Using Collateral to Secure Loans

BY YARON LEITNER

any businesses post collateral as security for theoretical work on collateral has been driven by economists’ desire to provide M loans. Collateral protects the lender if the explanations for the use of collateral borrower defaults. However, not all borrowers that are consistent with this empirical finding among others. put up collateral when taking out loans. There’s even some evidence that loans with collateral COLLATERAL AND BORROWERS’ INCENTIVES attached may be riskier for lenders. Why is collateral We start by focusing on the way used sometimes, but not others? And why does collateral collateral affects a borrower’s incen- tives to ensure the business’s success. potentially involve more risk? In this article, Yaron Consider a loan contract where an Leitner considers these questions. He looks at some of the individual borrows some money to explanations for using collateral, focusing on its benefits start a new business. The success of the business often depends on actions and drawbacks. the borrower takes after the loan is signed, for example, the way he allo- cates money among different activities, and the effort he expends in choosing Collateral is a contractual device finance their investments. low-cost/high-value alternatives. Ideal- used by borrowers and lenders around Understanding collateral is im- ly, the loan contract would specify all the world. Collateral has also been portant because it is a characteristic of these actions. However, in many around for a long time. In one famous feature of bank loans, which help to cases, this is impossible because some 1 example, a pound of Antonio’s flesh channel resources to their best use. of these actions may not be observable collateralized Shylock’s loan to Bas- While early research focused mainly to a third party or even to the lender; sanio in Shakespeare’s “Merchant of on how collateral affects the borrower’s for example, it may be difficult for the Venice.” Generally, the term collateral behavior, recent research has also bank to argue in court that a borrower refers to assets pledged by a borrower incorporated lenders’ behavior, for did not exert enough effort in choosing to secure a loan. The lender can seize example, how collateral affects lenders’ the best alternatives.2 these assets if the borrower does not incentives to take care in evaluating If the borrower and lender had make the agreed-upon payments on a business’s prospects. Economists the same objectives, the fact that the the loan, so the lender has some pro- have also examined the relationship borrower’s actions are not observable tection if the borrower defaults. There- between collateral and risk, empirically to others would not be a problem. fore, the use of collateral can make verifying bankers’ common wisdom it easier for firms to obtain loans to that collateralized loans are riskier for the bank than noncollateralized 2 The finance and economics literature refers loans. To a significant extent, recent to this hidden action problem as moral hazard. Yaron Leitner is a This term, which was coined in the insurance senior economist industry, captures the idea that an individual in the Research who has insurance is less likely to take actions 1 Department of According to the Federal Reserve’s Surveys to avoid problems. For example, if you have of Terms of Business Lending, more than 50 comprehensive car insurance with no deduct- the Philadelphia percent of the value of all commercial and ibles, you may be less careful about locking your Fed. This article industrial loans made by domestic banks in the car or parking it in a safe spot. More broadly, is available free U.S. is currently secured by collateral (based on the term moral hazard refers to any contracting of charge at www. the surveys for February 2005, May 2005, and problem where the actions of one party cannot philadelphiafed.org/ August 2005). be observed by others. econ/br/index.html. www.philadelphiafed.org Business Review Q2 2006 9 The borrower would take the actions according to the lender’s wishes, but collateral would normally sell for. In that are best for him, and these ac- when these actions cannot be verified addition, businesses in a given industry tions would also be best for the lender. in court, such a promise is just cheap often fail together. But when many However, in practice, the borrower and talk. lenders try to sell at the same time, lender often have different objectives. Collateral May Induce the Bor- the market gets flooded and