Lending relationships between banks and firms present an important dynamic in well-functioning capital markets. Many academic studies in accounting and finance have explored the effects of lending relationships where the focus is on the repeated interaction between a specific bank and a specific firm.[1] However, we still know relatively little about how these banking relationships form in the first place. In our study, we examine the role of common institutional investors in the formation of new lending relationships between borrowing firms and lenders as well as the associated loan contract terms. We provide evidence that banks extend their lending relationships along cross-ownership lines, partially answering how banks build new lending relationships. Further, our study provides evidence that lenders utilize information stemming from cross-ownership of portfolio borrowers when setting loan contract terms.[2]
Over the past several decades it has become more common for an institutional investor to simultaneously own large equity stakes in multiple firms in the same industry (i.e., institutional cross-ownership). He and Huang (2017) show that “the fraction of U.S. public firms held by institutional blockholders that simultaneously hold at least 5% of the common equity of other same-industry firms has increased from below 10% in 1980 to about 60% in 2014.” Recent studies try to investigate whether and how institutional cross-blockholders exert influence on their cross-held firms’ strategies and decision making.[3] However, there is limited evidence on whether and how institutional cross-ownership affects portfolio firms’ interaction with credit markets. Debt and equity holders have different objectives and risk preferences and because regulation on cross-ownership structures is sparse. For these reasons, understanding whether this increasing trend of institutional cross-ownership affects banks’ decisions to issue loans and the terms of the loan contracts issued is important and of interest to both academics and policy makers.
Consider the following scenario: Institutional Investor P obtains significant equity stakes in both Firm A and Firm B in the same industry (i.e., cross-ownership begins); Firm A has an existing loan from Bank M; and Firm B decides to enter the loan market to obtain financing after the cross-ownership begins. Here, our research questions are twofold: (1) given that Firms A and B are cross-owned by Investor P and that Bank M already has an existing loan with Firm A, is it more likely that Firm B will obtain their loan from Bank M (i.e., a “related bank”) than from another unrelated bank, say Bank N?; (2) given that Firm B indeed obtains loans from both Bank M and Bank N, are the loan terms it receives from Bank M more favorable than those from Bank N?
We predict that the answer to both questions is “yes” because a prior lending relationship with a cross-held firm helps a bank to build a new relationship with the potential borrower through the institutional cross-owner. In addition, cross-owners may play a more active monitoring role than normal institutional investors and encourage more collaboration and diffusion of information among portfolio firms. A prior lender understands the style of the cross-owner and the possible implications (e.g., whether and how they get involved in portfolio firms’ decision-making) better than other banks and this will reduce information frictions between a prior lender and a new potential borrower in the institutional cross-holder’s portfolio. Thus, we predict that relationship banking based on cross-ownership will help both banks and potential borrowers, leading to a higher probability of lending and better loan contract terms.
Empirically, we focus our analysis on the probability of loan issuance and the loan contract terms that arise after a borrowing firm is newly added to an institutional cross-owner’s portfolio. For borrowing firms that obtain a loan within three years of entering the cross-owner’s portfolio, we compare the probability of loan issuance and the loan contract terms from banks that have previously issued a loan to other firms held by the same cross-owner to those from other banks. This research design provides a strong identification strategy which first controls for a firm’s need for financing (and the intention to utilize private loan financing) when examining the probability of loan issuance, and second controls for the borrowing firm’s fundamentals when examining the characteristics of loan contract terms. We employ a sample of 13,298 loans, and we find that firms newly added to institutional cross-owner’ portfolios are 5.7% more likely to borrow from banks that have previously issued a loan to other firms held by the same cross-owner within the previous three years, compared with other banks active in the market. We also find that loans obtained from these related banks contain more favorable loan terms compared to loan terms offered by other banks that have not lent to a cross-owned firm. Specifically, a loan from these related banks is associated with a 9.39 basis point interest spread reduction.
We also explore the variation in the interest spread effect and find that the effect is driven by borrowers with high information asymmetry, low accounting quality, and more financial constraints. We also find that this effect is larger for dedicated institutional cross-owners, consistent with the lender valuing the institutional cross-owner’s monitoring role with respect to the borrower, and/or the lender considering the desire for future relationship-based transactions with the institutional investor-owned companies when setting loan terms.
