Financial debt contracting and managerial agency problems

Imbierowicz, Björn; Streitz, Daniel Article — Published Version Financial debt contracting and managerial agency problems Financial Management Provided in Cooperation with: John Wiley & Sons Suggested Citation: Imbierowicz, Björn; Streitz, Daniel (2024) : Financial debt contracting and managerial agency problems, Financial Management, ISSN 1755-053X, Wiley, Hoboken, NJ, Vol. 53, Iss. 1, pp. 99-118, https://doi.org/10.1111/fima.12444 This Version is available at: https://hdl.handle.net/10419/294013 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. http://creativecommons.org/licenses/by-nc-nd/4.0/ DOI: 10.1111/fima.12444 ORIGINAL ARTICLE Financial debt contracting and managerial agency problems Björn Imbierowicz1Daniel Streitz2,3 1Deutsche Bundesbank, Central Office, Research Centre, Frankfurt am Main, Germany 2Halle Institute for Economic Research, Halle (Saale), Germany 3Faculty of Economics and Business Administration, Friedrich Schiller University Jena, Jena, Germany Correspondence Björn Imbierowicz, Deutsche Bundesbank, Central Office, Research Centre, Mainzer Landstrasse 46, 60325 Frankfurt am Main, Germany. Email: [email protected] Abstract This paper analyzes if lenders resolve managerial agency problems in loan contracts using sweep covenants. Sweeps require a (partial) prepayment when triggered and are included in many contracts. Exploiting exogenous reductions in analyst coverage due to brokerage house mergers and closures, we find that increased borrower opacity significantly increases sweep use. The effect is strongest for borrowers with higher levels of managerial entrenchment and if lenders hold both debt and equity in the firm. Overall, our results suggest that lenders implement sweep covenants to mitigate managerial agency problems by limiting contingencies of wealth expropriation. KEYWORDS agency problems, covenant, loan contract, sweep provision 1INTRODUCTION The allocation of control rights between creditors and shareholders is the main object of interest in a large body of literature (see, e.g., Christensen et al., 2016 for an overview). In contrast, the potential conflict between creditors and the firm’s management is often neglected. However, actions by entrenched managers can have adverse consequences for shareholders and lenders alike, often without triggering standard financial covenants. For instance, assets sales or the misuse of corporate cash reserves may affect firms’ default risk or collateral value. In this paper, we document evidence that is consistent with sweep covenants—which are included in almost half of all loan contracts in our sample—being used to address potential adverse consequences of managerial actions on lenders. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. © 2024 The Authors. Financial Management published by Wiley Periodicals LLC on behalf of Financial Management Association International. Financial Management. 2024;53:99–118. wileyonlinelibrary.com/journal/fima 99 100 IMBIEROWICZ AND STREITZ Sweep clauses do not oblige firms to maintain certain balance sheet or profit and loss statement ratios. Instead, sweep covenants require the borrower to immediately repay a given percentage of the loan when certain cash proceeds become available (e.g., from asset sales, debt or equity issuance, or insurance proceeds). By requiring payouts to creditors, sweep clauses may help discipline management. First, sweeps restrict managers’ ability to accumulate excess cash flow and hence may reduce managers’ flexibility. Second, sweeps may have incentive effects as they reduce the benefits from strategic asset sales or security issues (Lang et al., 1995), as part of the proceeds must be used to pay down debt. This may limit managers’ incentives to engage in such behavior ex ante. Third, the misuse of cash windfalls, for example, from unexpected insurance proceeds, can be restricted (Blanchard et al., 1994; Glaser et al., 2013). Overall, the implementation of sweep covenants in loan contracts implies that managerial flexibility is curtailed ex ante, limiting potential adverse effects of managerial agency problems on debt holders. We use exogenous variation in outside monitoring to examine the link between managerial agency problems and loan contract design. Specifically, we utilize changes in analyst coverage of borrowers induced by brokerage house mergers or closures (see, among others, Hong & Kacperczyk, 2010). Following the merger of two brokerage houses, due to overlapping coverage, a redundant analyst is typically let go (Wu & Zhang, 2009). This results in a decrease in analyst coverage and accordingly a lower degree of outside monitoring of the firm, which is independent of any firm and manager characteristics.1Given that managerial agency problems are more severe for firms with less outside monitoring (e.g., Jensen & Meckling, 1976),2this creates a natural setting to study the effects of agency problems on loan contract design. We find that an exogenous decrease in outside monitoring increases the probability of including a sweep covenant in a loan contract by nine percentage points. This corresponds to a sizable increase of sweep use in treatment firms’ loan contracts of 21% relative to the unconditional mean. We also observe that the number of sweep covenants in a contract increases. In contrast, we do not observe an effect of the change in outside monitoring on financial covenants, which are typically used to address conflicts of interest between shareholders and debtholders. Further tests show that the parallel trends assumption holds; that is, there is no pre-trend in sweep use for treated versus control firms prior to a brokerage house merger or closure, providing support for our identification strategy. We provide a series of cross-sectional tests that corroborate our main result. First, we investigate the effect of outside monitoring for firms with different initial levels of analyst coverage. We hypothesize that effects should be more pronounced for more opaque firms, that is, firms with less initial coverage. Our results indicate that lenders implement sweep covenants following a reduction in analyst coverage, especially when firm transparency is low ex-ante. Second, we investigate the role of other corporate governance mechanisms. External monitoring and internal monitoring (e.g., board oversight) might be either substitutes or complements (e.g., Irani & Oesch, 2013). We therefore subdivide firms by their pre-treatment governance quality. For example, managerial agency problems