Institutional Investors, Dividend Policy, and Idiosyncratic Volatility: Evidence from European Equity Markets

Abstract

This paper investigates the relationship between institutional ownership and firm-level idiosyncratic volatility across European equity markets, with a particular focus on the moderating role of dividend policy. Using a sample of STOXX Europe 600 constituents from 2005 to 2025, we estimate idiosyncratic volatility via the Fama-French three-factor model and employ fixed-effects regressions with clustered standard errors. Our empirical results reveal a positive and statistically significant association between institutional ownership and idiosyncratic volatility, suggesting a destabilizing rather than stabilizing role in European markets. This volatility-enhancing effect is significantly more pronounced among dividend-paying firms and is primarily driven by transient institutional investors with high portfolio turnover. Furthermore, we find that: (1) larger firm size (market capitalization) and higher leverage (debt-to-capital ratio) are positively associated with heightened volatility; (2) growth-oriented firms (high market-to-book ratios) exhibit increased volatility, particularly among non-dividend payers; and (3) higher profitability (ROE) and favorable analyst coverage (buy recommendations) act as stabilizers, reducing idiosyncratic risk. These findings persist in both contemporaneous and lagged specifications. This study contributes to the literature by identifying dividend policy as a key channel through which institutional trading behavior amplifies firm-specific risk, providing novel evidence on the asset class effect within major European benchmark indices.

1. Introduction

Institutional investors have exerted considerable influence over corporate decision-making, price discovery, and market liquidity, emerging as the dominant actors in modern capital markets. Global equity markets are increasingly shaped by the expanding share of these sophisticated entities; while institutional ownership saw a drastic increase in the late 20th century (Khorana et al., 2005), this trend has accelerated and evolved over the last two decades. Recent evidence suggests that institutional holdings now represent the vast majority of equity value in developed markets, frequently exceeding 80% in major indices (Ben-David et al., 2021; Aretz & Arisoy, 2023). This ongoing structural shift carries profound implications for volatility transmission. Despite advancements in trading technology and information flows, the behavior of institutional investors, which ranges from long-term monitoring to rapid high-turnover trading, continues to either amplify or mitigate firm-specific and aggregate market risks in the current market environment (Bushee, 2004; Ferreira & Matos, 2008; Boehmer & Kelley, 2009; Glossner et al., 2025).

Even though previous research focused on the relationship between institutional investors and stock market volatility, the results remain contested, with both stabilizing and destabilizing effects. The stabilizing effects of institutional investors on stock market volatility are due to the improved information and monitoring management function, as well as reducing mispricing (Hartzell & Starks, 2003; Chen et al., 2007). Therefore, this point of view supports the fact that higher institutional ownership is associated with lower idiosyncratic volatility, considering that informed trading may diminish price noise and speculative activity (Boehmer & Kelley, 2009). On the contrary, another strand of the literature supports the destabilizing role of institutional investors, considering the herding behavior, correlated portfolio rebalancing that can induce an asset class risk, or a focus on short-term performance (Sias, 2004; Basak & Pavlova, 2013; Ben-David et al., 2021). Furthermore, under specific market conditions such as high uncertainty, institutional investors were empirically associated with increased volatility and contagion effects across equities (Glossner et al., 2025).

The heterogeneity regarding the impact of institutional investors on stock market volatility can be explained by the different categories of institutional investors. Divergent implications for volatility were identified based on the type of institutional investors. Transient investors are commonly engaged in momentum trading and pursue short-term returns, while dedicated investors provide stability by maintaining their stake in the company for a long period of time (Bushee, 2004; Ferreira & Matos, 2008). Furthermore, large trades conducted by institutional investors can amplify stock market volatility, and the impact could be persistent, especially in environments with low liquidity (Gabaix et al., 2006). Similar findings were obtained by Koch et al. (2016) and Lou (2012), which emphasize that mutual fund capital flows and mutual fund ownership can generate shocks across portfolios that, in turn, generate idiosyncratic volatility.

Another dimension that is insufficiently explored is the link between institutional investors and the payout policy, which concerns the impact on stock market volatility. The dividend distribution can either amplify or diminish stock market volatility based on the manner in which it shapes trading behavior and investor ownership. Prior studies argue that the volatility is increased by institutional investors for dividend-paying companies; yet, a negative relationship exists between institutional investors and volatility for companies that do not pay dividends (Rubin & Smith, 2009). This asymmetric behavior could potentially be explained by the trading rules and preferences of institutional investors, namely that institutional investors prefer firms that pay dividends (Grinstein & Michaely, 2005). Dividend-paying policies were associated with prudence and stability (Del Guercio & Hawkins, 1999). Therefore, it is expected that institutional investors should respect strict and prudent regulations with the clients’ money, therefore preferring equities with large market capitalization and good governance as well as financial stability (Ferreira & Matos, 2008). However, within dividend-paying companies, the trading exhibited by institutional investors may lead to increased turnover and stock price fluctuations due to benchmarking practices and portfolio rebalancing (Karpoff, 1987; Basak & Pavlova, 2013). A stronger association was identified between institutional investors and the turnover for the companies that pay dividends than for those that do not pay dividends (Azzam, 2010). Moreover, recent empirical evidence provided by Crane et al. (2016) shows that exogenous increases in institutional ownership cause economically significant rises in dividend payouts. Consequently, prior evidence from the literature provides support to the conjecture that the relationship between institutional investors and stock market volatility may differ significantly between dividend-paying companies and non-dividend companies.

The European stock market provides a unique setting to examine the link between institutional investors and stock market volatility, which has been less studied than the American market. Furthermore, the European markets have varying institutional participation rates and varying degrees of investor protection. For instance, the regulatory frameworks such as MiFID II and the EU Shareholder Rights Directive have impacted governance, disclosure, and trading behavior, which, in turn, may affect the channel through which institutional investors impact idiosyncratic volatility.

Considering the theoretical background, this study analyzes how institutional investors affect idiosyncratic volatility across the European stock markets and whether the dividend policy influences this relationship. We compute idiosyncratic volatility as the standard deviation of the residuals from Fama-French three-factor models in line with prior studies (Ang et al., 2006; Fu, 2009). Our dataset spans the 2005–2025 period, and we use a panel of constituents from the STOXX Europe 600 stock index. We employ firm-level fixed effects and standard errors that are clustered at the firm level in our regressions to account for within-firm correlation over time. The regressions are applied both contemporaneously and with a lag based on the conjecture that institutional investors can have a lagged impact on idiosyncratic volatility.

