proxy for personal risk-taking) have riskier capital structures, among the highest in firms, in terms of leverage and acquisition activity, which increase firm risk and stock return volatility (Cain & McKeon, 2016). Cronqvist et al. (2012) show that CEOs’ personal leverage preferences influence corporate capital structure decisions. The risk associated with such behaviors, particularly in weak governance firms, has significant implications for stock market behavior and financial outcomes. In addition, firms with conservative financial policies and high cash reserves are found to have superior liquidity because they are seen to have lower default risk (Bradley et al., 2016).

Business diversification, measured by the number of segments or industries in which a firm operates, is both a risk management tool and a potential source of liquidity challenges. Theoretically, diversification reduces bankruptcy risk and liquidity by attracting risk-averse investors, as argued by Amihud and Lev (1981). However, excessive diversification, particularly under overconfident CEOs, increases organizational opacity and information asymmetry, discouraging trading and reducing liquidity (Berger & Ofek, 1995; Cain & McKeon, 2016; Denis et al., 1997). While operational diversification reduces firm-specific risks (Jiang & Feng, 2021), its complexity often dilutes management focus, amplifies uncertainty, and impedes liquidity (Stulz, 1994). Moreover, Khanna and Palepu (2000) find that emerging markets’ diversified business groups overcome their agency and information problems better, resulting in better performance under particular governance conditions.

Cash policy, or the management of a firm’s cash reserves and liquidity, plays an important role in corporate risk and stock liquidity. Efficient cash management increases the capacity of a firm to fulfill operational and investment requirements, especially when economic times are challenging, which decreases the perceived risk and thereby attracts investors (Bates et al., 2009; Xu et al., 2023). Theoretically, Opler et al. (1999) stated that firms with effective cash management practices maintain optimal cash reserves, ensuring smoother operations and reducing financial distress risks. This financial stability reduces the risk premium demanded by investors, enhancing their willingness to trade the stock (Acharya et al., 2007). However, excessive cash reserves may lead to managerial complacency and suboptimal investment decisions, which may undermine financial performance, deter investor confidence and increase stock liquidity risk (Dittmar & Mahrt-Smith, 2007; Huang & Mazouz, 2018). Conversely, low cash holdings expose firms to increased financial risk, increasing uncertainty and reducing stock liquidity. Empirical evidence exists that shows prudent cash management firms, particularly those in low-growth industries, have higher stock liquidity as a result of reduced uncertainty and increased investor confidence (Gopalan et al., 2012).

Corporate policy risk is a major but complex aspect of market behavior and is a part of the overall corporate policy, including investment, financial, diversification, and cash management policy. The relationship between these interconnected elements and investor confidence, trading activity, and the cost of capital is the basis for corporate financial dynamics. The interplay is further complicated by the dual role of liquidity: it allows for risk-taking and market efficiency, but excessive risks from managerial decisions can amplify volatility and erode trust, which adversely impact liquidity (Hsu et al., 2018). This study, therefore, empirically examines this relationship by constructing an overall corporate policy risk index (PRI). Consequently, this framework leads to the formulation of our study’s hypothesis:

Hypothesis (H1): 

Corporate policy risk (PRI) is negatively associated with stock liquidity.

3. Empirical Research Methodology

3.1. Sample Selection

Our sample comprises all listed firms in the National Stock Exchange of India (NSE) spanning the period from the financial years (FYs) 2003–2004 to 2022–2023 (i.e., from 1 April to 31 March)3. Following Chordia et al. (2005) and Dang et al. (2022), we employed the following stock selection criteria in our study. First, we required that each stock be continuously traded throughout the sample period, ensuring the availability of daily trading information. Second, we excluded stocks that were not actively traded, as well as firms lacking relevant market and accounting variables throughout the entire sample period, which facilitated the construction of a balanced panel dataset, resulting in an initial pool of 544 firms out of a total of 2206 NSE active listed companies. Third, following the guidelines set forth by Brogaard et al. (2017), we further refined our sample by excluding financial companies due to their distinct capital structures, disclosure practices, and regulatory frameworks4. This stringent process culminated in a final dataset comprising 9320 firm-year observations across 466 Indian firms, spanning the FYs from 2003–2004 to 2022–2023. Definitions of all the variables are provided in

Table A1. Variables’ definitions.

Type of Variables	Variables	Description
Dependent variables	Amihud	The natural logarithm of (1+ average daily Amihud (2002) ratio over a year) × −1.
	HLS	(1+ average daily closing high low spread according to Corwin and Schultz (2012) over a year) × −1.
Independent variable	PRI	A firm’s overall corporate policies risk index (PRI) is computed by capturing risk embedded in managers’ different policy decisions, such as investment, financing, diversification, and cash management, by weighting each policy risk through the regression decomposition method as per Equation (7).
Control variables	RVOL	Return volatility in the year is measured as the standard deviation of monthly stock returns over the year.
	IPRC	The inverse of the average stock price over the year.
	Size	Natural log of total assets in a year.
	ROA	Net profit to total assets in a year.
	Tobin’s_Q	The ratio of the sum of the market value of equity and total liabilities (preferred stock and debt) to the total assets in a year.
	LEV	Total debt to total assets in a year.
	R&D	Research and development expenses investments scaled by total assets in a year.
	IO	Institutional ownership is calculated as the percentage of shares held by institutional investors over the year.
PRI constructing variables	StDev of Returns	This variable represents the standard deviation of market-adjusted monthly stock returns, also known as abnormal returns ($A{R}_{i,t})$, over a specified period. Abnormal returns are computed as $A{R}_{i,t}=R_{i,t}-R_{m,t}$ where $R_{i,t}$ is the stock $i$’s return during month $t$ and $R_{m,t}$ is the is the equally-weighted market return for the same month. The standard deviation of these abnormal returns is calculated over a 36-month (3-year) period to provide a measure of the stock’s volatility relative to the market.
	StDev of CFs	This variable represents the standard deviation of a company’s annual cash flows, calculated over a 3-year period.
	StDev of ROA	This variable represents the standard deviation of a company’s ROA, calculated over a 3-year period.
	Idios. Volatility	The stock’s idiosyncratic volatility is calculated using residuals derived from the Fama–French–Carhart 4-factor model. Monthly stock return data ${(R}_{it})$ is used alongside the risk-free rate ${(R}_{ft}$), represented by the return on 91-day Treasury bills. Factor returns, including the market risk premium $\left(R_{mt}-R_{ft}\right)$, size premium $\left(SM{B}_t\right)$, value premium ($HM{L}_t)$, and momentum premium ($UM{D}_t)$, are retrieved from the Agarwalla et al. (2014) research database available at IIM Ahmedabad website: https://faculty.iima.ac.in/iffm/Indian-Fama-French-Momentum/ (accessed on 10 August 2024). The following regression model is applied for each stock: $R_{it}-R_{ft}=a_i+{\beta }_i\left(R_{mt}-R_{ft}\right)+{\gamma }_{i}SM{B}_t+{\chi }_{i}HM{L}_t+{\mu }_{i}UM{D}_t+{\epsilon }_{it}$ where the residuals (${\epsilon }_{it})$ capture the portion of stock returns not explained by the four factors. The idiosyncratic volatility is computed as the standard deviation of these residuals over 36 months, providing a 3-year average measure of volatility for each stock.
	Investment Policy	$Log(1+{R\&D}_{i,t}/{CAPX}_{i,t}$) where CAPX denotes the firm’s capital expenditure.
	Capital Structure Policy	${Total debt}_{i,t}/{Total assets}_{i,t}$
	Business Diversification Policy	$Log({Total business segemnts}_{i,t}$)
	Cash Policy	Cash policy is represented by excess Cash holdings of the firm as per the Opler et al. (1999) measure: $\begin{array}{c}\\ \mathrm{Ln}\left(\mathrm{Cash}\right)={\beta }_1\ln \left({\mathrm{Assets} }_{it}\right)+{\beta }_2{\mathrm{C}\mathrm{F}}_{it}+{\beta }_{3}NW{C}_{it}+{\beta }_{4}MV{E}_{it}+{\beta }_5{ \mathrm{CAPX} }_{it}+\\ {\beta }_6{ \mathrm{LEV} }_{it}+{\beta }_{7}R{D}_{it}+{\beta }_{s}DI{V}_{it}+{\phi }_i+v+{\tau }_t+{\mu }_{it}\\ \end{array}$ where CF represents the operating cash flow scaled by total assets, NWC denotes net working capital, calculated as the difference between current assets and current liabilities, scaled by total assets, MVE is the market value of equity, obtained by multiplying the stock price with the total number of outstanding shares at the end of the financial year, CAPX refers to capital expenditure scaled by total asset during the year. LEV indicates the total debt to total assets ratio, RD stands for research and development expenditure scaled by total assets, and DIV is a dummy variable equal to 1 if the firm paid dividends during the year and 0 otherwise.
Additional variables	Complexity	The complexity of financial reporting is calculated as the sum of the word count for each complex word, provided by Loughran and McDonald (2023), to the total number of words in the annual report, expressed as a percentage.