are more severe for companies with excess funds or long CEO tenure (Jensen, 1986; Kalcheva & Lins, 2007;Lie,2000), while more institutional ownership reduces managerial entrenchment (Chava et al., 2010; Finkelstein & Hambrick, 1989). Our results show that the effect of a reduction of analyst coverage on sweep covenant use is particularly pronounced among firms with poor corporate governance. In contrast, the effect is limited for well-governed firms. Our results thereby complement the finding of Irani and Oesch (2013) that external monitoring by analysts and other governance mechanisms are substitutes. One potential concern might be that even if managerial agency problems affect debtholders, actions taken by equity holders might be better suited to discipline managers. For instance, shareholders might incentivize firms to use excess cash for payouts (dividends or share repurchases) or interest payments, that is, to lever up (Easterbrook, 1It is also independent of the individual analyst. For instance, Hong and Kacperczyk (2010) and Irani and Oesch (2013) provide evidence that analyst coverage reductions are concentrated at the target brokerage house, that is, the reduction of analysts follows a rule which is not related to skill. 2For instance, Kelly and Ljungqvist (2012) document that a decrease in analyst coverage increases information asymmetry (see also Brennan & Subrahmanyam, 1995; Ellul & Panayides, 2018). Irani and Oesch (2013) provide evidence that financial reporting quality is lower following a reduction in coverage. Dyck et al. (2010) show that information intermediaries are often among the first to detect managerial misbehavior. Chen et al. (2015) present results, which are consistent with the conjecture that managerial agency problems increase following analyst coverage reductions. IMBIEROWICZ AND STREITZ 101 1984; Jensen & Meckling, 1976). Such actions, however, might not be in the interest of debtholders. Increasing leverage increases default risk and may exacerbate risk-shifting incentives of shareholders (Chava et al., 2010; Maxwell & Stephens, 2003). Hence, debtholders may prefer other means of addressing manager misbehavior, such as sweep contracts.3 Furthermore, most firms operate well above their default barrier such that the contingencies that sweeps address are particularly relevant for equity holders. If sweeps are used by lenders to address managerial agency problems, we should therefore expect to observe a stronger increase in sweep use for lenders that simultaneously also hold equity of the borrowing firm (dual holding). In a third step, we therefore identify dual holders. Our results confirm that lenders increase sweep use more in loan contracts of firms in which they also hold equity. The work by Huang (2010) is closest to this paper. He argues that sweep provisions can mitigate conflicts of interest between creditors and shareholders. Sweeps may shorten the effective maturity of loans by requiring firms with excess cash flow to prepay their debt. This may force firms to return to the capital market more frequently, giving lenders more control over firms’ investment decisions in future contract negotiations. Thereby, the ability of firms to pursue investment strategies that benefit shareholders but hurt debtholders might be reduced. Consistent with this view, Huang (2010) observes a positive correlation between sweep covenants and firm leverage and institutional ownership. We complement the work by Huang (2010) by documenting an increased use of sweep provisions in loan contracts following an exogenous decrease in borrower transparency. This effect is particularly pronounced for more opaque borrowers and borrowers with a higher level of managerial entrenchment. This evidence suggests that managerial agency problems are a first-order determinant of sweep provisions.4 Our work also contributes to several other strands of literature. First, we add to the literature on the impact of managerial agency problems on debt contracting. Chava et al. (2010) document that factors associated with managerial entrenchment positively correlate with investment restrictions in bond contracts and negatively correlate with subsequent financing and dividend restrictions. Their findings suggest that managerial agency problems are a factor in debt contract design. Begley and Feltham (1999) document that managerial share ownership has a significant effect on the inclusion of bond covenants that restricts additional borrowing or dividend payouts. This literature exclusively focuses on bonds. We contribute to this literature by documenting that lenders implement sweep covenants in loan contracts to address contingencies of wealth expropriation by firm management. Our identification strategy thereby allows us to establish a causal relationship between the degree of outside monitoring, used as a proxy for the degree of managerial agency problems, and the use of sweep clauses in loan contracts. Second, we contribute to the literature on sweep covenants in loan contracts. Despite their frequent use, most of the literature does not specifically focus on sweep covenants. Instead, sweep covenants are often included in covenant intensity indices (Bradley & Roberts, 2015; Demiroglu & James, 2010) or investigated in relation to financial covenants (Christensen & Nikolaev, 2012). As discussed above, one notable exception is Huang (2010). We add to this literature by documenting an increased use of sweep provisions in loan contracts when borrower transparency decreases. Furthermore, we provide empirical evidence that this effect is particularly pronounced for more opaque borrowers and borrowers with a higher level of managerial entrenchment, suggesting that sweep provisions are used to address managerial agency problems. Finally, we also add to the larger literature on lender rights and control through loan contract design. Most prior work focuses on the allocation of control rights between creditors and shareholders via financial covenants. Roberts (2015), Li et al. (2016), and Nikolaev (2018) investigate the allocation of control rights within a loan, while Chava and Roberts (2008), Roberts and Sufi (2009a, 2009b), Nini et al. (2009, 2012), Demerjian and Owens (2016), and Freudenberg et al. (2017) investigate the effects of shifts in control rights to creditors on the firm level, and Demerjian (2011) 3Stulz (1988) argues that takeover threats can constrain mangers, which, however, might also not be beneficial for debtholders. For instance, the financial risk of the target firm increases if the takeover is accompanied by a large increase in leverage (Chava et al., 2009). 