Our contribution is fourfold. First, we bring Europe to the forefront by assembling a balanced panel of STOXX Europe 600 companies (2005–2025), a setting characterized by rapidly increasing but heterogenous institutional ownership across countries and sectors; we also differentiate investor types, and we include a transient ownership proxy in the spirit of Bushee (2004). Second, we examine how the dividend payout policy conditions the relationship between institutional ownership and firm-level idiosyncratic volatility, an aspect largely overlooked in prior work, by contrasting dividend-paying and non-paying firms. Third, we provide new empirical evidence to the debate regarding the stabilizing versus destabilizing role of institutional investors by estimating firm fixed-effects models with clustered errors and controls, isolating within-firm variation. Finally, we explicitly address the time-varying nature of idiosyncratic volatility (Fu, 2009; Herskovic et al., 2020) by reporting both contemporaneous and lagged specifications, showing that the main inferences do not rely on timing.

The remainder of the paper is structured as follows. Section 2 analyzes the theoretical and empirical literature on institutional ownership, volatility, and dividend policy. Section 3 describes the data, variable definitions, and econometric methodology. Section 4 presents and discusses the empirical results. Section 5 concludes the paper.

2. Literature Review

2.1. The Stabilizing vs. Destabilizing Role of Institutional Ownership

Not only do institutional investors have an important role in developed capital markets, but they also have an important role in developing markets as well (Khorana et al., 2005). Such importance could be derived from the influence of institutional ownership on corporate governance, as well as the resources and abilities institutional investors possess, often accounting for the majority of trading volume and liquidity provision. Previous evidence from the literature suggests that institutional investors are more informed than retail investors (Lin et al., 2007). Furthermore, another key factor is that the contemporaneous daily return patterns can be positively explained mostly by the institutional trades, considering the previous day’s excess returns (Griffin et al., 2003).

Empirical evidence on the impact of institutional investors on stock market volatility remains mixed, with both stabilizing and destabilizing effects. The results are mixed regarding the impact of institutional investors on volatility, considering the heterogeneity of institutional investors that affects the investment horizon, as well as generating constraints on the trading activity. On the one hand, the destabilizing effects of institutional investors were identified. Large institutional investors were associated with increased stock market volatility and noise concerning pricing, as well as fragility during a stock market crisis (Ben-David et al., 2021). The channel through which institutional investors might influence stock market volatility is highlighted in the previous study based on the mechanism relating to the granularity of large institutional investors, by which such institutions trade in larger volumes, with a greater impact on the stock prices. This granularity of large institutional investors is based on focused subunits within a company, which reduces diversification against idiosyncratic shocks and can amplify market fluctuations. On a similar note, different categories of institutional investors were identified in the literature based on trading behaviors and objectives, including transient investors, who are focused on the short-term performance and tend to increase stock volatility, while quasi-indexers and dedicated holders are oriented towards long-term objectives and are associated with lower volatility (Bushee, 2004).

Another mechanism through which institutional trading can lead to stock market volatility is the book-to-market effect, namely, institutional investors that respond to intangible information by buying shares when information is positive and selling when it is negative, generating price overreactions (Jiang, 2010). This effect is more pronounced with stocks that are heavily traded by institutions, while this effect is absent from shares that are only moderately traded by institutions. Therefore, institutional traders may contribute to market inefficiencies and greater volatility, leading to price overreactions and value premiums.

A positive contemporaneous relationship between institutional investors and stock market volatility was suggested by the empirical evidence from the literature. A theoretical framework was established that posits that large institutional investors can induce significant price shocks and generate excess volatility due to the trading volumes, even in the absence of a change in the fundamentals of the equities (Gabaix et al., 2006). Transactions in large volumes by institutional investors could exert a disproportionate price impact and lead to firm volatility. Recent evidence confirms that, as institutional funds have grown larger, this ‘granularity’ has become a more significant driver of non-diversifiable idiosyncratic shocks (Ben-David et al., 2021). Similarly, volatility spikes can be triggered by institutional ownership, as Sias (1996) identified, based on the institutional investors’ active trades and portfolio rebalancing, in which they adjust to new market information.

The amplifying mechanism of institutional investors during the crisis periods was reiterated in the literature by Glossner et al. (2025), who identified that, during the COVID-19 stock market crash, companies from the United States with higher institutional ownership had an inferior performance than those with lower institutional ownership and provided two possible arguments that could explain this effect. Firstly, institutions reacted quickly to the market conditions and withdrew capital rapidly from equities. Secondly, institutional investors shifted their investments towards assets that were more resilient to the pandemic shock. Therefore, the institutional investors’ behavior during tail risk could amplify price crashes during extreme market conditions. Similarly, it was identified that institutional investors’ sell orders may decrease the prices below the true value during negative market events, considering that institutional investors were found to react more strongly than retail investors to abnormal stock market returns, with the magnitude of the reaction and stock turnover depending on institutional ownership (Dennis & Strickland, 2002). Furthermore, a positive association was identified between institutional ownership and stock market volatility at quarterly frequency, after controlling for the company’s size, which could be related to the reduction in the informativeness of prices before announcing earnings (Potter, 1992).

Another factor that could explain the relationship between institutional ownership and stock volatility amplification is herding behavior. Even though it is reasonable to assume that institutional investors could be more sophisticated than retail investors, empirical evidence suggests the herding behavior of institutional trades. More specifically, Sias (2004) identified that institutional investors exhibit herding behavior, namely that the current demand for an equity is positively associated with the demand for the same equity from the previous quarter. This effect is not explained by the momentum trades but is more likely explained by institutional investors that follow other institutional investors and demand the same securities. While herding was traditionally attributed to human fund managers following the investment of peers, recent studies suggest that algorithmic and high-frequency trading may accelerate these synchronized movements in the modern era (Jarrow & Protter, 2012; Breckenfelder, 2024).

