How do leveraged buyouts affect industry peers' performance: Evidence from Europe
Abstract
EconStor is a publication server for scholarly economic literature, provided as a non-commercial public service by the ZBW.
Full text
Kathan, Manuel C. Article — Published Version How do leveraged buyouts affect industry peers' performance: Evidence from Europe Review of Financial Economics Provided in Cooperation with: John Wiley & Sons Suggested Citation: Kathan, Manuel C. (2025) : How do leveraged buyouts affect industry peers' performance: Evidence from Europe, Review of Financial Economics, ISSN 1873-5924, Wiley, Hoboken, NJ, Vol. 43, Iss. 4, pp. 519-547, https://doi.org/10.1002/rfe.70011 This Version is available at: https://hdl.handle.net/10419/329820 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/
Rev Financ Econ. 2025;43:519–547. | 519 wileyonlinelibrary.com/journal/rfe 1 | INTRODUCTION A growing body of literature examines the impact of private equity (PE) investments on target firms' industry peers. These papers document effects on several dimensions such as corporate governance structure (Harford etal.,2016; Oxman & Yildirim,2008), industry profitability (Aldatmaz & Brown,2020; Bernstein etal.,2017), investment strategies (Truong & Walz,2024), and firm valuation (Chevalier,1995a; Hsu etal.,2011; Slovin etal.,1991). Studies analyzing the spillover effects of leveraged buyouts (LBOs) on other firms have predominantly focused on the United States, resulting in a sparse body of research within the European context. However, a comparison of the ratios of PE deal value to the market capitalization of the S&P 500 in the United States and the STOXX Europe 600 in Europe reveals similar patterns over time, as illustrated in Figure1. The graphical illustration suggests that PE investments may also influence firms in the European market. Consequently, it is crucial to investigate in more detail whether LBOs harm or benefit target firms' industry peers in Europe. This paper contributes to the literature by analyzing the impact of LBOs on the profitability of industry peers in Europe and the channels through which profitability may be affected. As PE investors possess private information (e.g., Dittmar etal.,2012), their investments can generate information spillovers that other market participants may use in their decisionmaking. In this regard, LBOs might signal Received: 8 April 2025 | Revised: 8 April 2025 | Accepted: 21 July 2025 DOI: 10.1002/rfe.70011 ORIGINAL ARTICLE How do leveraged buyouts affect industry peers' performance: Evidence from Europe Manuel C.Kathan1,2 This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2025 The Author(s). Review of Financial Economics published by Wiley Periodicals LLC on behalf of University of New Orleans. 1University of Augsburg, Augsburg, Germany 2University of St. Gallen, St. Gallen, Switzerland Correspondence Manuel C. Kathan, University of Augsburg, Augsburg, Germany. Email: [email protected] Abstract This paper analyzes the impact of leveraged buyouts (LBOs) on the profitability of target firms' industry peers in Europe. To address the endogeneity of LBO activity, I employ a control function approach, using the European Takeover Directive as an instrumental variable. The results indicate that peers improve their profitability following LBOs, driven by improved asset utilization and enhanced cost efficiency. Unlike the findings in the USbased literature, my analysis reveals that positive future industry developments also contribute to the overall effect. These findings suggest that the impact of LBOs on industry peers varies to some extent in the European context. KEYWORDS control function approach, leveraged buyouts, peer firms, spillover effects JEL CLASSIFICATION G23, G34
520 | KATHAN information about followon acquisitions, future prospects, agency problems (Slovin etal.,1991), or changes in the competitive environment within the target firm's industry (Harford etal.,2016). A common feature of these informational aspects of LBOs is an increased threat of takeovers for industry peers. If an LBO announcement signals future prospects and followon acquisitions, the takeover risk for industry peers increases (Harford etal.,2016). Similarly, industrywide agency problems arising from the separation of ownership and control may increase the likelihood of takeovers, as inefficiently managed firms become more attractive acquisition targets (e.g., Jensen,1986, 1993). Finally, LBOs may affect competition (e.g., Bharath etal.,2014; Chevalier,1995a, 1995b; Kovenock & Phillips,1997) and increase the takeover threat, as industry peers respond by adjusting their businesses and engaging more actively in acquisitions (Harford etal.,2016). The increased likelihood of being taken over incentivizes managers of industry peers to operate their firms more efficiently to reduce the risk of acquisition, as they typically face the risk of losing their positions if their firms are acquired (Hartzell etal.,2004). Consequently, the impact of LBOs on the peer firms' profitability operates through various channels, and the extent of their adjustments may depend on the information content of the LBOs. However, the relationship between LBO activity and industry peers' profitability may be spurious. For example, an unobserved industry stimulus could simultaneously influence both LBO activity and industry peers' outcomes, leading to a misleading correlation. To address endogeneity issues, I employ a control function approach (CFA) that uses the European Takeover Directive (ETD) as an instrumental variable in the firststage regression. Specifically, I model LBO activity as a function of this instrumental variable, control variables, and fixed effects. I then incorporate the residuals from this regression into the secondstage regression, which analyzes peer firms' outcome variables as a function of LBO activity. By including the residuals from the first stage, this approach corrects for the endogeneity of LBO activity (Heckman & Robb,1985; Wooldridge,2015). The ETD aimed to harmonize takeover laws and facilitate takeovers across European Union (EU) countries (Clerc etal.,2012). The directive orients on the UK's takeover law, which, for some, but not all countries, meant significant changes to their takeover laws. Industry peers in countries that adjusted their takeover regulations due to the ETD experienced an increase in LBO activity and the takeover threat relative to those firms in countries without adjustments. Therefore, the ETD exogenously affects LBO activity, which, in turn, influences the behavior of industry peer managers and serves as an instrumental variable in my analysis. In the context of the CFA approach, an instrument is considered exogenous if it is uncorrelated with the error terms in both the firstand secondstage regressions. The ETD satisfies this condition due to its external imposition by the EU, which was independent of firmspecific or market factors. While individual EU member states had some flexibility in how they transposed the directive into national law and enforced its provisions, the primary goal of the ETD was to FIGURE 1 PE deal value in relation to the market value of indices (United States vs Europe). This figure displays the development of PE deal value relative to the total market capitalization of two major indices. The solid line represents the total value of US deals relative to the S&P 500, while the dashed line shows the European LBO deal value relative to the market capitalization of the STOXX Europe 600. The xaxis denotes the years from 1993 to 2021 (Sources: LSEG Eikon and CRSP).
| 521 KATHAN harmonize takeover regulations across the EU (Clerc etal.,2012). This harmonization reduces concerns about reverse causality and strengthens the instrument's validity. Using a sample of private and public LBO deals of the five largest European countries,i I find that industry peers' profitability improves significantly following LBOs. Further analysis shows that this effect results from better asset utilization of existing assets and enhanced cost efficiency. However, it also shows that PE investors possess the ability to identify industries when they are at the bottom of their profitability. As a result, these industries present growth opportunities, and investors may fail to fully consider the future operating performance of peer firms. Furthermore, the findings do not offer evidence that industry peers alter their internal corporate governance, investment strategies, or organizational structures. By conducting a subsample analysis that splits the sample at the median values of various variables, I further explore the heterogeneity of the effect. The findings reveal that smaller firms and firms with lower leverage ratios improve profitability following LBOs. Smaller firms possess higher growth opportunities, while lower leverage ratios provide financial flexibility to capture investment possibilities (e.g., Frank & Goyal,2009). Additionally, the analysis provides evidence that peers tend to be more sensitive to LBOs in country–industries with higher agency costs and greater growth potential, which relates to the information content of LBOs. The paper is most closely related to the crosscountry studies by Bernstein etal.