4Clearly, managerial agency problems and conflicts of interest between shareholders and debtholders are not mutually exclusive. That is, our evidence does not preclude that sweep provisions may also help mitigate creditor–shareholder conflicts of interest. 102 IMBIEROWICZ AND STREITZ and Christensen and Nikolaev (2012) distinguish between different types of covenants. Murfin (2012) and Demerjian and Owens (2016) develop and investigate aggregate measures of financial covenant strictness. Hallman et al. (2022) also examine loan contract terms around changes in analyst coverage of non-financial firms. They provide evidence that an increase in information asymmetry as a result of a reduction in analyst coverage is related to higher loan spreads, a reduction in credit supply, and more restrictive covenants. We add to this literature by emphasizing the differing role of sweep covenants compared to other financial covenants and loan contract restrictions. Sweep covenants have the potential to mitigate managerial agency problems by limiting contingencies of wealth expropriation. The remainder of this paper proceeds as follows. Section 2describes our identification strategy and the data. Section 3reports the results of the empirical analysis. Section 4concludes. 2 EMPIRICAL SETUP AND DATA 2.1 Identification strategy A simple empirical strategy to examine the link between managerial agency issues and loan contracting would be to regress a measure of sweep or financial covenant use on proxies for the degree of misuse of corporate resources by managers, such as measures of firm transparency. While intuitively appealing, the estimates from such regressions are hard to interpret due to endogeneity problems. For instance, it could be that firms with higher managerial agency problems require more outside control and hence may have higher transparency levels. That is, a simple regression of sweep use on transparency could indicate a misleading positive relationship, as both variables are endogenously determined by the (unobservable) degree of managerial agency problems. To overcome this issue, we utilize exogenous changes in outside monitoring, which directly affect the firm’s information environment. In particular, we follow Hong and Kacperczyk (2010) and Irani and Oesch (2013) and study brokerage house mergers and closures.5Wu and Zhang (2009) find that subsequent to a merger of two brokerage houses with an active equity research department, the resulting entity lays off analysts to mitigate redundancies caused by overlapping coverage. As a result, companies that were previously covered by both brokerage houses experience a decline in analyst coverage. Importantly, the reduction in coverage is independent of unobservable firm and manager characteristics and is neither determined by the individual analyst nor by the firm for which the coverage is reduced. Hong and Kacperczyk (2010) and Irani and Oesch (2013) provide evidence that the analyst coverage reduction follows a rule unrelated to skill—in most cases, the analyst of the target brokerage house is let go. Similarly, the closure of a brokerage house leads to a decline in analyst coverage for affected firms. Further, particularly important in our setting, brokerage houses are generally not financial institutions granting loans; that is, it is unlikely that a brokerage house merger or closure has a direct effect on credit supply to the firms covered by the entities.6 Several studies provide evidence consistent with the first step required for our argument; that is, the idea that a loss in analyst coverage indeed increases information asymmetry as outside monitoring and information production is reduced. Kelly and Ljungqvist (2012), for instance, document that stock market-based measures for information asymmetry (e.g., probability of informed trading, bid-ask spreads) worsen following losses of analyst coverage (see also Brennan & Subrahmanyam, 1995; Ellul & Panayides, 2018). Irani and Oesch (2013) provide evidence that financial reporting quality worsens. The next step, that is, the link between (external) information production, managerial entrenchment, and managerial agency problems, goes back to Jensen and Meckling (1976) and has been examined in 5While broker research reports primarily target equity investors, they are also an important source of information used by (prospective) lenders. For instance, Irani and Oesch (2013) show that a reduction in analyst coverage leads to an overall lower financial reporting quality and that reporting quality has been shown to affect loan terms (Graham et al., 2008). 6Some brokerage houses are affiliated with financial institutions that also have lending business. However, as discussed in the next section, we do not find evidence that our results are driven by events that involve a brokerage house that is affiliated with an active lender in the syndicated loan market. IMBIEROWICZ AND STREITZ 103 several studies since.7Our main hypothesis is on the final step of this causal chain; that is, that sweep provisions can be used to limit managerial discretion if external oversight is (exogenously) reduced and managerial entrenchment increased. We follow Hong and Kacperczyk (2010) and Irani and Oesch (2013) and screen the SDC Mergers and Acquisition database for mergers of two financial institutions and limit the sample to firms with SIC code 6211 (“Investment Commodity Firms, Dealers and Exchanges”). We require that both brokerage houses are disseminating estimates to the I/B/E/S database and that both brokerage houses have an overlapping coverage of at least two stocks. Given that I/B/E/S does not assign analysts to individual brokerage houses after 2006, we end up with the same 13 mergers as Hong and Kacperczyk (2010) and Irani and Oesch (2013). In addition to mergers, we further identify 11 brokerage house closure events in I/B/E/S following Kelly and Ljungqvist (2012).8Finally, we merge the I/B/E/S information to LPC DealScan, which contains detailed loan-level information. Sweep and covenant information relates to a loan package which oftentimes includes several loan facilities. As is common in the literature (e.g., Nini et al., 2009, 2012), we analyze the data at the facility level to be able to account for factors varying at this level such as the size and the maturity of a facility. We also include fixed effects for the specific type of facility in our analyses.9 For each brokerage house merger, we identify all stocks that are covered by both merging