the price The lender wants to make sure that rower to Exert Effort… Suppose the they can obtain decreases. Overall, the loan is paid in full; the borrower borrower posts his house or some of his economists call this loss in asset value cares about the profits left after pay- business assets as collateral to secure a deadweight loss because the lender ing the loan. The borrower may also the loan. This may induce him to put does not gain as much as the borrower care about some perks that benefit more effort into ensuring the business loses. Another deadweight loss involves him, but not the business as a whole; succeeds because if the business fails, transferring control of the collateral- for example, the borrower may enjoy the borrower loses his collateral. In ized assets, which often involves legal expensive business meals, a private jet, other words, collateral can give the and other administrative costs. There- and so forth. borrower the incentive to work harder. fore, there is a tradeoff: Collateral re- Consider the following as an ex- ample of a conflict of interests between Collateral reduces the cost of borrowing borrowers and lenders: A business can either succeed or fail. If it fails, because it gives the borrower incentives to the loan cannot be repaid, and both work hard, but it also increases the cost of the borrower and lender get nothing. If the business succeeds, the loan is borrowing because the collateral may be worth paid in full, and the borrower is left more to the borrower than to the lender and with the rest of the profits. Now sup- because transferring control imposes costs. pose that the borrower can take an action that has the following effect: When the borrower works harder, the duces the cost of borrowing because it If the business is a success, the action business is more likely to succeed, and gives the borrower incentives to work increases profits; however, the action the borrower is less likely to default. hard, but it also increases the cost of reduces the chances that the business But then the lender may be more will- borrowing because the collateral may will succeed.3 The borrower may be ing to lend his money and at a lower be worth more to the borrower than happy to take such an action because it interest rate. to the lender and because transferring increases the money left for him — re- …But Using Collateral Is Costly. control imposes costs. member, he gets paid only if the busi- The benefit above comes at a cost. A A Long-Term Relationship with ness succeeds.4 The lender, however, is business might fail even if the borrower a Bank Can Reduce the Need for unhappy because he is less likely to get exerts a lot of effort; the borrower may Collateral. In their paper, Arnoud his money back. have bad luck. In this case, the bor- Boot and Anjan Thakor suggest that Anticipating the conflict of inter- rower loses the collateral, which may long-term relationships between a ests above, the lender may demand a be worth more to him than it is to the borrower and a lender can reduce the higher interest rate on the loan, and lender. For example, if the borrower need for collateral. When the loan in some cases, he may not lend at all. has posted his house as collateral, be- contract is a one-time transaction for Of course, the borrower can promise ing able to continue living there is the bank and borrower, there are two to take some agreed-upon actions important to the borrower but not the ways to induce the borrower to exert lender. Or if the borrower has posted effort. his business assets, they may be worth The first is to require collateral, 3 An example of such an action is a business expansion. If the business succeeds, there are more to him, since he knows how to as discussed above. The second is to more profits. But because the firm spends re- use those assets to produce goods, lower the interest rate on the loan. A sources on the expansion, it has less to spend cultivating its old customers. and the lender does not. The lender lower interest rate leaves more profits may choose to sell the collateral to for the borrower and therefore induces 4 Of course, many businessmen and -women are someone else, but since the lender has him to exert effort to make the busi- motivated by ethical concerns and their reputa- tions. For the most part, we ignore these moti- an incentive to sell as quickly as pos- ness succeed. However, if the interest vations to highlight the role of collateral. sible, he may obtain less than what the rate needed to induce the borrower to