(2017) and Aldatmaz and Brown(2020), which document positive spillover effects on industry profitability. However, these studies analyze the impact of PE investments at the aggregate level, which may not reveal the complete picture. For example, the aggregation of variables does not control for firmlevel influences, which could create an omitted variable bias problem (e.g., Holderness,2016). Therefore, this study uses the individual peer level to explore the heterogeneity of industry peers and extends our understanding of spillover effects concerning leveraged buyouts (e.g., Feng & Rao,2022; Harford etal.,2016; Truong & Walz,2024). Moreover, the results of this study offer an alternative explanation for the positive effect on peers' profitability. Rather than attributing the observed effect primarily to an increase in competition (e.g., Aldatmaz & Brown,2020; Feng & Rao,2022), my findings suggest that, in addition to an improvement in efficiency, industry developments also contribute to the overall effect. Further, in contrast to the existing literature, European industry peers do not change their corporate governance or alter their investment strategies, unlike their US counterparts (e.g., Feng & Rao,2022; Harford etal.,2016; Oxman & Yildirim,2008). The United States, like the United Kingdom, operates under a commonlaw system, which is characterized by welldeveloped capital markets, investor protection, and effective corporate control markets (e.g., Goergen etal.,2005; La Porta etal.,1997, 1998). In contrast, most continental European countries are subject to civil law, which emphasizes stakeholder orientation through codified laws and is accompanied by large shareholders (blockholderbased system) (e.g., Goergen etal.,2005). European firms have a more concentrated ownership structure (e.g., Enriques & Volpin,2007; La Porta etal.,1997, 1998) and more familycontrolled firms (e.g., Faccio & Lang,2002; Guo etal.,2011). As a result, the managers of industry peers in Europe respond differently to LBO announcements, leading to distinct implications for the effects of LBOs on other firms. This study also contributes to the literature on the effects of the European Takeover Directive. To the best of my knowledge, this is the first paper to use the ETD in the context of LBOs. Previous empirical studies have primarily focused on the ETD's impact related to merger and acquisition (M&A) deals. While the empirical evidence on takeover efficiency gains is mixed (Dissanaike etal.,2021; HumpheryJenner,2012; Wang & Lahr,2017), I provide evidence that the ETD increased LBO activity in the industries of treated countries relative to those in control countries. The remainder of the paper is structured as follows. Section2 develops the study's underlying hypotheses. Section3 outlines the sample construction. Section4 details the empirical strategy, while Section5 presents the results. Section6 examines the heterogeneity of the LBO effect, and Section7 provides robustness tests. Finally, Section8 summarizes and discusses the findings within the broader LBO spillover literature. 2 | DEVELOPMENT OF HYPOTHESES Literature on the spillover effects of LBOs documents changes in the operations of target firms' industry peers (e.g., Feng & Rao,2022; Harford etal.,2016; Oxman & Yildirim,2008). These LBO spillover effects originate from releasing information through PE investment activities. PE investors are informed agents who possess private information (Dittmar etal.,2012) through their sophisticated due diligence process. This information is relevant not only for target firms but also for the entire industry. When they invest in a firm, parts of this information become public and may be utilized by other market participants. In particular, LBOs may signal followon acquisitions, future prospects, and agency problems within an industry (Slovin etal.,1991). Moreover, LBOs may also indicate changes in the competitive environment
522 | KATHAN (e.g., Bharath etal.,2014; Chevalier,1995a, 1995b; Kovenock & Phillips,1997). On average, PE investors enhance the profitability of their target firms (e.g., Acharya etal.,2013; Boucly etal.,2011; Guo etal.,2011), which in turn affects competition within an industry. The common feature of the potential information content of LBOs is an increased threat of takeovers. Regarding followon acquisitions and future prospects, existing literature shows that PEbacked industries grow more quickly (Bernstein etal.,2017; Boucly etal.,2011). Financial and strategic investors try to capture these growth opportunities by taking over firms in these specific industries (e.g., Slovin etal.,1991). As a result, the risk of being acquired increases for industry peers (Harford etal.,2016). If LBOs are associated with industrywide agency problems arising from separating ownership and control, the likelihood of takeovers might also increase. Firms with higher agency costs do not operate at their optimal efficiency level owing to managers' inefficient use of free cash flows (e.g., Jensen,1986, 1993). In general, corporate governance issues are correlated across firms within an industry. For example, firms in competitive industries tend to have lower agency costs (e.g., Chhaochharia etal.,2017; Giroud & Mueller,2010). Empirical literature indicates that firms with weak corporate governance structure show negative abnormal returns (e.g., Bebchuk etal.,2008; Gompers etal.,2003). Due to their poor performance, these firms may be easier to take over (e.g., Lel & Miller,2015). Moreover, PE investors are particularly interested in poorly performing firms because of their restructuring skills (e.g., Gorbenko & Malenko,2014), which contributes to increased industry takeover activity. LBOs may increase the competition within an industry (e.g., Feng & Rao,2022). Consequently, industry peers may alter their business and intensify their acquisition activity to reduce competitive pressure, leading to more takeovers (Harford etal.,2016). From a theoretical perspective, LBOs intensify takeover activity within an industry. Empirical evidence elucidates an increase in bidders after acquisitions of financial investors (Dittmar etal.,2012). Additionally, Harford etal.(2016) demonstrate that LBOs predict future takeovers and increase the takeover threat of industry peers. The increased threat of takeovers signaled by LBOs incentivizes the managers of industry peers to run their firms more efficiently, as these transactions often result in the replacement of the target firm's chief executive officers (CEO) (Hartzell etal.,2004; Kaplan & Stromberg,2009). In this context, corporate control markets can discipline managers (e.g., Denis & Serrano,1996; Holmstrom & Kaplan,2001; Kaplan,1989; Martin & McConnell,1991; Morck etal.,1989), particularly if firms operate inefficiently. Building on this, the incentive structure for managers should positively affect the profitability of industry peers. In the US context, studies indicate that industry peers become more profitable after LBOs (Feng & Rao,2022; Truong & Walz,2024) and hostile takeovers (Servaes & Tamayo,2014). A similar rationale should apply in the European context, leading to the following hypothesis. Hypothesis 1. LBOs signal an increased likelihood of takeovers for target firms' industry peers, incentivizing managers to enhance their firms' performance to reduce the threat of being taken over. If managers of industry peers improve their firms owing to the increase in the takeover risk following LBOs, it becomes essential to ascertain the alterations they implement to their firms associated with the LBO signal. One possible way is to use their assets more efficiently or to cut operational expenses. Through both ways, industry peers increase their efficiency to reduce future takeover threats and competitive pressure. Feng and Rao(2022) show positive spillovers on operating efficiency after PE investments. Thepresent study assumes similar effects on European industry peers. Thus, the next hypothesis is as follows. Hypothesis 2a. LBOs positively impact the operating efficiency of industry peers. LBOs might indicate industrywide agency problems (Slovin etal.,1991). To improve the internal corporate governance structure and create value in target firms, PE investors provide strong incentives to CEOs, reduce board size, and closely monitor the management (e.g., Cornelli & Karakaş,2012; Nikoskelainen & Wright,2007). Likewise, industry peers might implement these changes, resulting in a positive relationship. Oxman and Yildirim(2008) empirically confirm the positive association, whereas Harford etal.(2016) document a negative one. Given that LBOs increase the incentive to improve corporate governance by reducing the likelihood of a takeover, I propose the following hypothesis. Hypothesis 2b. LBOs positively impact the corporate governance of industry peers.
| 523 KATHAN However, European firms generally exhibit a more concentrated ownership structure than their US counterparts (e.g., Enriques & Volpin,2007). As a result, the separation between ownership and control may be less problematic in Europe, potentially diminishing the impact of LBOs on the corporate governance structure of industry peers. To mitigate competitive pressure and the risk of takeovers, industry peers may adjust their business operations to the new competitive environment (e.g., Feng & Rao,2022). At the aggregated level, in a crosscountry study, Aldatmaz and Brown(2020) show that investments within the industry increase following PE investments. These adjustments may also include increasing the R&D expenses of industry peers, as target firms tend to become more innovative (e.g., Lerner etal.,2011; Ughetto,2010). In addition, LBO targets often restructure their assets (e.g., Denis,1994; Muscarella & Vetsuypens,1990). Thus, industry peers may also refocus their organizational structure to become more efficient and competitive. From these arguments, I derive the following hypothesis. Hypothesis 2c. LBOs positively impact the adjustment of industry peers' businesses by increasing their investment or refocusing their organizational structure. PE investors may strategically select profitable industries. Slovin etal.(1991) suggest that LBOs indicate favorable industry prospects, and Harford etal.