parties in the year prior to the merger, that is, stocks with an “overlapping coverage.” Similarly, for closures, we identify all stocks that are covered by the brokerage house in the year prior to the closure. These firms are the focus of this paper and are in the following referred to as “treated.” We analyze all loan facilities contained in LPC DealScan in the symmetric 4-year window around each merger or closure, that is, a window consisting of 2 years before (720 days) the event and 2 years after the event.10 Note that in this setting, being treated is not a firm fixed effect; that is, each event affects a different set of firms. To construct symmetric windows and deal with overlapping events, we first construct separate samples for each event. These samples are then pooled. Accordingly, the same loan contract might be included in different windows when event windows are overlapping. We address this issue by including event (merger or closure) ×firm fixed effects in all our estimations, that is, focus on within-event variation across firms. Further, the use of a staggered design, that is, pooling event samples, addresses potential concerns otherwise associated with staggered difference-in-difference (DiD) frameworks (Goodman-Bacon, 2021). To account for systematic differences between treated and control firms, for each brokerage house merger or closure, we match untreated firms to each treated firm based on firm size (total assets).11 These firms form the control group. Deryugina et al. (2020) provide empirical evidence that this matching estimator generates more precise estimates than the standard DiD estimator. We further control for other differences across treated and control firms by including standard firm-level and loan contract control variables in our regressions, defined in more detail in the following section. To empirically implement our natural experiment and test how the use of sweep clauses (and financial covenants) changes following a shock to analyst coverage of the firm, we estimate versions of the following pooled 7Chen et al. (2015), for instance, provide evidence that following an exogenous decrease in outside monitoring, CEO compensation increases, the likelihood that management invests in value-destroying acquisitions increases, and managers are more likely to engage in earnings management activities. Irani and Oesch (2013) document that a reduction in outside monitoring reduces financial reporting quality; that is, management may strategically make financial statements opaquer to cover self-dealing. 8We identify fewer closure events compared to Kelly and Ljungqvist (2012), as we require firms that are covered by a brokerage house to also be active borrowers in the LPC DealScan database. 9One concern might be that results differ between term loans and revolvers. In unreported robustness tests, we also examine whether revolving loan facilities exhibit differential outcomes with respect to the implementation of sweeps in response to treatment and do not find this confirmed. Note that the trigger of a sweep typically implies a repayment of the loans included in a loan package in the following order: i. term loans, ii. cancellation of available revolving commitments, iii. prepayment and cancellation of used revolving commitments, and iv. repayment and cancellation of ancillary facilities. 10 Hong and Kacperczyk (2010) and Irani and Oesch (2013) analyze a 2-year window, whereas our window is 4 years. The reason is that their object of analysis is financial statement information, while we analyze loan issuances, which are in general less frequent to observe. That is, only a few firms issue a loan both in the year prior to the brokerage house merger as well as in the year afterward. The increase in time therefore allows for more statistical power. 11 We only match on total assets to increase the probability that a suitable control firm can be found for each treated firm. However, we confirm in Table 1 that after matching on total assets treated and control, firms are very comparable across most observable dimensions. 104 IMBIEROWICZ AND STREITZ panel (DiD) regression: SWEEPm,j,i,t =𝛼 m,i +𝛼 m,t +𝛽POST ×TREATEDm,i,t +𝜃 ′Yi,t +𝛿 ′Zm,j,i,t,(1) where SWEEPm,j,i,t is an indicator variable that equals one if loan jby firm iat time tin the estimation window around event (brokerage house merger or closure) mincludes a sweep clause, and zero otherwise. POSTm,t is a dummy variable that equals one in the period after the event m, and zero otherwise. TREATEDm,i is a dummy variable that equals one if firm iis part of the treatment sample for event m, and zero otherwise. am,i is a set of firm ×event fixed effects, am,t is a set of time ×event fixed effects, Yi,t is a set of firm characteristics, and Zm,j,i,t is a set of loan characteristics. Note that, following Irani and Oesch (2013), we do not include calendar year fixed effects as any period-specific effect will be captured by the merger (×time) fixed effects. The coefficient of interest is 𝛽, which captures the treatment effect. It shows the effect of the brokerage house merger/closure, and the associated reduction in analyst coverage, on the use of sweep covenants (or other outcome variables) in loan contracts. In all regressions, we report standard errors clustered at the firm level as treatment variation is mainly across firms.12 2.2 Sample selection and control variables We obtain data on security analyst coverage from I/B/E/S. For each event, we obtain all loans issued by public U.S. non-financial companies in a 720-day window before and after the brokerage house merger/closure date from LPC DealScan. We merge this sample with borrower balance sheet and income statement information from Compustat.13 Throughout the analysis, we control for basic firm characteristics. We control for firm size (log of total assets), leverage, market-to-book ratio, profitability, tangibility, interest coverage, current ratio, and credit rating. The latter is based on S&P and included via indicator variables for each rating notch. Further, we control for basic loan characteristics. While loan characteristics are important factors that can explain the use of financial covenants and sweep provisions, most loan terms are simultaneously determined, that is, endogenous. For instance, a borrower may pay a lower spread because a covenant is included in the contract. We therefore restrict our loan level control variables to these with a high likelihood of being independent of the decision to include a sweep covenant in the loan contract. We include the (log) loan size, (log) maturity, loan type, and loan purpose. The rationale is that these factors are in general determined by the firm prior to applying for a loan.14 Table 1shows descriptive statistics split by the treatment and control groups. The table shows that a high fraction of loans include sweep covenants: 46% (43%) of all loans in the control (treatment) sample include at least one sweep covenant. Sweeps can be classified by the source of the cash proceeds: (i) asset sale, (ii) debt issuance, (iii) equity issuance, (iv) excess cash flow, and (v) insurance proceeds. Table 1reports that on average, loans include 1.38 (1.27) sweep provisions. If we focus on the subset of loans that include at least one sweep, we find that on average, loans to treated (control) firms include 2.93 (3.02) sweep provisions (not tabulated). This indicates that usually a combination of sweep clauses is used. This should not be surprising given that firms are able to substitute between different sources of cash to some degree. For instance, if a loan includes an excess cash flow sweep but no security issuance sweeps, a manager could simply finance a project with debt or equity instead of using 12 In unreported robustness tests, we re-estimate our regression and cluster standard errors at several other levels. The results show that clustering does not seem to be a factor which substantially influences our results. We also estimate a Poisson model for our dependent variables number of sweeps and number of financial covenants following Cohn et al. (2022). Results are very comparable to the OLS estimates we provide in our tables. 13 We use Michael Robert’s Dealscan-Compustat Linking Database to merge Dealscan with Compustat (Chava & Roberts, 2008). We obtain borrower information from the last available fiscal year prior to the loan. 14 We acknowledge that these variables are not entirely independent of the loan contract design negotiations. For instance, a firm may require a “large” loan, but the exact size is determined by the design of the loan contract and an outcome of the negotiation with the lender. However, all our results remain virtually unchanged if we do not control for these factors. IMBIEROWICZ AND STREITZ 105 TABLE 1 Summary statistics for the treatment and control samples. Treatment group Control group N Mean Q1 Median Q3 SD N Mean Q1 Median Q3 SD diff. (p-value) Loan characteristics Facility amount (million USD) 1692 436.05 109.77 266.75 550.12 479.26 1849 454.47 115.00 274.42 575.00 516.49 18.42 (0.272) Maturity (months) 1671 41.67 12.00 36.00 60.00 26.85 1827 43.47 12.00 38.00 60.00 28.12 1.80 (0.053) SWEEP (0/1) 1692 0.43 0.00 0.00 1.00 0.50 1849 0.46 0.00 0.00 1.00 0.50 0.03 (0.227) #SWEEP 1692 1.27 0.00 0.00 3.00 1.68 1849 1.38 0.00 0.00 3.00 1.78 0.11 (0.079) FIN COV (0/1) 1692 0.94 1.00 1.00 1.00 0.23 1849 0.95 1.00 1.00 1.00 0.22 0.01 (0.684) #FIN COV 1692 1.99 1.00 2.00 3.00 1.08 1849 2.04 1.00 2.00 3.00 1.05 0.05 (0.198) #PERF COV 1692 1.56 1.00 1.00 2.00 1.18 1849 1.65 1.00 2.00 2.00 1.15 0.09 (0.018) #CAP COV 1692 0.43 0.00 0.00 1.00 0.55 1849 0.39 0.00 0.00 1.00 0.55 0.04 (0.012) Firm characteristics Total assets (million USD) 1690 5111.68 873.54 2031.54 5937.98 7394.65 1845 5062.0 822.67 2163.9 6104.0 6944.1 −49.62 (0.84) Leverage 1686 0.37 0.20 0.36 0.49 0.22 1839 0.37 0.22 0.37 0.50 0.21 0.00 (0.593) Market-to-book 1618 1.93 1.22 1.57 2.19 1.18 1735 1.74 1.13 1.45 1.95 1.00 −0.19 (0.00) Profitability 1686 0.19 0.09 0.15 0.26 0.16 1812 0.18 0.10 0.17 0.26 0.17 −0.01 (0.822) Tangibility 1681 0.40 0.17 0.37 0.59 0.25 1837 0.37 0.17 0.34 0.54 0.23 −0.03 (.000) Coverage 1648 10.71 2.62 4.90 10.56 19.80 1783 11.69 2.49 4.46 9.14 24.33 0.98 (0.197) Current ratio 1618 1.74 1.03 1.44 2.12 1.18 1678 1.55 0.98 1.38 1.95 0.86 −0.19 (0.000) Note: This table reports summary statistics for the sample of syndicated loans to non-financial North American borrowers. Statistics are reported separately for the treatment and control samples over the 4-year window surrounding brokerage house mergers. The treatment sample consists of loans obtained by firms that are covered by both merging brokerage houses in the year prior to the merger. The control sample consists of loans obtained by firms that are not affected byt he event. Treatment and control firms are matched based on firm size (total assets). All firm data are measured in real terms with 2000 as base year and are winsorized at the 1% and 99% levels. All variables are defined in Supporting Information Appendix A1.The last column shows the difference between control and treatment firms and the p-value for a t-test of its statistical significance. 106 IMBIEROWICZ AND STREITZ FIGURE 1 Use of sweep clauses over time. This figure shows the fraction of loan contracts that include at least one sweep clause. The sample comprises syndicated loans obtained by public U.S. non-financial firms over the 1996–2010 period. excess cash flow. Almost all loans in our sample (94%) include at least one financial covenant, in line with prior studies (see, e.g., Roberts & Sufi, 2009a). The average number of financial covenants is slightly higher in the control sample compared to the treatment sample (2.04 vs. 1.99). The average loan maturity is 42 months for the treatment sample and 43 months for the control sample. Loans by treated firms are slightly smaller than loans by control firms. The average loan size is 454 million USD for the control sample compared to 436 million USD for the treatment sample. The table shows that the average book value of assets is 5112 million USD in the treatment sample and 5062 million USD in the control sample. Treated firms have on average the same leverage as control firms (37%), have higher market-to-book ratios (1.93 vs. 1.74), and have a higher fraction of tangible to total assets (40% vs. 37%). Further, treated firms, on average, have a higher return on assets than control firms (0.19 vs. 0.18), smaller interest coverage (10.7 vs. 11.7), and larger current ratios (1.74 vs. 1.55). The descriptive statistics show that—despite only matching based on firm size—our propensity score matching approach performs reasonably well; that is, the remaining differences between the treatment and the control sample are minor. We depict in Figure 1the use of sweep covenants in loan contracts over time. Interestingly, the use of sweep clauses appears to be pro-cyclical. This suggests that sweeps are used especially in periods when the possibility of wealth expropriation due to excess funds to a firm is higher. 