10 Q2 2006 Business Review www.philadelphiafed.org exert effort is too low, the loan may the start of the relationship, and this COLLATERAL AND RISK not be profitable to the lender; he may compensates the bank for the loss of We have seen that collateral pro- be able to get a higher interest rate by profits later in the relationship. Over- vides incentives for the borrower to lending to other firms or individuals. all, the bank breaks even, and the cost avoid default. Collateral also reduces The result is that the lender may need of collateral is reduced because, at the the loss to the lender if a borrower de- to require collateral, and as we have start of the relationship, the promise of faults on a loan: If the loan is not paid, seen, this comes at a cost. better loan terms reduces the need for the lender can seize the collateral. One When the borrower and lender collateral, and when the relationship might conclude that secured loans are have a long-term relationship, the progresses, collateral is not needed. safer for the lender than unsecured bank has another way to induce the Boot and Thakor’s model predicts loans. The data, however, show the borrower to exert effort. The bank that borrowers with a longer banking opposite. can promise the borrower better terms relationship are less likely to pledge In their 1990 paper, Berger and on new loans in the future, once the Udell found that net chargeoffs (the business shows some signs of success.5 amount of a loan the bank cannot Better terms mean less collateral and When a borrower collect) are likely to be higher when a a lower interest rate. The borrower has posts collateral, loan is secured. They also found that an incentive to work hard even though borrowers who post collateral are more he pledges less collateral because the bank becomes likely to perform poorly; for example, working hard increases the chances less conservative in they are more likely to be late on that the business will succeed and the approving his loan. their payments. These two findings terms on future loans will improve. In suggest that secured loans are riskier the future, under the new loan terms, for the bank; this is consistent with the borrower has an incentive to work collateral. This prediction is consistent conventional wisdom in the banking hard because of the low interest rate; with what Allen Berger and Gregory industry.8 therefore, collateral is no longer need- Udell found in their 1995 paper. Us- A possible explanation is that ed to induce effort. ing data on collateral requirements banks require more collateral when But how can the lender afford on lines of credit issued to small busi- they perceive a loan to be riskier. to reduce the interest rate on future nesses, Berger and Udell found that Banks collect information about bor- loans? In a competitive loan market, firms that had long-term relationships rowers, for example, the borrower’s all lenders break even; they make with a lender were less likely to pledge income and performance with past enough money just to cover their costs. collateral.6 An additional 10 years of loans. Banks can use this informa- Thus, a lender that offers a lower in- bank-borrower relationship lowered tion to distinguish between borrowers terest rate and requires less collateral the probability of collateral’s being who are more risky (that is, borrowers than anyone else would lose money. pledged from 53 percent to 37 percent. more likely to default) and borrowers The lender can make up for this loss Boot and Thakor’s model also predicts who are less risky (those less likely to by charging a higher interest rate in that the interest rate on the loan will default), and they require more col- the initial periods. In other words, decline as the relationship progresses; lateral from the riskier borrowers. at the beginning of the relationship however, results regarding this predic- Even though seizing collateral when a with a borrower, before the business tion are mixed.7 borrower defaults reduces the bank’s shows signs of success, the lender must loss, this is not enough to compensate demand an interest rate that is higher than a break-even rate; later on, he 6 The data came from the 1988-89 Survey of requires a lower interest rate. In this Small Business Finance, conducted by the Federal Reserve Board and the Small Business 8 Ideally, the analysis would use data on indi- way, the bank makes a lot of profits at Administration. vidual loans. For example, the researcher would follow every loan to see if it was collateralized, 7 See Philip Strahan’s chapter for a survey of if the borrower paid on time, and what the net results from small-business loans around the chargeoff was. Since such data do not exist world. For the most part, the finding that col- outside bank loan files, Berger and Udell used 5 Such a promise might be believable because lateral requirements fall with the length of the data on chargeoffs and loans past due at the there is an explicit contract or maybe because relationship is replicated in a number of studies. bank level. They found that a bank with a larger the bank, which deals with many firms, cares The effect of relationships on loan rates varies share of collateralized loans has a larger number about its reputation for keeping its promises. widely across studies. of chargeoffs and loans past due.