(2019) argue that industries are undervalued during management buyouts. Thus, PE investors may leverage their industry expertise (Kaplan & Stromberg,2009) and private information (Dittmar etal.,2012) to time their entry into a particular industry. In this context, LBO activity coincides with the improvement in industry peers' profitability. The increased takeover activity results from less informed market participants entering a merger sequence, of which PE investors tend to be first movers (Harford etal.,2016). Therefore, the hypothesis reads as follows. Hypothesis 3. Private equity investors can identify industries where profitability improvements are likely to occur, regardless of their involvement. The central assumption of this study is that an increase in the takeover threat, signaled by LBOs—whether through information about industrywide agency problems, changes in competition, or followon acquisitions—incentivizes managers of industry peers to operate their firms more efficiently. While all hypotheses suggest an increase in takeover activity within an industry, they are not mutually exclusive. However, the implications of the takeover threat generated by LBOs are distinct from the observed actions of industry peers, which helps to disentangle the effect and uncover the underlying mechanism between LBOs and industry peers' profitability. 3 | DATA AND SAMPLE CONSTRUCTION I retrieve LBO deals from LSEG (formerly Refinitiv) Eikon. The sample covers private and public deals from 1993 to 2021 and focuses on the five largest European countries in terms of their GDPs, that is, France, Germany, Italy, Spain, and the United Kingdom. Other European countries are significantly smaller and have less developed PE markets. Therefore, it is more likely to observe effects on the profitability of industry peers in these countries. Furthermore, I exclude target firms belonging to the utility industry (SIC codes 4900–4999), the financial industry (SIC codes 6000–6999), and government entities (SIC codes >9000) (e.g., Leary & Roberts,2014). In addition, I ensure that PE funds have a majority interest (>50%) in the firm to implement their valuecreating strategies. After applying these filters, the sample contains 10,656 LBO deals. Table1 presents descriptive statistics for the LBO sample. Panel A shows that the United Kingdom accounts for a significant proportion of the sample, representing over 50% of the number of deals (No. of deals) and nearly half of the Total deal value. LBO deals in Germany are larger on average (Mean deal value).ii Italy and Spain exhibit considerably lower figures in terms of both Total deal value and No. of deals. The proportion of public deals (Ratio of public deals) ranges from 2% to 4% compared to the entire LBO sample, indicating that most PE funds acquire private targets. Panel B of Table1 illustrates the distribution of LBO deals across industries, revealing that PE investors primarily invest in the Manufacturing and Services industries within the sample countries. The firms in my sample (peer firms) are obtained from the Compustat Global database, comprising all nonfinancial, nonutility, and nongovernment firms from the five analyzed countries in this study between 1993 and 2021. Industries are classified using threedigit SIC codes (e.g., Grennan,2019) within each country. Based on this classification, I merge
524 | KATHAN the primary dataset with the LBO sample. The final sample consists of 5695 firms and 62,727 firmyears, of which 30,983 are impacted by LBO announcements.iii Peer firm characteristics are obtained from Compustat Global, Datastream, Capital IQ, and Boardex Europe. This study employs control variables similar to those in Harford etal.(2016), with all values converted to US dollars. Table2 presents summary statistics for the characteristics of peer firms. The columns labeled “All firms” report variable statistics for the entire sample.iv This study focuses on the perspective of peer firms and their response to LBO activity within their industry. In the panel setting, a peer firm may encounter LBOs in some years and none in others. Accordingly, the column labeled “LBO” provides variable statistics for years when a firm faces an LBO announcement within its countryindustry, while the “NonLBO” column reports statistics for years without such activity. The final column compares the mean differences between the “LBO” and “NonLBO” groups for statistical significance. At this point, I highlight a few observations. First, Table2 shows that firms' profitability (e.g., EBITDAtoassets) is significantly lower in country–industry–years with LBO activity compared to those without. Second, firms in the “LBO” column tend to be smaller (Log(1 + Assets)) but exhibit higher valuations (M/Bratio) and lower leverage ratios (Book leverage) than those in the “NonLBO” group. These initial findings reveal notable differences between the two groups. 4 | EMPIRICAL STRATEGY This study employs the following baseline model to examine the impact of LBOs on peer firms: where c denotes country, i the firm, j the industry, and t the year. The dependent variable yc,i,j,t + 1 represents firm i′s characteristics (e.g., profitability) in the subsequent period. The variable of interest, LBOc,j,t , captures the leveraged buyout activity in industry j and country c at time t . I proxy this variable as the number of leveraged buyouts divided by the number (1) yc,i,j,t + 1 = 𝛼 + 𝛽1LBOc,j,t + 𝛽2Xc,i,j,t + 𝛽3Ic,j,t + 𝛽4Mc,t + 𝛾c + 𝜂j + 𝛿t + 𝜀c,i,j,t TABLE 1 Descriptive statistics of target firms. Panel A—LBO characteristics Country Variables France Germany Italy Spain United Kingdom LBO sample Total deal value (bn $) 123.73 165.52 52.02 45.23 381.43 767.93 No. of deals 2100 1781 573 386 5816 10,656 Mean deal value (mio $) 283.77 656.84 376.98 373.82 153.31 223.56 Ratio public deals (%) 2.71 2.92 3.84 2.85 3.94 3.48 Panel B—No. of deals per industry Country Industry France Germany Italy Spain United Kingdom LBO sample Mining 0 3 1 0 41 45 Construction 60 23 5 8 149 245 Manufacturing 1007 1025 397 139 2358 4926 Transportation and public utilities 110 62 18 36 278 504 Wholesale trade 101 77 7 6 506 697 Retail trade 102 57 21 16 397 593 Services 720 534 124 181 2087 3646 Note: This table presents descriptive statistics for LBO targets in France, Germany, Italy, Spain, the United Kingdom, and the overall sample (“LBO sample”) from 1993 to 2021. Panel A shows the Total deal value (bn$), No. of deals, Mean deal value (mio $), and Ratio public deals (%). Panel B illustrates the No. of deals for different industries based on the major groups of the Standard Industrial Classification (SIC). LBO targets in the utility and financial industries, as well as government entities, are excluded.
| 525 KATHAN of firms within a country–industry–year to account for the intensity of the LBO signal. This definition captures the idea that more deals in a small industry have a different impact than a few deals in industries with many firms.v Control variables include firmlevel characteristics ( Xc,i,j,t ), industrylevel factors ( Ic,j,t ), and macroeconomic conditions ( Mc,t ), which vary over time across countries. Country ( 𝛾c ) and industry ( 𝜂j ) fixed effects account for unobserved timeinvariant differences across these dimensions, while year effects ( 𝛿t ) capture common trends and market conditions. These fixed effects help control for institutional differences and countryspecific confounders. Standard errors are clustered at the firm level to account for withinfirm correlation and heteroskedasticity (Petersen,2009). In Equation(1), a key concern is the potential endogeneity of the LBO variable. For example, a third variable, such as an unobserved industry stimulus, may affect both LBO activity and the outcome variables of industry peers. As a result, TABLE 2 Summary statistics of peer firms. Variables All firms LBO NonLBO LBO versus nonLBO Obs. Mean Median SD Mean Mean Diff. mean Profitability and control variables EBITDAtoassets 62,727 0.041 0.088 0.227 0.028 0.053 −0.025*** Operating margin 58,482 0.039 0.094 0.407 −0.013 0.063 −0.076*** Returnonassets 54,641 −0.011 0.028 0.164 −0.022 −0.008 −0.014*** M/Bratio 62,727 2.531 1.562 3.941 2.870 2.201 0.669*** Log(1 + Assets) 62,727 5.092 4.867 2.228 4.741 5.435 −0.694*** Net PPEtoassets 62,727 0.225 0.161 0.216 0.187 0.262 −0.075*** Book leverage 62,727 0.203 0.163 0.221 0.171 0.211 −0.040*** Log(1 + Cash) 62,727 2.896 2.597 2.035 2.706 3.082 −0.376*** Operating efficiency Turnover ratio 62,899 1.034 0.922 0.762 1.054 1.015 0.039*** Current asset turnover ratio 62,891 2.190 1.915 1.640 2.174 2.204 −0.030** Expense ratio 60,330 1.343 0.909 2.782 1.529 1.281 0.248*** SG&A margin 59,663 0.458 0.175 1.617 0.586 0.401 0.185*** Corporate governance Log(1 + Board size) 31,777 1.511 1.609 0.657 1.446 1.548 −0.102*** Fraction independent directors 9090 0.296 0.286 0.206 0.286 0.290 −0.004 Log(1 + Cash compensation) 35,744 0.634 0.484 0.669 0.604 0.617 −0.013* Stock optionstoSHO 31,062 0.010 0.000 0.020 0.010 0.008 0.002*** Investment and asset sales CAPXtoassets 57,198 0.048 0.031 0.054 0.044 0.052 −0.008*** R&Dtoassets 21,787 0.077 0.032 0.122 0.093 0.042 0.051*** Asset sales 31,288 0.010 0.002 0.024 0.009 0.011 −0.002*** Misvaluation and growth opportunities Timesseries industry error 26,483 −0.009 −0.007 0.516 0.001 −0.020 0.021*** Longrun to book value 26,483 0.676 0.694 0.639 0.766 0.463 0.303*** Firmspecific error 26,483 0.000 0.000 0.631 −0.001 0.010 −0.011 Note: This table presents summary statistics for all nonfinancial, nonutility, and nongovernment firms from the Compustat Global database between 1993 and 2021 for the five largest EU countries—France, Germany, Italy, Spain, and the United Kingdom—based on their GDP. The “LBO” column displays the mean values of various variables for firms exposed to an LBO announcement in their country–industry–year, while the “NonLBO” column shows the mean values of these variables for firms in country–industry–years without LBO exposure. The “LBO vs NonLBO” column reports the differences between these mean values. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively, based on twosided ttest. All variables are winsorized at the 1% level. For variable descriptions, see TableA1.