3RESULTS 3.1 Baseline results We first investigate the impact of a change in analyst coverage on the use of sweep covenants in loan contracts in general. That is, we examine the average treatment effect. This includes an analysis of the parallel trends assumption IMBIEROWICZ AND STREITZ 113 TABLE 5 Brokerage house mergers—Effect by corporate governance. SWEEP (0/1) (1) (2) (3) (4) (5) POST ×TREATED × NOT RATED 0.189*** (0.000) POST ×TREATED ×RATED 0.041 (0.267) POST ×TREATED × HIGH CASH 0.141*** (0.004) POST ×TREATED × LOW CASH 0.051 (0.180) POST ×TREATED × LONG TENURE 0.152** (0.035) POST ×TREATED ×SHORT TENURE 0.053 (0.163) POST ×TREATED ×HIGH CASH COMP 0.116** (0.039) POST ×TREATED × LOW CASH COMP 0.021 (0.585) POST ×TREATED ×LOW INST INV 0.241** (0.028) POST ×TREATED × HIGH INST INV 0.041 (0.340) Firm controls Yes Yes Yes Yes Yes Loan controls Yes Yes Yes Yes Yes Merger ×Post FE Yes Yes Yes Yes Yes Merger ×firm FE Yes Yes Yes Yes Yes Observations 2966 2966 1893 2966 2941 Adjusted R20.690 0.689 0.693 0.688 0.686 Statistical difference between coefficients Difference 0.148*** 0.090 0.099 0.095 0.199* Difference p-value 0.010 0.103 0.180 0.132 0.076 Note: This table reports results from the estimation of a pooled panel regression analyzing the use of sweep and financial covenants around brokerage house mergers. For each merger, we consider a 2-year window prior to the merger (pre-merger window) and a 2-year window after the merger (post-merger window). We construct an indicator variable (TREATED) for each merger, which is equal to one for each firm covered by both merging brokerage houses in the pre-merger window (treatment sample), and zero otherwise. For each merger, POST is a variable that is equal to one for the post-merger period and zero for the pre-merger period. Both variables are included as base effects in each regression. SWEEP (0/1) is a dummy variable, which equals one if the loan contract includes at least one sweep covenant, and zero otherwise. LN(#SWEEP) is the log of one plus the number of sweep covenants included in the loan contract. SWEEP RATIO is defined as the number of sweep covenants divided by the total number of sweep and financial covenants included in the loan contract. FIN COV (0/1) is a dummy variable, which equals one if the loan contract includes at least one financial covenant, and zero otherwise. LN(#FIN COV) is the log of one plus the number of financial covenants included in the loan contract. PERF COV RATIO is the number of performance-covenants divided by the total number of financial covenants (performance-covenants plus capital-covenants) in the loan contract. Financial covenants are divided into performance-covenants and capital-covenants following Christensen (Continues) 114 IMBIEROWICZ AND STREITZ TABLE 5 (Continued) and Nikolaev (2012). NOT RATED (RATED) is a dummy variable that equals one if the firm has no credit rating pre-merger, and zero otherwise. LOW (HIGH) CASH is a dummy variable that equals one if the firm is in the bottom (top) half of the cash ratio to total assets distribution pre-merger, and zero otherwise. LONG TENURE (SHORT TENURE) is a dummy variable that equals one if the CEO is (not) in the top decile in terms of tenure pre-merger, and zero otherwise. HIGH CASH COMP (LOW CASH COMP) is a dummy variable that equals one if the proportion of total compensation of the CEO paid through cash salary and bonuses is above (below) median pre-merger, and zero otherwise. LOW INST INV (HIGH INST INV) is a dummy variable that equals one if the number of institutional owners is below (above) median pre-merger, and zero otherwise. All regressions include merger ×firm as well as merger ×POST fixed effects. Further, the regressions include firm characteristics (log total assets, leverage, market-to-book, profitability, tangibility, coverage, current ratio, and rating fixed effects [notch level]) used with their previous year-end value and loan characteristics (log loan size, log maturity, and indicator variables for loan purpose and loan types). p-values (in parentheses) are determined using standard errors robust to clustering at the firm level. ***, **, and * denote 1%, 5%, and 10% statistical significance. All variables are defined in Supporting Information Appendix A1. 3.3 Dual holdings We argue that sweep provisions are used by lenders to address managerial agency problems. However, one concern might be that actions taken by equity holders might be better suited to discipline managers, which might not be in the interest of debtholders. Furthermore, most firms operate well above their default barrier. In this case, the contingencies that sweeps address might be especially relevant for equity holders. This suggests that sweep use increases more for lenders who simultaneously also hold equity of the borrowing firm (dual holding). Ferreira and Matos (2012) show that the benefit of dual holding mainly accrues to the bank. Chava et al. (2019) show that banks as dual holders are less likely to include a capital expenditure restriction in their loan contract. Peyravan (2020) investigates the impact of financial reporting quality on institutional investors’ dual holdings and finds that institutional investors are more likely to become dual holders in firms with low reporting quality. This suggests that dual holdings might also often be related to firms with a potentially higher degree of managerial entrenchment. We identify lenders who also hold equity of their borrowing firm using 13f filings. We investigate the use of sweep provisions separately for firms with and without a lender as dual holder. Table 6reports the results. In Panel A, we interact the treatment ×post indicator with a dual holding indicator that is equal to one if the (lead) banks in the loan syndicate hold at least 0.5% of the equity of the borrowing firm in the pre-merger period. In Panel B, we split the sample into dual holding and non-dual holding banks. The results indicate that banks implement sweep provisions after reductions in analyst coverage in particular when they hold both equity and debt of the same firm. In terms of economic magnitudes, the results indicate that it is about three times more likely that sweep provisions are included post-treatment in the presence of dual holders compared to situations where banks do not have equity holdings in the borrowing firms. Note that while the effect is stronger in the presence of dual holders, also non-dual holders are significantly more likely to include sweep provisions following a reduction in analyst coverage. Overall, the results indicate that lenders increase sweep use in particular when they are most exposed to managerial wealth expropriation. 