Electronic copy available at: http://ssrn.com/abstract=2700514

Co-Migration and the Benefits of Relationships in Bank Lending*

Urooj Khan

Columbia Business School Columbia University

Email: uk2117@gsb.columbia.edu

Xinlei Li Columbia Business School

Columbia University Email: xl2297@columbia.edu

Chris Williams Stephen M. Ross School of Business

University of Michigan Email: williacd@umich.edu

Regina Wittenberg-Moerman Marshall School of Business

University of Southern California Email: reginaw@marshall.usc.edu

Current version: December, 2015

* We appreciate the helpful comments of conference participants at the University of North Carolina Accounting Alumni Conference and workshop participants at Columbia University. We are grateful to Vincent Pham for his excellent research assistance. We thank the Thomson Reuters Loan Pricing Corporation for providing loan data. We gratefully acknowledge the financial support of Columbia Business School, Ross School of Business, University of Michigan, the University of Chicago Booth School of Business and the Marshall School of Business, the University of Southern California.

Electronic copy available at: http://ssrn.com/abstract=2700514

Co-Migration and the Benefits of Relationships in Bank Lending

Abstract

Established relationships with a borrower is one of the primary channels through which lenders reduce agency problems and information asymmetry in debt contracting. Given the critical role individual managers play in shaping corporate policies and performance, we investigate the importance of lenders’ relationships with individual managers and their impact on lending practices. Utilizing a setting of executive turnovers, we find that a lender is 1.6 times more likely to start a lending relationship with a firm when a manager with whom it has a prior lending relationship is hired by the firm as a top executive. We also find that the likelihood of a lender’s co-migration with a manager is higher when the firm is more opaque, consistent with the higher importance of prior relationships in poor information environments. We next show that co-migration benefits the borrowing firm through greater access to credit and a lower cost of financing. Further, a stronger lender-manager relationship increases the likelihood of a lender’s co-migration with the manager, especially when the lender is under pressure to expand her lending portfolio and seek new borrowers. Overall, our paper provides novel evidence on the importance of lender-manager relationships in credit markets.

Electronic copy available at: http://ssrn.com/abstract=2700514

1

1. Introduction

Economic theories suggest that market frictions such as information asymmetry and agency

costs can hinder the allocation of capital to positive NPV projects (e.g., Stiglitz and Weiss, 1981).

By ex-ante screening and ex-post monitoring of a borrower, banks reduce information asymmetry

and allow for the efficient allocation of capital to these projects (Diamond, 1984, and Fama, 1985).

One of the primary channels through which lenders reduce agency problems is by forging a

relationship with their borrowers. By developing a strong relationship through repeated

transactions and continuous interaction with the borrowing firm, lenders are able to gather relevant

information not only about the borrowing firm’s prospects, but also about the competence and

trustworthiness of its management (e.g., Rajan, 1992, and Schenone, 2010).1 However, prior

studies that examine relationship lending focus primarily on the relationship between lenders and

the borrowing firm, without addressing the role played by the relationship between lenders and

individual managers (e.g., Petersen and Rajan, 1994 and 1995, Degryse and Van Cayseele, 2000,

Bharath et al., 2011, Santos and Winton, 2009, and Schenone, 2010). In this paper, we investigate

the importance of the lender-manager relationship and its impact on lending practices.

Individual managers play an important role in shaping corporate behavior and performance

(Hambrick and Mason, 1984, and Hambrick, 2007). A manager’s characteristics and style

significantly affect a variety of firm practices and policies, including investment and financial

strategies (Bertrand and Schoar, 2003), disclosure and communication choices (Bamber, Jiang and

Wang, 2010, Ge, Matsumoto and Zhang, 2011, DeJong and Ling, 2013) and the extent of tax

aggressiveness (Dyreng, Hanlon and Maydew, 2010). Prior research also documents that

uncertainty about manager-specific characteristics increases investors’ assessment of a firm’s

1 See Boot (2000) and Elyasiani and Goldberg (2004) for a comprehensive review of the literature on relationship lending.

2

riskiness and increases its cost of debt (e.g., Pan et al., 2015a and 2015b). Institutional evidence

also suggests that credit rating agencies and loan officers consider a manager’s “character” as one

of the key factors in assessing a firm’s credit risk (Plath et al., 2008, and Koch and MacDonald,

2014). Because, through prior interactions with a manager, lenders can assess her talent, honesty,

strategic vision, motivation and risk attitude, we expect the lender-manager relationship to reduce

lenders’ information asymmetry about a borrowing firm’s creditworthiness and future

performance and thus significantly affect lending practices.

As our primary focus in this paper is lenders’ relationship with a manager, the key empirical

design challenge we face is how to disentangle this relationship from the one between lenders and

the borrowing firm. To accurately identify the lender-manager relationship and its consequences

on lending practices, we utilize a setting of executive turnover. Specifically, we examine how the

relationship a lender develops with a manager at the company the manager leaves (origin firm)

affects its lending to the company that the manager joins (destination firm). We identify 648 origin-

destination firm pairs that experience CEO or CFO turnover over the 1992-2014 period. To

accurately assess the effects of the lender-manager relationship, we also construct a matched

sample of pseudo origin-destination firm pairs (Brochet et al., 2014). We analyze these firms’

syndicated lending and thus focus on the lead arrangers of syndication, which take on the primary

information collection and monitoring responsibilities. 2 Based on this empirical setting, we

investigate three broad research questions: 1) do an origin firm’s lead arrangers follow the manager

to the destination firm (i.e., co-migrate with the manager)? 2) what factors increase the likelihood

of co-migration? and 3) are there benefits to co-migration for the destination firm?

We expect that lenders’ established relationship with a manager will decrease their

2We use lenders and lead arrangers interchangeably throughout the rest of the paper.

3

uncertainty about the destination firm’s risk taking, investment strategies and future financial

performance, thus reducing the risks involved in providing it credit funding. In addition, because

managers often provide private lenders with important non‐public information about their

financial position (e.g., Dennis and Mullineaux, 2000, Standard and Poor’s, 2007), lender-manager

relationship will enhance the reliability of such information and thus reduce lenders’ efforts to

verify its quality and accuracy. We therefore predict that lenders will have incentives to continue

utilizing their valuable relationships with the manager and will be likely to co-migrate with the

manager to the destination firm. We find evidence consistent with this prediction. A lead arranger

is 1.6 times more likely to start a lending relationship with a firm when a manager with whom it

has a prior lending relationship moves to the firm as a top executive. This finding emphasizes that

lenders create important relationships with the managers of a borrowing firm.

We supplement this analysis by testing whether the probability of a lead arranger co-

migration with the manager is higher for more informationally opaque borrowers. Lending to

opaque borrowers is associated with substantial agency costs and costly monitoring (e.g.,

Diamond, 1991, Petersen and Rajan, 1994 and 1995, Berger and Udell, 1995), suggesting that the

lenders’ relationship with a manager should be particularly valuable when the destination firm

operates in a poor information environment. We consider a destination firm to be more

informationally opaque when it is smaller (e.g., Bharath et al., 2007, and Wittenberg-Moerman,

2008), not rated by credit rating agencies (e.g., Sufi, 2007, and Kraft, 2014) or when it has low

analyst coverage (Güntay and Hackbarth, 2010, and Mansi et al., 2010). Across all information

opaqueness measures, we find that the likelihood of co-migration is significantly higher when the

destination firm is more opaque.

We next investigate whether the lenders’ co-migration benefits the destination firm. Building

4

on the prior relationship lending literature that documents that borrowers with close ties to lenders

have greater availability of credit (Petersen and Rajan, 1994, and Rajan and Zingales, 1998), we

predict that lender co-migration endows the destination firms with higher access to debt capital.

We assess our prediction in two ways. First, we examine whether a lender is more likely to co-

migrate with the manager when a destination firm has limited access to capital, which would

suggest that co-migration at least partially alleviates the firm’s capital constraints. Using the

number of the destination firm’s prior lead arranger relationships as a measure of its access to

credit (e.g., Houston and James, 1996, and Murfin, 2012), we find that the probability of co-

migration is significantly higher for firms with weak access. We supplement this evidence by

examining whether the probability of co-migration is higher when the lending standards in credit

markets are tight. Using changes in bank lending standards, as reported by the Federal Reserve

Board’s Senior Loan Officer Opinion Survey (e.g., Bassett et al., 2012), we find that lead arrangers

are more likely to co-migrate with a manager when credit standards are tight and firms in the

economy face greater credit constraints.

Second, we perform a more direct test of credit availability by examining whether co-

migration increases the extent of the destination firm’s syndicated loan financing. For the sample

of firms that experience executive turnover, we find that a lead arranger’s co-migration is

associated with an 18 percent higher syndicated loan borrowing for destination firms following

turnover relative to destination firms that do not experience lender co-migration. Overall, our

findings suggest that managers’ relationships with lenders benefit borrowers by enabling greater

access to credit.

We extend these analyses by exploring whether co-migration also has a favorable effect on

the cost of debt. It is possible that through their relationship with the manager, lenders will share

5

potential cost savings due to reduced information asymmetry with the destination firms. However,

prior literature provides mixed evidence with respect to whether relationship lending decreases the

cost of borrowing, sometimes even finding an increase in these costs due to the hold-up problem

in relationship lending (e.g., Petersen and Rajan, 1994, Degryse and Van Cayseele, 2000, Santos

and Winton, 2009, and Schenone, 2010). We find that the destination firms’ cost of borrowing is,

on average, 21.7 basis points lower when lenders co-migrate with managers, which represents 11.8

percent of the average interest spread for the sample destination firms.