www.philadelphiafed.org Business Review Q2 2006 11 the bank for the fact that the loan was collateral may decrease over the life of this happens, economists say that the riskier to begin with.9 the loan. Second, the “automatic stay” project has a positive net present value Berger and Udell provide evi- clause in the U.S. bankruptcy code of- (NPV).16 In Inderst and Müller’s mod- dence consistent with the explanation ten creates a significant delay between el, banks tend to be too conservative. above.10 Loosely speaking, they show the time the borrower defaults on the They refuse loans to projects that have in their 1990 paper that a collateral- loan and the time the lender can seize a positive but relatively low NPV. ized loan typically has a higher inter- the collateral. Even though the value In the model, a firm applies for a est rate. To correct for the fact that of the collateral is usually preserved, loan from a local bank. The local bank higher interest rates can reflect differ- the fact that the payment is delayed faces competition from other lenders, ent points in the business cycle, they imposes a cost on the lender.13 Accord- but it has an information advantage. subtract the interest rate on a Treasury ing to Andrea Eisfeldt and Adriano For firms located nearby, it can dis- security with the same duration to Rampini, the difficulty in repossessing tinguish between projects that have calculate the markup on the bank loan collateral explains why some firms may positive NPVs and projects that have and show that the collateralized loan prefer to lease their assets, rather than negative NPVs.17 To other lenders, all typically has a larger markup.11 Since to borrow money to purchase assets.14 projects look essentially the same, so Treasury securities are believed to be they must charge a higher interest rate default free, the markup is a measure COLLATERAL AND LENDERS’ than the local lender to compensate of how risky the loan is. If we assume INCENTIVES for losses from the possibility of financ- that a bank charges a higher markup Boot and Thakor’s model focused ing the negative NPV projects.18 when it perceives that a loan is riskier, on how collateral affects the borrower’s How can the local bank use its Berger and Udell’s result suggests that incentives to exert effort in ensuring information advantage? It can charge a bank requires more collateral when it that the loan is paid.15 Roman Inderst a high interest rate, but there is a limit. perceives a loan is riskier.12 and Holger Müller shift focus by deal- If the bank charges an interest rate Note that, in theory, the bank ing with the lender’s incentives. The that is too high, the firm would simply could eliminate the risk of default by problem in their model is that lenders go to the other lenders. This places a requiring more collateral. In practice, may choose not to finance some proj- however, the bank faces risk even if ects even though it is socially desirable the whole value of the loan is secured to undertake them. Inderst and Müller 16 One of the difficulties in saying whether a by collateral. First, the value of the show that using collateral can improve project creates value is that cash flows are re- ceived at different times; for example, a dollar the lender’s incentives to finance these you receive this year is worth more than a dollar projects. you receive in five years because you can invest it and start earning interest earlier. In addition, 9 Note that the fact that chargeoffs are higher Socially, it is desirable to under- cash flows can be uncertain; for example, they for riskier loans does not mean that a bank that take a project when consumers are can be high or low. The net present value takes makes these loans loses money. Not all borrow- willing to pay more than what the into account the timing and riskiness of all cash ers default. The bank can charge a higher inter- flows; it indicates the value of the project (to- est rate when it perceives a loan to be riskier. resources cost, that is, when the proj- day) net of the initial investment and net of all While the bank loses money on riskier borrow- ect creates value that can be shared future investments. ers who default on their loans, it makes money on those who pay in full. between owners and lenders. When 17 The local bank may have an information advantage because it is easier to monitor and 10 The data came from the Federal Reserve’s collect information about a firm located nearby. Survey of Terms of Bank Lending, which con- More generally, the “local” bank might refer to tains information on individual characteristics a bank with which the borrower has had prior of domestic loans. 13 For more details, read Chapter 10 in Gregory dealings. Udell’s book. 11 When payments are made before final ma- 18 The local bank has access to “hard” informa- turity, the duration of a security is less than its 14 Eisfeldt and Rampini focus on the following tion (for example, the firm’s books) as well as maturity. The duration of a security is shorter tradeoff: Leasing allows the firm to borrow more “soft” information (for example, information when a larger share of the total payments are because it is easier for the lender to repossess about the borrower’s managerial quality). The made earlier. the asset. However, leasing is costly because the other lenders have access only to hard informa- borrower (the lessee) has fewer incentives to tion; thus, they may not have a complete picture 12 A high interest rate on a loan can also reflect take appropriate care of the asset. of the firm. Rebel Cole, Lawrence Goldberg, a premium for additional collateral-related and Lawrence White provide evidence that monitoring costs or for the cost of evaluating 15 Examples of other papers that focus on collat- in approving small-business loans, large banks the loan as discussed in the next section. Yet, eral and borrower’s incentives are those by Yuk- tend to employ hard information, whereas small it is reasonable to believe that a higher interest Shee Chan and Anjan Thakor and by Arnoud banks are more likely to rely on soft informa- rate reflects more risk. Boot, Anjan Thakor, and Gregory Udell. tion.