A similar mechanism that reinforces herding dynamics is related to the tendency of institutional investors to assess the performance of their assets against a benchmark of equity indices. The preference of institutional investors for index constituents is increasing demand for those assets and is putting price pressure on their prices (Basak & Pavlova, 2013). Not only does the volatility increase for index constituents as compared to non-indexed equities, but it is also creating an asset class effect, namely, a stronger co-movement is created among stock index constituents. On a related note, Vayanos and Woolley (2013) suggest that institutional trades can generate momentum and price reversals in assets that are characterized by high idiosyncratic risk while also generating amplification effects and lead–lag dynamics.

Furthermore, the volatility generated by institutional investors could be related to the career concerns of fund managers, considering that managers could be fired by investors based on past performance and that some managers have better knowledge of default risks (Guerrieri & Kondor, 2012). A counter-cyclical reputation premium is created, namely, when the default risks are higher, then the bond returns are increased to compensate for the managers’ risk of being fired. Subsequently, the fluctuations related to the default risk are increasing market volatility, therefore linking institutional investors to amplified pricing volatility.

This section of the literature establishes that, while institutions have resources, their granularity, herding behavior, and career concerns often lead to price pressure and excess volatility. Based on these arguments, we state our first hypothesis:

Hypothesis 1 (H₁):

Institutional ownership is positively associated with idiosyncratic stock return volatility.

2.2. Investor Heterogeneity and the Stabilizing Role of Dedicated Holders

On the other hand, a growing number of studies argue that institutional investors can stabilize the stock market and contribute to market stability through several channels. Firstly, evidence was identified in the literature for the importance of attracting long-term investors, which, in turn, tends to lead to enhanced company value, especially for companies that are characterized by informational asymmetry (C. Y. Chung, 2025). In contrast to the short-term or transient investors that favor immediate returns, long-term institutional investors are more likely to be monitoring the company and to enhance corporate performance, which subsequently leads to value creation (Chen et al., 2007). Previous research highlights the active role of institutional ownership in corporate governance by direct negotiations and having proposals from shareholders during annual meetings (Smith, 1996; Carleton et al., 1998; Del Guercio & Hawkins, 1999). More recent studies emphasize that engaged or ESG-oriented institutional blockholders contribute to stronger governance and improved disclosure, mechanisms through which firm-specific risk may be reduced in modern markets (Appel et al., 2016; Dyck et al., 2019).

Therefore, long-term institutional investors were associated with value creation and stability, as mentioned earlier, while short-term investors were associated with negative consequences for shareholders, such as weaker corporate monitoring and value reduction acquisitions, or bargaining practices by the management for personal benefits (Gaspar et al., 2005). The monitoring activity of institutional investors discourages risky behaviors that could affect the company’s reputation (Aghion et al., 2013).

Moreover, there are other channels through which long-term institutional investors can influence stock volatility. Such long-term investors can influence managers to pursue strategies that aim to obtain long-term value, rather than to reap short-term benefits. For instance, long-term-oriented investors may be more prone to invest in projects that generate long-term returns and to maintain capital reserves (Harford et al., 2018), while companies that are owned by such investors (e.g., private equity funds with longer investment horizons) are more likely to be more innovative, as reflected by the number of patents owned (Barrot, 2017).

Other benefits that have been documented in the literature related to the influence of long-term institutional investors on corporate decisions highlight the governance role of institutional investors, which shape companies’ policies. Prior research indicates that large institutional ownership can reduce practices such as earnings management, namely, managers who report lower or higher earnings based on the desired level (R. Chung et al., 2002), and can contribute to the alignment of the pay level of the executive with actual performance (Hartzell & Starks, 2003). These large institutional investors may also influence the compensation structures depending on ownership levels and preferences (Janakiraman et al., 2010). Evidence was identified that institutional investors are affected by the payout policy and tend to favor dividend-paying companies over the companies that pay fewer dividends, as well as firms that engage in regular stock repurchases (Grinstein & Michaely, 2005). Similarly, the thorough monitoring of long-term institutions improves the financial fundamentals, measured through operational efficiency and improved F-scores, while transient investors trade on this information (C. Y. Chung et al., 2015).

Furthermore, another benefit of foreign institutional investors is the facilitation of cross-border mergers and acquisitions by reducing information asymmetries and transaction costs, with these benefits being more pronounced in environments with a weak governance structure (Ferreira et al., 2010). Furthermore, institutional investors may drive companies to be more active about corporate social responsibility and more engaged in sustainable activities. Previous studies identified that companies with stronger corporate social responsibility performance experience lower equity financing costs (El Ghoul et al., 2011). The effects of institutional ownership mentioned are just a few of the many areas of influence that institutional ownership exerts. Collectively, these factors may contribute to the reduction in idiosyncratic volatility and increase transparency, facilitating greater stock market stability.

Another argument that supports the stabilizing effect of institutional ownership on stock markets is the improvement in price efficiency after institutional trades. Studies argue that institutional ownership improves the efficiency of stock market prices by influencing the nature of the return distribution, such as skewness and kurtosis. Evidence supports a negative correlation between institutional investors and volatility of equity returns, measured through standard deviation, as well as an inverse relationship with kurtosis and skewness (Aggarwal & Rao, 1990). Therefore, institutional investors play an essential role in smoothing equity price fluctuation. Similarly, a more informational efficient pricing due to the institutional ownership was identified by Boehmer and Kelley (2009), which was measured as deviation from a random walk. This effect remains robust after controlling for factors that could influence the relationship, such as liquidity, analyst coverage, and short-selling activity. The improvement in efficiency of asset prices is present for both active and passive trades made by institutional investors. Other potential benefits exerted by institutional investors could be greater management transparency, improved liquidity, and more analyst monitoring, leading to lower informational asymmetry and lower transaction costs (Boone & White, 2015).

The literature suggests that the impact of institutions depends largely on their investment horizon. While long-term monitoring may provide stability, short-term trading behaviors typically drive idiosyncratic risk. Therefore, we suggest the following:

Hypothesis 2 (H₂):

The volatility-enhancing effect of institutional ownership is primarily driven by transient institutional investors.