Although our analyses so far focus on borrowers’ characteristics and financing needs, in our

last set of tests we examine what factors affect a lender’s decision to co-migrate with a manager.

We predict that the strength of the lender-manager relationship increases the probability of co-

migration. Because a stronger lender-manager relationship is likely to be associated with lenders’

greater trust in the manager’s integrity, talent and information that she provides, it should therefore

lead to a higher reduction in information asymmetry about the borrowing firm. We perform lender

level analyses for a sample of all lead arrangers that had syndicated loans to the origin firm over a

manager’s tenure. We designate lead arrangers that syndicated more than 75 percent of all loans

to the origin firm as having a strong lending relationship with the manager. We find that these lead

arrangers are 3.4 times more likely to follow a manager to the destination firm relative to other

lead arrangers of the origin firm.

We add to this analysis by examining whether the strength of the lender-manager relationship

has a more significant effect on the co-migration probability when lenders are under pressure to

expand their loan portfolio and seek new borrowers. We show that when lead arrangers with a

strong relationship with a manager experience relatively low loan growth, they are more likely to

expand their lending portfolio by extending credit to the destination firm. We also find that the

6

strength of the lending relationship has a higher impact on the probability of co-migration when a

lead arranger experiences low profitability, consistent with new lending relationships having the

potential to increase a lender’s future profitability via additional interest revenues, loan origination

fees or fees from other services provided to the destination firm.

Our paper contributes to the literature across several dimensions. First, we add to the

literature on relationship lending. Prior studies have documented that the relationship between

lenders and a borrowing firm plays a critical role in reducing information asymmetry and agency

problems in private lending (e.g., Rajan, 2002, Berger and Udell, 1995, Bharath et al., 2011, Santos

and Winton, 2009, and Schenone, 2010). Petersen and Rajan (1994, 1995), Boot and Thakor

(1994), Berger and Udell (1995) and Bharath et al. (2011), among others, also demonstrate the

benefits to borrowers of an established lending relationship. We extend these studies by showing

that lenders establish a valuable relationship not only with the borrowing firm, but also with its

managers. We also document that lender-manager relationships benefit the firm the manager joins

via enhanced access to credit and lower cost of debt financing.

Our findings also add to the emerging literature on the role of relationships in capital markets

in general, and in bank lending in particular. Engleberg et al. (2012) show that firms benefit from

lower interest spreads on their loans when firm and bank executives attended the same college or

previously worked together. Haselmann et al. (2013) report that German banks provide

significantly more credit to firms whose executives belong to the same social club as the bank’s

executives. Karolyi (2014) reports that following executive turnover, borrowers lose as much as

74% of the benefits from lenders with personal relationships with the former executive and

therefore tend to borrow from new lenders. We supplement these studies by documenting that

personal relationships between executives and lenders survive managerial turnover and carry over

7

to the firm the manager joins. We also show that a borrower’s information opacity and capital

needs influence the likelihood of lender co-migration.

Third, our study is also relevant to the literature on the impact of personal networks and

relationships on business and investment decisions. Brochet et al. (2014) find that executive

turnover triggers analysts following the company that the manager leaves to initiate coverage of

the company the executive joins. Cohen et al. (2008) show that personal relationships influence

the flow of information into asset prices. They find that mutual fund managers are more likely to

place larger bets on firms run by managers who graduated from the same institution and that this

practice allows them to earn higher average returns on their investments. Fracassi (2014) and

Schmidt (2015) show that social networks influence capital investments and mergers and

acquisition outcomes, respectively. We add to these studies by documenting that lenders’ personal

relationship with a manager significantly affects their decision to initiate a lending relationship

with the new firm. We also show that the strength of the lender’s prior relationship with a manager

increases the probability that the lender provides credit to the new firm the manager joins.

The rest of the paper is organized as follows. Section 2 discusses the related literature and

develop the hypotheses. Section 3 describes the sample selection and empirical design. Section 4

reports the empirical results and Section 5 concludes.

2. Related Literature and Hypotheses Development

2.1. Manager and Lender Co-Migration

Prior literature has extensively examined the importance of relationship lending.

Relationship lenders have deep knowledge of a borrower’s operations and well developed channels

of communication with the borrower’s managers (e.g., Rajan, 1992, Petersen and Rajan, 1994,

Bharath et al., 2007, and Schenone, 2010). Prior studies also suggest that borrowers are inclined

8

to reveal more information to lenders with whom they have an established lending relationship

and that relationship lenders have stronger incentives to invest in information production

(Greenbaum and Thakor, 1995, Boot, 2000, and Bharath et al., 2008). Relationship lenders thus

play an important role in reducing the agency problems associated with debt financing. There is

also ample evidence that relationship lending has important consequences for lending practices,

such as a higher availability of credit to borrowers with a strong relationship with their lenders

(e.g., Hoshi, Kashyap and Scharfstein, 1990a, 1990b and 1991, and Petersen and Rajan, 1994 and

1995).

For the most part, previous studies focus on the relationship at the institutional level between

lenders and the borrowing firm (e.g., Petersen and Rajan, 1994 and 1995, Degryse and Van

Cayseele, 2000, Bharath et al., 2011, Santos and Winton, 2009, and Schenone, 2010). For example,

Petersen and Rajan (1994) and Berger and Udell (1995) focus on lenders’ relationship with small

firms, for which data is available under the National Survey of Small Business Finances. Degryse

and Van Cayseele (2000) examine relationship lending in the context of small businesses in

Belgium. Bharath et al. (2011), Santos and Winton (2009) and Schenone (2010) utilize the

relationship between lenders and borrowing firms in the syndicated loan market.

However, individual managers play a crucial role in shaping firm strategies and performance

(Hambrick and Mason, 1984, and Hambrick, 2007). Manager characteristics, such as confidence,

risk attitude, intrinsic motivation and underlying talent are often collectively referred to as

“management style.” Prior research documents that management style impacts a wide spectrum

of organizational outcomes. Bertrand and Schoar (2003) report that management style explains a

significant amount of the heterogeneity in the investment, financial and organizational practices

of firms. The individual-specific characteristics of managers have also been shown to affect other

9

firm policies and practices, such as tax avoidance strategies (Dyreng et al., 2010), accounting

choices (Ge et al., 2011, and DeJong and Ling, 2013) and voluntary financial disclosure policies

(Bamber et al., 2010). We therefore explore whether lenders’ relationships with individual

managers significantly affect lending practices. 3

The lender-manager relationship is likely to reduce lenders’ uncertainty about the manager’s

underlying talent, personality, risk attitude and strategic vision, consequently reducing the risks

involved in funding the borrowing firm. Pan et al. (2015a) report that a firm’s stock return volatility

increases significantly around CEO turnover and then declines over the tenure of the CEO. They

interpret this finding as suggesting that over time, the market learns about a CEO’s individual-

specific characteristics and that this reduction in uncertainty results in a decline in a firm’s

riskiness. Similarly, Pan et al. (2015b) show that uncertainty about a firm’s management also

affects the cost of borrowing. They find that a firm’s CDS spreads, loan spreads and bond yield

spreads decline over the first three years of CEO tenure, holding other macroeconomic, firm and

security level factors constant. Further, evidence from practice also highlights the importance of

manager “character” in credit decisions. Koch and MacDonald (2014) report that it is standard

practice for banks to assess managers’ abilities and talents when evaluating credit requests. Credit

rating agencies also carefully analyze a manager’s impact on firm value when evaluating a firm’s

credit riskiness (e.g., Plath et al., 2008).

We expect the lender-manager relationship to be particularly important when lenders extend

credit to a new borrower, as adverse selection problems are the most severe in this case. Therefore,

to explore the effects of the lender-manager relationship on lending practices, we utilize an

empirical setting where a lender extends credit to a new borrower whose manager has an

3 It would also be interesting to examine the importance of a manager’s relationship with the loan officers assigned to its loans, but we do not possess data on manager-loan officer interactions.

10

established relationship with the lender. Specifically, we focus on a sample of firms that experience

executive turnover, where a lender has developed a relationship with a manager at the company

the manager leaves (origin firm) and may start a lending relationship with the new company a

manager joins (destination firm). Through prior transactions with the manager at the origin firm,

the lender has had an opportunity to assess her talent, trustworthiness, strategic vision, motivation

and risk taking behavior. This assessment will help the lender to decrease its uncertainty about the

destination firm’s investment and financial policies, future strategies and financial performance,

thus allowing a more accurate evaluation of its credit risk. In addition, because in private lending

managers often provide lenders with confidential information about their financial position and

projections (e.g., Dennis and Mullineaux, 2000, Standard and Poor’s, 2007), prior lender-manager

relationship will enhance the reliability of such information. This in turn will decrease the efforts

lenders need to exert to verify the quality and accuracy of this non-public information and will

further enhance lenders’ ability to correctly asses the risks involved in funding the borrower.

We therefore predict that lenders will have incentives to continue utilizing their valuable

information about the manager and well established channels of communication with her and will

be likely to follow the manager to the destination firm. Because we conduct our analyses in the

syndicated loan market setting, we focus our hypotheses on the lead arrangers of the syndication.

Lead arrangers establish a relationship with the borrower, negotiate the terms of the loan contract,

perform the primary screening and monitoring of the borrower, and recruit participants to join the

syndicate (e.g., Sufi, 2007). Accordingly, our first research hypothesis is as follows:

H1 – A lead arranger is more likely to start a lending relationship with a firm when a

manager with whom it has a prior relationship moves as a top executive in this firm.

