Unformatted text preview: Queensland University of Technology School of Economics & Finance Financial Institutions Lending Unit Workbook Dr W illiam W ild February 2010 Financial Institutions Lending Unit Workbook 1 Queensland University of Technology School of Economics & Finance This workbook supports a one semester undergraduate course in Financial Institutions Lending. It is divided into four sections, containing nine chapters. Section 1 Theoretical Background and Basic Loan Structure Chapter 1 Lending as Financial Intermediation Chapter 2 Loan Purpose and Utilization Section 2 Costs of Lending Chapter 3 Credit Loss Provision Chapter 4 Cost of Funds Chapter 5 Opportunit y Cost Section 3 Lender’s Compensation Chapter 6 Contractual Compensation Chapter 7 Relationship Value Section 4 Loan Terms and Conditions Chapter 8 Conditions Precedent, Representations and Undertakings Chapter 9 Events of Default, Remedies and Security This workbook provides a framework for students to understand and explore the imperatives that drive lending by financial institutions and the way in which these might be reflected in actual practice and historical events in the loan markets. It is, in a sense, a skeleton to which readings and lectures will add the flesh, and so definitely does not represent the entirety of the course. Financial Institutions Lending Unit Workbook 2 Queensland University of Technology School of Economics & Finance Loans made by financial institutions are one of the fundamentals of the global financial markets, utilized by a wide range of borrowers to meet a very diverse range of financing requirements. In this unit it is not possible to address in detail every specific form of loan market instrument, and so it focuses on the market for international corporate loans as representative of all forms. There are two main reasons for this choice. First, the international corporate loan market has well developed precedents and these have now, at least as far as possible, been standardized and documented by professional lenders’ associations such as the Loan Market Association (LMA) in London and the Asia Pacific Loan Market Association (APLMA), and in the work of financial institutions regulators and their association, the Basel Committee of the Bank for International Settlements (the BIS). Second, the balance of power between borrowers and lenders is more equal in the market for large corporate loans than for other loans (small business or consumer loans, for example), and so the practices and precedents that have developed in this market tend to reflect only the necessary and reasonable requirements of the parties. The major output of the Basel Committee is the Capital Accord, known generally as Basel II, and that is an important reference for this course. Another important reference is the document entitled “APLMA Multicurrency Term and Revolving Facilities Agreement” (the Facility Agreement). This is a primary document which sets out the generic terms and conditions of an international loan facility. It is produced by the Asia Pacific Loan Market Association and represents the consensus of the majorinternational and local lending institutions and legal practitioners in the region. It is also consistent with, and representative of, the documentation of loan facilities in other currencies and in other regions. It will be observed that the Facility Agreement is structured as a syndicated loan; that is, a loan provided jointly by two or more lenders, but this does not prevent the document being read as a generic single lender (bilateral) loan. For “Lenders” simply read “Lender” and, exercising judgement as to context, either ignore references to the “Agent” and “Arranger” or read them also as references to the “Lender”. Financial Institutions Lending Unit Workbook 3 Queensland University of Technology School of Economics & Finance Section 1 Theoretical Background and Basic Loan Structure Financial Institutions Lending Unit Workbook 4 Queensland University of Technology School of Economics & Finance Chapter 1 Lending as Financial Intermediation Financial Institutions Lending Unit Workbook 5 Queensland University of Technology School of Economics & Finance Financial institutions, such as banks, act as financial intermediaries in the sense that they are conduits between investors, who have surplus funds to invest, and firms, who at the same time have a requirement for funds. This raises an important question. W hy do financial intermediaries exist at all, given that it is possible for investors to effectively make loans directly to firms, including by purchasing their debt securities in a capital market? The question is significant in the context of this unit because it might reasonably be expected that loans made by financial institutions are structured, and loan market practice developed and loan markets events occurred, to reflect the reasons that financial institutions and their loans exist at all. There is a large literature that addresses the existence of financial intermediaries, but it is fair to say that it can be difficult for nonacademic to penetrate. It is possible, however, to draw out a few major explanations, and in this chapter I attempt to express those simply and specifically in the context of financial institutions lending. To begin, let’s make some assumptions about the subject. W e will consider a specific form of financial institution, a bank, which receives deposits from and issues other forms of senior debt to investors, and makes loans to firms. It also sells other financial products and services to firms. The questions to be considered can now be specified as follows: 1. W hy might investors choose to invest in deposits and other senior debt issued by a bank, rather than directly lending to firms or purchasing their debt securities? 2. W hy might firms choose to borrow from a bank in the form of loans rather than by borrowing directly from investors or issuing debt securities? I think there are five major arguments, which we will consider briefly. These are: A. Capture B. Transaction costs C. Asset transformation D. Information E. Delegated monitoring Financial Institutions Lending Unit Workbook 6 Queensland University of Technology School of Economics & Finance A. CAPTURE In the unfortunately idealized world of perfect and complete markets, what I describe as “capture” could not exist. In that world, all potential buyers and sellers, all potential investors and investees, have perfect knowledge of each other and their respective transaction preferences, and the final “choice” (although, given the classical assumptions, all choices are selfevident) is which of all the possible parties to transact with. In reality, however, an entity can never contemplate and select between all possible transactions with all possible counterparties. By “contemplation” I mean more than a simple awareness, although awareness is a necessary first step. I mean the full process of thought and communication that is the precursor to two or more entities jointly undertaking a transaction. In this context, then, I would say that an entity “captures” another when that other entity contemplates undertaking a transaction with it and not other entities with which it might have done so. It is plain that banks, as a class of institution, are overwhelmingly successful at capturing both potential depositors and investors, on the one hand, and potential borrowers on the other. Conversely, it is also plain that individuals and nonfinancial institutions are relatively unsuccessful at capturing other individuals and nonfinancial institutions as potential investors or borrowers. Banks have some clear advantages in capturing potential counterparties. 1. Generally, they are the only entities authorized by their governments to offer deposits to the public without a prospectus. Further than this, they are treated by governments as the primary channel for the provision of credit in the economy. 2. Further to 1, they have perceived credibility as counterparties, backed by their special status, which individuals and firms may not, or may have to establish at significant cost. 3. Their simple size and scope as financing specialists enables them to capture potential counterparties that are not regular borrowers or lenders. They have an advantage in advertising reach and multiple locations. Taking deposits and making loans is their ongoing business, so they are generally perceived to be always available as counterparty. 4. Further to 3, modern banks have had such a long head start in capturing potential counterparties such that their lead may now be unassailable. As an industry they may be in the position of winner takesall with respect to the provision of financial products. Financial Institutions Lending Unit Workbook 7 Queensland University of Technology School of Economics & Finance It will be pointed out that markets for debt have developed just to allow individuals and firms to transact directly with each other, and that is true. It even has a name, disintermediation, and not that many years ago disintermediation was being trumpeted as the death of banks. Nevertheless banks today are bigger than ever, and to the extent that traded markets have grown, it has not been at the expense of banks. In addition to bank’s continuing success in capturing potential counterparties, banks may distinguish themselves from capital markets in the ways described in the remainder of this section. Financial Institutions Lending Unit Workbook 8 Queensland University of Technology School of Economics & Finance B. TRANSACTION COSTS The transaction costs explanation of financial intermediation says that intermediation by a bank may lower the direct costs of financial transactions between investors and firms. It therefore makes sense for investors and firms to delegate the transaction process to banks to undertake on their behalf. To begin it is useful to appreciate that there are often substantial costs involved for the parties if they do not utilize banks as intermediaries, and instead transact directly in the capital markets. A security such as a bond is held by a large number of individual investors and so there are costs for the issuer of the bond in identifying, accessing and dealing with all of them. These investors are also, by definition, public, and so are protected by securities laws and regulations which have costs of compliance for issuing firms, particularly in the area of information disclosure and legal costs. Finally, bearer securities need to have mechanisms in place for their trading, and there are obviously costs in this as well. This can be readily contrasted with a bank loan which is a private contract between a firm and a single (or perhaps a small group of) banks. Banks are assumed to be professional and expert investors and the balance of power and knowledge between banks and firms to be relatively even, and so bank lenders do not require and are not granted the benefit of consumer protection in their dealings with firms. And while loans can be traded among banks there is no need for formal mechanisms and exchanges to facilities this process. From the investor side, most people would identify substantial differences in the cost and convenience (assuming that investors place a value on convenience) of investing in public securities and making a bank deposit. It should be noted, however, that the advantage in favour of bank deposits may have been somewhat eroded way in recent years with the massive developments in technology and the increased pace of financial market deregulation. For example, it is not uncommon for people to use their brokerage accounts in the same way as they use their bank accounts, and some brokerage firms even now offer chequeing services. And with the development in electronic trading and custody services, investors now rarely need to hold or trade physical securities. Even leaving aside these basic differences between loans, deposits and securities, the number and volume of transactions undertaken by banks can be assumed to be very large relative to the investments and borrowings of individual investors and firms. This clearly raises the potential for economies of scale in banks’ favour, and economies of scale are one of the first things to look for in any assessment of the relative costs of undertaking any economic activity. Financial Institutions Lending UnitWorkbook 9 Queensland University of Technology School of Economics & Finance The search for ever greater economies of scale was, in fact, a major motivation behind the flood of bank mergers that began in the 1990s and continues today, although the emergence of mega banking institutions has raised some serious questions as to whether there is a certain bank size beyond which the benefits of greater economies of scale are overcome by the costs of merging and managing institutions. Financial Institutions Lending Unit Workbook 10 Queensland University of Technology School of Economics & Finance C. ASSET TRANSFORMATION To understand the asset transformation explanations of bank intermediation it can be useful to contrast banks with another type of financial intermediary, the pure fund manager. A pure fund manager holds market securities issued by firms (primary securities) directly on behalf of investors. The investor’s returns are a direct function of the returns on those primary securities. A bank offers the majority of its investors a fixed claim (a promised return), in the form of short term demand deposit or senior debt, with a smaller group of investors taking a residual, usually equity, claim (offering an uncertain return). It then transforms these investors’ claims (secondary securities) into long term loans to firms (primary securities). I describe this as asset transformation. Simple Bank Balance Sheet Assets funded by Liabilities Capital Short term deposits/ senior debt Shareholder’s equity Long term loans to firms The asset transformation explanations say that there may be some advantage for investors from this asset transformation process. They generally begin with the famous MillerModigliani Irrelevance Proposition. The irrelevance proposition says that, in a perfect market, the combination of debt and equity with which a firm is funded (its capital structure) is irrelevant to the risk and return of its assets and therefore the returns, and the value of the firm, to its investors. Let’s now consider financial intermediaries in this framework, starting with fund managers. It is self evident that in a perfect market there could be no advantage from the existence of fund managers who simply hold securities on behalf of investors. W ith perfect information and no transaction costs, investors can create portfolios of investments equally as efficiently themselves. But what about banks, which do not simply hold securities on behalf of investors but are firms that engage in asset transformation? The irrelevance proposition says that despite this key difference, however, in a perfect market the asset transforming bank provides no advantage either. The bank’s assets are its loans to firms. The risk and return from those loan assets is unchanged irrespective of how the bank funds those assets; that is, it is independent of the combination of deposits, other senior debt, subordinated debt and/or equity that it raises from investors. Financial Institutions Lending Unit Workbook 11 Queensland University of Technology School of Economics & Finance Miller himself has made the point that there is, in the perfect market, no difference in the total risk and return for investors between a bank traditionally funded with a combination of demand deposits and equity and the same bank if it were funded entirely by equity. In the latter case the bank would, effectively, be acting as a fund manager for its equity holders, and we have already noted that it then would have no advantage in a perfect market. Of course real markets are not perfect, and the M&M irrelevance proposition is most commonly used to explain why capital structure might be relevant, and why there might be an optimal capital structure with which to fund assets, because of deviations from the perfect market. Applying this to the question of the existence of banks, it might be argued that the types of investments created by banks (deposits/debt and equity) might not be able to be replicated, or replicated at the same cost, in the capital markets through the interaction of investors and firms alone because of the presence of market imperfections or “frictions”. W e can identify four possible imperfections. First, there might be a market imperfection represented by a subsidy in the form of an implied and unpriced