2.3. Market Efficiency and the Moderating Channel of Dividend Policy

Another strand of papers from the literature examines the preference of institutional investors for equities from the point of view of volatility. Interestingly, evidence was identified that mutual funds’ equity holdings are averse to low idiosyncratic volatility stocks and prefer equities that are more visible and with lower transaction costs (Falkenstein, 1996). Furthermore, the demand for large capitalization stocks by institutional investors was reiterated by Gompers and Metrick (2001), indicating evidence that, within 16 years, the presence of institutional investors nearly doubled, which could explain the appreciation of the large capitalization stocks as opposed to smaller companies. Therefore, the increased demand for large capitalization stocks and visible stocks may drive up the volatility for those equities, especially if they are part of a benchmark index, as identified earlier by Basak and Pavlova (2013). The trading of short-term institutions is forecasting future returns, and this effect is more pronounced for small and growth equities, while the trading conducted by long-term institutions does not predict returns, reflecting that short-term institutional traders are acting more actively to exploit the informational advantage they possess (Yan & Zhang, 2009).

A key factor that could be influencing the relationship between institutional ownership and stock market volatility, apart from the types of institutional investors, could be the dividend policy. The volatility could be exacerbated or mitigated depending on the dividend policy, as evidence from the literature argues. Previous empirical results suggest that institutional ownership is positively associated with volatility among dividend-paying companies and negatively correlated with volatility for non-dividend firms (Rubin & Smith, 2009). The increase in volatility is due to the trading behavior exhibited for institutional ownership, and the results indicate an interaction between the preference for low-volatility equities of institutional traders. Even though non-dividend-paying companies are prevalent in the financial markets, financial institutions avoid investing in non-dividend firms (Grinstein & Michaely, 2005). One hypothesis through which institutional investors tend to increase volatility for dividend-paying companies is based on turnover, namely, institutional investors may be more prone to portfolio rebalancing than retail investors, and such portfolio turnover is associated with increased volatility (Karpoff, 1987). Institutional investors may herd for the constituents of an equity index and benchmark the performance of the funds they manage against an equity index, therefore driving the demand up for such securities (Basak & Pavlova, 2013). Other factors, such as capital inflows to a portfolio managed by an investment fund or withdrawal, may influence higher turnover for institutional investors, indicating a stronger association between institutional investors and turnover for the companies that pay dividends than for those that do not pay dividends (Azzam, 2010).

Institutional preferences for large-cap, visible, and dividend-paying stocks often lead to concentrated trading activity in those securities, potentially amplifying volatility. Based on these arguments, we suggest the following:

Hypothesis 3 (H₃):

The positive relationship between institutional ownership and idiosyncratic volatility is more pronounced in dividend-paying firms than in non-dividend-paying firms.

3. Data and Methodology

The link between institutional ownership and stock market volatility is examined in this study by focusing on European equity markets, focusing on a dynamic time-varying risk. Our sample includes the constituents of the Eurostoxx 600 Index, which includes 600 companies listed on the stock market from Eurozone countries, and we collect the daily equity prices that span from September 2005 to July 2025. All the variables were downloaded from the Refinitiv Workspace database, apart from the Fama-French three-factor model for European markets, which was sourced from Kenneth French Data Library (French, 2025).

Firstly, we compute the monthly logarithmic returns for each firm:

$$R_{i,t}=\ln ({\displaystyle \frac{P_{i, t}}{P_{i, t-1}}})$$

(1)

where P_i,t represents the monthly end price of stock i at time t. To account for the risk-free rate, we calculate excess returns by subtracting the one-month risk-free rate (R_f,t) from the monthly stock returns.

$${ExR}_{i,t}=R_{i,t}-R_{f,t}$$

(2)

To ensure methodological consistency with the Fama-French framework, the one-month risk-free rate (R_f,t) was obtained directly from the Kenneth R. French Data Library. This series is constructed using the US One-Month Treasury Bill rate, sourced from Ibbotson Associates for data through May 2024 and transitioned to the ICE BofA US One-Month Treasury Bill Index thereafter. This rate serves as the standard benchmark for calculating excess returns (R_m − R_f) within the Fama-French international datasets.

The procedure is applied for the entire sample that consists of the 600 companies using a looped procedure in Stata 18, which ensures consistency of the application for all the observations from our sample. We decompose the stock returns and isolate firm-specific volatility by employing the Fama-French three-factor model for European equities:

$${ExR}_{i,t}={\alpha }_i+{\beta }_{i,MKT}·{MKT}_t+{\beta }_{i,SMB}·{SMB}_t+{\beta }_{i,HML}·{HML}_t+{ϵ}_{i,t}$$

(3)

where MKT_t represents the excess market return, SMB_t captures the size effect, HML_t captures the value effect, and ϵ_i,t represents the idiosyncratic component of the return. The data for the Fama-French three-factor model are aligned to the monthly frequency of the stock returns.

The standard deviation of the residuals (ϵ_i,t) of the Fama-French three-factor regression represents the idiosyncratic volatility. To capture the dynamic nature, we implement a 12-month rolling window procedure, and we compute the standard deviation of the residuals within each window. The results yield a monthly time series of idiosyncratic volatility for each company, which enables us to study time-varying risk.

Furthermore, considering that a large share of the explanatory variables is disclosed by companies at a quarterly frequency (e.g., institutional ownership, leverage, profitability, market-to-book ratio, analyst recommendations), we align the returns series with the same quarterly temporal structure. For analyzing the stock returns at the quarterly frequency, we aggregate the monthly returns into quarterly returns to preserve the compounding effect of returns across time:

$$R_{i,t}^Q={\prod }_{m=1}^3(1+r_{i,m})-1$$

(4)

where $R_{i,t}^Q$ represents the compounded monthly return of firm i in quarter t, and r_i,m represents the monthly return of company i in quarter t. This procedure enables us to analyze the cumulative performance of each company over the quarter, rather than using an arithmetic average of monthly returns, which would not consider the compounding effects.

A similar procedure is conducted to establish the dependent variable from our analysis, that is, quarterly stock market idiosyncratic volatility, which is estimated using the daily stock market returns from within each quarter. This aggregation allows us to consistently estimate the relationship between institutional ownership and idiosyncratic volatility.