11

2.2. The Effect of a Borrower’s Information Opacity on the Probability of a Lender’s Co-Migration

with the Manager

We expect the lender-manager relationship to be especially valuable in lending to

informationally opaque borrowers. These borrowers rely primarily on relationship lending because

information asymmetry about their performance and creditworthiness deters prospective lenders

from extending them credit (e.g., Diamond, 1991, Petersen and Rajan, 1994 and 1995, and Berger

and Udell, 1995). Providing credit to informationally opaque borrowers requires lenders to exert

substantial screening and monitoring efforts. Further, assessing the reliability and accuracy of non-

public information provided by managers at loan initiation and within the course of the loan is

particularly important in lending transactions with these borrowers (Bharath et al, 2007 and

Bushman et al., 2010). In the syndicated loan market, loans to non-transparent borrowers are also

more costly for the lead arranger to syndicate as the arranger has to commit to holding a larger

proportion of the loans to attract syndicate participants (e.g., Sufi, 2007, and Ivashina, 2009). Thus,

when extending credit to borrowers that operate in a poor information environment, having an

established relationship with a manager should endow lenders with meaningful cost savings.

From the borrower’s perspective, opaque borrowers are also more likely to seek financing

from lenders that have established relationships with the new manager. Information asymmetry

increases loan pricing and has a detrimental effect on other contractual terms, such as maturity and

collateral requirements (e.g., Francis et al., 2005, Zhang, 2007, Bharath et al., 2008, and Costello

and Wittenberg-Moerman, 2011). Therefore, non-transparent borrowers are likely to benefit from

more lenient loan terms when information asymmetry is mitigated due to lender-manager

relationships. Combined, lender- and borrower-based arguments lead to our second hypothesis:

12

H2 – The likelihood of the lead arranger’s co-migration with a manager is higher when the

destination firm is more informationally opaque.

2.3. The Consequences of Lender-Manager Co-Migration: Benefits to the Destination Firm

Prior literature on relationship lending documents that a borrower’s close ties with its lenders

allow for greater access to credit (Petersen and Rajan, 1994, and Rajan and Zingales, 1998).

Petersen and Rajan (1995) extend this evidence and show that relationship lending is especially

beneficial to young, more credit rationed firms. Bolton et al. (2013) further highlight that

relationship lending is valuable when firms have limited access to credit. They show that banks

endowed their borrowers with higher credit availability during the recent financial crisis, although

they charge higher interest spreads in normal times.

We extend these arguments to the lender-manager relationship setting and predict that when

the lender co-migrates with the manager, the destination firm benefits from higher access to debt

capital. However, empirically assessing the change in credit availability to the borrower is always

challenging. Therefore, we assess our prediction in two ways. First, we examine whether a lender

is more likely to co-migrate with the manager when the destination firm has more limited access

to debt capital, thus relieving, at least partially, its capital constraints. Our third hypothesis is thus

stated as following:

H3a – The likelihood of the lead arranger’s co-migration with a manager is higher when the

destination firm is more capital constrained.

Second, we examine the association between a lender’s co-migration and the extent of the

destination firm’s syndicated loan borrowing following an executive turnover. Stated as a formal

hypothesis:

H3b – The extent of a destination firm’s syndicated loan financing following executive

13

turnover is higher when a lead arranger co-migrates with a manager relative to when the lead

arranger does not follow the manager to the destination firm.

The benefits of lender-manager relationship can also extend to less expensive credit. Lenders

may pass on some of the potential cost savings due to the reduced information asymmetry from its

established relationship with a manager to the destination firm. If this is the case, a lender’s co-

migration is likely to be associated with a lower interest spread. However, prior literature provides

mixed evidence with respect to whether relationship lending results in cheaper credit. Berger and

Udell (1995) show that borrowers with longer lending relationships experience lower credit costs

than do borrowers with shorter relationships. Similarly, Bharath et al. (2011) show that repeated

borrowing from the same lender results in lower spreads. In contrast, Degryse and Van Cayseele

(2000) find that interest rates increase with the length of the lending relationship, implying that

relationship banks exploit their information advantage. Consistent with the hold-up problem

associated with relationship lending (Rajan, 1992), Santos and Winton (2009) find that bank-

dependent borrowers pay higher spreads during a recession. Similarly, Schenone (2010) shows

that relationship loans become more expensive at high levels of relationship lending intensity.

Despite this conflicting prior evidence, we formulate the following hypothesis:

H4 – The cost of borrowing following executive turnover is lower for destination firms that

experience a lead arranger’s co-migration with the manager relative to destination firms where a

lead arranger does not follow the manager.

2.4 The Importance of the Strength of the Lender-Manager Relationship in Determining Lenders’

Co-Migration

Next, we focus on a lender’s incentives to co-migrate. We have assumed so far that all

relationships are equal; however, the strength of the lender-manager relationship can be an

14

important determinant of co-migration. Specifically, we expect a stronger lender-manager

relationship to be associated with the lender placing greater trust in a manager’s integrity, talent

and information that she provides. Consequently, this stronger relationship is likely to reduce

information asymmetry about the future performance and creditworthiness of the destination firm

to a greater extent. Building on these arguments, we state our next hypothesis:

H5a – The likelihood of the lead arranger’s co-migration with a manager to the destination

firm increases with the strength of the lender-manager relationship.

Although we predict that a strong relationship between the lender and the manager should

substantially reduce information asymmetry about the destination firm, adverse selection still

remains a significant concern when lenders extend credit to new borrowers. We therefore expect

the strength of the lender-manager relationship to have a more significant effect on the co-

migration probability when lenders are under pressure to expand their loan portfolio and thus seek

new borrowers. Lenders are more likely to extend their borrowing base when they experience slow

loan growth. Low profitability may also incentivize lenders to follow the manager to the

destination firm. New lending relationships can increase lenders profitability via additional interest

revenues and loan origination fees. Establishing a lending relationship with the destination firm

may also help a lender gain future underwriting of its public bonds and equity (e.g., Drucker and

Puri, 2005, Yasuda, 2005, and Brarath, 2007), thus increasing fee revenues. To reflect these

arguments, we formulate the following hypothesis:

H5b – The strength of the lead-arranger-manager relationship has a stronger effect on the

co-migration probability when a lead arranger experiences lower loan growth or lower

profitability relative to other lenders.

15

3. Sample Selection and Matching Procedure

3.1. Sample Selection

Because our study explores the co-migration of managers and lenders, we start our sample

selection process by identifying executives that switch companies. We collect executive turnovers

from the Execucomp database for the period from 1992 to 2014. We next search the names of the

executives joining and leaving Execucomp firms in the BoardEx database to construct their

comprehensive employment history. For each manager switching firms, we identify her origin and

destination firms. We require a manager to hold a senior executive position for a period of at least

three years in each firm. We focus on the turnovers of CEOs and CFOs because a firm’s top

management designs its policies and strategic goals and thus significantly affects its performance

and creditworthiness.

Specifically, we identify executives that held the position of a CEO or CFO at the origin firm

and migrated to the destination firm as a CEO or a CFO. As reported in Panel A of Table 1, we

are able to identify 723 instances of CEOs’ or CFOs’ migrations in Execucomp over our sample

period. We next obtain firm characteristics from Compustat and syndicated loan data from

DealScan. We require each origin and destination firm to have at least one syndicated loan during

the executive’s tenure at each firm. We also require that sample firms have sufficient data for our

analyses. These requirements reduce our sample to 648 executive turnovers.

In Panel B of Table 1, we present the comparison of origin and destination firms across a

number of characteristics, including size, profitability and leverage. Origin Firm Size is the natural

logarithm of total assets of the origin firm. Origin Firm Profitability is the ratio of the origin firm’s

net income before extraordinary items to total assets. Origin Firm Leverage is the ratio of the

origin firm’s long-term debt to total assets. Destination Firm Size, Destination Firm Profitability

16

and Destination Firm Leverage are defined analogously. All firm characteristics are measured at

the end of the most recent fiscal year preceding the executive’s turnover. We find that origin firms

are on average, larger, have higher leverage and are more profitable than destination firms. These

differences between the origin and destination firms’ characteristics are statistically significant,

although the difference between the two set of firms’ leverage is only significant at the 10% level.

3.2. Control Sample

To accurately assess the effects of lenders’ co-migration with the manager to the destination

firm, we construct a control sample of origin-destination firm pairs. The matched pairs are created

to mirror the treatment origin-destination firm pairs that experience executive turnover. Our

matching procedure follows Brochet et al. (2014) and employs the Propensity Score Methodology

(PSM), as in Rosenbaum and Rubin (1983). Specifically, we estimate a propensity score for the

origin as well as the destination firm and identify a matched pair of origin-destination firms. To

do so, we estimate the following Logit Model:

.

(1)

As described in Figure 1, the matching procedure involves a three step process. In the first

step (i.e., Step 1 in Figure 1), we pair the origin firm with all firms in the destination firm’s two-

digit SIC group and estimate Equation (1). The dependent variable, Turnover, equals one when

the executive’s origin firm is paired with its actual destination firm, zero otherwise. Therefore, in

this step, Equation (1) is modeling the executive’s decision to migrate from the origin to a

particular destination firm. Next, we use the estimated coefficients to find a matched “pseudo”

17

destination firm with the closest propensity score to the actual destination firm.

In Step 2, we re-estimate Equation (1) for a sample consisting of the executive’s actual

destination firm paired with all firms in its origin firm’s two-digit SIC group. For this sample, the

dependent variable, Turnover, equals one if the executive’s destination firm is paired with its actual

origin firm, zero otherwise. In this step, Equation (1) is modeling a destination firm’s decision to

hire an executive from a certain origin firm. Next, as in Step 1, we use the estimated coefficients

to find a matched “pseudo” origin firm with the closest propensity score to the actual origin firm.