Notes: This table reports descriptions of all the variables used in this study.

(Appendix A).

We obtained our sample data from different sources to construct all our variables. Specifically, (i) individual stock liquidity proxies and other market-related data originate from the Bloomberg database, (ii) firm-specific accounting and finance-based control variables were obtained from the Centre for Monitoring Indian Economy Prowess (CMIE Prowess) database.

3.2. Variable Description

3.2.1. Dependent Variable: Stock Liquidity

Stock liquidity is an elusive and hard-to-define concept. Liquidity is defined as the extent to which an investor can buy or sell substantial amounts of security without paying any huge trading cost and without causing an adverse change in the value of the asset (Amihud et al., 2006). We use two different measures of stock liquidity to examine the relationship between PRI and stock liquidity.

Our first proxy is the Amihud (2002) measure of liquidity (Amihud), which captures the price impact characteristics of liquidity. This is a widely used proxy of liquidity and has been used by the most prominent authors of the literature such as Fong et al. (2017), Brogaard et al. (2017), and Debata et al. (2018). The Amihud price impact is calculated as the logarithm of one plus the average ratio of the daily absolute return to the INR trading volume on day d for stock i over year t, where Dit is the number of trading days for stock i in year t following Cheung et al. (2015) and Dang et al. (2022). We then multiplied by −1 to convert illiquidity to liquid to make the analysis easily interpretable and noted as the following:

$${\mathrm{A}\mathrm{m}\mathrm{i}\mathrm{h}\mathrm{u}\mathrm{d}}_{it}=log\left(1+{\displaystyle \frac{1}{D_{it}}}{\sum }_{d=1}^{D_{it}}{\displaystyle \frac{\left|R_{idt}\right|}{{VOLD}_{idt}}} \right)\times -1$$

(1)

where $\left|R_{idt}\right|$ is the absolute stock return of stock i on day d of year t, ${VOLD}_{idt}$ is the trading volume of stock i on day d of year t, and $D_{it}$ is the number of days with available data for stock i in year t. We transform the ratio by taking its natural logarithm and then multiplying by −1 to convert illiquidity to liquidity.

Second, we calculate the high–low spread (HLS) of stock construed by Corwin and Schultz (2012) and used by Marshall et al. (2012), Cheng and Fang (2023), and Roy et al. (2022). In order to untangle the variance and spread parts of the high–low price, Corwin and Schultz (2012) compute the summation of the squared log price for two successive days:

$$\beta ={\displaystyle {\sum }_{d=0}^1}{\left[\mathit{ln}\left({\displaystyle \frac{H_{t+d}^O}{L_{t+d}^O}}\right)\right]}^2$$

(2)

$$\gamma ={\sum }_{d=0}^1{\left[\mathit{ln}\left({\displaystyle \frac{H_{t,t+1}^O}{L_{t,t+1}^O}}\right)\right]}^2$$

(3)

where $H^O$ represents the observed high price of a stock on day d, and $L^O$ represents the observed low price of a stock on day d. ln = natural logarithm.$H_{t,t+1}^O=$ is the high price over the two consecutive days t and t + 1. $L_{t,t+1}^O=$ is the low price over the two successive days’ t and t + 1. Corwin and Schultz (2012) derived a solution for the spread (HLS) by drawing from previous research conducted on high–low price ratios.

$${HLS}_{it}=\left({\displaystyle \frac{2 \left({e^{\alpha }}^{it}-1\right)}{1+{e^{\alpha }}^{it}}}\right)\ast -1$$

(4)

where

$$\alpha ={\displaystyle \frac{\sqrt{\left({2\beta }_{it}\right)}-\sqrt{{\beta }_{it}}}{3-2\sqrt{2}}}-\sqrt{{\displaystyle \frac{{\gamma }_{it}}{3-2\sqrt{2}}}}$$

(5)

The HLS estimation is calculated for each two-day interval by utilizing the daily high and low prices obtained from the Bloomberg database. For each year, the bid–ask spread for each sample stock is computed by averaging all spreads throughout all two-day intervals. The formula used for averaging is as follows:

$$\mathrm{Average} {\mathrm{H}\mathrm{L}\mathrm{S}}_i={\displaystyle \frac{{\sum }_{n=1}^N\hspace{1em}HL{S}_{it}}{N}}$$

(6)

where N is the number of two-day intervals with available trading data. If there are days with no trading activity, these intervals are excluded from the averaging calculation, reducing N accordingly. In order to convert the illiquidity measurement of spread to liquidity for a convenient way to interpret, it was multiplied by −1.

3.2.2. Independent Variable: Corporate Policy Risk

Our key independent variable is the policy riskiness index (PRI), which is measured to construct the aggregated riskiness of a firm’s key corporate policies, isolating policy-driven risk from exogenous industry and macroeconomic influences. This PRI is defined as a holistic measure of a firm’s risk profile, as the weighted impact of policy decisions on realized firm risk. This approach implements a regression-based decomposition of realized risk into separate policy components. Following Çolak and Korkeamäki (2021), we construct the PRI by following three major steps: the PRI identifies policy proxies, determines realized risk measures and estimates policy weights.

Step 1: Identifying Policy Proxies

Our PRI comprises four central aspects of corporate policies: investment, capital structure, business diversification, and cash policy, reflecting different components of corporate risk. Investment policy is first proxied with the firm’s research and development expenditures (R&D) to capital expenditures (Capex) ratio. Following Cassell et al. (2012), (Coles et al., 2006), and Kothari et al. (2002), we treat R&D spending as a more risky investment than Capex, which is typically more predictable in terms of cash flows. Skewness is accounted for by transforming the ratio of R&D to Capex logarithmically. Second, the capital structure policy is measured by the firm’s leverage (ratio of total debt to total assets). The risk-taking literature suggests that a CEO’s preference for financial risk is indicated by higher leverage (Faccio et al., 2016). Third, the business diversification policy is the number of industry segments a firm operates in, with more segments generally meaning lower risk exposure (Cassell et al., 2012; Jacquemin & Berry, 1979; Khanna & Palepu, 2000). We quantify this by using the logarithm of the number of segments as a diversification measure. Finally, cash policy is defined in Opler et al. (1999) as excess cash holdings influenced by dividend policy, short and long-term financing decisions, and current cash flows. This excess cash policy serves as a buffer, reducing operational risk, but can also represent concerns of the agency.