guarantee of deposits by governments (as in Australia up to the GFC, and to which Australia will shortly revert), or a mispriced explicit guarantee in the form of government deposit insurance (as in the US and now many other countries). Closely associated with this is the presence of explicit but unpriced government regulation of banks which has the express intention of reducing their risk of defaulting on obligations to depositors (such regulation is almost unique to financial institutions). This would enable banks to offer investors who deposit with banks greater return for the same risk, or less risk for the same return, compared with investing directly in firm securities without the benefit of government support. Second, there might be a market imperfection represented by the relative illiquidity in the capital market of firm securities compared with highly liquid bank deposits. Illiquidity of firm securities can result from such frictions as incomplete markets and information asymmetries. There are two reasons why bank deposits might be more liquid than any illiquid loans they fund. Banks are highly diversified, with relatively large balance sheets compared with the deposit of any individual investor, and so there is regular inflow of funds in the form of new deposits to match, in the ordinary course of business, the withdrawal of deposits. In addition, the partial funding of banks with “firstloss” equity capital, plus the benefits of the abovementioned government regulation and guarantees, means that banks’ senior debt has far lower risk than the portfolios of loans that they fund. The ability to offer close to riskfree returns means that banks Financial Institutions Lending Unit Workbook 12 Queensland University of Technology School of Economics & Finance have substantial capacity to raise new deposits or issue other types of senior debt securities to meet short term calls on their liquidity through excess withdrawal of deposits or utilization of loan commitments. Of course investors might be able to diversify and purchase firstloss insurance (or issue first loss securities, equivalent to bank capital, on their own), but these activities might be more efficiently and cheaply undertaken by banks because of the market imperfections identified above. Third, there might be a market imperfection represented by illiquidity in the supply of future funding for firms who attempt to issue new debt securities to meet unanticipated future funding requirements. For the same reasons that bank deposits are liquid, banks may be able to offer liquidity to firms in the form of loan commitments that allow the firm to access certain funding in the future. Fourth, and further to three, asset substitution may enable securities to be created which more closely match the individual risk and liquidity preferences of different investors, avoiding the need to compensate investors with higher returns for accepting investments with other, less preferred, risk and liquidity characteristics. Financial Institutions Lending Unit Workbook 13 Queensland University of Technology School of Economics & Finance D. INFORMATION Information as to the risk of an investment is central to any investment activity. It is the prime determinant of the return demanded by investors, and thus firms’ costs of capital. The literature on information in banking is often focused on the presence of asymmetries of information between borrowers and lenders, and the consequent issue of adverse selection in which there is, theoretically, no price at which borrowers and lenders will transact in the market. I think this kind of analysis can be somewhat contrived and confusing, however, and so we will address information and bank lending in much simpler and intuitive terms. Start with two simple assumptions. Assume that investors compete for the opportunity to invest in firms. Further assume that any uncertainty or lack of information about a firm is assumed to represent risk for investors, for which they will require additional or premium return to compensate1 . Now consider the case where competing investors have different levels of information about firms (we describe investors as being either informed or uninformed). Informed investors will be better able to identify firms that represent low risk and so would be able to accept a lower return on any investment in them, compared with uninformed investors who would require an uncertainty premium. It follows that the latter will not be able to transact with (i.e. invest in) these firms as they are outcompeted by the former. The informed investor will also be able to identify firms who pose substantial risks and so would demand higher return commensurate with that risk. The uninformed investor will not be able to do so, and will be unable to differentiate between firms who pose high risk and those that do not. It follows that uninformed investors will end up transacting only with the risky firms, but with insufficient return to compensate for actual risk, as they will outcompete the informed investors. This might explain the existence of financial intermediaries where they have an advantage over investors acting individually in acquiring or generating information about firms. This advantage might be that information is simply not available to public investors (it might be private information) or that the costs of generating information are lower for banks than individual investors. Consider the first possibility. Clearly there is an incentive for firms (or at least firms who do not actually pose substantial risks) to provide information about their activities to investors and so minimize the uncertainty premium and thus their cost of capital. W here this information is market sensitive, however, then 1 W e are implicitly assuming that investors are risk averse, which is a common and reasonable assumption in finance. Financial Institutions Lending Unit Workbook 14 Queensland University of Technology School of Economics & Finance firms may not wish it to be disseminated to the wide set of public investors. In such a case there may be an advantage for a financial intermediary which is able to maintain confidentiality of information and to whom firms are willing to disclose sensitive information. A bank is the classic example of such a financial intermediary. There are many obvious cases where this would provide an advantage. One would be where a firm is undertaking a hostile acquisition of another firm, for which the former requires funding to be in place prior to the launch of the bid. Clearly it would not be feasible for the firm to raise funds for this purpose through the issuance of public securities which would require public disclosure of its intentions. Next consider the cost of information. The specialization of banks in the information gathering process might also result in their having lower costs of acquiring information or acquiring superior information for the same cost. For example, banks may have information technology or other resources that can be used to generate information across a wide range of firms. To the extent information can be acquired at a cost, financial institutions may also enjoy economies of scale. They may generate these economies of scale by investing in a large number of firms in the same industry or region. Much of the academic literature on the cost of information in banking focuses, however, on one particular aspect. This is what is called “relationship banking” in which a bank undertakes multiple transactions with the same firm. Essentially the argument is that the bank can generate information about a particular firm at lower average cost when it undertakes multiple transactions with that firm. The literature on relationship banking doesn’t just limit itself to information acquired in anticipation of lending or investing. It also considers information acquired by banks over the life of a loan or investment. The bank, as a private lender, should have an advantage over public capital markets investors in acquiring private information about the borrower through direct interaction with the firm over the life of the loan or investment. This can provide an advantage to the bank in securing future lending or investment opportunities. It can also, however, provide an advantage in dealing with the changing risks of the borrower over the life of the current loan or investment for which it might, again, require lower uncertainty premium than capital market investors and so lower the borrowing firm’s cost of funds. This brings us directly to the final explanation, delegated monitoring. Financial Institutions Lending Unit Workbook 15 Queensland University of Technology School of Economics & Finance E. DELEGATED MONITORING 2 According to Diamond delegated monitoring describes the delegation of costly monitoring of loan contracts by investors to a bank to undertake on their behalf. Implied in this is that lenders/investors desire to monitor their investments in order to control, or take action against, the firm if necessary to protect the value of their investment (this is discussed in more detail in Chapter 3 Section 3B). In the absence of a delegated monitor, such as a bank, each investor has to either monitor the firm directly, “free rides” on the monitoring undertaken by other investors, or simply fails to monitor. Clearly this will be more costly to investors overall and, so, disadvantageous. From the borrowing firm’s point of view the more intrusive nature of bank monitoring compared with that of capital markets investors is often cited as a disadvantage, but this misses the point. The ability of banks to monitor can reduce the risk and thus the return required, which can be translated into a lower cost of firm funding under bank loans than the capital markets alternative. Furthermore the borrower under a bank loan has far greater flexibility in renegotiating the loan in response to its own changed circumstances compared with the difficulty or impossibility of renegotiating a public debt security held by thousands of anonymous investors. This is described as the contracting cost advantage of bank lending. In addition the bank, which often has a relationship with the borrower that extends beyond the provision of the loan, may have greater incentive to agree to the firm’s requests to renegotiate. 2 Financial Intermediation and Delegated Monitoring. Douglas W. Diamond. Review of Economic Studies (1984) LI, 393414 Financial Institutions Lending Unit Workbook 16 Queensland University of Technology School of Economics & Finance Chapter 2 Loan Purpose and Utilization Financial Institutions Lending Unit Workbook 17 Queensland University of Technology School of Economics & Finance In this chapter loans are compared with their major financing alternative, bond securities issued in the public capital markets. Bonds are wellunderstood, as much because of their relative simplicity as their widespread use. The major theme of this chapter is that loans are more complex than bonds, but it is because of this complexity that the former are able to meet a range of financing requirements that are beyond the scope of the latter. Ultimately, the explanation for the differences between bonds and loans is found in the nature of their providers or investors. Bonds are structured on the assumption that they are fixed debt claims purchased by capital markets investors with surplus funds. In terms of elementary finance theory investment in a bond is a mechanism for deferring consumption. By contrast loans are structured on the assumption that they are provided by banks. Banks are not assumed to lend surplus funds. Instead they are the classic financial intermediaries that raise the funds they lend by selling fixed debt claims, traditionally by accepting deposits from investors or in the interbank market. So what is it about a loan, structured on the assumption that it is provided by a true financial intermediary, which distinguishes it in its capacity to meet a firm's debt financing requirements? Foremost is in the way in which the funds are available to and utilized by the borrower. A bond issue is a series of identical debt claims that are initially sold by the firm to investors, and the purchase price paid by the initial investors generates the required funding to the issuer. The issuing firm will, therefore, receive its funding in one lump sum on the payment date of the bonds. After issuance the bonds may be freely traded so that, when they mature and the debt becomes due, the issuer pays out to the parties that hold the bonds at that time. Prepayment prior to the maturity date is not allowed or, if it is allowed, the issuer must compensate the then bondholders for the opportunity cost of the cash flows they will forego over the remaining original life of the bond. This is necessary because the bondholders will have purchased the bonds, and determined the price for them, in expectation that their cash flows will be received as scheduled. W hen a loan is executed and becomes binding on the parties it does not create an immediate debt claim. Instead a loan gives the borrower the right, subject to conditions, to require the lender to make cash advances to it upon request. So loans are more accurately described as loan facilities that set out the conditions on which the borrower may draw advances and the terms of those advances. This is possible because the lender does not raise its funding for an advance until it is required; it does not therefore have surplus funds sitting idle on which it requires an immediate return. Financial Institutions Lending Unit Workbook 18 Queensland University of Technology School of Economics & Finance In the academic literature a distinction has been made between what are called spot loans and loan commitments. Loan commitments are those loan facilities where the borrower's options to draw advances exist throughout the life of the facility, provided only that the total of advances outstanding at any time do not exceed the maxi mum facility amount. W here it is expected that the borrower will utilize a loan commitment regularly it is usually described as a revolving loan facility. Standby facilities are usually expected to be drawn only in specific, rare circumstances. W hile a loan commitment allows the borrower to request and maintain advances throughout the often lengthy life of the facility, each individual advance is actually repayable within a relatively short period; 1, 3 or 6 months being usual. These periods are called funding periods or interest periods. Loan commitments are structured in this way because of the underlying assumption that lenders are banks that raise most of their funding to meet loan advances in the form of short term deposits or issues of short term senior debt (see Chapters 1 and 3). The typical 1, 3 or 6 month advance periods are the most common periods for banks to raise funding in the deposit or interbank markets, but longer or shorter periods may be agreed between borrower and lender. If it wishes to keep the funds from an advance outstanding beyond the term of the current funding period the borrower simply redraws the advance for a further period and the lender matches this by refinancing its own funding for the same period. The other type of loan facility described in the academic literature is the spot loan. This is typically distinguished from the loan commitment on the basis that it is a kind of bond equivalent, in the sense that all the loan funds are fully drawn upon the execution of the loan and then repaid in full on some agreed maturity date. In reality, however, spot loans are not at all equivalent to bonds. All bank loans, including what have been described as spot loans, provide their borrowers with options in utilization and repayment of loan funds. The spot loans of the academic literature are more accurately called term loans and are distinguished from loan commitments only by certain limitations on, but not the extinguishment of, the options available to the borrower. The borrower under a term loan can only request advances during an initial availability period, although this may be many months or even years long. The facility amounts not drawn at the end of the availability period are automatically cancelled, and once repaid advances cannot be reborrowed. Term loans usually have a defined repayment or amortisation schedule, and provide for advances of often lengthy periods. These long advances are, however, still assumed to be funded by a succession of shortterm fund raisings by the lender, each new one effectively refinancing the previous one. Consequently, term loan advances are also segmented into consecutive Financial Institutions