12 Q2 2006 Business Review www.philadelphiafed.org ceiling on the local bank’s return from revenue is $120 million (½ x 200 + ½ fore, the borrower is more likely to making the loan, and the lender may x 40). If the probability is higher, the default. The model also predicts that choose not to finance the project even expected revenue is higher. For exam- borrowers who are more risky to begin though it has a positive NPV. ple, if the probability is 80 percent, the with will post more collateral and pay To see why, consider the follow- expected revenue is $168 million (0.8 x a higher loan rate (that is, a higher ing example: Suppose that because of 200 + 0.2 x 40). We saw earlier that in markup over the interest on Treasury competition from other banks the local the first case (revenue of $120 million), bills) than borrowers who are less risky. lender must leave the borrower with at the lender will reject the loan. In the Here the intuition is simple: When the least $15 million of revenues. Now sup- second case, the lender will approve bank faces a risky borrower, it takes pose the local lender estimates that the the loan because he will be left with more measures to protect itself. project will cost $110 million and the expected revenue of $153 million ($168 ...But Too Much Collateral May expected revenues will be $120 mil- million minus $15 million), which is Have a Negative Effect. In Inderst lion. Since the revenues are more than more than the initial cost. More gener- and Müller’s model collateral is good the cost, the project has a positive ally, the bank will approve the loan for society because it allows more proj- NPV of $10 million.19 Now suppose only if it thinks that the probability of ects that have a positive NPV to be that because the borrower has no cash, success exceeds some cutoff level. financed. Although the bank is less se- the local lender must provide all of the Now suppose that the borrower lective in approving projects (so there investment outlay. Since the borrower posts collateral. The bank seizes the is more default), the bank finances obtains $15 million, the lender is left collateral only if the project fails. only projects that have a positive NPV. with an expected revenue of $105 mil- Thus, if the project is very likely to In some cases, however, collateral- lion, an amount that is less than the succeed, collateral has a very small ized lending can actually be bad for initial investment. The local lender effect on the bank’s payoff. However, society. Indeed, if the borrower posts will reject the loan because if he does if the project has a lower probability a lot of collateral, the lender might be not, he loses $5 million.20 of success, the bank’s expected profits tempted to finance a project even if Collateral Can Improve Lenders’ increase significantly when the bor- he knows the project has a negative Incentives... To see how collateral can rower posts collateral. In other words, NPV. The lender may gain from such a improve the bank’s lending policy, it is collateral increases the bank’s payoff loan because he obtains the collateral helpful to think first about the bank’s mainly from projects whose probability whenever the loan goes bad. However, lending policy when collateral is not of success is relatively low. Thus, when society as a whole (in particular, the used. To do so we make the example a borrowers post collateral, the cut-off borrower) loses because of the dead- little more realistic by recognizing the (success) probability for approving a weight cost associated with collateral fact that the project can either succeed loan becomes lower.21 and because resources are spent on or fail. If the project succeeds, it yields Consistent with the empirical projects with a negative NPV.22 In their $200 million; if it fails, it yields only findings in the previous section, the working paper, Philip Bond, David $40 million. model associates collateral with more Musto, and Bilge Yilmaz use the term To determine whether the project risk. Intuitively, when a borrower posts predatory lending to refer to a situation is profitable, the lender needs to esti- collateral, the bank becomes less con- in which a lender knowingly makes a mate the probability that the project servative in approving his loan; there- loan that is harmful to the borrower.23 will succeed. For example, if the prob- But if the borrower is worse off, ability of success is half, the expected why would he agree to such a loan? 21 When the borrower posts collateral, the bank will require a lower interest rate; otherwise, the borrower will go to other lenders. Thus, 19 To make the example simple, I ignore the fact under the loan contract with collateral, the 22 This may suggest that, in some cases, society that revenues are not received at the same time bank obtains more if the project fails but less if as a whole can benefit by limiting the maximum as the investment is made. I also ignore the fact the project succeeds. In other words, collateral amount of collateral that can be posted in loan that revenues are risky. shifts the bank’s payoff from the good states contracts or by including bankruptcy exemp- (where the project succeeds) to the bad states tions and provisions that limit banks’ ability to 20After the local lender rejects a loan, other (where the project fails). Requiring a higher repossess collateral. lenders, who know that the loan was rejected interest rate would not improve the bank’s lend- by the local lender, will reject the loan too. The ing policy because a higher interest rate, which 23 The Bond, Musto, and Yilmaz model focuses reason is that other lenders know there is a is paid only if the project succeeds, improves the on one aspect of predatory lending. In practice, chance that the loan was rejected because the bank’s payoff mainly from projects that would there may be other important aspects not ex- project was found to be unprofitable. have been approved anyway. plored in this model.