526 | KATHAN LBOs may not cause changes in the operations of industry peers. To mitigate endogeneity concerns, I employ a control function approach. Specifically, I first model LBOc,j,t as a function of an instrument variable z , as well as control variables and fixed effects. Second, I use the residuals ( 𝜈 ) of the first stage and plug those in Equation(1) (structural equation) to correct for the endogenous variable LBO. To be a valid instrument in the CFA context, z must be correlated with LBO activity while satisfying the conditions E( z�𝜀 ) = 0 and E( z�𝜈 ) = 0 (Heckman & Robb,1985; Wooldridge,2015). These conditions establish the exogeneity of z , ensuring that it is uncorrelated with the error terms in both the firststage regression and the structural equation. Building on the argumentation in Section2, this study employs the European Takeover Directive as an exogenous shock to LBO activity and the takeover threat exerted by financial investors within an industry. The directive attempted to harmonize and improve takeover rules within the EU. As a result, some countries had to change their takeover regulations significantly, while other countries' regulations remained unchanged (Clerc etal.,2012; Dissanaike etal.,2021). To examine the effect of LBOs on industry peers, I use the ETD as an instrumental variable within the CFA framework. By exploiting ETD's exogenous variation in LBO activity, the CFA accounts for unobserved factors that may influence both LBO activity and industry peers' outcomes, thereby addressing potential endogeneity. The ETD was promulgated on April 21, 2004, with EU Member States required to implement it into national law by May 21, 2006. The directive set minimum takeover requirements,vi based on the UK's takeover regulations, intending to standardize rules across the EU (e.g., Clerc etal.,2012; Dissanaike etal.,2021; HumpheryJenner,2012). The specific objectives of the ETD are stated in Clerc etal.(2012, p. 11–12): (i) legal certainty on the takeover bid process and Communitywide clarity and transparency with respect to takeover bids; (ii) protection of the interests of shareholders, in particular, minority shareholders, employees, and other stakeholders when a company is subject to a takeover bid for control; and (iii) reinforcement of the freedom for shareholders to deal in and vote on securities of companies and prevention of management action that could frustrate a bid. These objectives should facilitate takeover bids, increase the takeover threat (i and iii), and protect firm stakeholders (ii). The ETD altered the takeover laws in the sample countries in different ways. France, Italy, and the United Kingdom experienced no significant changes, so firms in these countries are included in the control group, as the new regulation did not affect LBO activity. In contrast, Germany and Spain made substantial changes to their takeover laws. Both countries introduced the squeezeout and sellout rules (Clerc etal.,2012; Dissanaike etal.,2021). The squeezeout rule gives the controlling shareholder the right to squeeze out minority shareholders if the bidder acquires a certain percentage of a firm's shares. The sellout right gives minority shareholders the right to demand a fair price for their shares from the controlling shareholders, subject to the same conditions as in the squeezeout right (e.g., Goergen etal.,2005). From a PE perspective, the introduction of squeezeout rights is particularly relevant, as PE funds typically acquire all assets.vii Moreover, in Spain, the mandatory bid rule, which forces the bidder to make a binding bid to all shareholders at an equitable price, may not be as effective as in other countries. This is because shareholders can waive the mandatory bid rule (Dissanaike etal.,2021), and the private benefits of controlling shareholders, which amplify the cost of a mandatory bid rule, are considerably lower compared to other European countries (e.g., Italy) (Dyck & Zingales,2004). In Germany, the equitable price mitigated the strength of the crossshareholdings as a takeover defense tool (Dissanaike etal.,2021). Overall, the change in takeover rules facilitated acquisitions and reduced their costs in Germany and Spain, making firms in these countries part of the treatment group.viii To address endogeneity in the LBO variable, I exploit the relative change in LBO activity between firms in treated and control countries. In Figure2, I show how the ETD changed the LBO activity of industries in treated countries compared to industries of control countries by displaying yearly regression coefficient estimates. The dependent variable, LBO, is defined as the number of LBO deals divided by the number of firms within a country–industry–year. It is regressed on the binary variable Treat, which equals one for industries in treated countries and zero otherwise. Before the announcement of the ETD in 2004, industries in treated countries exhibited considerably lower LBO activity than those in control countries, illustrated by the negative coefficients around −0.06. Following the implementation of the new legislation in the treated countries, which occurred on July 8, 2006, in Germany and on April 13, 2007, in Spain,ix the disparity in LBO activity between the two groups diminished and leveled out to similar intensities. In this regard, Figure2 illustrates that the ETD positively affects LBO activity for firms in treated countries compared to those in control countries and demonstrates its relevance to PE investors and their takeover activity. Thus, I model LBO activity in the first stage of the CFA as follows: (2) LBOc,j,t = 𝛼 + 𝛽1Treatc,j × Postt + 𝛽2Mc,t − 1 + 𝛽3Ic,j,t − 1 + 𝛾c + 𝜂j + 𝛿t + 𝜈c,j,t
| 533 KATHAN TABLE 6 Corporate governance. Log(1 + Board size)t+1 Frac. ind. directorst+1 Log(1 + Cash comp.)t+1 Stock optionstoSHOt+1 Log(1 + Board size)t+1 Frac. ind. directorst+1 Log(1 + Cash comp.)t+1 Stock optionstoSHOt+1 Whole sample—OLS Restricted sample around ETD—CFA (1) (2) (3) (4) (5) (6) (7) (8) LBO 0.0288 −0.0148 0.0605*** −0.0015*** 0.0231 −0.0846** −0.0018 −0.0022** (1.35) (−1.43) (3.06) (−2.97) (0.62) (−2.10) (−0.05) (−1.99) Residuals (firststage) 0.0302 0.0909** 0.0745** −0.0007 (0.82) (2.38) (2.26) (−0.63) Controls Yes Yes Yes Yes Yes Yes Yes Yes Macro controls Yes Yes Yes Yes Yes Yes Yes Yes Country FE Yes Yes Yes Yes Yes Yes Yes Yes Industry FE Yes Yes Yes Yes Yes Yes Yes Yes Year FE Yes Yes Yes Yes Yes Yes Yes Yes Observations 31,777 9090 35,744 31,062 17,596 3995 19,440 17,211 AdjR20.395 0.366 0.350 0.247 0.342 0.406 0.306 0.259 Note: This table presents ordinary least squares (OLS) and control function approach (CFA) estimations, with the oneyear lead dependent variables Log(1 + Board size), Fraction independent (Frac. ind.) directors, Log(1+Cash component), and Stock optionstoshares outstanding (SHO). Columns (1)–(4) cover the whole sample from 1993 to 2021. Columns (5)–(8) focus on the period from 1999 to 2011, which is around the implementation of the ETD. The variable LBO is defined as the ratio of LBO deals to the number of firms within a country–industry–year. For the CFA estimations, the following firststage regression is applied: The residuals ( 𝜈c , j , t ) from this regression, referred to as Residuals (firststage), are included in the analysis of Columns (5)–(8) to correct for the endogenous variable LBO. The binary variable Post equals one for firms in France (Germany, Italy, Spain, and the United Kingdom) from the implementation date of May 20, 2006 (July 8, 2006; July 19, 2007; April 13, 2007; and May 20, 2006) to the end of the year 2011. It is zero from 1999 until the day before each country's ETD implementation date. The binary variable Treat equals one for firms in treated countries (Germany, Spain) affected by the ETD and zero for firms in control countries (France, Italy, and the United Kingdom). Treat × Post is the interaction of the two binary variables. Mc,j,t − 1 represent lagged macroeconomic variables, while Ic,j,t − 1 captures industryspecific variables. 𝛾c , 𝜂j , and 𝛿t denote country, industry, and time fixed effects, respectively (see Table3, Column (2) for the regression results). For variable descriptions, see TableA1. Firms in the utility and financial industries, as well as government entities, are excluded. All continuous variables are winsorized at the 1% level at both ends. Control and macro control variables correspond to those in Table4. All regressions include a constant. Standard errors are clustered at the firm level, and tvalues are reported in parentheses. *** and ** indicate statistical significance at the 1% and 5% levels, respectively. LBOc,j,t = 𝛼 + 𝛽1Treatc,j × Postt + 𝛽2Mc,t − 1 + 𝛽3Ic,j,t − 1 + 𝛾c + 𝜂j + 𝛿t + 𝜈c,j,t