Step 2: Realized Risk Measures

In this step, we examine how the above-mentioned various corporate policies translate into ex-post risk by analyzing four key risk measures over the three years following policy implementation. This approach assumes that CEO or top managers’ policy decisions influence subsequent risk outcomes, which becomes evident through these realized metrics. This measures provide a unique, comprehensive perspective on firm-level risk, enabling a thorough evaluation of policy impacts on cash flow stability, profitability, and market volatility. By capturing the long-term effects of policy decisions, this study offers valuable insights into the interplay between corporate strategies and financial risk.

The first measure, the standard deviation of cash flows (StDev of CFs), is a measure of volatility in operating cash flows, a measure of the stability of cash generation after policy actions (Minton & Schrand, 1999). When calculating the standard deviation of cash flows, we use data from multiple years to capture variability. Specifically, we compute the standard deviation across the three-year period post-policy implementation, thus capturing the inter-annual volatility. This approach accounts for fluctuations in cash flows over time, allowing for a meaningful assessment of cash flow stability. Similarly, the second measure, standard deviation of return on assets (StDev of ROA), is a measure of profitability volatility over the subsequent three years, indicating how consistently policies support returns on the firm’s asset, as per prior studies (Faccio et al., 2016; Ferris et al., 2017; John et al., 2008).

We follow Cain and McKeon (2016) in measuring market-based risk. The first is the return volatility, which is the standard deviation of abnormal stock returns (StDev of Returns) over a 3-year period, a measure of firm-specific return variability above and beyond expected market performance. We also calculate idiosyncratic stock volatility (Idios.Volatility) by using the Fama–French–Carhart four-factor model to measure risks associated with the firm’s policies that are unique to the firm over the 3 years. We measure Idios. Volatility as the annual average standard deviation of the monthly standard deviation of residuals from the Fama–French–Carhart four-factor model. This measure is consistent with prior literature in this context (Adams et al., 2005; Akbar et al., 2017; Ferris et al., 2017).

Step 3: Estimating Policy Weights and Constructing PRI

To create a comprehensive corporate policy risk index reflecting the risk contribution of each policy, we employ a “policy decomposition of realized risk”. This approach involves estimating the relative weight of each key policy decision—investment, capital structure, business diversification, and cash policies—on a firm’s realized risk by regressing these risk measures on the respective policy variables. Industry and time-fixed effects are included to control for sectoral and temporal influences, ensuring a clear focus on firm-level impacts. This method provides a nuanced understanding of how risk involved in different corporate policy decisions by a CEO contribute to overall firm risk. Our policy riskiness index (PRI) is constructed using the following regression model:

$$\begin{array}{cc}{Realized Risk}_{i,t+3}\hfill & \\ & ={\beta }_0+{\beta }_1{Investment policy}_{i,t}+{\beta }_2{Capital strcture policy}_{i,t}+{\beta }_3{Busines Diversification policy}_{i,t}\hfill \\ & +{\beta }_4{Cash policy}_{i,t}+{\displaystyle \sum _j}{\gamma }_j{Ind}_j+{\displaystyle \sum _n}{\theta }_n{Year}_n+{\in }_{i,t}\hfill \end{array}$$

(7)

This decomposition approach separates realized risk into three main elements: (1) industry fixed-effects dummies (${\sum }_j{\gamma }_j{Ind}_j$) to account for industry-specific risk, (2) time fixed-effects dummies (${\sum }_n{\theta }_n{Year}_n$) to control for macroeconomic or technological shocks, and (3) the primary focus of this study, the firm-specific policy risk, captured by the coefficients on the four policy variables (from ${\beta }_1 \mathrm{t}\mathrm{o} {\beta }_4)$. This component captures the effect of CEO-driven policy decisions, distinguishing PRI from other firm volatility measures that may be confounded by external factors.

Our PRI is calculated as the predicted value of realized risk using only the coefficients of the policy variables:

$$\begin{array}{c}{PRI}_{i,t}=\widehat{{\beta }_0}+\widehat{{\beta }_1}{Investment policy}_{i,t}+\widehat{{\beta }_2}{Capital strcture policy}_{i,t}+\widehat{{\beta }_3}{Busines Diversification policy}_{i,t}\hfill \\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}+\widehat{{\beta }_4}{Cash policy}_{i,t}\hfill \end{array}$$

$$\begin{array}{c}{PRI}_{i,t}=0.1799+0.0011{Investment policy}_{i,t}+0.0230{Capital strcture policy}_{i,t}\hfill \\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}-0.0087{Busines Diversification policy}_{i,t}-0.0018{Cash policy}_{i,t}\hfill \end{array}$$

(8)

The values used in our PRI calculation reflect the coefficients from Column (4) of

Table 1. Components of policy riskiness index.

	StDev of CFs	StDev of ROA	StDev of Returns	Idios. Volatility
Variables	(1)	(2)	(3)	(4)
Investment policy	−0.0008	0.0006 *	−0.0007	0.0011 *
	(−0.83)	(1.77)	(−0.42)	(1.78)
Capital structure olicy	0.0339 ***	0.0109 ***	0.0588 ***	0.0230 ***
	(10.69)	(7.94)	(10.75)	(4.79)
Business diversification olicy	−0.0018 *	−0.0010 **	−0.0036 *	−0.0087 ***
	(−1.75)	(−1.97)	(−1.84)	(−4.92)
Cash policy	0.0005	−0.0004	0.0002	−0.0018 **
	(0.88)	(−0.51)	(0.32)	(−1.99)
Constant	0.0606 ***	0.0248 ***	0.1623 ***	0.1799 ***
	(8.14)	(7.73)	(14.55)	(16.82)
Observations	9320	9320	9320	9320
Year FE	Yes	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes	Yes
Adj. R-squared	0.078	0.097	0.159	0.187

Notes: The t-statistics reported in parentheses are based on standard errors clustered at the firm level. *** p < 0.01, ** p < 0.05, * p < 0.1.

, where the standard deviation of future idiosyncratic stock volatility (Idios. Volatility) serves as the dependent variable. This specification aligns with prior findings. Notably, those of Çolak and Korkeamäki (2021) highlight the statistical significance of most policy coefficients in explaining idiosyncratic volatility, though some, such as investment and cash policies, display less significance in other specific risk measures like StDev of CFs and StDev of Returns.

Overall, this framework thus enables a refined analysis of combined policy-induced risk, isolated from industry and macroeconomic noise, providing insight into the strategic risk outcomes directly attributable to managerial decisions.