4CONCLUSION We investigate how changes in analyst coverage affect the implementation of sweep covenants in loan contracts. For this purpose, we utilize exogenous changes in analyst coverage resulting from brokerage house mergers or closures. We observe that an exogenous decrease in coverage results in a more intense use of sweep provisions in a loan contract. This effect is stronger for firms with poor corporate governance and for firms with lenders as dual holders. Overall, our results are consistent with lenders implementing sweep covenants in loan contracts to address managerial agency problems. IMBIEROWICZ AND STREITZ 115 TABLE 6 Brokerage house mergers—Effect by dual holdings. Panel A: Effect by dual holdings—interaction term Total sample Total sample Total sample (1) (2) (3) SWEEP (0/1) LN(#SWEEP) SWEEP RATIO POST ×TREATED 0.065*0.098** 0.058*** (0.052) (0.033) (0.008) POST ×TREATED ×DUAL 0.186*0.274*0.118 (0.095) (0.056) (0.118) Base effects Yes Yes Yes Firm controls Yes Yes Yes Loan controls Yes Yes Yes Merger ×Post FE Yes Yes Yes Merger ×firm FE Yes Yes Yes Observations 2966 2966 2919 Adj. R20.689 0.737 0.652 Panel B: Effect by dual holdings—sample split No dual No dual No dual Dual Dual Dual (1) (2) (3) (4) (5) (6) SWEEP (0/1) LN(#SWEEP) SWEEP RATIO SWEEP (0/1) LN(#SWEEP) SWEEP RATIO POST × TREATED 0.063* 0.094** 0.060*** 0.225** 0.337** 0.151** (0.059) (0.041) (0.007) (0.047) (0.031) (0.021) Firm controls Yes Yes Yes Yes Yes Yes Loan controls Yes Yes Yes Yes Yes Yes Merger ×Post FE Yes Yes Yes Yes Yes Yes Merger ×firm FE Yes Yes Yes Yes Yes Yes Observations 2638 2638 2592 326 326 326 Adj. R20.684 0.736 0.644 0.790 0.804 0.803 Note: This table reports results from the estimation of a pooled panel regression analyzing the use of sweep covenants around brokerage house mergers. For each merger, we consider a 2-year window prior to the merger (pre-merger window) and a 2year window after the merger (post-merger window). We construct an indicator variable (TREATED) for each merger, which is equal to one for each firm covered by both merging brokerage houses in the pre-merger window (treatment sample), and zero otherwise. For each merger, POST is a variable that is equal to one for the post-merger period and zero for the pre-merger period. Both variables are included as base effects in each regression. DUAL is defined as loans of lenders which simultaneously hold at least 0.5% of equity of the borrowing firm in the pre-merger period. POST ×DUAL is included as base effect in each regression in Panel A. SWEEP (0/1) is a dummy variable, which equals one if the loan contract includes at least one sweep covenant, and zero otherwise. LN(#SWEEP) is the log of one plus the number of sweep covenants included in the loan contract. SWEEP RATIO is defined as the number of sweep covenants divided by the total number of sweep and financial covenants included in the loan contract. All regressions include merger ×firm as well as merger ×POST fixed effects. Further, the regressions include firm characteristics (log total assets, leverage, market-to-book, profitability, tangibility, coverage, current ratio, and rating fixed effects [notch level]) used with their previous year-end value and loan characteristics (log loan size, log maturity, and indicator variables for loan purpose and loan types). p-values (in parentheses) are determined using standard errors robust to clustering at the firm level. ***, **, and * denote 1%, 5%, and 10% statistical significance. All variables are defined in Supporting Information Appendix A1. 116 IMBIEROWICZ AND STREITZ Our findings are important for loan contract design. Most of the existing literature focuses on financial covenants and, if at all, includes sweep covenants only as a by-product in the analyses. Our results indicate that sweep covenants serve an important role in addressing managerial agency problems. Financial covenants focus on borrower performance and leverage ratios and trigger a shift of control rights to lenders when the financial condition of the firm deteriorates. However, contingencies of lender wealth expropriation by management exist especially in situations with high managerial flexibility and hence need to be addressed through other means, such as sweep covenants. Contracting can increase the value of the firm by addressing potential agency conflicts. Accordingly, sweep covenants deserve much more attention than is given today. An analysis of sweep covenants and firm value might be a promising area for future research. ACKNOWLEDGMENTS We thank the editor, Kathleen Kahle, an anonymous referee, and Matthias Efing, Felix Noth, Emilia Garcia-Appendini, and participants at the third IWH-FIN-FIRE workshop, the 2017 German Finance Association (DGF) meetings, and the 2017 FIRS meetings for valuable comments and suggestions. The views expressed in this paper are those of the authors and do not necessarily represent those of the Deutsche Bundesbank or the Eurosystem. Open access funding enabled and organized by Projekt DEAL. REFERENCES Aghion, P., & Bolton, P. (1992). An incomplete contracts approach to financial contracting. Review of Economic Studies,59, 473– 494. Begley, J., & Feltham, G. A. (1999). An empirical examination of the relation between debt contracts and management incentives. Journal of Accounting and Economics,27, 229–259. Blanchard, O. J., Lopez-de-Silanes, F., & Shleifer, A. (1994). What do firms do with cash windfalls? Journal of Financial Economics, 36, 337–360. Bradley, M., & Roberts, M. R. (2015). The structure and pricing of corporate debt covenants. Quarterly Journal of Finance,5, 1–37. Brennan, M. J., & Subrahmanyam, A. (1995). Investment analysis and price formation in securities markets. Journal of Financial Economics,38, 361–381. Chava, S., Kumar, P., & Warga, A. (2010). Managerial agency and bond covenants. Review of Financial Studies,23, 1120–1148. Chava, S., Livdan, D., & Purnanandam, A. (2009). Do shareholder rights affect the cost of bank loans? Review of Financial Studies, 22, 2973–3004. Chava, S., & Roberts, M. R. (2008). How does financing impact investment? The role of debt covenants. Journal of Finance,63, 2085–2121. Chava, S., Wang, R., & Zou, H. (2019). Covenants, creditors’ simultaneous equity holdings, and firm investment policies. Journal of Financial and Quantitative Analysis,54, 481–512. Chen, T., Harford, J., & Lin, C. (2015). Do analysts matter for governance? Evidence from natural experiments. Journal of Financial Economics,115, 