Lending Unit Workbook 19 Queensland University of Technology School of Economics & Finance shorter funding or interest periods the length of each being, again, typically 1, 3, or 6 months long. W hen the term loan is repaid then the lender uses the proceeds to repay the current funding and does not refinance it. This assumption that a loan is match funded for short funding periods allows the borrower the right to prepay the term loan prior to its scheduled maturity or amortisation without cost or penalty, provided only that the prepayment occurs at the end of a funding period. If the borrower wishes to prepay during a funding period, then it will be required to pay the lender break costs, which compensate the latter for the carrying cost of its own funding. This is because banks cannot, typically, repay their own funding prior to its scheduled maturity. If they are required to reinvest funds that are repaid early by the borrower, then it is possible that they may not earn enough return to cover their own cost of their funds. This represents the break cost. This right to prepay is a valuable option to the borrower that is not available in the capital markets. So a major advantage of loans over bonds is that loans grant borrowers options in both the drawing and repayment of advances. A loan can provide a borrower with certainty in meeting an uncertain future funding requirement, whether the uncertainty is in the amount that will be required or its timing. Furthermore, the borrower can repay or prepay as cash flows or more attractive alternative funding becomes available or funding is no longer required, avoiding the carrying cost of maintaining a sinking fund to meet future bond maturities. Financial Institutions Lending Unit Workbook 20 Queensland University of Technology School of Economics & Finance Section 2 Costs of Lending Financial Institutions Lending Unit Workbook 21 Queensland University of Technology School of Economics & Finance This section begins to focus on an individual loan from the perspective of the bank lender. If we change perspective slightly and now consider the loan as a “product” that the bank sells to its customer, then we can say that rationally the loan should be structured and managed by the bank to minimize the it’s “costs of production” relative to the “price” at which it can be sold. In this section, then, we address the three major costs of production of a bank loan, and in the next section we address the price of the loan product, the compensation the lender receives from the borrower. Together these determine the loan’s profitability, and so this approach provides a useful framework for understanding the way a loan might be structure and managed to meet a bank’s presumed objective of maximizing its profit. Financial Institutions Lending Unit Workbook 22 Queensland University of Technology School of Economics & Finance Chapter 3 Credit Loss Provision Financial Institutions Lending Unit Workbook 23 Queensland University of Technology School of Economics & Finance W hat happens if a borrower fails to repay any part of the principal and interest on a loan when it is due (this is called a default)? Remember that the bank is an intermediary and so it is still obliged to repay its own funding and the interest thereon. Any shortfall in the amounts it receives from the borrower must therefore be made up by the bank from its own shareholder’s funds. This reduces the value of the bank (as a firm) to its shareholders. Such a shortfall is described as a credit loss. Banks typically have very large portfolios of loans and always expect that some of these loans will default. They therefore treat credit loss as a cost of undertaking a lending business. More specifically, banks are required by their regulators and auditors to make a provision for the credit loss across their portfolio in each period. As with any provision this is an expense and so it does formally represent a cost to the bank. Of course the actual credit loss on a loan portfolio in any period is random and could range, potentially, from zero if no borrower defaults up to the entire portfolio amount if all borrowers defaulted on the full loan amounts. The provision for credit loss is, however, just a single number. So banks almost universally estimate the expected credit loss on their portfolios and use this as their provision amount. The term expected is used here in its statistical sense, and so expected credit loss is the probability weighted average of all possible credit losses. As it is an expected value, then the expected credit loss across a bank’s entire portfolio of loans is simply the sum of the expected credit loss on each loan individually (you may recall this from portfolio theory). If the cost of having a portfolio of loans is the expected portfolio credit loss, then we can also say that its own individual expected credit loss represents a cost of making a particular loan. Let’s now conceptualize an approach to determining the expected credit loss on an individual loan. Assume that a bank, today, lends a firm $60. The loan matures and is to be repaid in full at the end of 30 months (for the moment, ignore interest on the loan). Immediately after borrowing the money the firm has an asset value of $100. Now the value of any firm’s assets change over time as it undertakes its business and realizes profits and losses. Let’s consider a possible future for this firm where it performs poorly and makes regular losses. This might be the case as shown in the Figure 3.1 below. Financial Institutions Lending Unit Workbook 24 Queensland University of Technology School of Economics & Finance 120 100 80 V alue, $ 60 40 20 F irm A s s et s F irm Debt 0 1 3 5 7 9 11 13 15 Ti m e 17 19 21 23 25 27 29 Figure 3.1 Observe that after 30 days when the loan is to be repaid the firm only has $40 worth of assets left. If this firm had no other debt then all of these assets would be claimed by the bank to repay its loan. The assets are insufficient, however, and so the bank will suffer a credit loss. The actual credit loss to the bank from this loan for the 30 day period is $20, or 33.33%. Now compare an alternative future for this firm; one where it performs relatively well and does generate some profit. This might be the case shown in the Figure 3.2 below, where at the end of 30 days the firm has assets with value of $115. In this case the actual credit loss to the bank will be zero as there are enough assets to repay the loan. 120 100 80 V al u e, $ 60 40 20 F irm A s s et s F irm Debt 0 1 3 5 7 9 11 13 15 Ti m e 17 19 21 23 25 27 29 Figure 3.2 W e can now suggest a theoretically ideal way to determine the expected credit loss on a loan. Identify every possible future for the firm over the given time period, determine the credit loss in each future, and then allocate a probability to each one occurring. This will give a distribution of credit losses, and the expected credit loss on the loan is then the weighted average across Financial Institutions Lending Unit Workbook 25 Queensland University of Technology School of Economics & Finance the distribution (its expected value). Identifying every possible future for the firm and the probability of each being realized is, however, a practically impossible task and so banks have developed an approach which is generally consistent but more practical to implement. From Figures 3.1 and 3.2 above we can see that it is possible to break down credit loss into two elements. The first is default itself, which describes the inability of the bank to repay its loan in full because the value of the assets is less than the amount of the loan. The second is the loss given default which is the amount by which the value of the assets is less than the loan amount, given that a default has occurred. This is how much the bank will actually lose, and can be expressed as a percentage of the loan amount. In a survey of bank practice in this area the Bank for International Settlements (BIS) observed that banks calculate the expected credit loss on an individual 3 loan as the product of these factors . That is: 1. Probability of default (PD): The probability that the borrower will fail to meet its payment obligations in full and on time; that is, default. 2. Expected loss given default (LGD): The expected proportion of the outstanding loan advances that the borrower ultimately fails to repay if it does default. Note that this is an expected value, because the range of loss could actually range from zero (if the bank recovers the entire loan) down to 100% of the loan amount (if the bank recovers nothing). Together the product of PD and LGD gives the expected credit loss expressed as a percentage of the outstanding loan amount. To convert this to a dollar provision amount it must be multiplied by the loan amount, and so the third factor is: 3. Expected credit exposure: The expected amount of loan advances outstanding at the time of default. It will be observed that this approach is not entirely consistent with the theoretically optimum approach of identifying the full distribution of future states of the firm because the three elements, which are in reality completely dependent on each other, are treated and estimated independently. This is, however, exactly why the approach is more practical to implement. The remainder of this chapter considers each of these elements, methods for their calculation and the major issues associated with them. 3 “Credit Risk Modelling: Current Practices and Applications.” 1999. http://www.bis.org/publ/bcbs49.pdf Financial Institutions Lending Unit Workbook 26 Queensland University of Technology School of Economics & Finance 3A. PROBABILITY OF DEFAULT Probability of default is the numerical representation of what most would consider as the credit risk of any loan. W hat is the risk that the borrower will not repay all the money that it owes when it is meant to? It is obvious that for the purposes of bank lending this risk must be described in numerical form, representing the probability of default required to calculate the credit loss provision. That is not the case for all assessments of credit risk, however. Consider the following examples. · A supplier is considering whether to allow a customer 30 day’s credit terms (i.e. 30 days in which to pay for the goods). It would most likely rely on general description of the credit risk of the buyer, such as high, moderate or low. In the past US regulators limited the type of investments in which insurance companies could invest their funds by specifying a minimu m credit rating on a scale of credit ratings issued by the major agencies such as Standard and Poor’s (S&P) or Moody’s. S&P uses a scale that ranges from AAA for the lowest risk to CCC for the highest. US insurance companies were only able to invest in securities with a BBB rating or higher, called investment grade securities. · The first example is of little value for our purposes but the second, the ratings granted by credit ratings agencies to firms for the purposes of issuing securities in the capital markets, is very useful. Let’s start by looking at the definitions of the various ratings categories by S&P: Financial Institutions Lending Unit Workbook 27 Queensland University of Technology School of Economics & Finance AAA “EXTREMELY STRONG capacity to meet its financial commitments” AA “VERY STRONG capacity to meet its financial commitments” A “STRONG capacity to meet its financial commitments but is somewhat more susceptible to the adverse effects of changes in circumstances and economic conditions than obligors in higherrated categories” BBB “ADEQUATE capacity to meet its financial commitments. However, adverse economic conditions or changing circumstances are more likely to lead to a weakened capacity of the obligor to meet its financial commitments” BB “is LESS VULNERABLE in the near term than other lowerrated obligors. However, it faces major ongoing uncertainties and exposure to adverse business, financial, or economic conditions which could lead to the obligor’s inadequate capacity to meet its financial commitments. Adverse business, financial, or economic conditions will likely impair the obligor’s capacity or willingness to meet its financial commitments” B “is MORE VULNERABLE than the obligors rated ‘BB’, but the obligor currently has the capacity to meet its financial commitments. Adverse business, financial, or economic conditions will likely impair the obligor’s capacity or willingness to meet its financial commitment on the obligation” CCC “is CURRENTLY VULNERABLE, and is dependent upon favorable business, financial, and economic conditions to meet its financial commitments” Financial Institutions Lending Unit Workbook 28 Queensland University of Technology School of Economics & Finance One obvious fact is that these ratings are written in highly subjective (or qualitative) ter ms. They do not refer to numerical probabilities of default. So clearly the process by which the agencies assess firms and generate their ratings is also, at least partly, qualitative. W e also observe that the qualitative assessment is couched in terms of the firm’s ability to withstand adverse business, economic and financial conditions. In terms of our understanding of default we could express this the following way. A firm is riskier where it is relatively closer to the point at which it will default than another firm. For a very risky firm, only a small deterioration in operating conditions, which could occur with relatively high probability, would lead to default. For a less risky firm, a much larger deterioration in operating conditions, which would occur with relatively lower probability, would be required to default. This distance to default, and the circumstances that would be required for the firm to travel it, is obviously the key to understanding credit risk. It can be seen very clearly in Figures 3.1 and 3.2. About 10 years ago a firm called KMV developed an approach to credit risk measurement in which the starting point was identifying the current market value of a firm’s assets, which it derived from its current share price and the amount of its liabilities. The difference between the asset value and its fixed debt obligations they described as the distance to default. This is in itself not particularly special. W hat was special about their approach was that they defined distance to default in terms of the volatility of the firm’s asset value which they implied by looking at the historical volatility in the firm’s share price. More specifically, they described the distance to default in terms of the number of standard deviations in the firm’s historical asset value. This was a practical application of Merton’s structural credit risk model (1974) and theoretically would allow the probability that the firm’s asset value would fall below the value of its fixed obligations in some given time period to be calculated directly using standard statistical techniques. In a different context you might recognize this approach as the basis of the famous BlackScholes Option Pricing Model. There are some serious problems in such an approach, however, including the assumptions of normality in the distribution of firm asset value and that historical volatility in the firm’s asset value is continued into the future, and problems in identifying the value of the firm’s liabilities and fixed debt obligations. These are sympto matic of the problems with all strictly quantitative approaches and I assume wer e part of the reason that KMV actually avoided putting their model to the test and calculating the probability of default directly from the model. Financial Institutions Lending Unit Workbook 29 Queensland University of Technology School of Economics & Finance Instead, and very interestingly, KMV calculated the probability of default (what they called expected default frequency) using a historical database of firms grouped according to their distances to defaults, by finding the historical rate at which the firms in each group defaulted. They marketed this database as proprietary information (which possibly provides the remaining explanation as to why they chose this approach). W hat might this tell us about bank credit risk analysis? First, it seems that most comprehensive attempts to assess the credit risk of firms eventually resort to grouping firms with others that are perceived, for whatever reason, to share the same level of risk or same characteristics that indicate risk. Significantly for our purposes, this is also the approach that is adopted by most banks. In 2000 the Bank for International Settlements released a paper entitled “Range of Practice in Banks’ Internal Ratings Systems.” Even