3.2.3. Control Variables

We control for market-related factors and firm-specific characteristics that may affect stock liquidity in line with previous studies. Market-related controls include stock return volatility (RVOL), inverse share price (IPRC), and Tobin’s Q. RVOL is a measure of firm uncertainty, which is essential for informed investors who are faced with a higher risk–reward trade-off in volatile stocks (Ellul & Panayides, 2018; Jegadeesh & Subrahmanyam, 1993). The IPRC controls for the higher risk and lower liquidity often associated with lower-priced stocks (Boubaker et al., 2019). According to Wang et al. (2022), Tobin’s Q is a measure of the firm’s market growth potential. The firm-level controls are return on assets (ROA), firm size (SIZE), leverage (LEV), research and development expenditure (R&D), and institutional ownership (IO). The firm’s operational efficiency and profitability are controlled by including ROA (Schoenfeld, 2017); size captures operational risk and information availability (Bhagat et al., 2015). Financial conditions are controlled by including LEV (Dang et al., 2022), while R&D controls innovation potential (Schoenfeld, 2017). The influence of institutional investors’ informational demands is analyzed by IO (Zhang et al., 2023).

3.3. Model Specification

In order to analyze the association between PRI and stock liquidity, we construct a panel fixed-effect regression model following the established literature (Mbanyele, 2023; Nadarajah et al., 2018; Wang et al., 2023), and as per the results of the Hausman (1978) specification test (reported in

Table A3. Hausman (1978) specification test.

For dependent variable Amihud:
	Coef.
Chi-square test value	148.73
p-value	0
For dependent variable HLS:
	Coef.
Chi-square test value	529.43
p-value	0

Note: The null hypothesis as panel random effect is appropriate.

); the model is outlined below:

$${LIQUIDITY}_{i,t}={\beta }_0+{{\beta }_{1}PRI}_{i,t}+{\beta }_2{CONTROLS}_{i,t}+FIXED EFCCTS+{\epsilon }_{i,t}$$

(9)

${LIQUIDITY}_{i t}$ is the stock liquidity of firm i and year t, and is the dependent variable. Our two proxies for assessing stock liquidity are the Amihud liquidity (Amihud) measure and the high–low spread (HLS). Our model’s key independent variable is the firm-specific policy risk index ${(PRI}_{i,t}$) that measures the cumulative risk exposure resulting from key corporate policy decisions. This index consists of four critical corporate policies such as investment, capital structure, business diversification, and cash policy, and is constructed by a regression decomposition approach that regresses these policies on the firm’s ex-post realized risk, namely idiosyncratic volatility, as expressed in Equation (2). ${CONTROLS}_{i,t}$ denotes the vectors of both firm and market-level control variables that may influence stock liquidity as mentioned in Section 3.2.3. Control Variables in the Methodology section. Furthermore, we incorporate year- and industry-fixed effects to mitigate omitted variable bias and to control the time trends and any sector-wide shocks that might drive liquidity over time. We also cluster the robust standard errors at the firm level. ${\epsilon }_{i,t}$ represents the error term. The detailed descriptions of all the variables are provided in Table A1 (Appendix A).

4. Empirical Results

4.1. Descriptive Statistics

Table 2. Descriptive statistics.

Variables	Obs	Mean	Std. Dev.	Min	Max
Amihud	9320	−0.014	0.022	−0.098	0.000
HLS	9320	−0.017	0.008	−0.051	−0.006
PRI	9320	0.186	0.011	0.156	0.229
RVOL	9320	0.223	0.092	0.088	0.711
IPRC	9320	0.051	0.129	0.004	0.80
SIZE	9320	9.432	1.804	5.363	15.114
ROA	9320	4.866	8.609	−28.201	29.610
Tobin’s_Q	9320	1.673	1.371	0.356	8.720
LEV	9320	0.198	0.664	0.012	5.719
R&D	9320	0.004	0.009	0.000	0.096
IO	9320	0.140	0.134	0.013	0.574

Notes: This table reports the summary statistics of all the variables. The sample consists of 9320 firm-year observations (balanced panel) with 466 NSE-listed unique firms from 2003–2004 to 2022–2023. The description of variables is available in Table A1 (Appendix A).

displays the descriptive statistics for the main variables of our study. The average values for the Amihud and HLS are −0.014 and −0.017, respectively. The average value for the PRI is 0.186, with a minimum value of 0.156 and a maximum value of 0.229. These values are comparable to a seminal study of Çolak and Korkeamäki (2021) corporate policy risk. The average value for RVOL is 0.367, IPRC is 21.918, size is 2.366, ROA is 0.458, Tobin’s_Q is 0.253, LEV is 0.03, R&D is 1.952. In addition, the average institutional ownership is 14.00%, indicating that institutional investors’ engagement is low for Indian firms.

Table 3. Pairwise correlations.

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	VIF
(1) Amihud	1.00
(2) HLS	0.36 *	1.00
(3) PRI	−0.18 *	−0.20 *	1.00									1.43
(4) RVOL	−0.59 *	−0.48 *	0.18 *	1.00								1.92
(5) IPRC	−0.51 *	−0.35 *	0.18 *	0.47 *	1.00							1.54
(6) Size	0.59 *	0.41 *	−0.14 *	−0.45 *	−0.32 *	1.00						2.87
(7) ROA	0.29 *	0.31 *	−0.32 *	−0.33 *	−0.24 *	0.12 *	1.00					1.69
(8) Tobin’s_Q	0.23 *	0.22 *	0.01	−0.29 *	−0.11 *	0.11 *	0.38 *	1.00				1.51
(9) LEV	−0.11 *	−0.17 *	0.34 *	0.11 *	0.13 *	0.02	−0.34 *	−0.01	1.00			1.33
(10) R&D	0.13 *	0.14 *	0.01	−0.14 *	−0.09 *	0.07 *	0.16 *	0.16 *	−0.06 *	1.00		1.21
(11) IO	0.38 *	0.36 *	−0.17 *	−0.34 *	−0.20 *	0.61 *	0.21 *	0.20 *	−0.08 *	0.12 *	1.00	2.07

Notes: This table reports the pairwise correlation of all the variables in our model. The sample consists of 9320 firm-year observations (balanced panel) with 466 NSE-listed unique firms from 2003–2004 to 2022–2023. The description of variables is available in Table A1 (Appendix A). * p < 0.01.

displays the pairwise correlations between all key variables in this study. As hypothesized, we find a robust negative relationship between PRI and stock liquidity measures (Amihud, HLS). Further, MS_BERT is positively related to both stock liquidity proxies, consistent with a link between manager sentiment and liquidity. The correlations of stock liquidity with size, ROA, Tobin’s Q, R&D, and IO are positive, while RVOL, IPRC, and LEV are negative with Amihud and ROA. Additionally, variance inflation factor (VIF) diagnostics (presented in the last column of Table 3) show that all variables have VIF values less than three, suggesting the model is not afflicted by multicollinearity.

4.2. Baseline Results

In our initial analysis, we regress stock liquidity (measured by Amihud and HLS) on PRI and additional control variables, with results reported in

Table 4. Baseline results: the impact of PRI on stock liquidity.

	Amihud	HLS
Variables	(1)	(2)
PRI	−0.1088 ***	−0.0536 ***
	(−2.95)	(−4.28)
RVOL	−0.0593 ***	−0.0145 ***
	(−11.50)	(−7.87)
IPRC	−0.0041 *	−0.0014 ***
	(−1.68)	(−3.02)
Size	0.0058 ***	0.0018 ***
	(13.68)	(11.94)
ROA	0.0021 ***	0.0005 ***
	(5.64)	(4.70)
Tobin’s_Q	0.0041 *	0.0052 ***
	(1.85)	(6.95)
LEV	−0.0002 ***	−0.0001 ***
	(−4.24)	(−4.39)
R&D	0.0482 ***	0.0116 ***
	(3.47)	(4.47)
IO	0.0166 ***	0.0017 *
	(5.33)	(1.72)
Constant	−0.0542 ***	−0.0126 ***
	(−7.21)	(−4.46)
Observations	9320	9320
Year FE	Yes	Yes
Industry FE	Yes	Yes
Adj. R-squared	0.623	0.630

Notes: The description of variables is available in Table A1 (Appendix A). The t-statistics reported in parentheses are based on standard errors clustered at the firm level. *** p < 0.01, * p < 0.1.