534 | KATHAN To test empirically whether the graphical illustration reflects growth opportunities or provides a beneficial industry valuation for PE investors, I employ the M/B decomposition of RhodesKropf etal.(2005). It decomposes the M/B ratio into three distinct components. The M/B ratio is often used as a proxy for a firm's growth opportunities (e.g., Frank & Goyal,2009) or firm valuation (e.g., Pástor & Pietro,2003).xiii The first component, Firmspecific error (FSE), measures firmspecific deviations from the fundamental value implied by industry multiples. The second component, Timeseries industry error (TSIE), examines shortterm industrylevel deviations from their longrun values. Both components capture the mispricing part of the M/B ratio. In particular, TSIE provides information on the misvaluation of firms regarding their industry valuation error. The third component, Longrun to book value (LRtB), indicates a firm's growth potential, as it measures the longrun average growth rates for an average firm in an industry. TABLE 7 Investments and asset sales. CAPXto −assetst+1 R&Dto −assetst+1 Asset salest+1 CAPXto −assetst+1 R&Dto −assetst+1 Asset salest+1 Whole sample—OLS Restricted sample around ETD—CFA (1) (2) (3) (4) (5) (6) LBO −0.0025** −0.0039 0.0003 −0.0009 −0.0214** −0.0012 (−2.54) (−1.42) (0.53) (−0.35) (−2.41) (−0.47) M/Bratio 0.0007*** 0.0022*** −0.0000 0.0006*** 0.0022*** −0.0000 (9.77) (5.25) (−0.37) (6.11) (3.94) (−0.02) Log(1 + Assets) −0.0045*** −0.0351*** −0.0010*** −0.0039*** −0.0346*** −0.0005* (−11.55) (−17.21) (−3.86) (−8.20) (−13.22) (−1.75) Net PPEtoassets 0.0938*** −0.0449*** 0.0089*** 0.0964*** −0.0382** 0.0116*** (27.99) (−3.45) (4.53) (23.20) (−2.12) (4.98) Book leverage −0.0069*** 0.0515*** 0.0095*** −0.0075*** 0.0523*** 0.0086*** (−3.41) (3.36) (5.94) (−2.88) (2.94) (4.58) Log(1 + Cash) 0.0042*** 0.0245*** −0.0005** 0.0042*** 0.0241*** −0.0006** (10.62) (13.71) (−2.16) (8.88) (10.44) (−2.19) Herfindahl index −0.0037* −0.0024 0.0000 −0.0026 −0.0113 0.0011 (−1.95) (−0.33) (0.01) (−1.07) (−1.17) (0.74) Residuals (firststage) −0.0031 0.0199** 0.0012 (−1.27) (2.05) (0.49) Macro controls Yes Yes Yes Yes Yes Yes Country FE Yes Yes Yes Yes Yes Yes Industry FE Yes Yes Yes Yes Yes Yes Year FE Yes Yes Yes Yes Yes Yes Observations 57,198 21,787 31,288 30,412 9997 17,897 AdjR20.250 0.339 0.086 0.241 0.318 0.086 Note: This table presents ordinary least squares (OLS) and control function approach (CFA) estimations, with the oneyear lead dependent variables CAPXtoassets, R&Dtoassets, and Asset sales. Columns (1)–(3) cover the whole sample from 1993 to 2021. Columns (4)–(6) focus on the period from 1999 to 2011, which is around the implementation of the European ETD. The variable LBO is defined as the ratio of LBO deals to the number of firms within a country– industry–year. For the CFA estimations, the following firststage regression is applied: The residuals ( 𝜈c,j,t ) from this regression, referred to as Residuals (firststage), are included in the analysis of Columns (4)–(6) to correct for the endogenous variable LBO. The binary variable Post equals one for firms in France (Germany, Italy, Spain, and the United Kingdom) from the implementation date of May 20, 2006 (July 8, 2006; July 19, 2007; April 13, 2007; and May 20, 2006) to the end of the year 2011. It is zero from 1999 until the day before each country's ETD implementation date. The binary variable Treat equals one for firms in treated countries (Germany, Spain) affected by the ETD and zero for firms in control countries (France, Italy, and the United Kingdom). Treat × Post is the interaction of the two binary variables. Mc,j,t − 1 represent lagged macroeconomic variables, while Ic,j,t − 1 captures industryspecific variables. 𝛾c , 𝜂j , and 𝛿t denote country, industry, and time fixed effects, respectively (see Table3, Column (2) for the regression results). For variable descriptions, see TableA1. Firms in the utility and financial industries, as well as government entities, are excluded. All continuous variables are winsorized at the 1% level at both ends. All regressions include a constant. Standard errors are clustered at the firm level, and tvalues are reported in parentheses. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. LBOc , j , t=𝛼+𝛽 1 Treatc , j×Postt+𝛽 2 Mc , t −1 +𝛽 3 Ic , j , t −1 +𝛾c+𝜂j+𝛿t+𝜈c , j , t
| 535 KATHAN Table8 uses these three components as dependent variables. However, instead of using continuous component variables, I create binary variables to categorize firms into growth and undervalued firms.xiv First, TSIE (binary), which equals one if shortterm levels are below their longrun values, indicates industry undervaluation. Second, LRtB (binary) equals one if the longrun values are above the current book value, implying growth opportunities for a firm. Finally, FSE (binary) is one if the market value is below the firm's fundamental value. These variables are equal to zero if the respective condition of the variable is not met. Importantly, I use these dependent variables contemporaneously because PE investors likely possess an information advantage. In the case of industry undervaluation, this advantage would likely diminish rapidly if markets are efficient and would not be observable in the subsequent period. Column (1) indicates that the likelihood of the shortterm value implied by industry multiples of an industry peer being below its longrun value increases significantly with an increase in LBO activity. Furthermore, industry peers also exhibit significantly higher growth potential (longrun values are above the current book value) when exposed to LBOs (Column (2)). However, in the CFA models, which correct for the endogeneity of LBO, only the effect on the undervaluation of an industry peer at the industry level (Column (4)) remains significant. The results in Table8 also suggest that the firmspecific undervaluation (market value of a peer is below its shortterm value implied by industry multiples) of industry peers (Columns (3) and (6)) does not play a role in this context. These findings support hypothesis 3, indicating that industry peers' performance improvement can also be attributed to industry developments, specifically industry undervaluation. In this regard, investors do not account fully for future operating performance, and PEs are able to select those industries. These results are consistent with rational models in which PE investors possess private information that is not part of the information set of other investors at time t (e.g., RhodesKropf & Viswanathan,2004). Existing literature contributions do not attribute the improvement in industry peers' profitability to the selection skills of PE investors (e.g., Aldatmaz & Brown,2020; Feng & Rao,2022). They argue that competitive spillovers cause industry peers to adjust their business. This section provides evidence that the timing of LBOs is a crucial factor in explaining the overall effect. 6 | HETEROGENEITY IN THE PROFITABILITY OF INDUSTRY PEERS 6.1 | Firmlevel In this section, I perform several subsample analyses to gain deeper insights into the positive impact of LBOs and to identify the firms that are most affected. To achieve this, I split the sample at the median of the variables Log(1 + Assets), M/Bratio, Book leverage, and Ownership concentration.xv More specifically, I examine whether firm size, growth opportunities, financial flexibility, and ownership structure play a role in this context. In Table9, the oneyear lead dependent variables are EBITDAtoassets and Operating margin. The analysis focuses on the “restricted sample” around the ETD, applying CFA regressions. In Panel A, although all coefficients of LBO are positive and statistically significant, the effect is more pronounced in magnitude for firms that are equal to or below the sample median of “Log(1 + Assets)” (Columns (1) and (2)), and for firms above the sample median of the “M/Bratio” (Columns (7) and (8)). FIGURE 3 Industry profitability. This figure displays the development of the average profitability within a country–industry–year. The xaxis indicates the years relative to the year of an LBO announcement. The solid (dashed) line depicts the average industry EBITDAtoassets (Operating margin). For variable descriptions, see TableA1.
536 | KATHAN Smaller firms tend to have higher growth opportunities, for which the results find some evidence of higher LBO activity in the previous section. Moreover, these firms are likely easier to take over because of their size. Thus, if LBOs signal followon acquisitions, managers of small industry peers and those with high growth opportunities may be more incentivized to reduce the takeover risk and run their firms more efficiently than managers of larger industry peers. Furthermore, smaller firms are less followed by analysts (e.g., Bhushan,1989; Lobo etal.,2012), which may support the undervaluation argument in Section5.3, as more analysts provide more meaningful information (e.g., Li etal.,2019). Firms with lower leverage ratios have greater financial flexibility, enabling them to capture investment possibilities (e.g., Frank & Goyal,2009). Moreover, a higher takeover threat may positively affect the leverage ratio since debt can be TABLE 8 LBOs and industry developments. TSIE (binary)t LRtB (binary)t FSE (binary)t TSIE (binary)t LRtB (binary)t FSE (binary)t Whole sample—OLS Restricted sample around ETD—CFA (1) (2) (3) (4) (5) (6) LBO 0.0403** 0.0327** − 0.0181 0.1476*** − 0.0018 − 0.0261 (2.00) (2.29) ( − 0.91) (3.84) ( − 0.07) ( − 0.63) Log(1 + Assets) − 0.0057** 0.0200*** − 0.0005 0.0018 0.0219*** 0.0021 ( − 2.20) (7.42) ( − 0.18) (0.55) (6.55) (0.58) Net PPEtoassets − 0.0035 − 0.0399** − 0.0136 0.0156 − 0.0488** − 0.0216 ( − 0.25) ( − 2.13) ( − 0.92) (0.89) ( − 2.22) ( − 1.16) Book leverage 0.0229 − 0.1552*** 0.0367** − 0.0356 − 0.1674*** 0.0241 (1.32) ( − 8.64) (1.98) ( − 1.55) ( − 7.56) (1.00) Log(1 + Cash) − 0.0038 − 0.0055** − 0.0048* − 0.0049 − 0.0061* − 0.0091** ( − 1.47) ( − 2.07) ( − 1.69) ( − 1.45) ( − 1.95) ( − 2.55) Herfindahl index 0.0069 − 0.0677*** − 0.0037 0.0416 − 0.0627** − 0.0098 (0.30) ( − 3.41) ( − 0.15) (1.23) ( − 2.18) ( − 0.29) Residuals (firststage) − 0.0769* 0.0260 0.0141 ( − 1.93) (0.94) (0.33) Macro controls Yes Yes Yes Yes Yes Yes M/B decomposition Yes Yes Yes Yes Yes Yes Country FE Yes Yes Yes Yes Yes Yes Industry FE Yes Yes Yes Yes Yes Yes Year FE Yes Yes Yes Yes Yes Yes Observations 28,848 28,848 28,848 15,256 15,256 15,256 AdjR20.575 0.459 0.577 0.596 0.466 0.587 Note: This table presents ordinary least squares (OLS) and control function approach (CFA) estimations, with binarydependent variables based on the M/Bdecomposition of RhodesKropf etal.(2005). TSIE (binary) equals one if a firm's implied shortterm value is below its implied longrun value, both based on industry multiples. LRtB (binary) equals one if the implied longrun value of a firm is above its current book value. FSE (binary) is equal to one if a firm's market value is below its shortterm value. These variables are zero if the conditions are not met. Columns (1)–(3) cover the whole sample from 1993 to 2021. Columns (4)–(6) focus on the period from 1999 to 2011, which is around the implementation of the ETD. The variable LBO is defined as the ratio of LBO deals to the number of firms within a country–industry–year. For the CFA estimations, the following firststage regression is applied: The residuals ( 𝜈c,j,t ) from this regression, referred to as Residuals (firststage), are included in the analysis of Columns (4)–(6) to correct for the endogenous variable LBO. The binary variable Post equals one for firms in France (Germany, Italy, Spain, and the United Kingdom) from the implementation date of May 20, 2006 (July 8, 2006; July 19, 2007; April 13, 2007; and May 20, 2006) to the end of the year 2011. It is zero from 1999 until the day before each country's ETD implementation date. The binary variable Treat equals one for firms in treated countries (Germany, Spain) by the ETD and zero for firms in control countries (France, Italy, and the United Kingdom). Treat × Post is the interaction of the two binary variables. Mc,j,t − 1 represent lagged macroeconomic variables, while Ic,j,t − 1 captures industryspecific variables. 𝛾c , 𝜂j , and 𝛿t denote country, industry, and time fixed effects, respectively (see Table3, Column (2) for the regression results). For variable descriptions, see TableA1. Firms in the utility and financial industries, as well as government entities, are excluded. All continuous variables are winsorized at the 1% level at both ends. All regressions include a constant. Standard errors are clustered at the firm level, and tvalues are reported in parentheses. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. LBOc,j,t = 𝛼 + 𝛽1Treatc,j × Postt + 𝛽2Mc,t − 1 + 𝛽3Ic,j,t − 1 + 𝛾c + 𝜂j + 𝛿t + 𝜈c,j,t