Because the selection of a firm into a cross-owner’s portfolio is not random, we focus on within-firm variation in bank choice and interest spread in our main analysis to mitigate potential endogeneity concerns. To further mitigate these concerns, we conduct two sets of tests. First, we add firm and year fixed effects as well as firm-year fixed effects in our main regression to mitigate the effect of unobservable firm-level variables. Our results hold in both specifications. Second, we exploit the quasi-natural experiment of financial institution mergers using a propensity score matching difference-in-differences approach. Merger decisions by financial institutions can be influenced by a variety of factors and are unlikely to be related to the portfolio holdings of the related parties, leading to a plausibly exogenous shock in equity cross-holding. Our treatment sample consists of firms that are newly influenced or controlled by the cross-owners becauseof the exogenous shock of a financial institution merger, and our control sample consists of other similar firms without actual cross-holding relationships. We find that treatment firms, relative to control firms, experience a significant increase in the probability of obtaining a loan from these “new” relationship banks and also receive more favorable loan terms when they do obtain a loan. Both identification tests suggest that endogeneity is unlikely to fully explain our results.
Finally, we also examine other loan contract terms in borrowers’ debt contracts. We find that loans issued by related banks are more likely to include interest-increasing performance pricing provisions, consistent with the idea that banks want to reduce the risk of shareholder rent extraction. We also find that loan amounts to cross-held firms are larger, consistent with the information asymmetry decrease facilitating the demand-side of borrowers wanting to take out larger loans.
Our paper contributes to the understanding of banking and relationship lending in several ways. First, we extendthe literature on how banks develop new lending relationships. Prior studies explore the matching of firms and banks in lending relationships based on several factors such as size, but they typically examine the general matching between two large categories of firms and do not attempt to address to which firm banks attempt to lend within each general category. In contrast, we provide evidence that banks extend their lending relationships along cross-ownership lines, partially answering how banks build new lending relationships.
Relatedly, we also contribute to the large literature investigating the effects of relationship banking. Our study provides evidence that lenders utilize information stemming from common ownership of portfolio borrowers, and that they consider such cross-ownership relationships when setting loan contract terms. Our study furthers the understanding of relationship banking beyond repeated interactions between a firm and a bank to the interaction between portfolio companies and banks.
Finally, we add to the developing literature on the role of equity cross-ownership in capital markets. Recentresearch has tried to investigate whether and how institutional investors’ cross-holdings can affect corporate strategies and decision-making processes. Additionally, there is current concern and ongoing discussion among researchers, practitioners, and policy makers about whether cross-held equities by institutional investors reduces competition and facilitates collusion in certain industries.[4] Our results indicate that institutional cross-ownership enables portfolio companies to obtain access to new financing relationships and borrow on more favorable loan contract terms. We thus contribute to the understanding of institutional cross-holding effects in the U.S., as well as providing new insights into the interaction they have with the equity and debt markets.
Zhiming Ma is an Associate Professor at Peking University, Guanghua School of Management
Edward Owens is an Associate Professor at the University of Utah, Accounting Department
Derrald Stice is an Associate Professor at the University of Hong Kong, HKU Business School
Danye Wang is a PhD student at New York University, Stern School of Business
This post is adapted from their paper, “Lending Relationships Along Ownership Lines: Institutional Cross-Ownership and Bank Loan Contracts”, available on SSRN.
[1] See e.g., Petersen and Rajan 1994; Puri 1996; Berger and Udell 1995; Cole 1998; Schenone 2010; Bharath, Dahiya, Saunders, and Srinivasan 2011; Dahiya, Saunders, and Srinivasan 2003; Bharath, Dahiya, Saunders, and Srinivasan 2007; Ivashina 2009.
[2] Throughout the paper we use the terms “cross-ownership”, “cross-blockholder”, and “cross-holding” to refer to cases where institutional investors hold equity stakes of at least 5% in at least two companies in the same industry.
[3] See e.g., Azar, Schmalz, and Tecu 2018; He and Huang 2017; Kennedy, O’Brien, Song, and Waehrer 2017; Koch, Panayides, and Thomas 2021; Lewellen and Lowry 2021.
[4] See e.g., Posner, Morton, and Weyl 2017; Rock and Rubinfeld 2017; Patel 2018. In addition, the Department of Justice (DOJ) in the U.S. and its antitrust counterparts in Europe (such as the European Commission) investigate institutional cross-blockholding when considering industry concentration issues (O’Brien and Waehrer 2017).