www.philadelphiafed.org Business Review Q2 2006 13 One possible explanation is that the whether the project is likely to be prof- COLLATERAL AND FIRMS’ borrower misunderstood the loan con- itable. INVESTMENT DECISIONS tract. Bond, Musto, and Yilmaz offer When the cost of evaluating the Until now, we have not been spe- another explanation. They show that project is lower than the cost of invest- cific about the type of collateral used. predatory lending may occur even if ing in a project with a negative NPV, Actually, there are two types: outside every borrower fully understands the society benefits if the bank evaluates collateral and inside collateral. Out- loan contract. each loan before approving it. Howev- side collateral refers to the case where For this to happen the lender er, since no one can verify how much the borrowing firm pledges assets not must be better informed than the effort the bank expended, the bank owned by the firm. For example, the borrower; only the lender knows that may be “lazy,” in Manove, Padilla, and firm’s owner might post his house as the borrower will be made worse off. Pagano’s terminology. In particular, if collateral for a business loan. Inside The bank (the lender) can assess the the bank is protected by collateral, its collateral refers to the case where the likelihood that the borrower will be incentive to exert effort in evaluating borrowing firm pledges assets it owns, able to repay the loan better than the loans is reduced because it can recoup such as machines and inventories. borrower, a plausible assumption since the value of the loan by seizing the col- Although some of the ideas discussed the bank has made many similar loans lateral. If, on the other hand, the bank earlier may apply to inside collateral, in the past and has followed many is not protected by collateral, the bank the models previously discussed are borrowers. The borrower in turn may evaluates the loan more carefully be- most convincing as explanations of overestimate his ability to repay the cause the bank does not obtain much outside collateral. loan because of lack of experience or if a firm’s project fails.25 The discussion in the next section maybe because of overconfidence. As in the model of Inderst and refers to inside collateral. When a bor- Of course, a borrower would never Müller, the use of collateral makes the rower posts collateral for a loan, such a apply for a loan if he knew that the bank more lenient in approving loans; loan is called secured debt. Implicitly, bank always exploited him. In Bond thus, collateral is associated with more a firm’s debt is secured by its assets and coauthors’ model, some bor- default. In Inderst and Müller’s model, because if the firm goes bankrupt, the rowers overestimate their likelihood being more lenient is good because the proceeds are used to pay the firm’s of repayment, and some borrowers bank approves more loans that have lenders.26 Therefore, most explanations underestimate. Only the bank knows positive NPVs. In contrast, in Manove, of debt secured by inside collateral whether a potential borrower is overly Padilla, and Pagano’s model, being depend on the firm’s having more than optimistic; nonetheless, the bank offers more lenient is bad because the bank one lender. Secured debt gives some the same contract to everyone. Thus, approves some negative NPV projects lenders priority over others for some the borrower cannot deduce the bank’s that would not be approved had the specific set of assets. information and predatory lending can bank conducted a careful evaluation. Collateral Can Overcome Un- occur.24 Moreover, their model does not predict derinvestment. In their article, René Collateral May Also Reduce In- that those who post collateral are bor- Stulz and Herb Johnson suggest that centives to Evaluate Loans. Michael rowers of low quality. In their model, issuing secured debt may allow a firm Manove, Jorge Padilla, and Marco firms have information about their to take advantage of investment op- Pagano explore another situation in own costs, and firms with low costs portunities with a positive NPV that it which the use of collateral may lead use collateral to communicate their otherwise could not. Taking advantage to a bad outcome. As in the previous information to the bank. (To learn of such investment opportunities is paper, the bank is better informed more, see Collateral Can Help the Bank desirable because it increases the firm’s than the borrower, but now the bank Distinguish Between Borrowers.) value; it increases the pie to be shared needs to incur some cost to obtain its among the firm’s shareholders and the information. In particular, by exerting firms’ debt holders (its lenders). some effort (for example, conducting 25 In Manove, Padilla, and Pagano’s model, col- The logic is as follows: Suppose an investigation), the bank can learn lateral reduces the bank’s incentives to evaluate the firm is considering borrowing to a project before a loan is approved. Raghuram Rajan and Andrew Winton explore how col- lateral affects the bank’s incentives to monitor a 24 Economists refer to this scenario, in which firm after the loan is approved. They show that 26 To be precise, some claimants, including law- the bank offers the same contract to all poten- collateral may actually increase banks’ incentive yers and the IRS, must be paid before lenders tial borrowers, as a pooling equilibrium. to monitor. receive anything.