| 537 KATHAN TABLE 9 Heterogeneity in industry peers' profitability—Firmlevel. Panel A EBITDAto −assetst+1 Operating margint+1 EBITDAto −assetst+1 Operating margint+1 EBITDAto −assetst+1 Operating margint+1 EBITDAto −assetst+1 Operating margint+1 Log(1 + Assets) M/Bratio ≤Median >Median ≤Median >Median (1) (2) (3) (4) (5) (6) (7) (8) LBO 0.0777*** 0.1158*** 0.0180*** 0.0265** 0.0213* 0.0481** 0.0650*** 0.0880*** (4.43) (2.78) (3.14) (2.37) (1.89) (2.17) (4.35) (2.90) Residuals (firststage) Yes Yes Yes Yes Yes Yes Yes Yes Controls Yes Yes Yes Yes Yes Yes Yes Yes Macro controls Yes Yes Yes Yes Yes Yes Yes Yes Country FE Yes Yes Yes Yes Yes Yes Yes Yes Industry FE Yes Yes Yes Yes Yes Yes Yes Yes Year FE Yes Yes Yes Yes Yes Yes Yes Yes Observations 16,152 14,237 16,246 16,146 17,236 16,334 15,158 14,046 AdjR20.277 0.179 0.144 0.108 0.247 0.115 0.288 0.202 Panel B EBITDAto −assetst+1 Operating margint+1 EBITDAto −assetst+1 Operating margint+1 EBITDAto −assetst+1 Operating margint+1 EBITDAto −assetst+1 Operating margint+1 Book leverage Ownership concentration ≤Median >Median ≤Median >Median (1) (2) (3) (4) (5) (6) (7) (8) LBO 0.0694*** 0.0910*** 0.0269*** 0.0328 0.0287** 0.0496 0.0494*** 0.0683** (4.52) (3.34) (2.86) (1.44) (2.36) (1.60) (3.38) (2.37) Residuals (firststage) Yes Yes Yes Yes Yes Yes Yes Yes Controls Yes Yes Yes Yes Yes Yes Yes Yes Macro controls Yes Yes Yes Yes Yes Yes Yes Yes Country FE Yes Yes Yes Yes Yes Yes Yes Yes Industry FE Yes Yes Yes Yes Yes Yes Yes Yes Year FE Yes Yes Yes Yes Yes Yes Yes Yes Observations 16,606 14,889 15,535 15,320 12,803 11,926 13,050 12,321 AdjR20.269 0.188 0.185 0.133 0.302 0.204 0.252 0.127 Note: This table presents subsample analyses using control function approach (CFA) estimations, with the oneyear lead dependent variables EBITDAtoassets and Operating margin. The analysis uses the restricted sample from 1999 to 2011, which is around the implementation of the ETD. The variable LBO is defined as the ratio of LBO deals to the number of firms within a country–industry–year. For the CFA estimations, the following firststage regression is applied: The residuals ( 𝜈c , j , t ) from this regression, referred to as Residuals (firststage), are included in the analysis to correct for the endogenous variable LBO. The binary variable Post equals one for firms in France (Germany, Italy, Spain, and the United Kingdom) from the implementation date of May 20, 2006 (July 8, 2006; July 19, 2007; April 13, 2007; and May 20, 2006) to the end of the year 2011. It is zero from 1999 until the day before each country's ETD implementation date. The binary variable Treat equals one for firms in treated countries (Germany, Spain) affected by the ETD and zero for firms in control countries (France, Italy, and the United Kingdom). Treat × Post is the interaction of the two binary variables. Mc , j , t − 1 represent lagged macroeconomic variables, while Ic , j , t − 1 captures industryspecific variables. 𝛾c , 𝜂j , and 𝛿t denote country, industry, and time fixed effects, respectively (see Table3, Column (2) for the regression results). Panel A splits the sample based on the median values of Log(1 + Assets) and the M/Bratio, while Panel B splits the sample at the median values of Book leverage and Ownership concentration. For variable descriptions, see TableA1. Firms in the utility and financial industries, as well as government entities, are excluded. All continuous variables are winsorized at the 1% level at both ends. Control and macro control variables correspond to those in Table4. All regressions include a constant. Standard errors are clustered at the firm level, and tvalues are reported in parentheses. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. LBOc,j,t = 𝛼 + 𝛽1Treatc,j × Postt + 𝛽2Mc,t − 1 + 𝛽3Ic,j,t − 1 + 𝛾c + 𝜂j + 𝛿t + 𝜈c,j,t
538 | KATHAN used as a defense tool (e.g., Garvey & Hanka,1999; Safieddine & Titman,1999) and improve profitability if firms use debt conservatively (e.g., Graham,2000). In line with this reasoning, Panel B indicates that the positive LBO effect is more pronounced for peer firms with equal or belowmedian “Book leverage.” European firms tend to have higher ownership concentration than US firms (e.g., Enriques & Volpin,2007), making them more difficult to take over (e.g., Holderness & Sheehan,1988; La Porta etal.,1999). However, the results in Columns (5)–(8) suggest that peer firms with larger shareholders (Columns (7) and (8)) benefit more from LBO activity. Large shareholders may enhance the monitoring of managers (e.g., Agrawal & Mandelker,1990; Shleifer & Vishny,1997), thereby incentivizing managers to improve firm performance. 6.2 | Industrylevel This section offers a more indepth analysis of the LBO channels and their impact on the takeover threat in relation to the profitability of industry peers. Similar to the subsample analysis in Section6.1, I divide the sample based on the median values of the following variables at the country–industry level: the No. of LBO deals, Turnover ratio, Herfindahl index, and M/B ratio. These variables serve as proxies for the various aspects of the takeover threat generated by LBO activity. Columns (1)–(4) of Panel A in Table10 account for the low or high levels of followon acquisitions, while Columns (5)–(8) capture these levels for industrywide agency problems. I use the turnover ratio at the countyindustry level as a proxy for the agency costs, as a higher turnover ratio indicates more efficient asset utilization and suggests an inverse relationship with agency costs (e.g., Ang etal.,2000). The results suggest that industry peers with fewer LBO deals in their industry tend to have a slightly stronger LBO effect on their profitability. Moreover, industry peers with turnover ratios equal or below their country–industry median, indicating higher agency costs, demonstrate a more pronounced effect with LBO activity. Even though Table6 does not indicate an improvement in the internal corporate governance structure, industry peers with higher agency costs are more sensitive to LBOs in enhancing their profitability. In Panel B of Table10, Columns (1)–(4) show similar effects on industry peers in the competitive and noncompetitive subsamples. The competition channel is associated with an increase in investment activity or changes in the organizational structure of industry peers, which Section5.2 does not support. Regarding the subsamples of the country–industry M/B ratio, the magnitude of LBO on the dependent variables appears to be more pronounced in the highM/B ratio subsample (Columns (7)–(8)), which aligns with the results in Table9. Overall, this section provides evidence that, first, firms characterized by small size, growth potential, financial flexibility, and concentrated ownership tend to gain mostly from LBO activity. Second, peer firms in industries with high agency costs and strong growth opportunities, which are also related to the potential signals of LBOs, tend to exhibit a more pronounced response to LBOs. 7 | ROBUSTNESS TESTS Using the ETD as an instrumental variable in the first stage of the CFA regressions corrects for the potential endogeneity of LBO activity, thereby helping to investigate the impact of LBOs on their industry peers. In addition to the argumentation in Section4 regarding the exogeneity of the ETD, it is essential to consider a firm's institutional and legal context (Karpoff & Wittry,2018) owing to omitted variables. In this context, it is possible that unrelated court decisions around the ETD in EU Member States may have influenced the institutional environment (Dissanaike etal.,2021) and, in turn, impacted the LBO market within these countries. However, the ETD offers several benefits regarding the institutional framework compared to changes in business combination laws in the US context. First, the legislative power of court decisions in the EU is comparatively weaker than in the United States. Second, the ETD covers several different regulatory elements, while these elements are distinct laws in the United States that are more challenging to account for (Dissanaike etal.,2021; Karpoff & Wittry,2018). Furthermore, the empirical models used in the firststage regressions include country, industry, and time fixed effects, which account for institutional and countryspecific differences. To empirically test whether other events or court decisions around the ETD influence the performance of firms, I run two placebo tests in Panel A of TableA3 (AppendixA). More specifically, the implementation dates in the countries were shifted to both ends within my sample. In Columns (1)–(3), the placebo ETD covers the years 1993 to 2002 (“Before”), and Columns (4)–(6) cover the years from 2012 to 2021 (“After”).