In Step 3, the matched pseudo origin and pseudo destination firms from Steps 1 and 2,

respectively, are paired together for each actual origin-destination firm pair to create a matched

control pair. Essentially, using propensity score matching we are able to identify matched origin-

destination control firm pairs that are similar to the actual origin-destination firm pairs across a

variety of firm characteristics.

We perform the matching based on the following origin and destination firm characteristics.

Origin Firm Size, Origin Firm Profitability, Origin Firm Leverage, Destination Firm Size,

Destination Firm Profitability and Destination Firm Leverage are as previously defined. All firm

characteristics are measured at the end of the most recent fiscal year preceding the executive’s

turnover. Whether origin and destination firms operate in the same industry is another important

component of the matching model. Same Industry is an indicator variable that equals one if the

origin and destination firms are in the same four-digit SIC group, zero otherwise. We also include

variables reflecting the extent of the origin and destination firms’ lending relationships in the

model. Banks Lending to Origin Firm is the number of lead arrangers that syndicated loans to the

origin firm during the executive’s tenure. Banks Lending to Destination Firm is the number of lead

arrangers that syndicated loans to the destination firm during the three years prior to the executive’s

18

appointment. Finally, we address the possibility that the same lead arrangers may be syndicating

loans to both the origin and destination firms prior to the executive’s turnover. Bank Overlap is an

indicator variable that equals one if at least one lead arranger lending to the origin firm during the

executive’s tenure also syndicated loans to the destination firm during the three years prior to the

executive’s appointment to the destination firm, zero otherwise. We measure our lead arranger

lending-based variables for the destination firm over the three-year period prior to the executive’s

turnover. However, our primary findings do not change if we extend this estimation period to five

years. Detailed variable definitions are provided in Appendix A.

Based on our three step matching procedure, we identify 648 control firm pairs for our

CEO/CFO turnover sample. Table 2 reports the summary statistics for the comparison between the

turnover and propensity-score-based matched samples. The differences in variable means between

these samples are insignificant for all variables, except for Bank Overlap, which is significant at

the 10% level. This evidence mitigates the concern that origin and destination firm-specific

characteristics may affect our findings.

4. Empirical Analyses

4.1 Test of Co-Migration

We begin our analysis by investigating the determinants of the co-migration of lenders and

managers between origin and destination firms using the following logistic regression –

.

(2)

Migrate is an indicator variable that equals one if at least one lead arranger that syndicated

loans to the origin firm during the executive’s tenure starts syndicating loans to the destination

19

firm within three years of the executive’s appointment, zero otherwise. We consider the lead

arranger as initiating a lending relationship with the destination firm if it did not arrange its

syndicated loans over the three-year period prior to the executive’s appointment at the destination

firm. We obtain similar results if we require the lead arranger to not have a lending relationship

with the destination firm over the five years prior to the executive’s appointment. Executive

Turnover is an indicator variable that equals one if the CEO or CFO of the origin firm migrates as

a CEO or CFO to the destination firm, zero otherwise (i.e., Executive Turnover equals one for the

actual origin-destination firm pairs and zero for the pseudo pairs).

We expect that lenders are more likely to migrate with a manager if the origin and destination

firms are more similar. Therefore, we control for differences in size, profitability and leverage

between the origin and destination firms. We measure Size Difference as the difference between

Destination Firm Size and Origin Firm Size. Profitability Difference and Leverage Difference are

defined analogously. If lenders have industry expertise, they are more likely to migrate with the

manager if she switches to another company in the same industry. Therefore, we include the Same

Industry indicator variable in the model (defined as in Model 1).

We also control for the number of the origin firm’s lead arrangers during the executive’s

tenure (Banks Lending to Origin Firm), as a larger number of lead arrangers with established

relationships with an executive should be associated with a higher probability of at least one of

them migrating to the destination firm. In contrast, a greater number of lead arrangers with

established relationships with the destination firm (Banks Lending to Destination Firm) is likely

to increase competition on syndicating its loans (Rajan, 1992), thus decreasing the likelihood of

the origin firm lead arrangers’ migrating with the executive. We also account for the possibility

that lead arrangers lending to the origin firms have a lending relationship with the destination firm

20

prior to the executive’s turnover (Bank Overlap). Banks Lending to Origin Firm, Banks Lending

to Destination Firm and Bank Overlap are defined as in Model 1.

We present our findings about the determinants of the co-migration of managers and lenders

in Table 3. We find a significant and positive coefficient on Executive Turnover, consistent with

our predictions that lenders are likely to follow a manager to the destination firm. The coefficient

of 0.451 on the Executive Turnover variable indicates that a lender is 1.6 times more likely to start

a lending relationship with a firm when a manager with whom it has a prior lending relationship

moves to the firm. The evidence of lender-manager co-migration suggests that personal

relationships between lenders and managers survive and carry over to the new firm following

managerial turnover.

In terms of control variables, as expected, we find a significant and positive coefficient on

the Banks Lending to Origin Firm variable. Bank Overlap also significantly increases the

probability of co-migration. This result is intuitive as the overlap in lead arrangers between the

origin and destination firms suggests that these borrowers are likely to have similar credit profiles,

increasing the probability that the origin firm’s lenders consider the destination firm to be an

attractive borrower.

We supplement these tests with untabulated analyses using a sample of firms that experience

turnover of non-CEO-CFO top officers who are disclosed in SEC filings as senior executives (e.g.,

Chief Risk Officer, Chief Operating Officer, Executive Vice President, Corporate Marketing and

Communications, Executive Vice President and General Counsel, or Executive Vice President,

Administration).4 Although relative to CEOs and CFOs these executives play a lesser role in

4The firms are required to provide information about their top 5 executive in SEC filings. Therefore, we are able to identify the turnover of non-CEO-CFO executives based on the Execucomp and BoardEx databases.

21

establishing a firm’s policies and strategies, lenders’ familiarity with their managerial style and

trust in their integrity may increase the likelihood of the lenders initiating lending relationships

with the destination firm. We identify 2,008 non-CEO-CFO executive moves and follow the same

procedure as with the CEO-CFO sample to create a control sample of origin-destination firm pairs.

We re-estimate Model 2 for the non-CEO-CFO sample and continue to find a positive and

significant coefficient on the Turnover variable. This evidence further supports our hypothesis that

lenders’ prior relationship with the origin firm’s top executives reduces information asymmetry

about the destination firm, increasing the probability that they will provide debt capital to the

destination firm.

4.2 Co-Migration and the Destination Firm’s Information Opacity

We next turn to testing and exploring cross-sectional differences in lender-manager co-

migration. As discussed in Section 2, we expect the likelihood of co-migration to be higher when

the destination firm is characterized by a poor information environment. We examine three

different proxies for the destination firm’s information environment. First, consistent with prior

research, we consider smaller firms to be more informationally opaque (e.g., Bharath et al., 2007,

and Wittenberg-Moerman, 2008). We assign a borrower to the small (large) firm subsample if its

long-term assets, measured at the end of the most recent fiscal year preceding the executive’s

turnover, are below (above) the sample median. Because credit rating agencies provide

assessments of a borrower’s creditworthiness that are valuable to lenders (e.g., Sufi, 2007, and

Kraft, 2014), the second proxy we use for a firm’s information environment is whether the firm is

rated by a credit rating agency. We obtain information on the credit ratings of the sample borrowers

from the S&P and Moody’s historical credit ratings databases. We assign a borrower to the non-

rated subsample if it is not rated by either S&P or Moody’s when a firm experiences the executive’s

22

turnover. Otherwise, the borrower is included in the rated subsample. We further assess a

borrower’s information environment by the extent of its equity analysts’ coverage (Güntay and

Hackbarth, 2010, and Mansi et al., 2010). We obtain analysts’ coverage data from the I/B/E/S

database. We assign a borrower to the low (high) analyst coverage subsample if the number of

analysts covering the firm in the year preceding the year of the manager’s move to the destination

firm is below (above) the sample median.

We report the results of these analyses in Table 4. As evidenced from Panel A, while the

coefficient on Executive Turnover is not statistically significant for the larger firm partition, a lead

arranger is 1.95 times more likely to start a lending relationship with a smaller borrower when a

manager with whom it has a prior lending relationship moves to this firm. The magnitude of the

Executive Turnover coefficient is also significantly larger for the small size partition (the

difference in coefficients on Executive Turnover across partitions is significant at the 10% level).

For non-rated borrowers, when an executive moves to the destination firm, the probability that one

of the origin firm’s lenders will start lending to the destination firm is higher by 2.3 times relative

to control firms. The effect of executive turnover is not significant for the rated partition, with the

difference in the coefficients on Executive Turnover between the two partitions being highly

significant. These findings are in line with Bharath (2007) findings that small and non-rated

borrowers are significantly more likely to use their relationship banks for future loans.

We obtain similar results for our last information opacity measure. Relative to more

transparent borrowers, when a borrower has low analyst coverage, lenders are significantly more

likely to co-migrate. For the low analyst coverage partition, the probability that one of the origin

firm’s lenders starts lending to the destination firm is 2.0 times higher when a manager with whom

it has a prior lending relationship moves to the firm. Overall, the results in Table 4 provide strong

23

support for our prediction that the occurrence of co-migration substantially increases when the

destination firm is more informationally opaque.