14 Q2 2006 Business Review www.philadelphiafed.org Collateral Can Help the Bank Distinguish Between Borrowers

ichael Manove, Jorge Padilla, and Marco be approved, they know they are the ones subsidizing the Pagano’s model illustrates what econo- high-cost firms. mists call the screening role of collateral. To avoid this, low-cost firms may try to distinguish M In their model, collateral helps the bank themselves from high-cost ones by offering to post col- distinguish between firms that are likely lateral. An economist would say that the low-cost firm is to have positive net present value (NPV) using collateral to signal its information to the bank. Post- projects and firms that are likely to have negative NPV ing collateral is costly to the firm because the firm loses it projects. if its project fails. However, since the firm’s costs are low, Suppose there are two types of firms: firms with high it knows the project is very likely to succeed and the risk operating costs and firms with low operating costs. When of losing collateral is not large. a firm applies for a loan, it knows its operating cost, so it However, low-cost firms can signal their information has an idea of whether its project is likely to be successful using collateral only if high-cost firms find it unprofitable and have a positive NPV. But since there are other factors to mimic low-cost firms by posting collateral, too. This is affecting the project’s success, the firm cannot know for the case if the high- and low-cost firms differ enough. For sure. The bank can find out whether the firm has high a high-cost firm, the cost of putting up collateral is much costs or low costs as well as other information about the higher than for a low-cost firm because the firm knows it firm’s project, but only after some investigation. Before is more likely to default. The result is that low-cost firms the bank investigates, all firms look identical to the bank. post collateral and high-cost firms do not. To recoup the cost of evaluation the bank must The bank can then distinguish between the two charge some fee. To make sure it puts the appropriate firms. If a firm is willing to post collateral, the bank con- amount of effort into evaluating the loan, the bank charg- cludes that the firm has low costs and approves the firm’s es only those firms whose loans are approved. Otherwise, project without an evaluation; in this case, a careful eval- the bank can make money by charging a fee without do- uation is not likely to change the bank’s decision. If a firm ing an evaluation and then rejecting all applicants.a In is not willing to post collateral, the bank concludes the turn, firms whose loans are approved end up subsidizing firm has high costs and evaluates the project; in this case, the firms whose loans are not approved. But since the the bank’s evaluation may indicate that the firm’s project low-cost firms are the ones whose loans are more likely to has a positive NPV, even though the firm has high costs.b

a In the real world a bank that acted this way would develop a bad reputation and lose loan applicants. The reader should interpret the story in the model as a stark version of the real-world problem that if all applicants are charged a fee upfront, the bank will have an incentive to exert too little effort in monitoring.

b Economists refer to this scenario, where one firm distinguishes itself from another firm, as a separating equilibrium. Note that if separation works, the firm can avoid investigation by posting less collateral than in the case where all firms behave the same. Since the bank concludes that a firm that posts collateral has low cost, further investigation is not likely to change the bank’s decision. Helmut Bester first introduced the idea that a borrower who thinks his project is likely to succeed prefers to pledge more collateral than a bor- rower was thinks his project is likely to fail. One of the problems with this type of model is that the “inherently good” borrowers (for example, those with low cost) are the ones who post more collateral. This seems inconsistent with the empirical evidence and with the common wisdom in the banking industry.

finance a new investment project because its past investments may do new debt but not enough to pay both that has a positive NPV and is very poorly. If, instead, the firm undertakes the new and the existing debt? In this low risk. Further, suppose the firm the new project, the firm is less likely case, the firm goes bankrupt, and the already has relatively risky debt in to default on its existing debt because cash flows from the new project are place. In other words, if the firm it can use the cash flow from the new shared between the existing debt hold- does not undertake the new proj- project to pay existing debt holders. ers and the new debt holders; thus, ect, there is a significant likelihood But what if the cash flows from the the new debt holders get paid less than it will default on its existing debt new project are just enough to pay the what was promised to them. If, how-