| 539 KATHAN TABLE 10 Heterogeneity in industry peers' profitability—Industrylevel. Panel A EBITDAto −assetst+1 Operating margint+1 EBITDAto −assetst+1 Operating margint+1 EBITDAto −assetst+1 Operating margint+1 EBITDAto −assetst+1 Operating margint+1 Country–industry – No. of LBO deals Country–industry – Turnover ratio ≤Median >Median ≤Median >Median (1) (2) (3) (4) (5) (6) (7) (8) LBO 0.0388*** 0.0780*** 0.0465*** 0.0460 0.0462*** 0.1264** 0.0344*** 0.0197* (3.02) (3.20) (3.25) (1.40) (2.62) (2.25) (3.27) (1.74) Residuals (firststage) Yes Yes Yes Yes Yes Yes Yes Yes Controls Yes Yes Yes Yes Yes Yes Yes Yes Macro controls Yes Yes Yes Yes Yes Yes Yes Yes Country FE Yes Yes Yes Yes Yes Yes Yes Yes Industry FE Yes Yes Yes Yes Yes Yes Yes Yes Year FE Yes Yes Yes Yes Yes Yes Yes Yes Observations 16,980 15,955 15,421 14,431 14,446 12,786 17,942 17,588 AdjR20.239 0.141 0.297 0.200 0.229 0.120 0.279 0.183 Panel B EBITDAto −assetst+1 Operating margint+1 EBITDAto −assetst+1 Operating margint+1 EBITDAto −assetst+1 Operating margint+1 EBITDAto −assetst+1 Operating margint+1 Country–industry – Herfindahl index Country–industry – M/Bratio ≤Median >Median ≤Median >Median (1) (2) (3) (4) (5) (6) (7) (8) LBO 0.0481*** 0.0735** 0.0446*** 0.0561*** 0.0312*** 0.0354** 0.0667*** 0.1023*** (3.06) (2.41) (4.02) (2.58) (2.91) (2.29) (4.00) (2.72) Residuals (firststage) Yes Yes Yes Yes Yes Yes Yes Yes Controls Yes Yes Yes Yes Yes Yes Yes Yes Macro controls Yes Yes Yes Yes Yes Yes Yes Yes Country FE Yes Yes Yes Yes Yes Yes Yes Yes Industry FE Yes Yes Yes Yes Yes Yes Yes Yes Year FE Yes Yes Yes Yes Yes Yes Yes Yes Observations 17,453 16,150 14,944 14,235 17,271 16,672 15,124 13,708 AdjR20.239 0.141 0.297 0.200 0.229 0.120 0.279 0.183 Note: This table presents subsample analyses using control function approach (CFA) estimations, with the oneyear lead dependent variables EBITDAtoassets and Operating margin. The analysis uses the restricted sample from 1999 to 2011, which is around the implementation of the ETD. The variable LBO is defined as the ratio of LBO deals to the number of firms within a country–industry–year. For the CFA estimations, the following firststage regression is applied: The residuals ( 𝜈c , j , t ) from this regression, referred to as Residuals (firststage), are included in the analysis to correct for the endogenous variable LBO. The binary variable Post equals one for firms in France (Germany, Italy, Spain, and the United Kingdom) from the implementation date of May 20, 2006 (July 8, 2006; July 19, 2007; April 13, 2007; May 20, 2006) to the end of the year 2011. It is zero from 1999 until the day before each country's ETD implementation date. The binary variable Treat equals one for firms in treated countries (Germany, Spain) affected by the ETD and zero for firms in control countries (France, Italy, and the United Kingdom). Treat × Post is the interaction of the two binary variables. Mc , j , t − 1 represent lagged macroeconomic variables, while Ic , j , t − 1 captures industryspecific variables. 𝛾c , 𝜂j , and 𝛿t denote country, industry, and time fixed effects, respectively (see Table3, Column (2) for the regression results). Panel A splits the sample based on the country–industry median values of No. of LBO deals and the Turnover ratio, while Panel B splits the sample at the country– industry median values of the Herfindahl index and the M/Bratio. For variable descriptions, see TableA1. Firms in the utility and financial industries, as well as government entities, are excluded. All continuous variables are winsorized at the 1% level at both ends. Control and macro control variables correspond to those in Table4. All regressions include a constant. Standard errors are clustered at the firm level, and tvalues are reported in parentheses. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. LBOc,j,t = 𝛼 + 𝛽1Treatc,j × Postt + 𝛽2Mc,t − 1 + 𝛽3Ic,j,t − 1 + 𝛾c + 𝜂j + 𝛿t + 𝜈c,j,t
540 | KATHAN The results display insignificant coefficients for Treat × Post in all columns except for Column (1), which shows a significantly negative coefficient. This suggests that other events surrounding the ETD do not significantly impact the behavior of firms. Therefore, the positve and significant effect observed in the main results reflect the ETD's impact on industry peers' profitabilty by altering LBO activity and the takeover threat in treated countries. Furthermore, LBO activity consistently shows a positive and statistically significant relationship with the dependent variables across all models. Panel B of TableA3 presents alternative proxies for the LBO variable. Columns (1)–(3) employ Vactivity, defined as the log of one plus the total deal value within a country–industry–year (e.g., Haddad etal.,2017). In Columns (4)–(6), VBactivity is used, which is the ratio of Vactivity to the number of firms within a country–industry–year. Both proxies capture, to some extent, the value component of LBO deals, with larger deals likely having stronger implications for industry peers. In all models, I employ CFA regressions, using the residuals from the first stage of the respective LBO proxy. The findings from this analysis further demonstrate a significant effect of these alternative proxies on the profitability of industry peers, reinforcing the main results presented in Section5.1. I also analyze M&A dealsxvi instead of LBOs, applying the same definition for M&A activity as for LBO activity. Since LBOs may signal the start of a merger sequence (Harford etal.,2016), which could also contribute to an increased takeover threat within an industry. Consequently, industry peers might respond to M&A activity. However, I do not find a positive relationship between M&A activity and the profitability of industry peers (TableB3, AppendixA in AppendixS1). Additionally, I investigate whether the ETD influences M&A activity, but I find no evidence to suggest that the ETD impacts M&A activity (TableB2, AppendixA in AppendixS1). This finding may help explain the mixed evidence in the literature regarding ETD and takeover efficiency gains (Dissanaike etal.,2021; HumpheryJenner,2012; Wang & Lahr,2017). 8 | CONCLUDING REMARKS AND DISCUSSION This study examines the impact of leveraged buyouts on industry peers within the European context. Currently, only crosscountry studies on the aggregate industry level (Aldatmaz & Brown,2020; Bernstein etal.,2017) exist that investigate the research question in more detail. I provide new evidence at the individual peer level, demonstrating that LBO announcements positively influence industry peers' profitability. Using CFA regressions, I employ the ETD as an instrument in the firststage regression and use the residuals to correct for the endogeneity of LBOactivity. The findings suggest that the improvement in profitability is attributed to industry peers' more efficient asset utilization and increased cost efficiency. However, it also shows that the timing of PE investors plays a crucial role as they select industries when they are at the bottom of their average profitability. In this context, industry developments, such as growth opportunities and especially undervaluation, are important factors in explaining the overall effect. In particular, the attribution of the selection channel differs from other studies in the LBO context, which primarily link the improvements to positive competitive spillovers (e.g., Aldatmaz & Brown,2020) that also strengthen corporate governance (e.g., Feng & Rao,2022). However, this study finds no evidence that European peer firms adjust their operations in response to variables that are more susceptible to these channels. European firms generally have a less dispersed ownership structure than US firms (e.g., Enriques & Volpin,2007). Consequently, the problem of separation between ownership and control is less pronounced in Europe; therefore, peer firms do not need to improve their internal corporate governance structures. Furthermore, the subsample analyses in Section6 indicate that firms with larger shareholders benefit more from LBOs, suggesting that enhanced monitoring activities may incentivize managers to run their firms efficiently. While industry peers, on average, do not make specific adjustments to their corporate governance, they tend to show greater improvements when operating in a country–industry with high agency costs. This suggests that the enhancements observed among industry peers may stem from efforts to reduce agency costs. Additionally, European LBO deals tend to be smaller than US LBO deals,xvii which could influence peers differently. First, larger deals are more likely to impact competition within an industry. Second, the takeover threat is probably limited to smaller firms. The subsample analysis reveals that the positive effect is more pronounced in the sample of firms with belowmedian size. However, there is no evidence of a reduced skillset of PE investors in Europe compared to the US. Thus, PE investors can exploit favorable market and industry conditions. In addition to the different sample focus (United States vs Europe) and using individual peers instead of aggregation, the difference could also result from analyzing all LBO deals rather than focusing exclusively on publictoprivate