383–410. Christensen, H. B., & Nikolaev, V. V. (2012). Capital versus performance covenants in debt contracts. Journal of Accounting Research,50, 75–116. Christensen, H. B., Nikolaev, V. V., & Wittenberg-Moerman, R. (2016). Accounting information in financial contracting: The incomplete contract theory perspective. Journal of Accounting Research,54, 397–435. Cohn, J. B., Liu, Z., & Wardlaw, M. I. (2022). Count (and count-like) data in finance. Journal of Financial Economics,146, 529–551. Del Guercio, D., & Hawkins, J. (1999). The motivation and impact of pension fund activism. Journal of Financial Economics,52, 293–340. Demerjian, P. R. (2011). Accounting standards and debt covenants: Has the “Balance Sheet Approach” led to a decline in the use of balance sheet covenants? Journal of Accounting and Economics,52, 178–202. Demerjian, P. R., & Owens, E. L. (2016). Measuring the probability of financial covenant violation in private debt contracts. Journal of Accounting and Economics,61, 433–447. Demiroglu, C., & James, C. M. (2010). The information content of bank loan covenants. Review of Financial Studies,23, 3700– 3737. Deryugina, T., MacKay, A., & Reif, J. (2020). Thelong-run dynamics of electricity demand: Evidence from municipal aggregation. American Economic Journal: Applied Economics,12, 86–114. Dyck, A., Morse, A., & Zingales, L. (2010). Who blows the whistle on corporate fraud? Journal of Finance,65, 2213–2253. IMBIEROWICZ AND STREITZ 117 Easterbrook, F. H. (1984). Two agency-cost explanations of dividends, American Economic Review,74, 650–659. Ellul, A., & Panayides, M. (2018). Do financial analysts restrain insiders’ informational advantage? Journal of Financial and Quantitative Analysis,53, 203–241. Ferreira, M. A., & Matos, P. (2012). Universal banks and corporate control: Evidence from the global syndicated loan market. Review of Financial Studies,25, 2703–2744. Finkelstein, S., & Hambrick, D. C. (1989). Chief executive compensation: A study of the intersection of markets and political processes. Strategic Management Journal,10, 121–134. Freudenberg, F., Imbierowicz, B., Saunders, A., & Steffen, S. (2017). Covenant violations and dynamic loan contracting. Journal of Corporate Finance,45, 540–565. Gillan, S. L., & Starks, L. T. (2000). Corporate governance proposals and shareholder activism: The role of institutional investors. Journal of Financial Economics,57, 275–305. Glaser, M., Lopez-de-Silanes, F., & Sautner, Z. (2013). Opening the black box: Internal capital markets and managerial power. Journal of Finance,68, 1577–1631. Goodman-Bacon, A. (2021). Difference-in-differences with variation in treatment timing. Journal of Econometrics,225, 254– 277. Graham, J. R., Li, S., & Qiu, J. (2008). Corporate misreporting and bank loan contracting. Journal of Financial Economics,89, 44–61. Hallman, N., Howe, J. S., & Wang, W. (2022). Analyst coverage and syndicated lending. Review of Accounting Studies,28, 1531– 1569. Hartzell, J. C., & Starks, L. T. (2003). Institutional investors and executive compensation. Journal of Finance,58, 2351–2374. Heath, D., Ringgenberg, M. C., Samadi, M., & Werner, I. M. (2023). Reusing natural experiments. Journal of Finance,78, 2329– 2364. Hong, H., & Kacperczyk, M. (2010). Competition and bias. The Quarterly Journal of Economics,125, 1683–1725. Huang, R. (2010). Creditor control of free cash flow (Working paper No. 09–30). https://www.philadelphiafed.org/-/media/frbp/ assets/working-papers/2009/wp09-30.pdf Irani, R. M., & Oesch, D. (2013). Monitoring and corporate disclosure: Evidence from a natural experiment. Journal of Financial Economics,109, 398–418. Jensen, M. C. (1986). Agency costs of free cash flow, corporate finance, and takeovers. American Economic Review,76, 323–329. Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics,3, 305–360. Jensen, M. C., & Murphy, K. J. (1990). Performance pay and top-management incentives. Journal of Political Economy,98, 225– 264. Kalcheva, I., & Lins, K. V. (2007). International evidence on cash holdings and expected managerial agency problems. Review of Financial Studies,20, 1087–1112. Kelly, B., & Ljungqvist, A. (2012). Testing asymmetric-information asset pricing models. Review of Financial Studies,25, 1366– 1413. Lang, L., Poulsen, A., & Stulz, R. (1995). Asset sales, firm performance, and the agency costs of managerial discretion. Journal of Financial Economics,37, 3–37. Li, N., Vasvari, F. P., & Wittenberg-Moerman, R. (2016). Dynamic threshold values in earnings-based covenants. Journal of Accounting and Economics,61, 605–629. Lie, E. (2000). Excess funds and agency problems: An empirical study of incremental cash disbursements. Review of Financial Studies,13, 219–248. Maxwell, W. F., & Stephens, C. P. (2003). The wealth effects of repurchases on bondholders. Journal of Finance,58, 895–919. Murfin, J. (2012). The supply-side determinants of loan contract strictness. Journal of Finance,67, 1565–1601. Nikolaev, V. V. (2018). Scope for renegotiation in private debt contracts. Journal of Accounting and Economics,65, 270–301. Nini, G., Smith, D. C., & Sufi, A. (2009). Creditor control rights and firm investment policy. Journal of Financial Economics,92, 400–420. Nini, G., Smith, D. C., & Sufi, A. (2012). Creditor control rights, corporate governance, and firm value. Review of Financial Studies, 25, 1713–1761. Peyravan, L. (2020). Financial reporting quality and dual-holding of debt and equity. The Accounting Review,95, 351–371. Roberts, M. R. (2015). The role of dynamic renegotiation and asymmetric information in financial contracting. Journal of Financial Economics,116, 61–81. Roberts, M. R., & Sufi, A. (2009a). Control rights and capital structure: An empirical investigation. Journal of Finance,64, 1657– 1695. Roberts, M. R., & Sufi, A. (2009b). Renegotiation of financial contracts: Evidence from private credit agreements. Journal of Financial Economics,93, 159–184. 118 IMBIEROWICZ AND STREITZ Stulz, R. M. (1988). Managerial control of voting rights: Financing policies and the market for corporate control. Journal of Financial Economics,20, 25–54. Wu, J. S., & Zhang, A. Y. (2009). What determines financial analysts’ career outcomes during mergers? Journal of Accounting and Economics,47, 59–86. SUPPORTING INFORMATION Additional supporting information can be found online in the Supporting Information section at the end of this article. How to cite this article: Imbierowicz, B., & Streitz, D. (2024). Financial debt contracting and managerial agency problems. Financial Management,53, 99–118. https://doi.org/10.1111/fima.12444