Furthermore, the relationship between institutional ownership and idiosyncratic volatility is investigated by employing panel data models with firm fixed effects. This enables us to control unobserved heterogeneity, which may influence simultaneously both ownership structures and volatility patterns. The models were applied for the entire sample, for the dividend-paying companies, and for the non-dividend companies. The contemporaneous baseline regression specification is as follows:

$${Volatility}_{i,t}=\alpha +\beta {InstOwn}_{i,t}+{\mathsf{\gamma }}^{\prime }{X}_{i,t}+{\mu }_i+{\tau }_t+{ϵ}_{i,t}$$

(5)

where ${Volatility}_{i,t}$ represents the quarterly idiosyncratic volatility of stock returns computed by using the Fama-French three-factor model, $X_{i,t}$ represents the set of control variables (i.e., firm size, market-to-book ratio, leverage, profitability, dividend yield, analyst buy recommendations, and turnover) while ${\mu }_i$ captures firm fixed effects, and ${\tau }_t$ represents time fixed effects, implemented as a vector of year-quarter dummy variables. Conceptually, these capture ‘time shocks’, which are exogenous macro-level events that affect all firms in the Eurostoxx 600 simultaneously. The economic rationale for including ${\tau }_t$ is to account for unobserved temporal factors such as the 2008 Global Financial Crisis, the Eurozone sovereign debt crisis, and the COVID-19 pandemic, as well as shifts in European Central Bank (ECB) monetary policy.

Econometrically, the inclusion of both firm fixed effects (${\mu }_i$) and time fixed effects (${\tau }_t$) constitutes a two-way fixed-effects (TWFE) framework. This specification provides a robust identification strategy in which the relationship between institutional ownership and volatility is estimated using only within-firm variation beyond common market trends. By absorbing these aggregate shocks, the coefficient β is not biased by overall market volatility or business cycle fluctuations. Moreover, as detailed in Appendix A, this approach mitigates concerns related to time-invariant unobserved heterogeneity, while the data exhibit sufficient within-firm variation in institutional holdings to limit the attenuation bias typically associated with fixed-effects estimators.

Moreover, the standard errors are clustered at the company level to account for heteroskedasticity and serial correlation. The firm size is measured with the log of the market value of the firm and represents one control variable that is commonly used in the literature, considering that smaller firms may be more concentrated concerning operations, while larger companies may be more diversified (Campbell et al., 2001). We account for a profitability measure using return on equity (ROE), considering that the uncertainty regarding the future profitability of a firm normally drives volatility (Wei & Zhang, 2006). Another determinant of stock market volatility is leverage, and we control for the use of debt, which could, in turn, affect the profitability of the company and drive uncertainty (Pastor & Veronesi, 2003). Leverage is measured as debt divided by the total value of the capital. We have included a dividend dummy to account for the dividend payout policy of the firm during that quarter (Rubin & Smith, 2009). Furthermore, we control for positive analyst coverage by including the number of buy recommendations from analysts. Turnover is included as a control variable since it is highly associated with idiosyncratic volatility (Kim et al., 2023). The growth prospects of a company may be proxied with a measure such as the market-to-book ratio (Cao et al., 2008). A similar control variable for growth is included, namely, the market value of equity divided by the book value.

We then estimate a specification that relates the idiosyncratic volatility to lagged institutional ownership by one quarter to address potential simultaneity bias and address reverse causality concerns, which were identified previously in the literature (Dey et al., 2024):

$${Volatility}_{i,t}=\alpha +\beta {InstOwn}_{i,t-1}+{\mathsf{\gamma }}^{\prime }{X}_{i,t-1}+{\mu }_i+{\tau }_t+{ϵ}_{i,t}$$

(6)

Similar to the previous model, this specification is estimated for the full sample, as well as separately for companies that pay dividends and those that do not. This subsample enables us to examine the impact of dividend policy on the effect of institutional ownership on volatility.

Furthermore, we also investigate whether short-term-orienteered investors (i.e., transient investors) drive volatility, considering the classification of investors from the literature (Bushee, 2004). We have proxied for transient ownership with a dummy variable (Transient) that takes the value one if institutional turnover is above the sample median. This specification is also applied to the entire sample, as well as the subsamples, depending on whether the company is a dividend payer. This model captures the additional effect of transient trading behavior beyond the level of institutional ownership and is specified as follows.

$${Volatility}_{i,t}=\alpha +{\beta }_1{Transient}_{i,t}+{\beta }_2{InstOwn}_{i,t}+{\mathsf{\gamma }}^{\prime }{X}_{i,t}+{\mu }_i+{\tau }_t+{ϵ}_{i,t}$$

(7)

The descriptive statistics for the variables in the empirical analysis are reported in

Table 1. Descriptive statistics.

	Full Sample (1)
	Mean	Standard deviation	Min	Max
Idiosyncratic Volatility	0.196	0.051	0.000	0.376
Market Value (natural log)	3.973	0.642	0.937	6.495
Market-to-Book	2.731	54.348	−6139.213	1090.020
Return on Equity (%)	16.997	52.402	−411.010	2409.860
Debt-to-Capital	41.253	124.717	−9762.300	7864.710
Turnover Ratio	81,813.088	309,924.788	0.018	23,302,854.648
Dividend Yield (%)	2.938	2.713	0.000	137.949
No. Buy Recommendations	0.457	0.511	0.000	6.890
Institutional Holdings (natural log)	2.042	0.655	0.511	4.304
Observations	45,117
	Non-Dividend Firms (2)
	Mean	Standard deviation	Min	Max
Idiosyncratic Volatility	0.195	0.050	0.000	0.372
Market Value (natural log)	3.775	0.634	0.937	5.480
Market-to-Book	−1.717	158.374	−6139.213	80.789
Return on Equity (%)	7.187	46.933	−406.610	1301.830
Debt-to-Capital	35.922	298.901	−9762.300	298.570
Turnover Ratio	57,980.087	216,447.769	0.018	12,479,931.079
Dividend Yield (%)	0.000	0.000	0.000	0.000
No. Buy Recommendations	0.447	0.543	0.000	4.682
Institutional Holdings (natural log)	2.154	0.698	0.511	4.304
Observations	4467
	Dividend Firms (3)
	Mean	Standard deviation	Min	Max
Idiosyncratic Volatility	0.196	0.051	0.056	0.376
Market Value (natural log)	3.996	0.639	1.295	6.495
Market-to-Book	3.238	20.701	−1188.398	1090.020
Return on Equity (%)	18.092	52.865	−411.010	2409.860
Debt-to-Capital	41.849	85.507	−353.220	7864.710
Turnover Ratio	84,599.704	318,957.186	6.592	23,302,854.648
Dividend Yield (%)	3.278	2.664	0.001	137.949
No. Buy Recommendations	0.458	0.507	0.000	6.890
Institutional Holdings (natural log)	2.028	0.648	0.511	4.220
Observations	40,650

Source: Authors’ own work.