| 541 KATHAN transactions. Research on the spillover effects of LBOs has primarily focused on publictoprivate transactions (e.g., Feng & Rao,2022; Harford etal.,2016; Oxman & Yildirim,2008; Truong & Walz,2024). However, private deals may have different implications for industry peers. For instance, studies find a positive impact on the valuation of target firms' peers when the sample includes only publictoprivate LBOs (e.g., Feng & Rao,2022; Slovin etal.,1991), but a negative impact when the majority of deals are privatetoprivate transactions (e.g., Hsu etal.,2011; Kathan & Tykvová,2024). Studies that include private deals, such as Kathan and Tykvová(2024), find a negative operating performance of industry peers after LBOs. Their analysis is limited to a restrictive sample of US deals, excluding many transactions. This restriction may understate the positive implications of LBO activity on industry peers. Moreover, their findings suggest that the negative effect becomes considerably smaller when these restrictive assumptions are relaxed. A potential limitation of my study is the relatively limited data coverage for European firms compared to that for US firms. However, I address some of these data limitations by focusing on the largest European economies, which account for most of Europe's GDP. Nevertheless, this approach may restrict the generalizability of the results. DATA AVAILABILITY STATEMENT The data that support the findings of this study are available from WRDS and LSEG Eikon. Restrictions apply to the availability of these data, which were used under license for this study. Data are available from the author(s) with the permission of WRDS and LSEG Eikon. ACKNOWLEDGMENT Open Access funding enabled and organized by Projekt DEAL. ORCID Manuel C. Kathan https://orcid.org/0000-0001-6369-3404 Endnotes i The study includes the five largest countries in terms of their GDPs in Europe: France, Germany, Italy, Spain, and the United Kingdom. They also provide the most buyout capital invested in Europe (Aldatmaz & Brown,2020). Additionally, the data coverage for the variables used in this study is significantly more complete for the selected countries compared to other European countries. ii Note, a substantial proportion of private deals do not disclose their deal value. Table1 reports the average deal value of the available data, which is driven by public and large private deals. iii Descriptive statistics on the distribution of peer firms across countries and industries are provided in the AppendixA, TableA2. iv The number of observations varies across variables due to differences in the availability of data for each respective variable. v For the empirical analysis, I use LBO for LBO activity because many private deals do not provide a deal value. vi Minimum rules concerning mandatory bid rule, breakthrough rule, board neutrality rule, squeezeout right, and sellout right. For more information on the minimum rules, see Goergen etal.(2005) and Clerc etal.(2012). vii PE funds aim to acquire full ownership in the majority of deals within the sample, resulting in an average ownership stake of 96.78%. viii I closely follow Dissanaike etal.(2021), who rely on the classification of Clerc etal.(2012) to map ETDrelated changes to the countries and specify the treatment and control group accordingly. ix For the control countries, the implementation dates were May 20, 2006, in France and the United Kingdom, and July 19, 2007, in Italy (Wang & Lahr,2017). x The variables Treat and Post are collinear with the fixed effects. xi In untabulated results, I also perform the Oster(2019) diagnostic test for unobserved factors and coefficient stability in the secondstage regression. The “Breakdown Delta” for an R2 of 1.00, which is 71.6%, provides evidence that the findings are not sensitive to omitted variable bias. xii Corporate governance variables are retrieved from Capital IQ and Boardex Europe. Note that Boardex Europe does not cover firms from the United Kingdom. xiii I show the formal derivation of the M/B decomposition in the AppendixB in AppendixS1. xiv The components show the deviations between a firm's actual and implied value, which can be above or below the actual value. However, this analysis focuses on firms categorized as growth or undervalued firms. xv I retrieve ownership data from Datastream. xvi TableB1 and AppendixA in AppendixS1 provides descriptive statistics for M&A target firms. xvii In untabulated results, the mean deal value of US targets is 690.24 mio $ compared to 223.56 mio $ in Table1.
542 | KATHAN REFERENCES Acharya, V. V., Gottschalg, O. F., Hahn, M., & Kehoe, C. (2013). Corporate governance and value creation: Evidence from private equity. The Review of Financial Studies, 26(2), 368–402. Agrawal, A., & Mandelker, G. N. (1990). Large shareholders and the monitoring of managers: The case of antitakeover charter amendments. Journal of Financial and Quantitative Analysis, 25(2), 143–161. Aldatmaz, S., & Brown, G. W. (2020). Private equity in the global economy: Evidence on industry spillovers. Journal of Corporate Finance, 60, 101524. Ang, J. S., Cole, R. A., & Lin, J. W. (2000). Agency costs and ownership structure. The Journal of Finance, 55(1), 81–106. Axelson, U., Jenkinson, T., Strömberg, P., & Weisbach, M. S. (2013). Borrow cheap, buy high? The determinants of leverage and pricing in buyouts. The Journal of Finance, 68(6), 2223–2267. Bebchuk, L., Cohen, A., & Ferrell, A. (2008). What matters in corporate governance? Review of Financial Studies, 22(2), 783–827. Bernstein, S., Lerner, J., Sorensen, M., & Stromberg, P. (2017). Private equity and industry performance. Management Science, 63(4), 1198–1213. Bharath, S., Dittmar, A., & Sivadasan, J. (2014). Do goingprivate transactions affect plant efficiency and investment? The Review of Financial Studies, 27(7), 1929–1976. Bhushan, R. (1989). Firm characteristics and analyst following. Journal of Accounting and Economics, 11(2), 255–274. Boucly, Q., Sraer, D., & Thesmar, D. (2011). Growth LBOs. Journal of Financial Economics, 102(2), 432–453. Chevalier, J. A. (1995a). Capital structure and productmarket competition: Empirical evidence from the supermarket industry. The American Economic Review, 85(3), 415–435. Chevalier, J. A. (1995b). Do LBO supermarkets charge more? An empirical analysis of the effects of LBOs on supermarket pricing. The Journal of Finance, 50(4), 1095–1112. Chhaochharia, V., Grinstein, Y., Grullon, G., & Michaely, R. (2017). Product market competition and internal governance: Evidence from the Sarbanes Oxley Act. Management Science, 63(5), 1405–1424. Clerc, C., Demarigny, F., Valiante, D., & Aramendia, M.d. M. (2012). External study on behalf of the European Commission. The takeover bids directive assessment report. A legal and economic assessment of European Takeover Regulation. Marcus Partners and Centre for European Policy Studies. Coles, J. L., Daniel, N. D., & Naveen, L. (2006). Managerial incentives and risktaking. Journal of Financial Economics, 79(2), 431–468. Cornelli, F., & Karakaş, O. (2012). Corporate governance of LBOs: The role of boards. Unpublished working paper, London Business School. Denis, D. J. (1994). Organizational form and the consequences of highly leveraged transactions: Kroger's recapitalization and Safeway's LBO. Journal of Financial Economics, 36(2), 193–224. Denis, D. J., & Serrano, J. M. (1996). Active investors and management turnover following unsuccessful control contests. Journal of Financial Economics, 40(2), 239–266. Dissanaike, G., Drobetz, W., Momtaz, P. P., & Rocholl, J. (2021). The economics of law enforcement: Quasiexperimental evidence from corporate takeover law. Journal of Corporate Finance, 67, 101849. Dittmar, A., Li, D., & Nain, A. (2012). It pays to follow the leader: Acquiring targets picked by private equity. Journal of Financial and Quantitative Analysis, 47(5), 901–931. Dyck, A., & Zingales, L. (2004). Private benefits of control: An international comparison. The Journal of Finance, 59(2), 537–600. Enriques, L., & Volpin, P. (2007). Corporate governance reforms in Continental Europe. Journal of Economic Perspectives, 21(1), 117–140. Faccio, M., & Lang, L. H. (2002). The ultimate ownership of Western European corporations. Journal of Financial Economics, 65(3), 365–395. Feng, H., & Rao, R. P. (2022). The positive externalities of leveraged buyouts. Journal of Banking & Finance, 135, 106360. Frank, M. Z., & Goyal, V. K. (2009). Capital structure decisions: Which factors are reliably important? Financial Management, 38(1), 1–37. Garvey, G. T., & Hanka, G. (1999). Capital structure and corporate control: The effect of antitakeover statutes on firm leverage. The Journal of Finance, 54(2), 519–546. Giroud, X., & Mueller, H. M. (2010). Does corporate governance matter in competitive industries? Journal of Financial Economics, 95(3), 312–331. Goergen, M., Martynova, M., & Renneboog, L. (2005). Corporate governance convergence: Evidence from takeover regulation reforms in Europe. Oxford Review of Economic Policy, 21(2), 243–268. Gompers, P., Ishii, J., & Metrick, A. (2003). Corporate governance and equity prices. The Quarterly Journal of Economics, 118(1), 107–155. Gorbenko, A. S., & Malenko, A. (2014). Strategic and financial bidders in takeover auctions. The Journal of Finance, 69(6), 2513–2555. Graham, J. R. (2000). How big are the tax benefits of debt? The Journal of Finance, 55(5), 1901–1941. Grennan, J. (2019). Dividend payments as a response to peer influence. Journal of Financial Economics, 131(3), 549–570. Guo, S., Hotchkiss, E. S., & Song, W. (2011). Do buyouts (still) create value? The Journal of Finance, 66(2), 479–517. Haddad, V., Loualiche, E., & Plosser, M. (2017). Buyout activity: The impact of aggregate discount rates. The Journal of Finance, 72(1), 371–414. Harford, J., Stanfield, J., & Zhang, F. (2019). Do insiders time management buyouts and freezeouts to buy undervalued targets? Journal of Financial Economics, 131(1), 206–231. Harford, J., Stanfield, J. R., & Zhang, F. (2016). How does an LBO impact the target's industry? Unpublished working paper, University of Washington. Hartzell, J. C., Ofek, E., & Yermack, D. (2004). What's in it for me? CEOs whose firms are acquired. The Review of Financial Studies, 17(1), 37–61. Hausman, J. A. (1978). Specification tests in econometrics. Econometrica, 46(6), 1251–1271.