from the title it is clear that banks also rate their loans in the same way (although not necessarily with the same descriptors) as the professional rating agencies who assess firms for public securities issues. In that document the BIS said: “An internal rating refers to a summary indicator of the risk inherent in an individual credit. Ratings typically embody an assessment of the risk of loss due to failure by a given borrower to pay as promised, based on consideration of relevant counterparty and facility characteristics. A rating system includes the conceptual methodology, management processes, and systems that play a role in the assignment of a rating.” Later in this section we will provide an example of the way such a bank rating process – the conceptual methodology might work. But for the moment we need to consider the second lesson to be drawn from the KMV approach; that any description of a firm’s risk (such as a rating or distance to default, for example) needs to be translated somehow into a numerical probability. And again, the general method adopted by KMV is similar to the method adopted by banks. The basic idea is this. Look at all past loans/firms that had similar “risk” at some point in time. Count them. Now count the number of those that defaulted within a certain time period afterward, and express that number as a percentage of the total. This mortality analysis approach gives a numerical percentage that can be used as the probability of default for loans with that risk. The result of such a process is demonstrated in the following table of one year default rates from 1920 to 1996, which is drawn from "Moody's Rating Migration and Credit Quality Correlation". Moody's Investor Service, July, 1997. Financial Institutions Lending Unit Workbook 30 Queensland University of Technology School of Economics & Finance Moody’s Rating Aaa Aa A Baa Ba B CaaC 1 yr Default Rate 0.00% 0.06% 0.12% 0.28% 1.11% 3.49% 12.41% As with all such approaches there are problems in implementation. Having a sample of sufficient size is very important, and is a significant problem for some institutions that do not have long records of default history. One way of dealing with this problem is to translate the bank’s internal ratings into ratings of the public credit rating agencies, such as Moody’s, and then use the large Moody’s sample. However attempts to maximize sample size by using older data also generate problems in data relevance. Does the older data relate to a different economic environment, for example? In econometric terms, is it stationary? And then there are issues with the comparability of firms in the same ratings category. Apart from the difficult assumption that very different types of firms can be directly compared in terms of risk, having only a few rating categories necessarily implies that there will be variation in risk in firms even within the same categories. The alternative, increasing the number of ratings categories, may not be an improvement as it may exceed the ability of the analysis to distinguish between the risks of different firms, and makes the problem of generating samples of sufficient size even worse. You may, at this point, be thinking how far the practical approach has moved from the theoretically perfect approach we described earlier of identifying, and allocating probabilities to the occurrence of, every possible future state of the firm. This is a major problem with credit risk assessment. Now we have seen that KMV used the distance to default expressed in terms of asset volatility to group firms together in risk categories, and that banks categorize firms on the basis of an allocated rating. This suggests that the bank rating is also an indicator of distance to default and the likelihood that this firm will travel this distance. However the fact that banks use a descriptive rating rather than some purely quantitative measure, like distance to default, suggests that they believe qualitative factors are important in identifying these. KMV was subsequently acquired by the credit rating agency, Moody’s. It is very relevant at this point to note that Moody’s went to great pains to explain to their customers, those that rely on the ratings to make their investment decisions, that the KMV model would not now for m the whole basis of their Financial Institutions Lending Unit Workbook 31 Queensland University of Technology School of Economics & Finance own risk assessments. It would simply be another element of it, together with their traditional qualitative analysis. And this is reflected in bank practice and the recommendations of bank regulators. In the paper “Sound Credit Assessment and Valuation for Loans” (Bank for International Settlements Consultative Document, Nov 2005) the BIS says that … “5. A bank’s use of experienced credit judgment and reasonable estimates are an essential part of the measurement and recognition of loan losses.” W hy might qualitative elements continue to be seen as important? I think the best answer is this. Relying exclusively on quantitative analysis loses information. Let’s demonstrate by considering another quantitative approach. Financial Institutions Lending Unit Workbook 32 Queensland University of Technology School of Economics & Finance In 1968 Edward Altman used regression analysis of data on a large sample of publicly traded manufacturing firms to develop his “ZScore” model. The value of Z indicates the likelihood of firm default and is calculated by the following model. Z = 1.2(X1) + 1.4(X2) + 3.3(X3) + 0.6(X4) + 1.0(X5) where X1= working capital/total assets X2 = retained earnings/total assets X3 = EBIT/total assets X4 = market value of equity/book value of longterm debt X5 = sales/total assets As with any regression model, the appearance of precision can be deceptive. W e need to consider the power of the model, and this will be a function not just of the size of the sample that is used to generate it but also the information that it excludes. This model predicts default solely on the basis of the 5 ratios set out above. Presumably there is substantial other information about the firm that will also be relevant to its credit risk. Purely quantitative approaches are useful, however, where the cost of acquiring further information is prohibitive. For example, the credit risk of a credit card or home mortgage portfolio. It is unlikely to be economic for the bank to investigate the specific circumstances of each individual borrower, and so instead the bank relies on simple data provided with the loan applications, which it then uses to “score” each one. The large number of small and homogenous transactions makes such an approach realistic in these cases. W hen dealing with inhomogeneous corporate loans, however, the costs of investigation will be relatively small compared with the loan size and the consequences of misanalysis of individual loans relatively high. The following page provides an example of the type of analytical process often used in qualitative bank credit risk analysis. Financial Institutions Lending Unit Workbook 33 Queensland University of Technology School of Economics & Finance Financial Institutions Lending Unit Workbook 34 Queensland University of Technology School of Economics & Finance The importance of information or its lack thereof leads to another important observation about assessing credit risk. As was noted in Chapter 1 a lack of information about a firm, which we can also call uncertainty, also represents risk. It is self evident that the further into the future we are considering, the less information we will have about the circumstances that will exist. In terms of assessing credit risk for any particular period, the further into the future we go, all else being equal, the more risk there will be. This is well recognized and the usual assumption that the risk of loans and debt securities increases with the time to maturity is known as the term structure of credit risk. There is an exception. Consider a firm which is very close to default today. If it does not default within a short period of time then we might assume that it recovers and so its probability of defaulting in some future period is actually lower. It should be noted, however, that this is not the type of firm to whom banks will typically lend, and so this exception is not really relevant to us. Financial Institutions Lending Unit Workbook 35 Queensland University of Technology School of Economics & Finance 3B. LOSS GIVEN DEFAULT If the borrower defaults on its obligation to repay in full, how much does it actually repay to the bank? This amount is said to be recovered by the bank and is usually expressed as a % of the Credit Exposure. As such it is often called the Recovery Rate. Loss given default is then 1recovery rate. “Shock for HIH creditors” The Age, 21 May 2001 Unsecured creditors of HIH Insurance were likely to recover only half, and probably less, of what they were owed, provisional liquidator Tony McGrath signalled yesterday. In his most pessimistic assessment yet of the collapse of Australia's second largest general insurer, Mr McGrath said the final tally of losses was likely to be much higher than recent estimates of $3 billion. His previous estimate of HIH's losses was $1 billion. The mounting losses meant a grim outlook for thousands of unsecured creditors, including policyholders. "It is likely that the ultimate return to creditors across the board will be within that range of 50 cents in the dollar or less," Mr McGrath said. To help understand loss given default I’m going to start with the example of another type of lending, margin lending. Margin lending allows individuals and firms to borrow funds to invest in shares and commodities, for example. The bank lends them a certain fraction of the value of the shares (say 95%) and the borrower makes up the difference with what is called the margin. The bank holds the shares or commodities as security and has the right to sell them in certain circumstances and keep the proceeds to repay the loan. The bank’s objective is to monitor the market value of the security so that if it falls to the level of the loan then it can sell the security and get the full amount of the loan back. The borrower can avoid this by giving the bank more money to rebuild the margin. If the bank monitors the value of the security continuously and there is a liquid market for so that it can sell it quickly and without loss of value, then the bank should be able to avoid any loss given default. Now think of a corporate loan in the same terms. Recall from Chapter 1 that one of the explanations for the existence of banks is that they can monitor the borrower in ways that investors in the markets cannot (delegated monitor). W e can now say that the bank’s objective should be to monitor the value of the firm’s assets and liabilities in order to identify when the former might be getting close to the latter. At this point the bank will want to exercise its rights Financial Institutions Lending Unit Workbook 36 Queensland University of Technology School of Economics & Finance under the loan (see Chapter 9) to call the loan in early while the firm still has assets to repay it in full. Assessing the expected loss given default can thus be seen as an assessment of why the bank might not be able to do this successfully. Consider the differences between a corporate loan and margin lending. It is possible to observe the current market value of shares and commodities on a continuous basis. It is not possible to do so in relation to a firm’s assets. They are often highly illiquid and difficult to dispose of. Determining the true market value of a firm’s assets is a complex problem, and one that has seriously perplexed the accounting profession. And in a practical sense, firms are complex and change continuously. W hile records of their assets are captured in their financial statements these are certainly not real time documents. Instead they are snapshots produced only periodically, generally each 6 months. Furthermore, the value of the firm’s liabilities can also change dramatically with time, which is not the case for margin lending. Now even if a bank suspects that the value of the firm’s assets has deteriorated or its liabilities have increased and steps in to call the loan back, the lack of information as to the true value of the assets or liabilities means that by the time the bank can force the firm to liquidate its assets they might be worth less than the liabilities. This was clearly the case for the investors in the box “Shock for HIH creditors” above. One way that banks attempt to overcome these difficulties is by taking security over specific, liquid assets which can be identified and valued. This is the same as the way in which banks take security over a property when making a mortgage loan. If the value of the property is greater than the loan then the bank does not really have to worry about the borrower’s financial position. If there is a default then the bank has access to an asset which is worth more than the loan. Unfortunately for banks, however, large corporations are generally unwilling to grant security over their assets. W e might conclude, then, that expected loss given default is a function of the following factors: · · · · · · The presence of any security and its quality. The bank’s ability to monitor the value of the firm’s assets and liabilities. The level of volatility in the value of the firm’s assets and liabilities. The liquidity of the firm’s assets and liabilities. The bank’s ability to protect the value of the firm’s assets and control the creation of firm liabilities. The bank’s ability to liquidate the firm’s assets to repay the loan early if necessary. Financial Institutions Lending Unit Workbook 37 Queensland University of Technology School of Economics & Finance 3C. EXPECTED CREDIT EXPOSURE Expected credit exposure is very significant in the context of loans because, as was the major subject of Chapter 2, the amount of a loan which is actually borrowed or drawn by the borrower at any particular time can vary substantially over the life of the facility. Clearly the credit loss sustained by a bank if a borrower defaults is a function of the amount it actually has advanced to it at the time of default. “WorldCon” International Financing Review, Issue 1440, Jun 29, 2002 Banks are thanking their lucky stars that WorldCom had drawn down only a fraction of its US$8bn of bank lines before revealing last week that its books had been cooked. … Bondholders, meanwhile, have been devastated. As of June 26, the first business day after WorldCom revealed that it had inflated EBITDA for the past five quarters, the company had drawn down just US$2.6bn of the US$8bn available under three credit lines. Consider a revolving credit facility. The borrower can draw down and repay advances at its option, up to the maxi mum facility amount. The obvious difficulty for the bank in making a credit loss provision is estimating how much of the revolving credit might be drawn down at the time the borrower defaults. The BIS recognizes that this is an area of some difficulty for banks, and that they are often forced to rely on simplifying assumptions. For example, some banks make the severely simplifying assumption that expected credit exposure at default is equal to the total loan facility amount, on the basis that a firm that gets into financial difficulty and has increasing requirements for cash will draw as much as possible under its loan facilities. The problem with this is that it assumes away the possibility that the firm can get into difficulty so quickly that it does not have time to do so (see “W orld Con” box above, for example) or that the bank might not allow the firm to do so. The ways the bank might do so is covered in Chapters 8 and 9 but, to summarize here, the loan contract almost always gives the bank the right to not provide further advances or to demand repayment of existing advances early where it suspects the borrower might be in difficulty. This highlights a vital point. The way the loan is structured by the bank will directly affect the exposure at default credit loss provision and so the cost of making and holding the loan. Of course the way the borrower chooses to utilize the loan will also directly affect the exposure at default and this will, to the extent it is not controlled by the bank through the loan agreement, be outside of its control and thus uncertain. Financial Institutions Lending Unit Workbook 38 Queensland University of Technology School of Economics & Finance Chapter 4 Cost of Funds Financial Institutions Lending Unit Workbook 39 Queensland University of Technology School of