www.philadelphiafed.org Business Review Q2 2006 15 ever, the firm did not have the risky CONCLUSION lateral may also induce the borrower to debt in place, it could pay its new debt Even though collateral has been exert more effort to ensure the loan is holders in full. Accordingly, any new around for a very long time, research repaid. This is good because borrowers unsecured debt holders would supply into economic factors underlying the with good (positive NPV) investment funds only at a very high interest rate, use of collateral has been particularly opportunities can obtain credit more perhaps so high that the investment active in the past few years. Econo- easily. would be unprofitable for the firm. mists have deepened their understand- However, the use of collateral Now suppose the new debt is se- ing of the reasons some firms post comes at some cost. Transferring con- cured by the new assets purchased with collateral (and others don’t) and of trol may be costly, and the lender may the borrowed funds. Then if the firm’s society’s costs and benefits from collat- not value the collateral as much as the initial project fares poorly and the firm eralized lending. borrower does. In addition, a lender goes bankrupt, the new assets posted Using collateral protects the protected by collateral may exert too as collateral are transferred to the new lender if the borrower defaults. Col- little effort in evaluating projects; he debt holders rather than shared among may even be induced to engage in all creditors, new and old. Since the predatory lending. This is bad from new debt holders obtain more when society’s standpoint because firms ob- the firm goes bankrupt, they are will- 27 While Stulz and Johnson emphasize priority tain loans for projects that are likely to issues, Udell’s book on asset-based finance em- ing to provide funds at better terms phasizes the informational value of monitoring waste resources. A long-term relation- (a lower interest rate). This, in turn, inside collateral (inventory and accounts receiv- ship between a borrower and a lender increases stockholders’ profits from able). A recent working paper by Loretta Mes- can reduce the need for collateral and ter, Leonard Nakamura, and Micheline Renault 27 making the new investment. lends empirical support to Udell’s perspective. save on some of these costs. BR

REFERENCES

Berger, Allen N., and Gregory F. Udell. Boot, Arnoud W. A., Anjan V. Thakor, Manove, Michael A., Jorge Padilla, and “Collateral, Loan Quality, and Bank Risk,” and Gregory F. Udell. “Secured Lending Marco Pagano. “Collateral Versus Project Journal of Monetary Economics, 25 (1990), and Default Risk: Equilibirum Analysis, Screening: A Model of Lazy Banks,” Rand pp. 21-42. Policy Implications and Empirical Results,” Journal of Economics, 32 (2001), pp. 726-44. Economics Journal, 101 (1991), pp. 458-72. Berger, Allen N., and Gregory F. Udell. Mester, Loretta J., Leonard I. Nakamura, “Relationship Lending and Lines of Credit Chan, Yuk-Shee, and Anjan V. Thakor. and Micheline Renault. “Transactions Ac- in Small Firm Finance,” Journal of Business, “Collateral and Competitive Equilibria counts and Loan Monitoring,” Review of 68 (1995), pp. 351-81. with Moral Hazard and Private Informa- Financial Studies (forthcoming). tion,” Journal of Finance, 42 (1987), pp. Bester, Helmut. “Screening vs. Rationing 345-63. Rajan, Raghuram, and Andrew Winton. in Credit Markets with Imperfect Infor- “Covenants and Collateral as Incentives to mation,” American Economic Review, 75 Cole, Rebel A., Lawrence G. Goldberg, Monitor,” Journal of Finance, 50 (1995), pp. (1985), pp. 850-55. and Lawrence J. White. “Cookie Cutter vs. 1113-46. Character: The Micro Structure of Small Bond, Philip, David Musto, and Bilge Business Lending by Large and Small Strahan, Philip E. “Bank Structure and Yilmaz. “Predatory Lending in a Rational Banks,” Journal of Financial and Quantita- Lending: What We Do and Do Not World,” Working Paper 06-2, Federal Re- tive Analysis, 39 (2004), pp. 227-51. Know,” in Arnoud Boot and Anjan Tha- serve Bank of Philadelphia, 2005. kor, eds., Handbook of Financial Intermedia- Eisfeldt, Andrea L., and Adriano A. tion. Amsterdam: North Holland (forth- Boot, Arnoud W. A., and Anjan V. Tha- Rampini. “Leasing, Ability to Repos- coming). kor. “Moral Hazard and Secured Lending sess, and Debt Capacity,” Working Paper, in an Infinitely Repeated Credit Market Northwestern University. Stulz, René M., and Herb Johnson. “An Game,” International Economic Review, 35 Analysis of Secured Debt,” Journal of Fi- (1994), pp. 899-920. Inderst, Roman, and Holger M. Müller. “A nancial Economics, 14 (1985), pp. 501-21. Lender-Based Theory of Collateral,” Work- ing Paper. Udell, Gregory F. Asset-Based Finance: Proven Disciplines for Prudent Lending. The Commercial Finance Association, 2004.

16 Q2 2006 Business Review www.philadelphiafed.org