4.3. Benefits of Co-Migration to the Destination Firm

In this section, we investigate the benefits of co-migration to destination firms. An important

potential benefit to destination firms of hiring a manager with an established lender relationship is

access to financing through the relationship lender. We expect this access to relieve the capital

constraints that a destination firm may be facing. We investigate this conjecture by two sets of

tests. We start by examining whether the probability of co-migration is associated with a

destination firm’s access to debt capital. We predict that the likelihood of co-migration is higher

when the destination firm has limited access to capital. Next, we perform more direct analyses of

the benefits of co-migration by examining co-migration’s effects on the destination firm’s amount

of syndicated borrowing and its pricing.

4.3.1 Co-migration and access to debt capital

We employ two proxies for the destination firm’s access to debt capital. First, we assume

that a higher number of lead arrangers syndicating the destination firm’s loans (Banks Lending to

Destination Firm) indicates its better access to syndicated loan financing (Houston and James,

1996, and Murfin, 2012). We assign the destination firm to the weak access to credit category if it

had no lending relationships in the syndicated loan market in the three years prior to the executive

turnover. Otherwise, the destination firm is assigned to the strong access to credit category. In

untabulated analyses, we assign destination firms to the access to credit categories based on the

median number of lead arrangers and find similar results. Second, we assess the strength of the

destination firm’s access to debt capital based on the overall state of the debt markets. We assume

that the destination firm’s access to credit is weaker when the overall lending standards in credit

24

markets are tight. We use changes in bank lending standards, as reported by the Federal Reserve

Board’s Senior Loan Officer Opinion Survey on Bank Lending Practices (SLOOS) and which is

available from the Federal Reserve Bank of St. Louis, to identify changes in banks’ lending

standards (e.g., Bassett et al., 2012). We assign the destination firm to the weak access to credit

category when lending standards are tightening in the quarter of the loan’s origination, and to the

strong access to credit category otherwise.

Panel A of Table 5 reports the results from the estimation of Model 2 for subsamples based

on the destination firm’s number of lead arrangers. We find that the coefficient on Executive

Turnover is only significant in the weak access to credit subsample. Also, the coefficient is

significantly larger for destination firms with weak access relative to destination firms that have

better access to credit (the difference in coefficients between the weak and strong access to credit

partitions is significant at the 1 percent level). These findings suggest that a lead arranger’s co-

migration with the manager likely alleviates the destination firm’s credit constraints.

The results reported in Panel B further support these inferences. We find that the coefficient

on Executive Turnover is significant and positive only in periods of tight credit supply. Further,

the probability of a lead arranger starting a lending relationship with the destination firm is

significantly higher in periods of tight credit supply, compared to periods in which the credit supply

is relatively loose (the difference in coefficients between the weak and strong access to credit

partitions is significant at the 1 percent level). Thus, the destination firm benefits from access to

capital through the manager’s lending relationships when credit is tight and firms in the economy

are more likely to face credit constraints. Overall, the evidence presented in Table 5 is also

consistent with the relationship lending literature that demonstrates that the benefits of relationship

lending are particularly important when a borrower has limited access to debt capital (e.g., Rajan,

25

1992, Petersen and Rajan, 1994 and 1995) or during periods when bank credit is relatively scarce

(e.g., Bolton et al., 2013).

4.3.2 Co-migration and the extent and pricing of syndicated lending

In this section, to supplement our findings of lender-manager co-migration when the

destination firm is in need of debt capital, we more directly investigate the benefits of lenders’ co-

migration to the destination firm by focusing on the extent of the destination firm’s loan financing.

For the sample of firms that experienced executive turnover, we examine whether the destination

firm’s syndicated loan borrowing is more extensive when the lender co-migrates with the executive

relative to other destination firms where the lender does not co-migrate. We estimate the following

model using ordinary least square regressions:

. (3)

Amount is the destination firm’s syndicated loan borrowing, estimated as the average of the annual

syndication loan issuance divided by total assets over the first three years following the executive’s

turnover. Untabulated analysis indicates that the destination firms rely heavily on syndicated

lending. On average, these firms’ syndicated borrowing represents 34% of their long-term assets.

Migrate is defined as in previous analyses. A positive coefficient on Migrate will be consistent

with our expectation that destination firms have greater access to debt capital when the lead

arranger with an established relationship with the new manager co-migrates. We control for a

number of variables that are likely to be related to a destination firm’s financing needs. We capture

firm growth opportunities by the market-to-book ratio (Destination Firm Market-to-Book),

measured as the sum of market value of equity and total debt divided by the book value of assets.

26

We also include growth in the destination firm’s sales relative to the prior year (Destination Firm

Sales Growth) as another growth measure. Destination Firm Size, Destination Firm Profitability

and Destination Firm Leverage are defined as in previous analyses. All firm level characteristics

are measured in the year prior to the executive’s turnover.

We report the results of these tests in Panel A of Table 6. Consistent with our expectation,

we find a positive and significant (at the 10 percent level) effect of a lender’s co-migration on the

destination firm’s syndicated loan borrowing. The coefficient of 0.176 on the Migrate variable

suggests that lender co-migration results in an 18 percent higher syndicated loan borrowing for

destination firms relative to destination firms that do not experience lender co-migration. This

evidence of higher syndicated loan borrowing for destination firms when lead arrangers follow the

manager complements our findings in Table 5 that lead arrangers are more likely to co-migrate

when the destination firm is in need of relatively more debt capital.

We further extend the benefits of our co-migration analyses by exploring whether co-

migration is also associated with less expensive credit. As we discuss in Section 2, while lenders

may pass on some of the potential cost savings associated with lender-manager relationships to the

destination firm, the prior relationship lending literature provides mixed evidence with respect to

whether relationship lending decreases the cost of borrowing (e.g., Petersen and Rajan, 1994,

Degryse and Van Cayseele, 2000, Schenone, 2010, Bharath et al., 2011). We employ Model 4

above to examine whether loan pricing is associated with lender-manager co-migration, with

Spread as the dependent variable. Spread is the average interest rate spread on all syndicated loans

issued by the destination firm over the three-year period following the executive’s turnover. Note

that we focus on the average interest spread on all syndicated loans instead of the interest spread

on loans issued by the migrating banks only. Availability of credit from the migrating bank and

27

potentially more favorable pricing terms that it offers are likely to put competitive pressure on the

pricing of loans provided by other lenders (e.g., Bushman et al., 2015), thus making the average

loan pricing comparison more appropriate. Because loan pricing analyses focus on a sample of

treatment firms that experience executive turnover, this approach also allows us to assess the effect

of co-migration on loan pricing by comparing the cost of syndicated loan financing for firms where

the lender co-migrates with the manager to the cost of syndicated loan financing for firms that do

not experience co-migration. Lastly, we augment the model with the Amount variable to control

for the extent of syndicated lending.

Panel B of Table 6 presents the results of this analysis. We find that the coefficient on

Migrate is negative and significant (at the 10 percent level). This suggests that destination firms’

cost of borrowing is, on average, 21.7 basis points lower when lenders co-migrate with the

managers. Economically, this effect represents 11.8 percent of the average interest spread for the

sample destination firms. The coefficients on the control variables are also consistent with

expectations: larger and more profitable firms as well as firms with higher growth options

experience lower interest spreads, while firms with higher leverage and a higher extent of

syndicated lending pay higher spreads. Overall, interest spread analyses further support our

inference that lender co-migration brings important benefits to the destination firm.

4.4. The Strength of Lender-Manager Relationship and Co-Migration

Our analyses so far have focused primarily on the characteristics and financing needs of the

destination firm. In this section, we shed light on what affects a lender’s decision to co-migrate

with a manager. We predict that a stronger lender-manager relationship reduces information

asymmetry about the destination firm to a greater extent and thus is associated with a higher

probability of lenders’ co-migration. To investigate this prediction, we introduce a new

28

methodology where our level of observation is the lender and we consider each of the lead

arrangers that syndicated loans to the origin firm over the manager’s tenure in this analysis. The

sample comprises only those observations where Executive Turnover is equal to 1 (i.e., our

treatment sample from the analyses above where the manager moves from the origin firm to the

destination firm). We estimate the following lender level logistic regression:

.

(4)

Migrant is an indicator variable that equals one if the lead arranger migrates from the origin to the

destination firm, zero otherwise. We find that ten percent of the lead arrangers of the origin firms

follow the executive to the destination firm (untabulated). Primary Lender is our main variable of

interest and it reflects the strength of the lender-manager relationship. It is an indicator variable

that equals one if the lead arranger syndicated more than 75 percent of all loans to the origin firm

during the executive’s tenure, zero otherwise. We predict a positive coefficient on this variable.

We control for a number of variables that may affect the lead arranger’s likelihood of co-

migration. Banks Lending to Destination Firm is defined as in previous tests. A higher number of

lead arrangers with relationships with the destination firm is likely to increase the competition for

its loans (e.g., Rajan, 1992), reducing the probability of co-migration. Lending to Origin Firm

reflects the importance of the origin firm’s business to the lead arranger. It is defined as the ratio

of the number of the lead arranger’s syndicated loan deals to the origin firm divided by the total

number of deals syndicated by the lead arranger, estimated over the three-year period prior to the

executive turnover. The higher importance of the origin firm for the lead arranger’s syndication

activity may reduce its probability of co-migrating with the manager. We also control for a number

29

of the lead arranger’s additional characteristics that may affect its likelihood of initiating a new

lending relationship. In particular, we control for the lead arranger’s size (Lender Size), measured

as the natural logarithm of its total assets, its profitability (Lender Profitability), measured as the

ratio of the lender’s net income to shareholder equity, and capital ratio (Lender Capital), measured

as shareholder’s equity divided by total assets. These variables are estimated in the year preceding

the manager’s turnover. Finally, we add a control for the lead arranger’s loan growth (Lender Loan

Growth), measured as the annual percentage change in the lender’s loan portfolio as of the end of

the most recent fiscal year preceding the executive’s move to the destination firm.