Economics & Finance A bank is a firm like any other and so the cost of the funds with which it makes its loans, which are its assets, is simply its cost of capital. Capital cost is probably more significant to banks than most firms, however, because it provides, as described in Chapter 1, a major rationale for their existence at all. Essentially, a bank exists (at least according to the asset transformation explanation) because the return required by its investors, its cost of capital, is less than the return those investors would have required if they had instead invested directly in the firms to whom the bank lends. This is attributed to banks’ abilities to create liquid and low risk forms of capital, namely deposits and other forms of senior debt, which cannot be replicated for the same return by investors in the capital markets. The bank has advantages in diversifying due to its large size, low transaction costs, the benefit of direct government regulation and express or implied government guarantees. You will also recall that the creation of liquid and low risk deposits and other senior debt also requires that the bank funds part of the assets with firstloss equity capital, and so the bank’s cost of capital will reflect the weighted average of the low cost deposits and other debt, and the higher cost of equity. The advantage for banks is that the relatively low cost of deposits and other senior debt means that some combination of the former and equity has a lower cost of capital than funding with equity alone. You might recognize that this is the same type of argument as the one which says that there is more favourable tax treatment of debt capital than equity capital, and so all firms should rationally maximize the amount of debt in their capital structures. This is the first “friction” in M&Ms perfect world. It is not observed in practice, however, that nonbanking firms do maxi mize debt, and a number of explanations for this have been offered (bankruptcy costs, reform of tax regimes, etc). By contrast it is observed in practice that banks do have very high levels of senior debt, and particularly deposits, in their capital structures. In the context of this chapter we can say that a very high proportion of every one of a bank’s loan assets is funded by deposits and other senior debt, and so it is the cost of these that comprises the major part of the bank’s cost of providing each loan. The remainder of its cost of funds is found in the cost of the equity which provides the balance of its capital, and a small group of instruments that have characteristics of both debt and equity. So in this chapter, then, we focus on the factors that influence the cost of a bank’s deposits and other senior debt, how to identify this cost for a loan manager, and explore why this might vary between banks that are competing for loans in the market. W e then consider the cost of equity capital to a bank and how a bank decides how much equity (and equity equivalent) capital it should hold. Financial Institutions Lending Unit Workbook 40 Queensland University of Technology School of Economics & Finance At this point it is very important clear up a serious confusion in the terminology. You will be familiar with the use of the term capital to describe all forms of liabilities and equity with which a firm funds its assets. However in the banking literature and industry, capital is used to describe only equity and equity equivalents and not its liabilities. In the remainder of this chapter we will use this banking specific definition. Financial Institutions Lending Unit Workbook 41 Queensland University of Technology School of Economics & Finance 4A. DEPOSITS AND SENIOR DEBT It is commonly accepted that the relatively cheapest form of funding for banks is in the retail deposit market, where the distortion of market frictions on the risk/return tradeoff of investors, and where the bank’s comparative cost advantage, is greatest. In a competitive banking market, however, the relative attractiveness of deposits might be expected to generate significant competition between banks to secure them. To the extent that there is limited supply of deposits, then some of their relatively low cost might be competed away. To this we might add that there may also be a significant cost in the infrastructure required to raise deposits, in the form of physical branches that may not be required for other forms of funding. Despite competition between banks, deposits do still have a relatively low cost. On this basis it might seem rational for banks to raise debt only in this form. Yet this is never the case. Banks always try to raise some part of their debt as longer fixed term deposits and even long term debt in the capital markets where they compete with other debt securities. One of the reasons is that the greater the reliance on cheap demand deposits and other shortterm debt, the greater the risk of a liquidity crisis for the bank. A liquidity crisis is where a large number of depositors and short term debt holders simultaneously demand their funds back, which demand cannot be met by the bank as it has invested those funds in long term and relatively illiquid loans. There is another reason that banks might prefer to have some longer maturity debt, which is the risk of repricing. W hen debt matures and is repaid then, assuming the bank wants to maintain its assets (loans), the debt has to be replaced. The new debt that refinances it will have a cost that reflects the new market interest rate. Repricing risk is the risk that market interest rates have increased. It follows that banks might prefer to raise some of their debt as longer term debt which does not have to be refinanced, and so repriced, until further into the future. This is even though the cost of that longer term debt is usually higher than the short term debt at the time it is issued (where there is a positive yield curve). So this means that there is a tradeoff between the current cost of debt capital and liquidity and repricing risk, which means that most banks have some combination of debt capital of different maturities and cost. The mix, however, will be unique for each bank. Now let’s bring this down to the level of an individual loan. Banks are continuing entities, and each new loan and its funding represent additions to a preexisting balance sheet. So in this sense the cost of funds of any new loan is actually the bank’s marginal cost of capital; that is the cost of the new capital required to fund the new loan, not the bank’s whole cost of capital Financial Institutions Lending Unit Workbook 42 Queensland University of Technology School of Economics & Finance which includes capital raised previously in possibly different interest rate environments and which, could be argued, funds the bank’s existing loans. Now think that a typical bank enters into many thousands of new loans of different maturities every day, and raises funds from many hundreds of thousands of individual deposits of different maturities and amounts. Think also that a bank’s funding of its assets is the responsibility of a central treasury department, not the individual loan manager. Lending and funding are separate functions within the bank. Treasury typically manages all the bank’s funding requirements on a consolidated basis, and does not raise funding for each loan individually. In fact treasury usually does not match the maturity or basis of the bank’s funding directly to the cash flows from its loans and other assets, thereby creating liquidity and repricing risk independent from the risk of the loans themselves. How, then, can the cost of funds for any individual loan be identified if we are the loan manager attempting to structure and manage this loan for our customer? Fortunately this can be done by making a simple assumption and the use of simple market benchmarks. The assumption is one that we have already met in Chapter 2; it is that each loan is assumed to be funded by the bank for a series of consecutive short interest periods. The second element, the benchmark, is an index of the interest rate at which large and high quality banks in the market can raise funds for periods of similar length at the same time. Determining the cost of funds for a particular loan is then a case of determining the bank’s ability to raise funds in that market at the benchmark rates, should it so desire to do so. If it cannot do so, then the bank’s cost of funds for its loans is said to be the market benchmark rate plus a certain funding premium. A good example of one of these benchmarks is the Australian dollar Bank Bill Rate. Generated by the Australian Financial Markets Association, it is an index of the rates on a particular day for new deposits in A$ of different maturities. A typical day’s rates are shown below. Financial Institutions Lending Unit Workbook 43 Queensland University of Technology School of Economics & Finance Figure 4.1 The most well known benchmark is the London Interbank Offered Rate, or LIBOR, which is compiled by the British Banker’s Association for London deposits in US$. There are other benchmark rates for other currencies. So we can express a particular bank’s cost of funds on its loans as being some premium (or discount to) a relevant benchmark rate. For example, BBR + 0.10% would mean that the bank would have to pay BBR plus a 10 basis point premium to raise A$ deposits in the interbank market. Clearly this bank would be at a competitive disadvantage, having higher costs of making loans, compared with banks that could raise funds at BBR flat. The question, then, is why would any bank not be able to raise funds at the benchmark rate? There are two scenarios. 1. Normal market for funding, in which bank specific factors will be determinative. 2. Market disruption, where the benchmark may not reflect any bank’s actual costs of funds. The last 15 years have provided the most perfect examples of each scenario that anyone could hope to see; the Japanese premium and the GFC. These will be considered extensively in class. Financial Institutions Lending Unit Workbook 44 Queensland University of Technology School of Economics & Finance 4B. COST OF CAPITAL If deposits and other forms of senior debt are relatively low cost for ms of funding, and it is observed that, because of this, banks do have very high levels of liabilities in their balance sheets, then the question might be asked as to why banks have anything other than negligible levels of equity in their balance sheets at all. A simple answer is that they are forced to by their regulators. The more accurate answer is that, for most healthy banks, their management desires to. Let’s start with the first. You will recall from Chapter 3 that banks suffer credit losses; the losses that occur when the bank’s borrowers fail to repay loans. In that chapter we described the provision that the bank makes for expected credit losses, which is a cost of lending and should therefore be covered by the price at which the bank provides its loans to its customers. In other words, and as you will see in Chapter 6, it is the borrowers who actually pay for the credit loss provision in their interest rate on the loan. Now it is assumed that over the long term a bank's accumulated actual credit loss should converge to its accumulated provision for expected credit loss (presuming, of course, that the expected credit loss is an unbiased estimate). This means that over the long term the bank’s revenue from its loans should be sufficient to cover all its costs of lending and the bank will be profitable. In any short period, however, the provision and the loan revenue that provides it may be inadequate to cover actual portfolio credit loss. This is because the provision is only a point estimate, the expected credit loss, and this expected value may be substantially less than the actual credit loss the bank actually suffers. Any actual portfolio credit loss in excess of provision still needs to be funded because the depositors and other senior debt holders who provide the funding for the bank’s loans need to be repaid. This excess or unexpected credit loss can only be met from the banks equity capital, its shareholder’s funds. These do not have to be repaid if the bank has insufficient funds to meet them, and so they can absorb unexpected credit losses. The difficulty is that, as we have seen, banks desire to minimize the amount of equity in their balance sheets because it is relatively expensive. And the less equity the bank has then the smaller the amount of unexpected loss it can withstand before all the equity is wiped out and the bank becomes insolvent; where the bank has insufficient assets to pay its depositors and debt holders in full. In response to this, bank regulators in most developed countries require banks to hold sufficient capital to absorb unexpected credit losses up to a Financial Institutions Lending Unit Workbook 45 Queensland University of Technology School of Economics & Finance certain defined amount. 4 It will be obvious that this regulatory capital must be freely available to absorb credit losses; that is, it must not represent a fixed claim on the assets of the bank. Equity, which is only a residual claim on the bank's assets and so which does not need to be repaid if the bank has insufficient assets, is usually treated as the definitive form of capital. So in the simplest terms, these capital regulations require that a certain proportion of the bank’s loan assets are funded not with deposits or other senior debt but with equity or equity equivalent forms of funding that do not have to be repaid if the bank has insufficient funds to do so because of credit losses. Most national prudential regulations are based on the international Basel Capital Accord (Basel II) and specify the amount of capital to be held against each individual loan as a proportion of the loan advances outstanding. The capital requirements differ according to the type of loan and borrower according to a formula, but the regulators targeted an average across the portfolio of 8%. In other words the bank must fund, on average, 8% of every corporate loan with capital or, in practice, make sure it has equity and equity equivalent on its balance sheet to the value of 8% of its corporate loan assets. Even without the compulsion of regulatory capital requirements, bank management may desire to hold a certain level of capital to maintain the bank’s probability of becoming insolvent at a level they believe to be optimum, for example to achieve a target credit rating. This might not necessarily match the minimum level under the Basel Accord, which does not (and cannot) target an explicit probability of insolvency and, in any event, simply aims to provide a basic minimum level of capital for all institutions. The relationship between the bank’s capital and loan assets also affects the volatility of its shareholder’s returns, which is also an important issue for bank management. So it is fair to say that the relatively crude regulatory capital requirement will not necessarily reflect bank management’s objectives in relation to the optimum level of capital that they desire to hold. The idea behind minimum capital requirements is demonstrated in Figure 4.2, below. 