. The full sample is presented in Panel (1), while Panels (2) and (3) illustrate the descriptive statistics for non-dividend companies and dividend-paying companies. The average idiosyncratic volatility is associated with moderate variation across companies, with values that range from 0 to 0.376. The average market capitalization and market-to-book ratio indicate a relative heterogeneity concerning firm size and growth opportunities, even though the companies are constituents of the same stock market index. The large standard deviation values for return and equity (ROE) and leverage (debt-to-capital) indicate the presence of companies with extreme negative and positive values. The dividend yield is 2.94% for the full sample and higher for dividend payers (3.28%), while it is zero for non-dividend-paying companies. Institutional ownership is measured as the natural logarithm of institutional holdings and has a slightly higher value for companies that are not dividend payers. Overall, the descriptive statistics indicate that heterogeneity is present among the sample regarding size, profitability, and market characteristics, even though the companies are constituents of the STOXX Europe 600 index.

To assess the appropriateness of the fixed-effects estimator, we decomposed the variation of institutional ownership using a within–between transformation. The results show a within-firm standard deviation of 0.507, which exceeds the between-firm standard deviation of 0.454. This substantial temporal variation within firms ensures that our two-way fixed-effects (TWFE) estimates are well identified and not subject to the attenuation bias or weak identification issues that typically affect time-invariant regressors in panel data.

4. Empirical Results and Discussions

Table 2. Explaining idiosyncratic volatility: contemporaneous specification.

	(1)	(2)	(3)
	All Firms	Non-Dividend Firms	Dividend Firms
Market Value	0.008 **	0.015	0.010 ***
	(0.003)	(0.011)	(0.004)
Market-to-Book	0.028 ***	0.031 ***	−0.082 *
	(0.005)	(0.001)	(0.044)
Return on Equity	−0.047 ***	−0.213 **	−0.019
	(0.011)	(0.092)	(0.014)
Debt-to-Capital	0.637 ***	0.430	0.659 ***
	(0.100)	(0.329)	(0.110)
Turnover Ratio	0.035 ***	0.048	0.033 ***
	(0.008)	(0.063)	(0.007)
Dividend Yield	0.003 ***	-	0.004 ***
	(0.001)		(0.001)
No. Buy Recommendations	−0.007 ***	−0.002	−0.008 ***
	(0.001)	(0.003)	(0.001)
Ln(Institutional Holdings)	0.003 *	0.001	0.004 **
	(0.002)	(0.006)	(0.002)
Observations	14,545	1621	12,924

Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01. Source: Authors’ own work.

reports the results of the contemporaneous specification of the fixed-effects regression that examines the relationship between institutional ownership and idiosyncratic volatility. All specifications control for firm characteristics commonly used in the literature that are linked to volatility, such as return on equity (ROE), market-to-book ratio, debt-to-capital ratio, turnover, market value (log), dividend yield, and analyst buy recommendations. The models include firm-level fixed effects and standard errors that are clustered at the firm level to account for within-firm correlation over time. The sample consists of STOXX 600 Europe constituents, containing 20 years of daily data, which were transformed into quarterly data (2005Q3–2025Q3). The response variable in all regressions is idiosyncratic volatility (computed by using the Fama-French three-factor model with European data). Results are reported for all firms (1), non-dividend firms (2), and dividend firms (3). The coefficients for the various independent variables were rescaled to ensure formatting consistency and to improve readability without affecting the statistical significance of the results. More specifically, the coefficients for turnover ratio are reported in millions, while market-to-book, return on equity, and debt-to-capital ratios are scaled by a factor of 10³.

The results depict significant relationships that are mostly in line with the theoretical expectations, while also revealing heterogeneity between dividend-paying companies and non-dividend-paying companies. The results indicate evidence of a positive and statistically significant relationship between institutional ownership and idiosyncratic volatility, which is consistent with the conjecture that institutional ownership may promote trading activities and contribute to the price discovery process, leading to firm-specific risk. The model applied to the full sample of companies indicates that the coefficient of institutional ownership is positive at the 10% threshold, suggesting that a one-unit increase in institutional holdings is associated with a significant spike in firm-level idiosyncratic volatility. This effect is stronger for the sample consisting only of dividend-paying companies, statistically significant at the 5% percent threshold, whereas, for the non-dividend-paying companies, the result is not statistically significant, and the coefficient is smaller, implying that the dividend policy could shape the relationship between institutional ownership and volatility. Overall, the results in Table 2 provide initial empirical support for H₁, indicating a general positive relationship between institutional presence and firm-specific risk. Furthermore, the findings align with H₃, as the volatility-enhancing effect is statistically significant and larger in magnitude for dividend-paying companies compared to their non-dividend-paying counterparts.

The positive contemporaneous relationship between the volatility of stock returns and institutional ownership was previously documented in the literature; prior evidence suggests that an increase in institutional ownership can trigger volatility spikes (DeVault et al., 2019). Institutional investors adhere to prudent regulations regarding the clients’ money, and the literature also documents their preference for companies with large market capitalization, stable financial fundamentals, and good governance (Ferreira & Matos, 2008). Such prudence is commonly associated with dividend-paying policies, which have historically been considered a signal of stability (Del Guercio & Hawkins, 1999).

Our results are reinforced by the empirical evidence from the literature that considers the behavior of institutional investors concerning benchmark indices. In such models, institutional investors compare the performance obtained to an index and prefer index constituents (Basak & Pavlova, 2013). This rebalancing mechanism is putting price pressure on the companies that are included in the stock market index and is amplifying the volatility compared to stocks that are not included in the market index. Furthermore, the same paper identified that this trading behavior generates excessive correlations among stocks that are included in the benchmark index, which reinforces an asset class effect.

Our result is partially supported by another paper in which institutional ownership is examined with respect to volatility and dividends. Rubin and Smith (2009) found that the correlation between institutional ownership and stock market volatility depends on the dividend policy of the company, namely, there is a positive relationship between institutional ownership and volatility for dividend-paying companies, while the relationship is a negative relationship between institutional ownership and stock market volatility for those firms that are not paying dividends. Another example of institutional investors that leads to stock market volatility is supported by a theory that indicates that excess volatility is due to large institutional investors in illiquid markets without considerable changes in financial performance (Gabaix et al., 2006). Institutional investors were identified to trigger momentum and reversal for high idiosyncratic risk assets as co-movement and lead–lag effect and amplification (Vayanos & Woolley, 2013).