We present our findings in Table 7. Consistent with our prediction about the importance of

the strength of the lender-manager relationship, we find a positive and significant coefficient on

the Primary Lender variable. Economically, the coefficient on Primary Lender indicates that

primary lead arrangers are 3.4 times more likely to co-migrate with the manager relative to the

origin firm’s other lead arrangers. In terms of control variables, a negative coefficient on Banks

Lending to Destination Firm is consistent with competition over the destination firm’s loans

decreasing the probability of co-migration. We also find that larger and more profitable lead

arrangers are more likely to co-migrate, potentially due to their stronger financial position, which

allows them to extend lending.

In the last set of analyses, we examine whether the strength of the lender-manager

relationship has a more significant effect on the probability of co-migration when lenders are under

pressure to expand their loan portfolio and seek new borrowers. In Panel A of Table 8, we present

the results of estimating Model 5 for the subsamples of low and high loan growth, based on the

sample median value of the Lender Loan Growth variable. We find that although the coefficient

on Primary Lender is significant for both subsamples, it is significantly higher for the low loan

30

growth lenders (the difference in coefficients across the low and high growth subsamples is

significant at the five percent level). This finding supports our prediction that relatively low loan

growth induces lead arrangers that have strong relationships with executives to expand their

lending portfolio by extending credit to destination firms.

The analyses based on the lender’s profitability-based classification are presented in Panel

B. We assign lenders to the low versus high profitability partition based on the sample median

value of Lender Profitability. The coefficient on Primary Lender is positive and significant only

in the low profitability partition, with the difference in the coefficients on this variable being

significantly different at the one percent level between the two partitions. This finding suggests

that low profitability incentivizes primary lead arrangers to follow managers to destination firms,

as this new lending relationship may increase their future profitability via additional interest

revenues, loan origination fees or fees from other services provided to the destination firm.

5. Conclusion

An extensive literature examines the nature and consequences of relationship lending. The

majority of prior studies focus on the relationship at the institutional level between lenders and the

borrowing firm (e.g., Petersen and Rajan, 1994 and 1995, Degryse and Van Cayseele, 2000,

Bharath et al., 2011, and Schenone, 2010). Because managers significantly affect a firm’s

corporate policies, strategies and performance, in this study we pose the question of whether

lenders develop relationships not only with a borrowing firm, but also with its managers, and

examine how these lender-manager relationships affect private lending. We posit that lenders’

relationships with managers reduce information asymmetry about the borrowing firm’s

creditworthiness and future performance and thus significantly affect lending practices.

Consistent with this prediction, using a setting of executive turnovers to identify lender-

31

manager relationships, we find that a lender is 1.6 times more likely to commence a lending

relationship with a firm when a manager with whom it has a prior relationship joins the firm as a

top executive. Lenders’ relationships with managers are particularly valuable when borrowers are

informationally opaque - the likelihood of lenders co-migrating is significantly higher when the

destination firm is smaller, not rated by a credit rating agency or when it has low analyst following.

Further, co-migration benefits destination firms by providing them with greater access to credit

and at lower costs. We find that lenders are more likely to follow managers when the firms they

join have weaker access to debt capital. We also show that lenders’ co-migration is associated with

an 18 percent higher syndicated loan borrowing for destination firms following executive turnover

relative to destination firms that do not experience lender co-migration. On average, the destination

firms’ cost of borrowing is 21.7 basis points lower when lenders co-migrate with managers. With

respect to the factors that affects lenders’ decision to co-migrate, we find that co-migration is more

likely when lenders have a stronger prior relationship with the manager. Finally, the strength of

the lender-manager relationship has a significantly stronger influence on the decision to co-migrate

when lenders are under pressure to expand their loan portfolios and are seeking new clients.

We contribute to the relationship lending literature by highlighting the importance of the

relationship between lenders and managers. We also show that this relationship significantly

benefits borrowers via enhanced access to credit and a lower cost of debt financing. We extend the

developing literature on the role of personal relationships in capital markets. Our findings suggest

that the lender-manager relationship carries over to the new firm that hires the manager as a top

executive. We further contribute to the literature that examines how personal networks and

relationships affect business and investment decisions. We show that lenders’ relationship with a

manager incentivizes lenders to initiate a lending relationship with the firm that the manager joins.

32

References

Bamber, L., J. Jiang, and I. Wang. 2010. What’s my style? The influence of top managers and their personal backgrounds on voluntary corporate financial disclosure. The Accounting Review 85 (4): 1131-1162.

Bassett, W., M. Chosak, J. Driscoll, and E. Zakrajsek, 2012. Changes in bank lending standards and the macroeconomy. Finance and Economics Discussion Series Divisions of Research & Statistics and Monetary Affairs Federal Reserve Board, Washington, D.C.

Berger, A. N., and G. F. Udell. 1995. Relationship lending and lines of credit in small firm finance. Journal of Business 68: 351-382.

Bertrand, M., and A. Schoar. 2003. Managing with style: The effect of managers on firm policies. The Quarterly Journal of Economics 118 (4): 1169-1208.

Bharath, S., S. Dahiya, A. Saunders, and A. Srinivasan. 2007. So what do I get? The bank's view of lending relationships. Journal of Financial Economics 85 (2): 368-419.

Bharath, S., S. Dahiya, A. Saunders, and A. Srinivasan. 2011. Lending relationships and loan contract terms. Review of Financial Studies 24: 1141-1203.

Bharath, S., J. Sunder, and S. V. Sunder. 2008. Accounting quality and debt contracting. The Accounting Review 83 (1) :1-28.

Bolton, P., X. Freixas, L. Gambacorta, and P. Mistrulli. 2013. Relationship and transaction lending in a crisis. National Bureau of Economic Research Working Paper.

Boot, A. 2000. Relationship banking: what do we know? Journal of Financial Intermediation 9 (1): 7-25.

Boot, A., and A. Thakor. 1994. Moral hazard and secured lending in an infinitely repeated credit market game. International Economic Review 35 (4): 899-920.

Brochet, F., G. Miller, and S. Srinivasan. 2014. Do analysts follow managers who switch companies? An analysis of relationships in the capital markets. The Accounting Review 89 (2): 451-482.

Bushman, R., A. Smith, and R. Wittenberg-Moerman. 2010. Price discovery and dissemination of private information by loan syndicate participants. Journal of Accounting Research 48: 921- 972.

Bushman, R., B. Hendricks, and C. Williams. 2015. Bank competition: measurement, decision-making and risk-taking, Journal of Accounting Research, forthcoming.

Cohen, L., A. Frazzini, and C. Malloy. 2008. The small world investing: Board connections and mutual fund returns. Journal of Political Economy 116 (5): 951-979.

Costello, A. M., and R. Wittenberg-Moerman. 2011. The impact of financial reporting quality on debt contracting: Evidence from internal control weakness reports. Journal of Accounting Research 49 (1): 97-136.

33

Degryse, H., and P. Van Cayseele. 2000. Relationship lending within a bank-based system: Evidence from European small business data. Journal of Financial Intermediation 9 (1): 90-109.

DeJong, D., and Z. Ling. 2013. Managers: Their effects on accruals and firm policies. Journal of Business, Finance and Accounting 40 (1-2): 82-114.

Dennis, S., and D. Mullineaux. 2000. Syndicated loans. Journal of Financial Intermediation 9: 404-426.

Diamond, D. 1984. Financial intermediation and delegated monitoring. Review of Economic Studies 51: 393-414.

Diamond, D. 1991. Monitoring and reputation: The choice between bank loans and directly placed debt. Journal of Political Economy 99: 688-721.

Drucker, S., and M. Puri. 2005. On the benefits of concurrent lending and underwriting. The Journal of Finance 60 (6): 2763-2799.

Dyreng, S. D., M. Hanlon, and E. L. Maydew. 2010. The effects of managers on corporate tax

avoidance. The Accounting Review 85 (4): 1163-1189.

Elyasiani, E., and L. G. Goldberg. 2004. Relationship lending: A survey of the literature. Journal of Economics and Business 56 (4):315-330.

Engleberg, J., P. Gao, and C. A. Parsons. 2012. Friends with money. Journal of Financial Economics 103: 169-188.

Fracassi, C. 2014. Corporate finance policies and social networks. Working Paper.

Ge, W., D. Matsumoto, and J. Zhang. 2011. Do CFOs have styles of their own? An empirical investigation of the effect of individual CFOs on financial reporting practices. Contemporary Accounting Research 28 (4): 1141-1179.

Greenbaum, S., and Thakor, A. V. 1995. Contemporary Financial Intermediation. Dryden Press, New York.

Güntay, L., and Hackbarth, D. 2010. Corporate bond credit spreads and forecast dispersion. Journal of Banking & Finance, 34(10): 2328-2345.

Hambrick, D., 2007. Upper echelons theory: An update. Academy of Management Review 32 (2): 334-343.

Hambrick, D. C., and P. A. Mason. 1984. Upper echelons: The organization as a reflection of its top managers. Academy of Management Review 9 (2): 193-206.

Haselmann, R., D. Schoenherr, and V. Vig. 2013. Lending in social networks. Working Paper.