4 It should be noted, of course, that it is not possible to hold sufficient capital to meet the maximum possibl e portfolio credit loss without funding the bank entirely by capital. This is not the intention of regulators, however, and so there remains, even with regulator y capital requirements, some low risk of bank insolvency from credit losses. Financial Institutions Lending Unit Workbook 46 Queensland University of Technology School of Economics & Finance Figure 4.2 Probability Density Function P rovision for Expected Cred it Loss Eco nomic Ca pital Portfolio Cred it Loss T arget P robab ility of Insolvency The internal modelling approach to determining the optimum level of bank capital is usually undertaken by determining ValueatRisk; that is, identifying the maximum portfolio credit loss that will occur with given probability. This is then covered by the combination of provision and capital. As can be seen in the figure, the form of the probability density function and the target probability of insolvency, the latter being the probability that aggregate portfolio credit losses exceed the bank's economic capital, together determine the amount of economic capital that the bank will desire to hold against its existing portfolio over any given time horizon. The economic capital requirement attributable to any individual loan is then determined as the difference between the total economic capital requirement calculated without and then with the particular loan in the portfolio. Notwithstanding criticisms of the m, however, regulatory capital requirements are compulsory and regulators do not yet allow institutions to calculate credit risk capital for regulatory purposes using their own internal economic models. So even if an institution determines and maintains its own internal economic capital requirement it will still be required to hold the regulatory capital requirement if that is greater in amount. The cost of equity is not directly observable, in the way that the yield on deposits or other forms of debt is. For practical purposes it is usually set by the bank’s management as a target return on equity, which is ultimately driven by shareholder’s expectations and the market price of the bank’s equity. Financial Institutions Lending Unit Workbook 47 Queensland University of Technology School of Economics & Finance Chapter 5 Opportunity Cost Financial Institutions Lending Unit Workbook 48 Queensland University of Technology School of Economics & Finance 5.1 LENDING LIMITS Extract from PRINCIPLES FOR THE MANAGEMENT OF CREDIT RISK Basel Committee on Banking Supervision. September 2000. Full text available at http://www.bis.org/publ/bcbs75.pdf “Concentrations 2. Concentrations are probably the single most important cause of major credit problems. Credit concentrations are viewed as any exposure where the potential losses are large relative to the bank’s capital, its total assets or, where adequate measures exist, the bank’s overall risk level. Relatively large losses may reflect not only large exposures, but also the potential for unusually high percentage losses given default. 3. Credit concentrations can further be grouped roughly into two categories: Conventional credit concentrations would include concentrations of credits to single borrowers or counterparties, a group of connected counterparties, and sectors or industries, such as commercial real estate, and oil and gas. Concentrations based on common or correlated risk factors reflect subtler or more situationspecific factors, and often can only be uncovered through analysis. Disturbances in Asia and Russia in late 1998 illustrate how close linkages among emerging markets under stress conditions and previously undetected correlations between market and credit risks, as well as between those risks and liquidity risk, can produce widespread losses. ….. 5. The recurrent nature of credit concentration problems, especially involving conventional credit concentrations, raises the issue of why banks allow concentrations to develop. First, in developing their business strategy, most banks face an inherent tradeoff between choosing to specialise in a few key areas with the goal of achieving a market leadership position and diversifying their income streams, especially when they are engaged in some volatile market segments. This tradeoff has been exacerbated by intensified competition among banks and nonbanks alike for traditional banking activities, such as providing credit to investment grade corporations. Concentrations appear most frequently to arise because banks identify “hot” and rapidly growing industries and use overly optimistic assumptions about an industry’s future prospects, especially asset appreciation and the potential to earn aboveaverage fees and/or spreads. Banks seem most susceptible to overlooking the dangers in such situations when they are focused on asset growth or market share. 6. Banking supervisors should have specific regulations limiting concentrations to one borrower or set of related borrowers, and, in fact, should also expect banks to set much lower limits on singleobligor exposure. Financial Institutions Lending Unit Workbook 49 Queensland University of Technology School of Economics & Finance Most credit risk managers in banks also monitor industry concentrations. Many banks are exploring techniques to identify concentrations based on common risk factors or correlations among factors. W hile small banks may find it difficult not to be at or near limits on concentrations, very large banking organisations must recognise that, because of their large capital base, their exposures to single obligors can reach imprudent levels while remaining within regulatory limits.” 5.2 LENDING LIMITS AS RATIONING The objective of lending limits is to reduce concentration of credit risk and thus, ultimately, the bank’s volatility of returns and implicitly its probability of insolvency. As such they act to hard ration the loans that a bank can make, potentially forcing the bank to reject certain lending opportunities. In line with standard finance theory, however, a loan that substantially increases concentration of credit risk in a bank’s portfolio should never be rejected per se. The increased risk of any loan concentration should be reflected in a higher rate of return demanded by the bank’s investors who provide the funds with which the loans are funded. That is, the bank’s cost of capital will increase. If the loans generate sufficient return to cover the higher marginal cost of capital, however, then the bank should undertake them. In this sense lending limits as rationing would reduce the value of the bank, as a firm, to its investors. Financial Institutions Lending Unit Workbook 50 Queensland University of Technology School of Economics & Finance From FUNDAMENTALS OF FINANCIAL MANAGEMENT rd 3 Ed. James C. Horne. 1977. PrenticeHall. pp 262263. “Capital rationing occurs anytime there is a budget ceiling, or constraint, on the amount of funds that can be invested during a specific period of time, such as a year. … W ith a capitalrationing constraint, the firm attempts to select the combination of investment proposals that will provide the greatest profitability. The cost to the firm of a budget ceiling might be regarded as the opportunity foregone on the next most profitable investment after the cutoff. … Although all cashflows are discounted at the required rate of return, we do not necessarily accept proposals that provide positive net present values. Acceptance is determined by the budget constraint, which tells us which proposals can be accepted before the budget is exhausted. … we may reject proposals that provide positive net present values … Capital rationing usually results in an investment policy that is less than optimal. In some periods, the firm accepts projects down to its required rate of return; in others, it may reject projects that would provide returns substantially in excess of its required rate. If it rations capital and does not invest in all projects yielding more than the required rate, is it not foregoing opportunities that would enhance the market price of its stock?” Financial Institutions Lending Unit Workbook 51 Queensland University of Technology School of Economics & Finance The same kind of argument is also made against regulatory capital requirements, which you came across in Chapter 4. These are requirements for banks to fund a defined minimum proportion of their loan assets with equity (and equity equivalent forms of capital). Effectively these attempt to mitigate the risk of the bank’s assets by limiting the amount of financial risk the bank also takes (financial risk is a function of firm leverage and is additional to the risk of its assets). There is no reason, however, that an externally imposed capital requirement will coincide with the optimum capital structure that would maxi mize the value of the bank to investors in the absence of regulation. So we saw that many banks actually do determine a parallel internal or economic capital requirement at which, presumably, they believe they maxi mize the value of the bank to its investors. It might be argued, therefore, that lending limits are really a third level of control over the composition of a bank’s portfolio, which determines the volatility of its earnings and ultimately probability of insolvency. The first is the market discipline of the bank’s investors who will require that riskier marginal projects generate greater returns. The second is minimum capital requirements that impose a ceiling on the financial risk the bank can take. Pursuing this line of argument, lending limits might therefore be seen as unnecessary or redundant, with their continued existence implying a lack of faith by bank regulators and management in the operation of market discipline and/or the models by which they determine minimum capital requirements. So why might regulators and bank management still rationally utilize lending limits? This question can be reframed as follows: what are the shortcomings of market discipline and minimum capital requirements that require the further control of lending limits? Effective market discipline requires information. Interestingly, the international bank regulations, Basel II, includes a fundamental “pillar” which aims to increase the amount and quality of information that banks disclose to their investors specifically in order to enhance the operation of market discipline. Nevertheless Basel II still retains at its core minimum capital requirements and strong direct regulatory oversight. Implicitly this is a recognition that market discipline is still insufficient of itself to adequately control bank risks, and one of the major reasons is that investors have insufficient information about banks’ risks. It can be argued, in fact, that investors can never have sufficient information for market discipline to be entirely effective on its own, because the fundamental nature of banks is to be informationally opaque. Recall, from Chapter 1, that an important explanation for the existence of banks is that they can acquire information about the firms to whom they lend that is not available Financial Institutions Lending Unit Workbook 52 Queensland University of Technology School of Economics & Finance to investors in the market. Full availability of this information to investors would remove, arguably, their rationale for existing. On a practical level, the large size and scale of bank’s activities might also impose a limit on the ability of investor’s to understand their risks fully. For regulators there is another potential difficulty in relying on market discipline to control bank risk taking. Investors are only interested in the risk of their investments. Banks, however, play a central role in the economy that other types of firms do not. Banks effectively provide social services for small depositors and borrowers, and in many economies they are they are the major channel of credit and the mechanism by which governments manage monetary policy. The effects of bank failure in terms of these externalities are not accounted for by the bank’s investors when they set the bank’s cost of capital to reflect their perceived risk. This provides another explanation for the need to have further control of bank risk taking. The second level of control, capital requirements, can also be considered in terms of information. Not information known by investors, however, but information known by bank regulators and bank management itself. Perfectly informed regulators and bank management would be able to set capital requirements that perfectly reflected their objectives in terms of bank earnings volatility and probability of insolvency. Clearly, however, there is also an information gap here as well. Both regulatory and internal capital requirements are determined by models, and there will certainly be a lack on information at the level both of the model inputs (estimating risks and their correlations, for example) and in the validity of the model itself. Any regulatory model must be sufficiently simply and tractable to be practically implemented across a range of banking institutions and to be enforceable by regulators. The current regulation set out in Basel I are, but as a consequence have been severely criticized for oversimplification leading to fundamental flaws in the model. But even in the case of internal modelling by sophisticated banks there remains the potential for significant model error which must limit the confidence placed in its output and thus the capital requirement regime. This, then, provides a rationale for the continued use of lending limits. Financial Institutions Lending Unit Workbook 53 Queensland University of Technology School of Economics & Finance 5.3 OPPORTUNITY COST A bank is forced to reject a lending opportunity because it has no room within its lending limit. The loan would have been profitable for the bank by any measure. The profit foregone is the opportunity cost. Let’s consider this further by way of a concrete example. Say a bank has a lending limit of $20 million for a particular borrower, which is determined as 10% of its total capital base of $200 million. That limit is being fully utilized with existing loans to that borrower. Now the borrower comes to the bank with an opportunity to make another loan of $10 million. The bank estimates that the loan would have generated economic value added of $1 million given the return that investors would have demanded on the marginal funds required to make it. However the bank must reject this opportunity and so suffer $1 million opportunity cost. Is there nothing the bank can do? If the lending limit is set as a % of capital base (i.e. equity), then why can’t the bank just raise more equity and increase its capital base sufficiently to fit in the new loan? Let’s look at the consequences of doing so. To make the new $10 million loan the bank would have to raise $100 million in new equity, so that its total exposure to the borrower (now $30 million) would not exceed 10% of total capital (now $300 million). Recall, however, that the bank does not fund its loan entirely with equity. Let’s assume that this bank’s balance sheet is leveraged 9:1 so that it funds its loan assets with 90% debt and10% equity. So to make the new $10 million loan the bank will raise $9 million in new deposits or interbank borrowings and use only 5 $1 million of the new equity . The problem is that the remaining $99 million in new equity now has to be invested in order to generate a return to the investors that provided it. Given the bank’s leverage of 9:1, this means it needs to find another $990 million in profitable loans to at least another 99 separate firms (so as not to exceed the lending limits for any of these either). This is, obviously, a severe imposition just to allow the bank to do the single transaction. Another alternative is to undertake the new loan but to try to sell it or transfer the credit risk away, and so create room within the lending limit, should more profitable opportunities arise in the future. There are some difficulties with this, however. The first is that loans are, unlike bonds, not tradeable instruments by their nature (see Chapter 2). This does not mean that they cannot be traded at all, 5 Why can the bank not fund the loan entirely with the new equity? Recall that market frictions mean that equity is relativel y more expensive than debt as a source of funding, which explains why banks are so highly leveraged. A loan funded entirely by equit y would be uncompetitive in terms of the pricing the bank would have to charge the borrower. Financial Institutions Lending Unit Workbook 54 Queensland University of Technology School of Economics & Finance but that the process is more complex and there is usually less liquidity in the market for loans. The legal process is different because a loan is a contract between a defined lender and borrower with privacy of contract. The ability of a lender to transfer a loan, and the obligations it imposes on the lender, to another party must be specifically contemplated in the loan contract and may be subject to the approval of the borrower. And while there are some markets for loan transfers, they are certainly far, far less liquid and developed than markets for bonds. The second difficulty goes back to the reasons that banks and loans exist. Chapter 1 described the importance of information. The bank that is successful in making a loan to a borrower might be assumed to have been successful in securing that opportunity because it has superior information about that borrower compared to its bank competitors and the capital markets. For many of a bank’s loans, the bank may, in fact, be the only institutions with sufficient information to make the loan at all. The bank might also make a loan as part of a borrower relationship in which it generates income from a range of transactions with the borrower, and in some cases the loan might have been a lossleader to securing other far more profitable business with it. This is related to banks enjoying economies of scope. The consequence is that