Control variables are aligned with the previous literature. The log of market capitalization is statistically significant and has a positive coefficient, indicating that companies with larger market capitalization exhibit higher volatility. This result could be potentially explained by the state-dependent nature of volatility and stock co-movement, namely that conditional variances and correlations tend to increase after negative shocks, which can lead to heightened volatility for highly traded large capitalization companies (Engle & Ng, 1993; Cappiello et al., 2006). It was identified that shocks that lead to volatility were more persistent for larger companies rather than for smaller ones (Chelley-Steeley & Steeley, 1995). The market-to-book ratio is positively correlated with volatility, which is consistent with the hypothesis that growth explains the book–market value ratios (Fama & French, 1992). The return on equity represents a profitability measure and is negatively related to volatility, which is aligned with the rationale that more profitable companies have lower risk and stability concerning cash flows. On a similar note, leverage (debt-to-capital ratio) has a positive relationship with idiosyncratic volatility, reflecting the underlying financial risk. Furthermore, evidence was obtained that supports a positive relationship between trading turnover and volatility in line with theories linking speculative trading and liquidity to price modifications (Chordia et al., 2001). The analysts buy stock recommendations to reduce idiosyncratic volatility, indicating that favorable analyst coverage acts as a stabilizer for stock returns. Finally, a statistically significant and positive relationship was identified between dividend yield and stock market volatility, which may convey a sign about the outlook of the company.

As a robustness analysis to examine whether the relationship between institutional investors and idiosyncratic volatility is contemporaneous, we have applied a lagged specification model.

Table 3. Explaining idiosyncratic volatility: lagged specification.

	(1)	(2)	(3)
	All Firms	Non-Dividend Firms	Dividend Firms
Lagged Market Value (t − 1)	0.011 ***	0.014	0.013 ***
	(0.003)	(0.011)	(0.004)
Lagged Market-to-Book (t − 1)	0.024 ***	0.027 ***	−0.096 **
	(0.005)	(0.001)	(0.039)
Lagged Return on Equity (t − 1)	−0.046 ***	−0.175 *	−0.016
	(0.010)	(0.100)	(0.012)
Lagged Debt-to-Capital (t − 1)	0.675 ***	0.507	0.688 ***
	(0.097)	(0.313)	(0.108)
Lagged Turnover Ratio (t − 1)	0.048 ***	0.062	0.045 ***
	(0.014)	(0.052)	(0.013)
Lagged Dividend Yield (t − 1)	0.003 ***	0.002	0.003 ***
	(0.001)	(0.001)	(0.001)
Lagged No. Buy Recommendations (t − 1)	−0.006 ***	−0.005	−0.007 ***
	(0.001)	(0.003)	(0.001)
Lagged Ln(Institutional Holdings) (t − 1)	0.003 **	−0.000	0.004 **
	(0.002)	(0.006)	(0.002)
Observations	14,685	1602	13,083

Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01. Source: Authors’ own work.

reports the results of the model with lagged specification, where the idiosyncratic volatility in quarter t is explained by institutional ownership and firm characteristics at quarter t − 1. Therefore, this approach enables us to examine the predictive relationship between institutional ownership and future volatility while reducing the simultaneity concerns. To ensure consistency with the results presented in Table 2 and to ensure a clearer comparison of coefficients, the results have been rescaled for the following predictors: turnover ratio is expressed in millions (multiplied by 10⁶), while the coefficients for the following variables are expressed in thousands (multiplied by 1000): market-to-book, return on equity, and debt-to-capital.

The results illustrate that the lagged institutional holdings variable remains statistically significant and with a positive coefficient for the full sample and for dividend-paying firms, while the effect is not statistically significant for non-dividend-paying companies. The results reiterate the findings of the contemporaneous specification, indicating that institutional ownership is not only correlated with idiosyncratic volatility, but also predicts higher idiosyncratic volatility values, especially for dividend-paying companies. This empirical finding is consistent with DeVault et al. (2019), suggesting that volatility spikes can be induced by institutional trading, as well as the institutional investors’ preference for index or benchmark constituents that amplify volatility dynamics and generate an asset class correlation effect (Basak & Pavlova, 2013). The persistence of these results in the lagged specification confirms that institutional ownership serves as a predictor of future idiosyncratic volatility, further reinforcing the validity of H₁ and H₃ across different time horizons.

Furthermore, the control firm-level variables also display consistent effects that are aligned with the theoretical framework. Evidence obtained from the literature suggests that the firms are more vulnerable to idiosyncratic risk, and there is a positive cross-sectional association between volatility and institutional ownership (Xu & Malkiel, 2003). Furthermore, for companies that do not pay dividends, the market-to-book ratio is positively associated with idiosyncratic volatility and negatively associated with dividend-paying firms. This result suggests that the growth opportunities are driving the risk mainly in companies with a non-dividend payout policy. Similarly to the contemporaneous specification, the profitability proxy (ROE) is consistently reducing volatility, which could be related to the earnings-smoothing hypothesis, namely that managers are reporting higher than economic earnings to reduce the perceived volatility after a negative shock (Goel & Thakor, 2003). The debt-to-capital ratio proxies for leverage and increases volatility, indicating a shift in the capital structure choice that leads to risky financial behavior. Idiosyncratic volatility is positively predicted by lagged turnover, which is consistent with idiosyncratic volatility that is highly associated with turnover (Kim et al., 2023). Furthermore, higher trading intensity may represent a signal for market participants and new information arrival, which could lead to firm-specific risk. Dividend yield holds robust for this regression, indicating the positive relationship with volatility for dividend-paying companies. Therefore, a high dividend payout ratio does not insulate companies from institutional ownership-induced risk, but rather increases it. Volatility is negatively predicted by analyst recommendations, indicating that a stronger analyst coverage can temper idiosyncratic volatility, with the effect still being statistically significant after one quarter for the entire sample, as well as for dividend-paying companies. The stabilizing role of security analysts was highlighted previously in the literature (Song & Zheng, 2014).