. The results in Table 4 confirm the negative effect of PRI on Amihud and HLS at the 1% significance level, even after including control variables in our model. Using the results in Column (1)–(2) as an example, we can argue that the coefficient magnitudes are economically significant. Specifically, an increase of one standard deviation in PRI (1.10%) results in a decline of approximately 0.12 percentage points (=0.011 × (0.1088)) in Amihud and 0.06 percentage points (=0.011 × (0.0536) in HLS. The stated reductions can be analyzed as approximately 8.55% (=(0.011 × (0.1088))/−0.014) and 3.47% (=(0.011 × (0.0536))/−0.017) of the average Amihud and HLS values for the firms within the sample, respectively.

For the robustness of this baseline results, we alternatively measured the PRI as per the market-based realized risk metrics only due to the non-availability of profitability-based measured monthly data, such as StDev of Returns and Idios. Volatility, over a one-year period using monthly market data and report the regression decomposition results in

Table A4. Robustness test: components of policy riskiness index.

	StDev of Returns	Idios. Volatility
Variables	(3)	(4)
Investment policy	−0.0006	0.0009 *
	(−0.71)	(1.76)
Capital structure policy	0.0752 ***	0.0444 ***
	(11.12)	(3.55)
Business diversification policy	−0.0024	−0.0065 ***
	(−1.62)	(−4.92)
Cash policy	0.0011 *	−0.0032 **
	(1.82)	(−1.99)
Constant	0.1623 ***	0.1245 ***
	(14.55)	(16.82)
Observations	9320	9320
Year FE	Yes	Yes
Industry FE	Yes	Yes
Adj. R-squared	0.159	0.225

Notes: The t-statistics reported in parentheses are based on standard errors clustered at the firm level. *** p < 0.01, ** p < 0.05, * p < 0.1.

(Appendix A). We then reassess our baseline model’s Equation (9) by incorporating this alternative PRI, derived from Idios. Volatility, given its relevance across all four policy areas. The results, as detailed in

Table A5. Robustness test: baseline result.

	Amihud	HLS
Variables	(1)	(2)
PRI	−0.0724 **	−0.0354 **
	(−2.05)	(−2.22)
Constant	−0.0611 ***	−0.0157 ***
	(−5.25)	(−3.25)
Observations	9320	9320
Baseline control	Yes	Yes
Year FE	Yes	Yes
Industry FE	Yes	Yes
Adj. R-squared	0.575	0.608

Notes: The description of variables is available in Table A1 (Appendix A). The t-statistics reported in parentheses are based on standard errors clustered at the firm level. *** p < 0.01, ** p < 0.05.

(Appendix A), remain consistent with our initial findings of Table 4, reinforcing the validity of our PRI measure.

Our study confirms that corporate policy risk significantly influences stock liquidity, aligning with recent empirical evidence. Strategic R&D investments improve liquidity through enhanced transparency and innovation potential; however, high-risk ventures can increase volatility and stock illiquidity (Xiang et al., 2020). High leverage diminishes liquidity due to elevated financial risk (Cain & McKeon, 2016; Nadarajah et al., 2018). Business diversification can reduce idiosyncratic risk and therefore may decrease (Jiang & Feng, 2021), while prudent cash management boosts liquidity by reducing perceived risk (Xu et al., 2023). These findings reflect the nuanced relationship between corporate policies and liquidity, supporting insights from Hsu et al. (2018).

This study advances the empirical literature by establishing corporate policy risk as a crucial firm-specific determinant of stock liquidity. Unlike prior research focused on isolated corporate decisions, we introduce a comprehensive measure of corporate policy risk, capturing its multifaceted nature (Bernile et al., 2018; Cain & McKeon, 2016; Ferris et al., 2017).

Our findings reveal how managerial decision-making influences stock liquidity, bridging gaps in research on managerial risk attitudes and market microstructure (Coles et al., 2006; Hsu et al., 2018). Practically, these insights guide managers in developing balanced strategies to mitigate risks and enhance market efficiency. Policymakers can use this evidence to foster market transparency while investors gain tools to assess liquidity risks. This study provides a robust foundation for future research on corporate strategies and market dynamics.

With regard to control variables, we observe that size, ROA, Tobin’s_Q, R&D, and IO are positively and significantly associated with stock liquidity proxies, whereas RVOL, IPRC, and LEV are significantly negatively related. The plausible reason is that higher RVOL and IRPC dampen liquidity by depressing stock prices, increasing trading costs and risks, and reducing market participation (Boubaker et al., 2019; Ellul & Panayides, 2018; Jegadeesh & Subrahmanyam, 1993). On the other hand, size positively affects liquidity, as larger firms are more attractive to trading activity because of lower operational risk and more information availability (Dang et al., 2022). ROA and Tobin’s_Q also enhance a firm’s market growth and operational efficiency, which enhance stock turnover (Schoenfeld, 2017; Wang et al., 2022). R&D also boosts liquidity as investors like innovation and will be more attracted to trading these growth potential stocks (Dang et al., 2022; Schoenfeld, 2017). Similarly, IO enhances liquidity by stabilizing the market price through long-term investment and higher information demand (Zhang et al., 2023).

4.3. Role of Information Environment

This sub-section provides the results of the influencing role of the information environment on PRI-LIQUIDITY analysis. In order to analyze this mechanism, we utilize three indicators of information asymmetry: firm size, stock volatility, and the complexity of financial reports. The aforementioned measures have been extensively utilized in existing literature to assess the quality of the information environment at the firm level (Loughran & McDonald, 2023; Nagar et al., 2019; Wang et al., 2022; Zhang, 2006).

We tabulated these results in

Table 5. Role of the information environment.

Panel-A Dependent Variable: Amihud
	Small Size	Large Size	High Volatility	Low Volatility	High Complexity	Low Complexity
Variables	(1)	(2)	(3)	(4)	(5)	(6)
PRI	−0.2977 ***	−0.0389 *	−0.2331 ***	−0.0617 **	−0.1248 ***	−0.0829
	(−5.22)	(−1.72)	(−4.42)	(−2.10)	(−2.83)	(−1.58)
Constant	0.0227 **	−0.0071	−0.0844 ***	−0.0466 ***	−0.0588 ***	−0.0497 ***
	(2.15)	(−1.49)	(−8.37)	(−7.63)	(−6.76)	(−4.37)
Observations	4660	4660	4660	4660	3855	5465
Baseline control	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes	Yes	Yes	Yes
Adj. R-squared	0.571	0.309	0.632	0.455	0.652	0.552
Panel-B Dependent Variable: HLS
	Small Size	Large Size	High Volatility	Low Volatility	High Complexity	Low Complexity
	(1)	(2)	(3)	(4)	(5)	(6)
PRI	−0.0725 ***	−0.0520 ***	−0.0677 ***	−0.0509 ***	−0.0736 ***	−0.0308 **
	(−4.20)	(−4.46)	(−3.82)	(−4.22)	(−4.29)	(−2.13)
Constant	0.0075 **	0.0003	−0.0154 ***	−0.0160 ***	−0.0080 **	−0.0155 ***
	(2.27)	(0.14)	(−3.82)	(−6.52)	(−1.96)	(−5.11)
Observations	4660	4660	4660	4660	3855	5465
Baseline control	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes	Yes	Yes	Yes
Adj. R-squared	0.610	0.556	0.582	0.549	0.610	0.689

Notes: The description of variables is available in Table A1 (Appendix A). The t-statistics reported in parentheses are based on standard errors clustered at the firm level. *** p < 0.01, ** p < 0.05, * p < 0.1.