the loan may not have the same value to another lender as it does to the bank that currently owns it. The bank may, therefore, realize less value from the loan if it sells it than if it retains it. This increases the cost of transferring or selling it away to make room for a new loan and, thus, the opportunity cost to the bank of the lending limit that constrains it. So how do lending limits impact on the costs of lending at the level of the individual loan? The person within the bank responsible for the bank’s relationships with a client or a group of clients, say within an industry or country, may need to choose between different lending opportunities for those clients where it is not possible to fit all possible transactions within a lending limit. If they are so constrained only by the bank’s internal limits then they may seek to have it increased, but this is usually a very difficult and time consuming process. Absent this, then some potential transactions must be rejected. In practice, however, a bank rarely has a set of similar lending opportunities from which it must choose at any given time. It is much more common for the relationship manager to face a choice between a current transaction opportunity and a possible future transaction opportunity. Financial Institutions Lending Unit Workbook 55 Queensland University of Technology School of Economics & Finance A loan will utilize the scarce resource of a lending limit over its entire life, and so preclude the bank entering into other loans during this period. For this reason relationship managers are usually very wary of undertaking loans that utilize their lending limits for a long period, even if there are no certain alternatives currently available. Financial Institutions Lending Unit Workbook 56 Queensland University of Technology School of Economics & Finance Section 3 Lender’s Compensation Financial Institutions Lending Unit Workbook 57 Queensland University of Technology School of Economics & Finance Chapter 6 Contractual Compensation Financial Institutions Lending Unit Workbook 58 Queensland University of Technology School of Economics & Finance This section describes and considers the lender’s compensation under the facility agreement. It is important to recognize, however, that the lender’s compensation is not its profit. Rather, it is the revenue or income which offsets the lender’s costs of providing the facility (which were the subject of the previous section). Profit is determined by the relationship between revenue and costs. Compensation is of two major types and this chapter describes the first, compensation directly under the terms of the facility agreement. This is described by the general term “contractual compensation”, as it is recoverable by the lender under the terms of the legal contract which is the facility agreement. The other major type, noncontractual compensation, which the lender does not have a legal right to recover under the terms of the facility agreement, is the subject of the next chapter. It is important to also distinguish contractual compensation from the borrower’s enforceable legal obligation to repay the loan principal. Flows of loan principal are capital flows and not revenue/cost flows. The former exist as assets, liabilities and equity in the balance sheet and determine a bank’s net worth at a given point in time. The latter exist as elements of the income statement and determine a bank’s profitability over a given period of time. In this chapter four major elements of contractual compensation are identified. These are the base rate component of the interest rate, the margin component of the interest rate, the various facility fees and the additional costs provisions. Financial Institutions Lending Unit Workbook 59 Queensland University of Technology School of Economics & Finance 6A. BASE RATE Interest is the most obvious form of contractual compensation for a lender. In a loan facility, however, interest has a specific structure that gives rise to some i mportant consequences. The Facility Agreement defines the interest rate: “The rate of interest on each Loan for each Interest Period is the percentage rate per annum which is the aggregate of the applicable: (a) Margin; and (b) The Base Rate.” The first observation that can be made is that the rate of interest relates to a “Loan”, which we have already seen is not the facility itself but rather an individual advance drawn by the borrower according to the terms of the facility. So interest is paid by the borrower only on funds which are drawn. Further, where more than one advance is outstanding at a given time then each will have its own interest rate. The second observation is that there is an interest rate for each loan for each Interest Period. Recall that each loan advance, whether under a loan commitment or a term loan, is assumed to be provided by the lender only for a short Interest Period of, say, 1, 3 or 6 months. This is to reflect the assumption that the lender funds the loan in the deposit or interbank market only for such short periods which it either repays at the end of the interest period or refinances, if the borrower wishes to keep the funds outstanding, by raising the funding again for another interest period. It would seem to follow, then, that one of the major purposes of the interest rate on a loan advance is to compensate the lender for its own cost of raising that loan advance; in other words, the interest rate that it pays on its own funding. This would have two direct consequences. The first is that interest on a loan advance would have to be payable at the end of each interest period so as to match the ti ming of the interest payment by the lender on its own funding, which is usually payable at the maturity of that funding. And this is in fact the way the Facility Agreement is structured (see Clause 9.2). The second, and probably more important, consequence would be that the interest rate for any interest period of a loan advance could only be determined at the beginning of the interest period, as the interest rate on the lender’s own funding could only be determined at the time it is raised. And, as market interest rates vary over time, this means that the interest rate on any loan advance would have to be reset at the start of every interest period. Financial Institutions Lending Unit Workbook 60 Queensland University of Technology School of Economics & Finance Now consider the elements of the interest rate. The Margin we can, at this stage, make no comment about. The second element, the Base rate, should be familiar, however, from Chapter 4. where we saw base rates such as BBR, the Australian dollar Bank Bill Rate, which is the average interest rate at which a panel of banks selected by the Australian Banker’s Association would accept Australian dollar deposits from each other. As such it is an index of the current market cost to these banks of short term A$ debt capital. Other examples, including Libor, the US$ reference rate. In fact, where a loan is denominated in US$ then it is usual for the interest rate to be specified as the sum of the Margin and Libor. It would be correct to conclude, therefore, that the interest rate on an advance is the sum of a Margin and the relevant reference rate for the loan currency. In general terms the reference rate, when used this way as an element of a loan interest rate, is called the Base Rate. That the base rate on loan advances varies with time is demonstrated in the following table which shows daily quotes for US$ Libor rates over the course of a week in January 2006. It can be seen, for example, that for a 3 month rd interest period commencing on 3 January, the relevant base rate would have been 4.54438%p.a.. If, however, a 3 month interest period commenced on 11th January, the relevant base rate would have been 4.58%p.a.. Historic US Dollar Libor Rates, from January 2006 3Jan USD s/no/n 1w 2w 1m 2m 3m 4m 5m 6m 7m 8m 9m 10m 11m 12m 4.32500 4.32125 4.32938 4.39750 4.48313 4.54438 4.60063 4.65375 4.71000 4.74025 4.77000 4.80000 4.81725 4.83588 4.85000 4Jan 4.31313 4.31625 4.32625 4.40000 4.48000 4.54063 4.59000 4.63375 4.68000 4.70850 4.73000 4.75000 4.77000 4.78313 4.79625 5Jan 4.31000 4.31563 4.32750 4.41875 4.49000 4.55000 4.59025 4.63038 4.68000 4.70375 4.72438 4.74563 4.76438 4.77438 4.78688 6Jan 4.28875 4.31250 4.32688 4.42000 4.49000 4.55000 4.59100 4.63275 4.68000 4.70725 4.73000 4.75000 4.76788 4.77900 4.79000 9Jan 4.30063 4.31500 4.32688 4.43688 4.50063 4.56000 4.60525 4.65000 4.70000 4.72125 4.74375 4.76375 4.78063 4.79500 4.80813 10Jan 4.29625 4.31750 4.32750 4.44000 4.50625 4.56850 4.61000 4.65113 4.70000 4.72438 4.74788 4.76813 4.78188 4.79813 4.81000 11Jan 4.29375 4.31688 4.33188 4.44163 4.51000 4.58000 4.62463 4.66913 4.72000 4.75000 4.77663 4.80000 4.81838 4.83475 4.85000 Source: http://www.bba.org.uk/content/1/c4/66/72/Jan06.xls Financial Institutions Lending Unit Workbook 61 Queensland University of Technology School of Economics & Finance In our Facility Agreement, the Base Rate is defined for different currencies. The location of the daily BBR, for example, is specified (Reuters screen BBSY page) and it is confirmed that the relevant base rate is the BBR as at the first day of the interest period for a term equivalent to the period. There is also a procedure outlined in the event that no rate is displayed, and which requires that three Reference Banks provide quotes specifically for the purpose of determining the base rate. So the purpose of the base rate does seem to be the compensation of the bank for its cost of funds at current market rates. However where a market index like BBR or Libor is used as a base rate then the base rate alone will almost certainly not reflect the bank’s actual cost of funds, even if the bank were to exactly match fund the advance. There are two reasons. First, these indices reflect the cost of interbank deposits, which is debt funding. However no bank funds a loan entirely with debt capital. Chapter 4 described how bank regulators and bank management set capital requirements which determine, effectively, that a certain proportion of each loan must be funded with equity. Obviously equity is far more expensive than debt capital. So, on the proportion of the loan funded with debt capital, the base rate will certainly be insufficient to compensate the bank for its cost of capital. Second, the bank may not be able to raise debt capital at the base rate, and the funding premium that results was also described in Chapter 4. W hy, then, would a loan be structured this way? The reason would seem to be that, in a competitive loan market, borrowers are able to demand it. As the interest rate is variable it means that, from the borrower’s perspective, it is unknown at the time the borrower enters into the facility. Any reasonable borrower would be reluctant to give the lender total discretion to set the interest rate from time to time, as would have to be the case if the lender was free to determine its own actual cost of funds for each interest period of each advance. The use of the external reference rate as the variable component of interest gives the borrower comfort that the variation in the interest rate on loan advances over time does reflect real changes in market interest rates. This leaves open the question of how the lender recoups its additional costs of equity capital and funding premium. The simple answer is that it must build those into the Margin component of the interest rate, and this is described in the next section. Before moving onto that, however, it is useful to complete this section by further considering the variable, or floating, nature of the interest rate. There is no scope within the standard form Facility Agreement for the interest rate to be fixed over the term of the loan, and this is typical of practice in the market for international corporate loans. Variable or floating rates are universal in these loans. If the borrower desires to have a fixed interest rate, Financial Institutions Lending Unit Workbook 62 Queensland University of Technology School of Economics & Finance then does this mean it is excluded from using a loan? The answer is that it does not. The base rate is set for each interest period, and so effectively the interest rate is fixed for the term of each interest period. A borrower who is concerned about rising interest rates could reduce its interest rate risk by selecting longer interest periods, meaning that the interest rate would be reset less regularly. The limitation for the borrower in this approach is that it cannot prepay any loan advance during its life; that is, it can only prepay at the end of an interest period. Longer interest periods therefore reduce its flexibility in utilizing the loan. If the borrower desires to fix the interest rate for the entire length of the facility then it can do so, but not within the terms of the facility itself. The simplest solution is to enter into another financial contract, either with the lender or even with another, unrelated institution, called an interest rate swap. An interest rate swap is an agreement between the firm and a financial institution (the swap counterparty) to exchange a series of payments according to a fixed schedule over a fixed period. W here the firm wants to use the swap to fix, or hedge, the interest rate on a loan, it agrees to periodically pay to the counterparty an amount equivalent to a fixed interest rate on a notional principal amount. In turn the counterparty agrees to pay to the firm, at the same times, an amount equivalent to the current market base rate on the same notional principal amount. If the notional principal of the swap exactly matches the amount and timing of loan advances and interest periods under the facility then the amount the firm receives under the swap will exactly match the amount it has to pay out as the base rate component of interest on the loan advances. For example, the swap counterparty might agree to pay to the firm Libor each 3 months on a notional principal amount of US$100 million. If the firm has a loan advance of US$100 million with 3 month interest periods which exactly match the timing of the payment under the swap, then the firm will have fully hedged its interest rate risk. In return the firm pays to the counterparty a fixed interest rate on the notional principal amount. Continuing the example, say the firm agreed to swap the floating or variable rate into a fixed rate of 5%p.a. So every time it receives the payment from the counterparty of Libor on the notional principal amount it has to pay 5%p.a. of $100 million in return. This effectively turns the US$100 million loan into a fixed rate loan with an interest rate of 5%p.a. plus the Margin. Note that the firm still has to pay the Margin on the loan as the loan itself is independent of and unaffected by the swap. In practice, the swap parties do not actually swap the funds in this way. At the time of each payment they set off the amount that each owes the other so that Financial Institutions Lending Unit Workbook 63 Queensland University of Technology School of Economics & Finance only one party makes a payment of the net outstanding amount. In our example, say that Libor in a particular period is 7%p.a. Then the net effect is that the counterparty just pays to the firm 2%p.a. on the notional amount of $100 million. A firm could enter into an interest rate swap even if it doesn’t have an underlying loan contract. In that case it would not be hedging the loan but would be speculating on interest rates. The firm will also be speculating on interest rates even if it does have an underlying loan but the amount of the notional swap principal and/or the timing of swap payments do not exactly match those of the loan advances. Sometimes this might be deliberate on the part of the firm. For example, it might enter into a swap in notional principal amount equal to half the amount of outstanding loan advances, but with identical timing of payments, in which case it will have hedged half the interest rate risk and left half unhedged. The firm can use swaps (or other derivatives contracts such as interest rate options) to take any level of interest rate risk it desires. In fact, and as will be discussed in Chapter 7, selling the firm an interest rate derivative to go with the loan is a very common form of noncontractual compensation for the lender. There is another very important consequence of the fact that failure to match the amount and timing of the swap principal with the loan advances exposes the firm to interest rate risk. The firm cannot exercise its options under the facility to vary the timing and amount of loan drawings and repayments