Table 4. Explaining idiosyncratic volatility: the impact of transient institutional ownership.

	(1)	(2)	(3)
	All Firms	Non-Dividend Firms	Dividend Firms
Transient Ownership Dummy (Above-Median Turnover)	0.008 ***	0.004	0.008 ***
	(0.003)	(0.006)	(0.003)
Market Value	0.005	0.013	0.008 **
	(0.003)	(0.012)	(0.004)
Market-to-Book	0.029 ***	0.031 ***	−0.083 *
	(0.005)	(0.001)	(0.045)
Return on Equity	−0.047 ***	−0.212 **	−0.019
	(0.011)	(0.091)	(0.014)
Debt-to-Capital	0.630 ***	0.420	0.651 ***
	(0.100)	(0.333)	(0.110)
Turnover Ratio	0.033 ***	0.047	0.031 ***
	(0.008)	(0.064)	(0.007)
Dividend Yield	0.003 ***	-	0.004 ***
	(0.001)		(0.001)
No. Buy Recommendations	−0.007 ***	−0.002	−0.008 ***
	(0.001)	(0.003)	(0.001)
Ln(Institutional Holdings)	0.003 *	0.001	0.004 **
	(0.002)	(0.006)	(0.002)
Observations	14,545	1621	12,924

Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01. Source: Authors’ own work.

investigates whether the volatility effect of institutional ownership is heterogeneous depending on the type of institutional ownership based on the level of transience. The transient ownership is proxied with a dummy that equals one if the institutional ownership is associated with an above-median turnover. This classification is consistent with prior work in which institutional investors were grouped in multiple categories: transient investors, dedicated holders, and quasi-indexers (Bushee, 2004). The empirical evidence obtained suggests that transient institutional investors amplify idiosyncratic volatility. The transient dummy is highly statistically significant and positive across the entire sample, with a pronounced effect for dividend-paying firms. The effect is insignificant for the non-dividend-paying companies.

Transient institutions are characterized by shorter investment horizons and higher portfolio turnover. Prior studies identified that short-term institutions have an informational advantage and trade actively to generate superior returns with the information advantage that they possess (Yan & Zhang, 2009). This behavior may, in turn, exacerbate stock market volatility due to the responsiveness to short-term signals and trading strategies adopted by transient institutions. Contrastingly, institutional investors with longer horizons and those that are directly engaged in the governance of the companies are mostly associated with stability (Bushee, 2004). The coefficients for turnover, market-to-book, return on equity, and debt-to-capital ratios were rescaled as in previous tables. The high statistical significance of the transient dummy across the full sample and the dividend payer subsample confirms H₂. This suggests that the previously observed volatility spikes are indeed driven by the short-term trading strategies and high portfolio turnover of transient institutional investors.

The effect among dividend-paying companies is more pronounced and is noteworthy. Even though dividend-paying companies are normally considered stable firms, such firms are more likely to be part of a stock index, and it was previously identified in the literature that such firms may amplify volatility dynamics by attracting institutional investors (Basak & Pavlova, 2013). Furthermore, the mitigating role of dividends may be undermined due to attracting transient investors. This result reiterates the argument proposed by DeVault et al. (2019), according to which volatility spikes can be attributed to the institutional trading activity.

To ensure our findings are not sensitive to the specific choice of a median threshold, we re-estimated Equation (7) using a more restrictive cutoff to isolate high-intensity trading behavior. By defining transient investors as those in the top tercile (33 percent) of the institutional turnover distribution, we examine whether the observed volatility-enhancing effect is concentrated among the most active participants. The results confirm our primary findings and suggest that the relationship is driven by the tail of the distribution. For the full sample, the coefficient for this high-turnover classification is 0.0099 (p = 0.002), indicating that high-turnover institutions significantly amplify idiosyncratic risk.

Furthermore, the subsample analysis reveals that this effect is entirely driven by dividend-paying companies, where the coefficient increases to 0.0118 (p = 0.000). In contrast, the effect remains statistically insignificant for non-dividend-paying firms (p = 0.304). This suggests that the impact of short-term institutional trading on idiosyncratic volatility is robust to alternative classifications and is most potent among the most active participants within the dividend-paying segment of the market.

The set of control variables maintains consistency with previous specifications of the regressions. For instance, the market-to-book ratio remains statistically significant and positively associated with volatility for non-dividend firms and negatively associated with dividend-paying firms. This result illustrates the different risk profile of income-oriented stocks as opposed to growth stocks. Similar to the previous regressions, profitability reduces volatility, while leverage is positively correlated with volatility, illustrating that the market perceives the leverage as a risk-amplifying signal. The monitoring role of analyst coverage is maintained by the negative relationship between analyst buy recommendations and idiosyncratic volatility.

5. Conclusions

Our paper examines the manner in which institutional investors influence firm-level idiosyncratic volatility in European stock markets and how dividend policy impacts this relationship. Using data for the STOXX Europe 600 index for the period 2005 to 2025, we estimated idiosyncratic volatility from Fama-French three-factor model residuals. Our empirical analysis provides consistent support for our formulated hypotheses.

Specifically, we find that institutional ownership is positively associated with firm-specific risk, supporting H₁. However, this relationship is highly dependent on the firm’s payout behavior; idiosyncratic volatility is amplified by institutional investors primarily for dividend-paying companies, confirming H₃. In contrast, the effect is not statistically significant for non-dividend firms. These results remain robust for both contemporaneous and lagged model specifications, indicating that dividend policy is a key channel through which institutional investors impact stock price volatility in the European market.

Furthermore, by classifying institutional investors based on their investment horizon, we find evidence supporting H₂. Our results indicate that the observed volatility amplification is driven by transient institutional ownership, characterized by short-term trading strategies. By providing new European evidence and identifying the moderating role of dividends, we contribute to the ongoing debate regarding the stabilizing versus destabilizing role of institutional investors. To sum up, the results carry implications for regulators and corporate managers, indicating that the interaction between ownership structure and dividend payout rates can significantly modify a company’s exposure to idiosyncratic risk.

While our study provides robust evidence for the European market using the STOXX Europe 600, future research could extend this methodology to other major markets. Specifically, examining these dynamics in the United States or Far-East Asian markets would clarify whether the moderating role of dividend policy and the impact of institutional transience are universal or dependent on regional regulatory and governance frameworks.