(Panel-A for Amihud and Panel-B for HLS). First, we split our sample by firm size. In general, smaller firms exhibit higher information asymmetry due to their lower transparency, limited analyst coverage, and higher operational risks, while larger firms tend to have lower asymmetry with lower-yield spreads, better financial reporting, and analyst attention (Bhattacharya et al., 2013; Herskovic et al., 2016; Kim et al., 2021; Lu et al., 2010).

We report the results in Table 5 (Panel-A for Amihud and Panel-B for HLS). Columns (1)–(2) reveal that the negative effect of PRI on stock liquidity is significant between both small-size stocks and large-size stocks, while it seems to have a more pronounced impact on small-size stocks.

Next, we split the sample into two high and low stock volatility categories. Table 5, columns (3)–(4) indicate that the adverse impact of PRI on stock liquidity is significant and more pronounced in the context of high-volatility stocks, whereas it appears to be of lower magnitude for low-volatility stocks. This suggests that companies exhibiting higher volatility, which links with greater information asymmetry, experience more negative impacts from manager-induced corporate policy risk. Investors confronted with such firm stairgate uncertainty tend to seek higher returns or might opt to refrain from trading in illiquid and unpredictable stocks (Brogaard et al., 2017; Nagar et al., 2019). In contrast, companies characterized by lower volatility demonstrate a less pronounced or negligible negative association, indicating more stable and foreseeable stock price movements.

Finally, we classify the sample based on the complexity of the financial reporting (Complexity). We compute novel Complexity by leveraging the complexity lexicon developed by Loughran and McDonald (2023) and employing sophisticated NLP techniques on companies’ annual reports. The formula for this is the proportion of complex words in an Annual report relative to its total word count. Complex financial reports pose challenges for investors, particularly those with limited expertise or resources, in comprehensively grasping a manager’s future outlook toward different corporate policy decisions. This results in disparities in information access between insiders, such as managers and market participants, thereby exacerbating information asymmetry. Complex reports may conceal essential information or elevate the expenses associated with obtaining and analyzing information, potentially placing confident investors at a disadvantage due to their distrustful reaction to these obfuscation reports (Bloomfield, 2002; You & Zhang, 2009). Thus, we consider financial reporting linguistic complexity as an information asymmetry proxy (Bushee et al., 2018).

The findings are presented in Table 5 (Panel-A for Amihud and Panel-B for HLS). The findings presented in columns (5)–(6) indicate that the negative impact of PRI on stock liquidity is significant, especially pronounced in high-complexity firms, while it appears to exert a less significant effect on low-complexity firms.

Overall, the findings suggest that the adverse impact of a PRI on stock liquidity is prominent for hard-to-value firms and those involved in obfuscation reporting, which implies firms that exhibit high information asymmetry relative to their peers.

4.4. Role of Financial Crisis

In this sub-section, we examine the role of the financial crisis (Crisis) in explaining the association between firm corporate policy risk (PRI) and stock liquidity. To empirically control for this crisis impact on our baseline model, we split our sample into two subcategories (Crisis and non-Crisis) based on India’s recession or crisis period as provided in the Federal Reserve Bank of St. Louis (FRED) database.

The subsample analysis of this finding based on the Crisis and Non-Crisis periods is shown in

Table 6. Role of financial crisis.

	Amihud		HLS
Variables	Crisis	Non-Crisis	Crisis	Non-Crisis
	(1)	(2)	(3)	(4)
PRI	−0.1405 ***	−0.0930 **	−0.0612 ***	−0.0498 ***
	(−2.94)	(−2.39)	(−3.38)	(−4.09)
Constant	−0.0602 ***	−0.0238 ***	−0.0135 ***	−0.0172 ***
	(−6.54)	(−2.79)	(−3.84)	(−5.23)
Observations	4185	5115	4185	5115
Baseline controls	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes	Yes
Industry fixed effects	Yes	Yes	Yes	Yes
Adj. R-squared	0.665	0.550	0.657	0.606

Notes: The description of variables is available in Table A1 (Appendix A). The t-statistics reported in parentheses are based on standard errors clustered at the firm level. *** p < 0.01, ** p < 0.05.

. For both subsamples (Crisis and Non-Crisis), we find the coefficient of PRI to be negative for the Amihud and HLS proxies. Nevertheless, the magnitudes of the PRI coefficients are much smaller for firms in the Non-Crisis subsample models.

These results imply that systemic risk, reduced investor confidence, and constrained market liquidity (Brunnermeier, 2009; Suardi et al., 2022) amplify the negative effects of a firm’s corporate policy risk on stock liquidity during financial crises such as the Global Financial Crisis (GFC) and the COVID-19 pandemic. During crises, heightened uncertainty increases investor sensitivity to firm-specific risks (e.g., high leverage or speculative investments), and thus, bid–ask spreads are wider, trading volumes are lower, and market depth is reduced (Acharya & Viswanathan, 2011; Mazur et al., 2021). During crises, firms also deploy conservative liquidity management strategies as well as ill-fated investment decisions, with preferences over cash reserves rather than investments that lead to dampened trading activity as well as exacerbated liquidity challenges (Claessens et al., 2012; Derrien & KecskÉS, 2013). Additionally, the crisis led to structural changes in corporate governance mechanisms, which exacerbated the adverse impact of corporate policy risk by increasing information asymmetry eroding investor trust, and impairing liquidity (Conyon et al., 2011; Kirkpatrick, 2009; Zattoni & Pugliese, 2021).

Moreover, financial crises like the GFC and the COVID-19 pandemic magnify liquidity commonality and transaction costs and hit firms with high-risk policies disproportionately (Chung & Chuwonganant, 2023; Smales, 2024). Sell-offs are exacerbated by systemic liquidity shortages, which reduce market depth, and by investor herding behavior that leads to feedback loops of declining liquidity and rising volatility (Brunnermeier, 2009; Chordia et al., 2008). The finding that crises worsen stock liquidity reinforces the need to employ prudent risk management to temper the adverse effect of crises on liquidity.

4.5. Disaggregation of Industries

The impact of PRI on stock liquidity differs among industries due to factors such as corporate risk culture, stakeholder expectations, regulatory environments, and market dynamics (Cain & McKeon, 2016; Malik & Kashiramka, 2024; Pan et al., 2017). Therefore, examining the PRI–stock liquidity relationship within specific industries is crucial rather than treating them as a whole. Utilizing the Global Industry Classification Standard^®, provided by Morgan Stanley Capital International (MSCI) and Standard & Poor’s, a comprehensive breakdown of various industries was used to explain and understand the baseline findings thoroughly. Accordingly, the sample companies have been segregated into eight broad categories: consumer staples, consumer discretionary, IT and telecom, industrials, real estate, materials, healthcare, energy, and utility.