Hoshi, T., A. Kashyap, and D. Scharfstein. 1990a. Bank monitoring and investment: Evidence from the changing structure of Japanese corporate banking relationships. In R. Glenn Hubbard (ed.): Asymmetric Information, Corporate Finance and Investment. University of Chicago Press, Chicago.

34

Hoshi, T., A. Kashyap, and D. Scharfstein. 1990b. The role of banks in reducing the costs of financial distress. Journal of Financial Economics 27: 67-88.

Hoshi, T., A. Kashyap, and D. Scharfstein. 1991. Corporate structure, liquidity and investment: Evidence from Japanese industrial groups. The Quarterly Journal of Economics 106: 33-60.

Houston, J., and C. James. 1996. Bank information monopolies and the mix of private and public debt claims. Journal of Finance: 1863-1889.

Ivashina, V. 2009. Asymmetric information effects on loan spreads. Journal of financial Economics 92 (2): 300-319.

Karolyi, S. A. 2014. Personal lending relationships. Working Paper, Carnegie Mellon University.

Koch, T., and S. MacDonald. 2014. Bank management. Cengage Learning.

Kraft, P. 2015. Rating agency adjustments to GAAP financial statements and their effect on ratings and credit spreads. The Accounting Review 90 (2): 641-674.

Mansi, S., W. Maxwell, and D. Miller. 2011. Analyst forecast characteristics and the cost of debt, Review of Accounting Studies 16: 116-142.

Murfin, J. 2012. The supply‐side determinants of loan contract strictness. The Journal of Finance 67 (5): 1565-1601.

Pan, Y., T. Y. Wang, and M. S. Weisbach. 2015a. Learning about CEO ability and stock return volatility. Review of Financial Studies 28: 1623-1666.

Pan, Y., T. Y. Wang, and M. S. Weisbach. 2015b. Management risk and the cost of borrowing. Working Paper, Fisher College of Business.

Petersen, M. A. 2004. Information: Hard and soft. Working Paper, Northwestern University.

Petersen, M., and R. Rajan. 1994. The benefits of relationships: Evidence from small business data. The Journal of Finance 49: 1367-1400.

Petersen, M., and R. Rajan. 1995. The effect of credit market competition on lending relationships. The Quarterly Journal of Economics 110 (2): 407-443.

Plath, C. 2008. Analyzing credit and governance implications of management succession planning. Working Paper.

Rajan, R. 1992. Insiders and outsiders: The choice between informed and arm’s length debt. Journal of Finance 47: 1367-1400.

Rajan, R. G., and L. Zingales. 1998. Which capitalism? Lessons form the East Asian crisis. Journal of Applied Corporate Finance 11 (3): 40-48.

Rosenbaum, P., and D. Rubin. 1983. The central role of the propensity score in observational studies for causal effects. Biometrika 70 (1):41-55.

Santos, J., and A. Winton. 2013. Bank capital, borrower power, and loan rates. EFA 2009 Bergen Meetings Paper

Schenone, C. 2010. Lending relationships and information rents: Do banks exploit their information advantages? The Review of Financial Studies:1149-1199.

35

Schmidt, B. 2015. Costs and benefits of friendly boards during mergers and acquisitions. Journal of Financial Economics 117 (2): 424-447.

Sharpe, S. 1990. Asymmetric information, bank lending and implicit contracts: A stylized model of customer relationships. Journal of Finance 45: 1069-1087.

Standard & Poor’s. 2007. A Guide to the Loan Market (Standard & Poor’s, New York).

Stiglitz, J., and A. Weiss. 1981. Credit rationing in markets with imperfect information. American Economic Review 71: 393-410.

Sufi, A. 2007. Information asymmetry and financing arrangements: Evidence from syndicated loans. The Journal of Finance 62 (2): 629-668.

Wittenberg-Moerman, R. 2008. The role of information asymmetry and financial reporting quality in debt trading: Evidence from the secondary loan market. Journal of Accounting and Economics 46 (2): 240-260.

Yasuda, A. 2005. Do bank relationships affect the firm's underwriter choice in the corporate‐bond underwriting market? The Journal of Finance 60 (3): 1259-1292.

36

Appendix A: Variable Definitions Variable Definition Amount

Banks Lending to Origin Firm

The destination firm’s syndicated loan borrowing, estimated as the average of the annual syndication loan issuance divided by total assets over the first three years following the executive’s turnover (DealScan and Compustat).

The number of lead arrangers that syndicated loans to the origin firm during the executive’s tenure (DealScan).

Banks Lending to Destination Firm

The number of lead arrangers that syndicated loans to the destination firm during the three years prior to the executive’s appointment (DealScan).

Bank Overlap

An indicator variable equal to one if at least one lead arranger lending to the origin firm during the executive’s tenure also syndicated loans to the destination firm during the three years prior to the executive’s appointment at the destination firm, zero otherwise (DealScan).

Destination Firm Leverage

Destination Firm Market-to-Book

Destination Firm Profitability

Destination Firm Sales Growth

Destination Firm Size

The ratio of the destination firm’s long-term debt to total assets, measured at the end of the most recent fiscal year preceding the executive’s turnover (Compustat).

The sum of the destination firm’s market value of equity and total debt divided by the book value of assets, measured at the end of the most recent fiscal year preceding the executive’s turnover (Compustat).

The ratio of the destination firm’s net income before extraordinary items to total assets, measured at the end of the most recent fiscal year preceding the executive’s turnover (Compustat).

The destination firm’s sales growth relative to the prior year, measured at the end of the most recent fiscal year preceding the executive’s turnover (Compustat).

A natural logarithm of the destination firm’s total assets, measured at the end of the most recent fiscal year preceding the executive’s turnover (Compustat).

37

Executive Turnover

Lender Capital

Lender Loan Growth

Lender Profitability

Lender Size

Lending to Origin Firm

Leverage Difference

Migrant

Migrate

Origin Firm Leverage

An indicator variable that takes the value of one if the CEO or CFO of the origin firm migrates to the destination firm as a CEO or CFO, zero otherwise (Execucomp and Bordex).

Lead arranger’s shareholder equity divided by total assets, measured at the end of the most recent fiscal year preceding the executive’s turnover (Bank Compustat).

The annual percentage change in the size of the lead arranger’s loan portfolio, measured at the end of the most recent fiscal year preceding the executive’s turnover (Bank Compustat).

The ratio of a lead arranger’s net income to shareholders’ equity (Bank Compustat), measured at the end of the most recent fiscal year preceding the executive’s turnover (Bank Compustat).

The natural logarithm of the lead arranger’s total assets, measured at the end of the most recent fiscal year preceding the executive’s turnover (Bank Compustat).

The ratio of the number of syndicated loan deals of the lead arranger to the origin firm divided by the total number of deals syndicated by the lead arranger, estimated over the three-year period prior to the executives’ turnover (DealScan).

The difference between Destination Firm Leverage and Origin Firm Leverage.

An indicator variable that takes the value of one if the lead arranger migrates with the manager to the destination firm, zero otherwise (DealScan).

An indicator variable that takes the value of one if at least one lead arranger that syndicated loans to the origin firm during the executive’s tenure starts syndicating loans to the destination firm within three years of the executive’s appointment, zero otherwise. We consider the lead arranger as starting the relationship with the destination firms if it had not arranged its syndicated loans over the three-year period prior to the executive turnover (DealScan).

The ratio of the origin firm’s long-term debt to total assets, measured at the end of the most recent fiscal

38

Origin Firm Profitability

Origin Firm Size

year preceding the executive’s turnover (Compustat).

The ratio of the origin firm’s net income before extraordinary items to total assets, measured at the end of the most recent fiscal year preceding the executive’s turnover (Compustat).

A natural logarithm of the origin firm’s total assets, measured at the end of the most recent fiscal year preceding the executive’s turnover (Compustat).

Primary Lender

Profitability Difference

Same Industry

Size Difference

Spread

An indicator variable that takes the value of one if the lead arranger syndicated more than 75 percent of all loans to the origin firm during the executive’s tenure, zero otherwise (DealScan).

The difference between Destination Firm Profitability and Origin Firm Profitability.

An indicator variable equal to 1 if the origin and destination firms are in the same four-digit SIC group, and 0 otherwise (Compustat).

The difference between Destination Firm Size and Origin Firm Size.

An average interest rate spread on all syndicated loans issued by the destination firm over the three-year period following the executive’s turnover (DealScan).

Turnover

For Step 1 of the matching procedure, an indicator variable that takes the value of one when the executive’s origin firm is paired with its actual destination firm, zero otherwise. For Step 2 of the matching procedure, an indicator variable that takes the value of one when the executive’s destination firm is paired with its actual origin firm, zero otherwise (Execucomp and Boardex).