Table 7. Industry-wise impact of PRI on stock liquidity.

Panel-A Dependent Variable: Amihud
	Consumer Staples	Consumer Discretionary	IT and Telecom	Industrials	Real Estate	Materials	Healthcare	Energy and Utility
Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
PRI	−0.0528 *	−0.1122 *	−0.1628 ***	−0.0927 **	−0.1245 ***	−0.0642 *	−0.0685 ***	−0.0855 *
	(−1.72)	(−1.93)	(−5.11)	(−2.35)	(−3.53)	(−1.88)	(−3.53)	(−1.88)
Constant	−0.0244 **	−0.0115 *	−0.0244 **	−0.0247 **	−0.0744 ***	−0.0558 *	−0.0477 ***	−0.0545 **
	(−2.44)	(−1.91)	(−2.37)	(−2.15)	(−3.36)	(−3.55)	(−4.77)	(−2.35)
Observations	1080	1540	1100	1840	320	1660	900	880
Baseline control	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Adj. R-squared	0.544	0.604	0.594	0.666	0.522	0.611	0.652	0.594
Panel-B Dependent Variable: HLS
	Consumer Staples	Consumer Discretionary	IT and Telecom	Industrials	Real Estate	Materials	Healthcare	Energy and Utility
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
PRI	−0.0244 **	−0.0590 **	−0.0755 ***	−0.0509 **	−0.0603 **	−0.0344 **	−0.0272 ***	−0.0437 *
	(−2.11)	(−2.22)	(−3.44)	(−2.39)	(−2.11)	(−2.24)	(−2.96)	(−1.84)
Constant	−0.0125 ***	−0.0024	−0.0043 *	−0.0155 **	−0.0240 ***	−0.0051 ***	−0.0067 **	−0.0084 **
	(3.22)	(1.44)	(−1.82)	(−2.32)	(−3.77)	(−4.44)	(−2.14)	(−2.47)
Observations	1080	1540	1100	1840	320	1660	900	880
Baseline control	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Adj. R-squared	0.580	0.656	0.602	0.632	0.571	0.629	0.577	0.565

Notes: The description of variables is available in Table A1 (Appendix A). The t-statistics reported in parentheses are based on standard errors clustered at the firm level. *** p < 0.01, ** p < 0.05, * p < 0.1.

provides the findings of the industry-specific PRI impact on stock liquidity. The results show that the PRI has a significantly negative impact on stock liquidity in both proxies, Amihud and HLS. The findings indicate that consumer discretionary, IT and telecom, and real estate industries display stronger negative PRI coefficients, primarily due to high volatility, leverage, and sensitivity to external shocks. For instance, fluctuations in discretionary spending amplify the liquidity challenges when firms implement aggressive investment or capital structure policies. IT and telecom’s reliance on innovation heightens cash flow uncertainty, while real estate’s asset-heavy, leverage-driven structure exacerbates liquidity concerns under policy instability. Moderate coefficients are observed in industrials and energy and utility, where capital-intensive operations and regulatory oversight provide stabilizing effects. In contrast, consumer staples, materials, and healthcare exhibit lower coefficients, benefitting from stable demand, predictable cash flows, and regulatory protection.

5. Endogeneity Tests

5.1. Propensity Score Matching

In this sub-section, we implement the propensity score matching (PSM) of Rosenbaum and Rubin (1983) to deal with the omitted variable bias and sample self-selection bias. We conduct a logistic regression to calculate propensity scores, treating the dependent variable as binary: the firms in the treated group (firms with higher policy risk) are assigned as 1, and those in the control group (firms with lower policy risk) are assigned as 0. All the co-variables used in this model are the same as our control variables in the baseline regression model. Next, we use nearest neighbor matching with a caliper of 1% to select a matched control group for the treated group.

The regression results for the matched sample are presented in

Table 8. Endogeneity test: propensity score matching (PSM).

	Amihud	HLS
Variables	(1)	(2)
PRI	−0.0537 **	−0.0439 ***
	(−2.25)	(−3.16)
RVOL	−0.0541 ***	−0.0174 ***
	(−8.49)	(−7.89)
IPRC	−0.0127 ***	−0.0020 *
	(−3.34)	(−1.92)
Size	0.0057 ***	0.0018 ***
	(12.79)	(10.45)
ROA	0.0023 ***	0.0010 ***
	(5.07)	(4.08)
Tobin’s_Q	0.0051 *	0.0054 ***
	(1.90)	(5.61)
LEV	−0.0001 **	−0.0001 ***
	(−2.07)	(−4.16)
R&D	0.0382 ***	0.0105 ***
	(3.15)	(4.01)
IO	0.0147 ***	0.0014
	(4.22)	(1.37)
Constant	−0.0677 ***	−0.0137 ***
	(−9.64)	(−4.36)
Observations	6044	6044
Year FE	Yes	Yes
Industry FE	Yes	Yes
Adj. R-squared	0.624	0.619

Notes: The description of variables is available in Table A1 (Appendix A). The t-statistics reported in parentheses are based on standard errors clustered at the firm level. *** p < 0.01, ** p < 0.05, * p < 0.1.

. Columns (1) and (2) show that PRI continues to negatively affect stock liquidity. The PSM approach isolates the effect of policy risk on stock liquidity and controls sample selection bias, strengthening the validity of our analysis and supporting the hypothesized relationship.

The balance test results are presented in

Table A2. Balance tests.

Panel A: Balance Test (Pre-Matching)
		Mean		t-test
Variables	Treated	Control	t	p > t
RVOL	0.232	0.213	9.85	0.000
IPRC	0.078	0.061	2.10	0.036
Size	9.155	9.697	−14.61	0.000
ROA	3.712	6.019	−13.01	0.000
Tobin’s_Q	1.674	1.674	0.00	1.000
LEV	3.031	0.863	16.20	0.000
R&D	0.005	0.004	3.15	0.002
IO	0.121	0.160	−14.61	0.000
Panel B: Balance Test (Post-Matching)
		Mean		t-test
Variables	Treated	Control	t	p > t
RVOL	0.219	0.221	−0.93	0.352
IPRC	0.052	0.054	−0.36	0.716
Size	9.373	9.292	1.82	0.068
ROA	5.502	5.288	1.06	0.289
Tobin’s_Q	1.666	1.675	−0.27	0.789
LEV	1.338	0.983	4.30	0.000
R&D	0.004	0.004	−0.55	0.584
IO	0.139	0.136	0.62	0.534

Notes: The description of variables is available in Table A1.

(Appendix A). Before matching, all covariates have p-values less than 0.05 in the t-test results. However, the standard deviation of each covariate drops dramatically, and the p values of the t-tests are larger than 0.05 after PSM. This outcome is further illustrated in Figure 1, which shows the kernel density function (KDF) distributions of propensity scores before and after matching. Post-matching, the distributions of the treated and control groups almost perfectly overlap. This suggests that post-matching, there are no significant differences between the treated and control groups, which implies that PSM has successfully balanced the covariates between groups.

5.2. Two-Stage Least Squares Instrumental Variable Approach

To further solve the endogeneity, particularly the reverse causality and measurement error issues, we employ the two-stage least squares (2 SLS IV) regression analysis. Following the previous literature (Hasan et al., 2022; Xu et al., 2023), we use the average PRI of each firm with the same industry (two-digit NIC codes) in each year (Mean_PRI_Industry) as our instrumental variable. We employ these instruments based on prior research evidence suggesting that firms in the same industry exhibit common corporate policy risks. For example, Leary and Roberts (2014), Machokoto et al. (2021), and Chen and Ma (2017) contend that a firm’s corporate risk is not only a function of its own policy decisions but also a function of its industry peers’ actions.

Table 9. Endogeneity test: two-stage least squares instrumental variable approach.

	First-Stage	Second-Stage
	PRI	Amihud	HLS
Variables	(1)	(2)	(3)
PRI_Industry_Avg	0.7617 ***
	(12.98)
PRI_Instrumented		−0.1835 **	−0.1426 ***
		(−2.25)	(−3.16)
RVOL	0.0041 ***	−0.0591 ***	−0.0142 ***
	(2.85)	(−11.43)	(−7.72)
IPRC	0.0013 *	−0.0040 *	−0.0012 ***
	(1.78)	(−1.88)	(−2.63)
Size	−0.0006 **	0.0057 ***	0.0017 ***
	(−2.22)	(13.26)	(11.09)
ROA	−0.0004 ***	0.0019 ***	0.0003 **
	(−8.97)	(4.03)	(2.25)
Tobin’s_Q	−0.0011 ***	0.0048 *	0.0066 ***
	(−4.63)	(1.91)	(7.42)
LEV	0.0002 ***	−0.0002 ***	−0.0001 ***
	(7.69)	(−3.31)	(−2.73)
R&D	−0.0079	0.0479 ***	0.0110 ***
	(−0.71)	(3.33)	(4.33)
IO	−0.0085 ***	0.0171 ***	0.0009
	(−4.47)	(5.10)	(1.04)
Constant	0.0473 ***	−0.0428 *	−0.0041
	(4.37)	(−1.78)	(−0.47)
Observations	9320	9320	9320
Year FE	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes
Underidentification test:Anderson canon. corr. LM statistic	413.697
Chi-sq. (1) p-value	0.000
Weak identification test:Cragg–Donald Wald F statistic	429.536
Stock–Yogo critical value [at 10 percent]	16.38
Adj. R-squared	0.321	0.624	0.628

Notes: The description of variables is available in Table A1 (Appendix A). The t-statistics reported in parentheses are based on standard errors clustered at the firm level. *** p < 0.01, ** p < 0.05, * p < 0.1.

reports the 2SLS-IV estimation results. Column (1) shows the first-stage regression, where the PRI serves as the dependent variable. Consistent with expectations, the coefficient on the Mean_PRI_Industry is positive and highly significant at the 1% level, indicating the significant association between the instrument and the endogenous variable. In the second-stage regressions, displayed in columns (2) and (3), the instrumented corporate policy risk index (PRI_Instrumented) is positively and significantly associated with both liquidity measures, reinforcing a robust link between PRI and stock liquidity.

5.3. Two-Step System GMM

To further address endogeneity in our analysis of corporate policy risk (PRI) and stock liquidity, we apply the two-step system GMM estimator, following Arellano and Bover (1995) and Blundell and Bond (1998). System GMM is particularly appropriate here because it effectively manages the three primary sources of endogeneity: simultaneity, unobservable heterogeneity, and dynamic endogeneity (Roodman, 2009; Wooldridge, 2010). For instance, managers of highly liquid firms may be more conservative in policy risk, take on less risky projects, and be more responsive to pay for performance when they hold undiversified wealth (Coles et al., 2006; Fang et al., 2014).

Table 10. Endogeneity test: two-step system GMM.

	Amihud	HLS
Variables	(1)	(2)
PRI	−0.1478 **	−0.0555 ***
	(−2.23)	(−2.69)
RVOL	−0.0146 ***	−0.0145 ***
	(−2.72)	(−6.24)
IPRC	−0.0065 *	−0.0053 ***
	(−1.91)	(−5.06)
Size	0.0027 ***	0.0012 ***
	(5.31)	(6.43)
ROA	0.0010 **	0.0003 *
	(2.00)	(1.87)
Tobin’s_Q	0.0060 *	0.0059 ***
	(1.68)	(5.14)
LEV	−0.0002 **	−0.0001 ***
	(−2.25)	(−3.69)
R&D	0.0327 **	0.0138 *
	(2.04)	(1.72)
IO	0.0041	0.0016
	(0.81)	(1.43)
Lag. dependent	0.4059 ***	0.1430 ***
	(4.80)	(3.74)
Constant	−0.0170 *	−0.0148 ***
	(−1.88)	(−2.69)
Observations	8854	8854
Year FE	Yes	Yes
Firm FE	Yes	Yes
AR (1)	0.001	0.001
AR (2)	0.366	0.264
Hansen	0.195	0.272

Notes: The description of variables is available in Table A1 (Appendix A). The t-statistics reported in parentheses are based on standard errors clustered at the firm level. *** p < 0.01, ** p < 0.05, * p < 0.1.

provides the results of the two-step system GMM to confirm the hypothesized negative relationship between stock liquidity and PRI. The coefficients of PRI are significant and negative in columns (1) and (2) for both liquidity measures, Amihud and HLS. These results are consistent with the baseline regression results in Table 4.

6. Concluding Remarks and Policy Implications

The objective of this study is to examine the association between corporate policy risk and stock liquidity by constructing the novel corporate policy riskiness index (PRI), an aggregate measure of the risk associated with the different corporate policies. Using a robust sample of 466 non-financial firms listed on the National Stock Exchange of India (NSE), we find a significant negative association between PRI and stock liquidity. These results suggest that the riskiness of corporate policies (e.g., investment, financial, diversification, and cash management) is an important determinant of stock market outcomes. The results also confirm further that corporate policy risk has a greater adverse effect on stock liquidity for firms with opaque information environments, particularly for small-size, highly volatile, and higher complexity reporting firms. Additionally, we employ several robust endogeneity tests, including propensity score matching (PSM), instrumental variable regression (2 SLS IV), and two-step system generalized method of moments (System GMM), which are robust to our baseline conclusions.

This research makes several significant contributions to the literature. First, it bridges a gap in the intersection of corporate finance and market microstructure by providing an integrated, multidimensional measure of corporate policy risk through the construction of the PRI. This index goes beyond conventional one-dimensional approaches to incorporate the interdependence of policy decisions and their cumulative effect on firms’ risk profiles. Second, the study contributes to the theoretical understanding of how risk preferences affect managerial decisions toward different policies and how these shape firm-level outcomes, particularly stock liquidity. It adds to the growing literature on sustainable financial management by linking management practices to market behavior.

The results are essential for policy implications. The study shows that it is crucial for managers to adopt balanced and transparent corporate policies to mitigate risk-induced uncertainty and build investor trust. Investors can use this novel PRI to evaluate firms’ aggregate risk profiles, which enables them to make more informed decisions. Policymakers can use these insights to develop frameworks that foster good corporate governance practices, thereby improving market stability and liquidity. While this study provides valuable insights, it has certain limitations. The findings are specific to NSE-listed firms in India, which may limit their generalizability to other emerging markets. The corporate policy risk index (PRI) used captures overall corporate policy risk but does not fully account for dynamic changes in managerial risk perception, behavioral factors, or varying market conditions. Future research could refine this framework to incorporate these elements for a more nuanced understanding. Additionally, applying this model to global markets or different economic and regulatory contexts would further validate its robustness and enhance its relevance. These extensions would help uncover broader implications for corporate policy risk and its impact on stock liquidity across diverse environments.