39

Figure 1: The Illustration of the Matching procedure Step 1 X= every pseudo destination firm that has the same 2-digit SIC code as destination firm B. b = the pseudo destination firm that has the closest propensity score to the actual destination firm. Step 2 X=every pseudo origin firm that has the same 2-digit SIC code as origin firm A. a= the pseudo origin firm that has the closest propensity score to the actual origin firm. Step 3

Treatment sample

Turnover pair Origin firm A &

Destination firm B

1 to N matching Pseudo turnover pairs

Origin firm A & Pseudo destination firm X

 1 to 1 Matched sample

Pseudo turnover pair Origin firm A &

Pseudo Destination firm b

Treatment sample

Turnover pair Origin firm A &

Destination firm B

1 to N matching Pseudo turnover pairs Pseudo Origin firm X &

Destination firm B

1 to1 Matched sample

Pseudo turnover pair

Pseudo Origin firm a & Destination firm B

Treatment sample Turnover pair Origin firm A &

Destination firm B

1 to 1 Matched sample Pseudo turnover pair Pseudo origin firm a &

Pseudo destination firm b

40

Table 1:Sample Selection and Origin and Destination Firm Characteristics This table presents the sample selection process (Panel A) and descriptive statistics for origin and destination firms (Panel B). Variables are defined in Appendix A. Panel A: Sample Selection

Sample selection step CEO-CFO moves Unique executive moves in Execucomp & Boardex firms 723 Merged with Compustat, Dealscan and I/B/E/S 648

Panel B: Firm Characteristics of Origin and Destination Firms Variable Mean P-value for the difference Origin Firm Size 7.5463 Destination Firm Size 7.3554 Difference 0.1909 0.0409

Origin Firm Profitability 0.0170 Destination Firm Profitability -0.0011 Difference 0.0181 0.0285

Origin Firm Leverage 0.2228 Destination Firm Leverage 0.2053 Difference 0.0175 0.0916

41

Table 2: Treatment and Control Samples and Descriptive Statistics This table provides descriptive statistics for the treatment and control samples of origin and destination firms. There are 648 observations in each sample. Variables are defined in Appendix A.

Variable Treatment Sample

Mean Control Sample

Mean P-Value for Differences

Size Difference -0.1173 -0.1456 0.7794 Profitability Difference

-0.0220 -0.0120 0.4883

Leverage Difference -0.0087 -0.0116 0.8706 Same Industry 0.2047 0.1767 0.2024 Banks lending to original firm

2.4605 2.5256 0.7446

Banks lending to destination firm

1.2264 1.3984 0.1538

Bank overlap 0.1209 0.0915 0.0861

42

Table 3: Do Lenders Follow the Manager to the Destination Firm?The table presents the analysis of the effect of the executive turnover on the probability that at least one of the lead arrangers of the origin firm initiates a lending relationship with the destination firm. The dependent variable is Migrate, which takes the value of one if at least one lead arranger that syndicated loans to the origin firm during the executive’s tenure starts syndicating loans to the destination firm within 3 years of the executive’s appointment, zero otherwise. All other variables are defined in Appendix A. The t-statistics are reported in parentheses. ***, ** and * indicates significance at the 0.01, 0.05, 0.10 level, respectively.

Variables Dependent Variable: Migrate Executive Turnover 0.451** (2.38)Size Difference 0.072 (1.24) Profitability Difference 0.385 (0.87) Leverage Difference 0.043 (0.15) Same Industry 0.253 (1.13) Banks Lending to Origin Firm 0.100*** (4.43) Bank Lending to Destination Firm 0.032 (0.85) Bank Overlap 1.832*** (8.20) Model Logit Pseudo R2 N

0.1423 1,296

43

Table 4: Lenders’ Co-migration and the Destination Firm’s Information Opacity This table presents the analyses of the effect of the executive turnover on the probability that at least one of the lead arrangers of the origin firm initiates a lending relationship with the destination firm, conditional on the destination firm’s information opacity. Panel A presents the analyses based on a destination firm’s size, Panel B presents the analyses based on whether a borrower is rated by a credit rating agency and Panel C presents the analyses based on the extent of the destination firm’s analyst coverage. The dependent variable is Migrate, which takes the value of one if at least one lead arranger that syndicated loans to the origin firm during the executive’s tenure starts syndicating loans to the destination firm within three years of the executive’s appointment, zero otherwise. All other variables are defined in Appendix A. The t-statistics are reported in parentheses. ***, **, * indicates significance at the 0.01, 0.05, 0.10 level respectively. ###, ##, # indicates that the difference across destination firm size, rating status or analyst coverage partitions is significant at the 0.01, 0.05 and 0.10 level respectively. Panel A: Lenders’ co-migration with the executive, conditional on a destination firm’s size Dependent Variable: Migrate Variable Small Large Executive Turnover 0.667* 0.358 (1.83) (1.58)# Controls Included Included N 648 648

Panel B: Lenders’ co-migration with the executive, conditional on whether a destination firm is rated Dependent Variable: Migrate Variable Not Rated Rated Executive Turnover 0.840*** 0.027 (3.05) (0.09)### Controls Included Included N 853 443

Panel C: Lenders’ co-migration with the executive, conditional on the intensity of a destination firm’s analyst coverage Dependent Variable: Migrate Variable Low High Executive Turnover 0.690** 0.230 (2.42) (0.87)## Controls Included Included N 652 644

44

Table 5: Lenders’ Co-migration and a Destination Firm’s Access to Debt Capital This table presents the analyses of the effects of the executive turnover on the probability that at least one of the lead arrangers of the origin firm initiates a lending relationship with the destination firm, conditional on the destination firm’s access to debt capital. In Panel A, we measure a destination firm’s access to credit based on the number of lead arrangers that syndicated its loans over the three-year period prior to the executive’s turnover. In Panel B, we measure access to credit based on the tightness of the lending standards in credit markets in the quarter of a loan’s origination. The dependent variable is Migrate, which takes the value of one if at least one lead arranger that syndicated loans to the origin firm during the executive’s tenure starts syndicating loans to the destination firm within three years of the executive’s appointment, zero otherwise. All other variables are defined in Appendix A. The t-statistics are reported in parentheses. ***, **, * indicates significance at the 0.01, 0.05, 0.10 level respectively. ###, ##, # indicates that the difference across access to credit partitions is significant at the 0.01, 0.05 and 0.10 level respectively. Panel A: Measuring a destination firm’s access to credit based on the number of its prior banking relationships Dependent Variable: Migrate Variable Weak access Strong access Executive Turnover 1.990*** -0.037 (4.04) (-0.17)### Controls Included Included N 498 798

Panel B: Measuring a destination firm’s access to credit based on the tightness of lending standards in credit markets Dependent Variable: Migrate Variable Weak access Strong access Executive Turnover 0.694*** 0.317 (2.20) (1.31)### Controls Included Included N 525 771

45

Table 6: Co-Migration and the Extent and Pricing of Syndicated Lending This table presents the analyses of the effect of lenders’ co-migration with the manager on the syndicated lending amount (Panel A) and its pricing (Panel B). In Panel A, the dependent variable is Amount, which is the destination firm’s syndicated loan borrowing, estimated as the average of the annual syndication loan issuance divided by total assets over the first three years following the executive’s appointment. In Panel B, the dependent variable is Spread, estimated as the average interest rate spread on all syndicated loans issued by the destination firm over the three-year period following the executive’s turnover. All other variables are defined in Appendix A. The t-statistics are reported in parentheses. ***, **, * indicates significance at the 0.01, 0.05 and 0.10 levels, respectively. Panel A: Syndicated loan amount Variables Dependent Variable: Amount Migrate 0.176* (1.73)Destination Firm Market-to-Book -0.013 (-0.70) Destination Firm Sales Growth 0.005 (0.04) Destination Firm Size -0.057*** (-3.07) Destination Firm Profitability 0.285* (1.96) Destination Firm Leverage 0.459*** (3.42) Fixed Effects Industry, Year Model OLS R2 0.093 N 560

46

Table 6: Co-migration and the extent and pricing of syndicated lending (continued) Panel B: Syndicated loan pricing Variables Dependent Variable: Spread Migrate -21.700* (-1.68) Destination Firm Market-to-Book -21.526*** (-3.17) Destination Firm Sales Growth 7.163 (0.22) Destination Firm Size -17.919*** (-4.90) Destination Firm Profitability -269.673*** (-5.26) Destination Firm Leverage 81.114*** (2.97) Destination Firm Amount 47.396*** (6.28) Fixed Effects Industry, Year Model OLS R2 0.448 N 407

47

Table 7: Co-migration and Lending Intensity This table presents the analysis of the effect of the strength of lender-manager relationship on the lender’s probability of co-migrating with the manager. The dependent variable is Migrant, which is an indicator variable that takes the value of one if the lead arranger migrates with the manager to the destination firm, zero otherwise. All other variables are defined in Appendix A. The t-statistics are reported in parentheses. ***, **, * indicates significance at the 0.01, 0.05, 0.10 level respectively.

Variables Dependent Variable: Migrant Primary Lender 1.233*** (3.95) Bank Lending to Destination Firm -0.066*** (-3.30) Lending to Origin Firm -0.712 (1.40) Lender Size 0.356*** (4.16) Lender Loan Growth -0.507 (0.62) Lender Profitability 8.602*** (1.31) Lender Capital 4.484 (3.89) Model Logit N 1,895

48

Table 8: A Lender’s Incentives to Co-Migrate – Loan Growth and Profitability This table presents the analyses of the effects of a primary lender’s loan growth and profitability on its probability of co-migrating with the executive. Panel A presents the analyses based on the growth of the lead arranger’s loan portfolio, while Panel B presents the analyses based on a lead arranger’s profitability. The dependent variable is Migrant, which is an indicator variable that takes the value of one if the lead arranger migrates with the manager to the destination firm, zero otherwise. All other variables are defined in Appendix A. The t-statistics are reported in parentheses. ***, **, * indicates significance at the 0.01, 0.05, 0.10 levels respectively. ###, ##, # indicates that the difference across the loan growth or profitability partitions is significant at the 0.01, 0.05 and 0.10 level respectively. Panel A: Loan Growth Dependent Variable: Migrant Variable Low High Primary Lender 1.648*** 0.818** (2.95) (2.01)## Controls Included Included N 959 936

Panel B: Profitability Dependent Variable: Migrant Variable Low High Primary Lender 1.959*** 0.257 (4.06) (0.58)### Controls Included Included N 1,002 893