without taking interest rate risk. The fixed payment that the firm must make on the interest rate swap is determined by the swap counterparty on the basis of the scheduled notional principal amount of the swap. This notional principal amount does not need to be constant throughout the life of the swap, and the parties are free to negotiate any form of schedule for the notional swap principal that they like. The firm who desires to fully hedge its interest rate risk might have the notional swap principal reducing according to a schedule that matches the scheduled repayments under a term loan, for example. If the firm then does not utilize the loan in exactly the way it had anticipated, for example by varying the timing or amount of drawings or repaying the loan early, there will be a mismatch with the interest rate swap. Firms do not have the option, however, to vary the national swap principal to match the actual loan advances, or at least they cannot do so without cost. If the firm prepays the loan early then the only way it can reduce its swap obligations is to enter into another swap which will offset the original swap, at least to the extent of the amount it wishes to reduce the loan principal. There is no free option under a swap to cancel or vary the notional principal amount from that scheduled. Financial Institutions Lending Unit Workbook 64 Queensland University of Technology School of Economics & Finance Let’s continue with the previous example of a swap with notional principal that exactly matches loan advances of US$100 million. This swap has a fixed component, the amount the firm pays, of 5%p.a. Now also assume a scheduled term for the loan and the swap of 2 years. Say that after one year the firm decides to prepay the loan entirely. The swap, however, has one year left to run. The firm decides to remove its remaining swap obligations by entering into another interest rate swap in a notional principal amount of US$100 million with a term of 1 year (matching the remaining life of its current swap) but this time it pays out Libor and receives a fixed rate. It can do this second swap with the same counterparty as the first swap or another counterparty entirely. Effectively the firm will now periodically receive Libor on US$100 million under the first swap and periodically pay out Libor on $100 million under the second swap. These cancel each other out. It will also continue to periodically pay the fixed component of 5% on US$100 million on the first swap and will periodically receive a fixed component on the second swap. The risk for the firm, however, is that the fixed component on the second swap will be determined at the time the swap is entered into, and there is no reason to expect that it will equal 5%p.a. The fixed component is theoretically determined by the swap counterparty based on its cost of hedging out the risk of the swap, usually in the interest rate futures market, and so will reflect the current structure of market interest rates at the time it is entered into. If interest rates change such that the fixed component of the second swap is, say, 4%p.a., then the cost to the firm of closing out the first swap will be, effectively, 1%p.a. on US$100 million for one year, or US$1 million. Of course the fixed component of the second swap might be greater than 5%p.a. in which case the firm will make a windfall gain, but the fact that this is uncertain represents risk to the firm. Financial Institutions Lending Unit Workbook 65 Queensland University of Technology School of Economics & Finance 6B. MARGIN The second component of the interest rate on a loan advance in an interest period is the Margin. The Margin is defined very simply in the Facility Agreement as follows: “Margin means [ ] per cent. per annum.” W hen defined in this way the margin, unlike the base rate, on any loan advance in any given period is always the same throughout the life of the loan facility. This means that, to the extent the margin covers the lenders costs of providing loan advances, the lender must try to predict what these will be in relation to any future loan advance interest period. W here these costs are random then the lender will be taking risk that the margin might be insufficient to cover them. As interest is payable on loan advances it seems reasonable to expect that the margin covers costs of making loan advances specifically, and not costs of providing the facility more generally. In the previous section it has already been noted that it is the margin that must cover the lender’s funding premium and cost of equity capital with which the loan advance is funded. Before moving to the other costs to be covered by the margin it is useful to explore these in more detail. The funding premium was described in Chapter 4 as representing a lender’s costs of funds in the interbank market in excess of the benchmark rates. Clearly not all lenders will be subject to a market premium, or the same market premium. The second cost to be covered by the margin is the cost of the equity capital with which the loan is assumed to be funded. Recall that regulators and bank management require that the bank hold a certain amount of equity capital against its loan portfolio. Each loan is allocated a share of that equity capital with which it is assumed to be funded. The manner in which banks allocate their capital requirement among individual loans can vary but generally this should reflect the risk of the loan, with more risky loans being subject to higher capital requirement. Irrespective of the manner in which it is done, the loan will ultimately be allocated a capital requirement expressed as a percentage of loan advances, which we can treat as meaning that that proportion of the loan is funded with equity. Recall, however, that the base rate only compensates the bank for the cost of loan advances as if they were entirely funded in the interbank deposit market. Assuming equity has higher cost than debt, then the bank will need further compensation and it has already been noted that the bank must recoup this through the margin. Financial Institutions Lending Unit Workbook 66 Queensland University of Technology School of Economics & Finance The actual cost can also be expressed as a percentage of loan advances, the capital charge, which is the product of the capital requirement and the carrying cost of capital. The latter is the lender’s cost of equity less the base rate and less the funding premium. Again it can be seen that, where the margin is set as a constant throughout the life of the facility, the lender must estimate the future capital charge and carrying cost of capital. W here these are random then, with one exception which is described in the next section, the lender bears the risk that the margin will be insufficient to actually cover them. In a competitive banking market it makes sense for the lender to bear the risks of funding premium and capital charge as these are specific to each bank and so a source of competition between lenders in winning loan business. Ultimately they represent differences between lenders in their marginal cost of funds and the amount of risk a lender is willing to take on changes to its marginal costs of funds over the life of the facility. By contrast the base rate is a function of market interest rates and so it is reasonable that borrowers accept this risk (subject to their ability to hedge). Perhaps the most important, and most widely recognized, cost that the interest margin must cover is the cost of the credit loss provision on the loan advance. In colloquial terms this means that the margin will reflect the loan’s credit risk. Chapter 3 described the major components of the credit loss provision, namely the probability of default and the loss given default, the product of which gives the credit loss provision expressed as a percentage of the outstanding loan advances in any period. The credit risk of a loan is determined from information available to the lender at the time it enters into the loan. Clearly this determination will be most accurate for the initial period of a loan and become less accurate for later periods of the loan. For this reason banks always formally reassess the risk of their loans regularly, often annually or even semiannually, to take account of new information and allow their credit loss provision to be updated. It will be apparent, however, that if the margin is constant throughout the life of the loan the lender again faces risk that the credit risk of the loan might increase in future periods and the margin will again be insufficient to cover it. The lender might address this by setting the margin at a level above that required to cover the initial credit risk so as to cover possible future increases in credit risk. In this case the borrower is penalized with unnecessarily high interest margin in earlier years of the facility against the benefit of possibly insufficient interest margins (compared with the risk it represents) in later years. Of course the ability of the lender to set aside part of the margin in early years against possibly higher credit loss provision in later years will depend on how the loan is utilized by the borrower. For example, if the borrower unexpectedly does not utilize the loan in early years but utilizes it heavily in later years, then setting the margin to cover the average expected credit loss provision might not be effective. Financial Institutions Lending Unit Workbook 67 Queensland University of Technology School of Economics & Finance The credit loss provision that is built into the margin is a major source of competition between lenders, and is the source of competitive advantage for lenders who have superior information about borrowing firms (as described in Chapter 1) and for lenders who are willing to bear greater credit risk than others. This competition and the risk of deterioration in borrower credit risk have led to some alternative solutions to the structuring of interest margin over the life of a facility. The first alternative is used especially with facilities of long maturity. It is to have the margin step up over time. For example, the margin might be 1.00% for years 12, 1.25% for years 35, 1.75% for years 610. Another alternative is to have the margin tied to some measure of the borrower’s actual credit risk at the time of the loan advance, usually the leverage ratio or even an external credit rating. This is described as grid pricing, and an example is as follows: Leverage Ratio (Debt/EBITDA) > 3.25/1 > 2.75/1 <= 3.25/1 > 2.25/1 <= 2.75/1 > 1.75/1 <= 2.25/1 <= 1.75/1 Interest Rate Libor + 3.25% Libor + 3.00% Libor + 2.75% Libor + 2.50% Libor + 2.25% W hen considering the margin on a loan it is important to recognize that the borrower will usually monitor, throughout the life of a loan, the margin it could negotiate on a new loan. If the margin that it is contractually bound to pay under its existing loan is higher than it could negotiate under a new loan (after accounting for differences in fees and transaction costs of a new loan) then it could prepay the existing loan and replace it with a new, cheaper one. Of course the corollary is not true, and if the margin on an existing loan is less than the lender could demand to make a new loan to that borrower, the lender does not have the option to call the existing loan in early. This asymmetry explains why the option to prepay is so valuable for a borrowing firm. The final cost the margin must cover is the opportunity cost of lending, described in Chapter 5 as resulting from the imposition of lending limits that Financial Institutions Lending Unit Workbook 68 Queensland University of Technology School of Economics & Finance ration a lender’s ability to enter into transactions for given borrowers, industries or geographic regions. If forced to ration, then the return available on alternative investments with the same risk will set a minimum benchmark for the margin (in combination with the upfront fee and noncontractual compensation, as will be discussed later) that the lender will demand. This is where supply and demand in the market for loans directly impact on the margin, and thus the costs of the borrowing firm and the profitability of the lender. In a highly competitive market (colloquially called a “hot” market) where there are many lenders chasing few lending opportunities, the lender might only get a margin that is the bare minimum necessary to cover its costs, and often the lender must take greater risk on changes in its costs over the life of the facility. In such markets the lender’s profit will be very low. Conversely, where there is more demand than supply of loans (a “credit crunch”) the lender will be free to set the margin above the level of its costs, with the maxi mum determined by that which borrowers are willing to pay to other lenders. In this case a margin which only just covered the lender’s direct costs might represent opportunity cost in that the loan might more profitably have been made elsewhere. It is useful to also consider liquidity. In the bond markets the price of a bond (the return demanded by bondholders) will substantially reflect the liquidity of the bond; that is, the ability of the bondholder to liquidate the bond quickly without loss of value. Investors will demand higher return for bonds of low liquidity, and vice versa. Traditionally, however, banks made loans with the intention of holding them to maturity and so liquidity was irrelevant. In fact banks are even defined by some authors by their ability to make loans to firms whose debt is illiquid because of information asymmetries. In recent years banks have dramatically increased their capacity to trade their loans prior to maturity and actively manage their loan portfolios, and so it might be expected that the margin on loans should also reflect their liquidity. The hype surrounding the emergence of the leveraged loan market in the early mid2000s, the emergence of loan trading desks and nonbank lenders suggests that liquidity might be a more i mportant issue, at least in this area. It is difficult, however, to find any evidence of this and it remains an open question as to whether banks actively price liquidity into their margins on loans, and certainly the collapse of the leveraged market during the global financial crisis has calmed the hype about loan liquidity significantly. The answer might be that it depends sensitively on the type of loan, bank’s intention in entering into it, and whether the portfolio management techniques available to the banks actually require that loans be tradeable in the market. Liquidity premia might also reflect whichever accounting regime happens to hold sway at the time, as there is a huge and ongoing debate about whether bank loans should be accounted for at historical value or marktomarket value. Financial Institutions Lending Unit Workbook 69 Queensland University of Technology School of Economics & Finance Financial Institutions Lending Unit Workbook 70 Queensland University of Technology School of Economics & Finance 6C. FEES There are two major fees contemplated in the Facility Agreement. The first is an UpFront Fee or Arrangement Fee. This fee is usually calculated as a fixed percentage of the total facility amount and is payable either upon the signing of the Facility Agreement or, less usually, upon the first drawing of loan advances. W hat is usually described as the pricing of the facility is the sum of the margin and the annualized upfront fee. For example, a 2 year facility with a margin of 1.00%p.a. and an upfront fee of 0.50% would be said to have allin pricing of 1.25%. There is an obvious tradeoff between the upfront fee and margin, so that it is not at all unusual for the margin on a loan to be insufficient to cover the direct costs of providing loan advances but for this anticipated shortfall to be recouped by the lender through the upfront fee. Lenders might, in fact, be expected to prefer receiving a greater proportion of their compensation in the form of the upfront fee as they retain this irrespective of any early prepayment or cancellation of the loan, and irrespective of how much the loan is actually utilized by the borrower. In addition the upfront fee has greater present value as it is received by the lender early than any other form of compensation. The other major fee is the Commitment Fee. This provides for the borrower to periodically (usually every 3 months) pay a fee calculated as a fixed percentage of the available commitment, which is the daily facility amount that was available to be drawn by the borrower but has not drawn by it during the relevant period. The commitment fee covers two costs for the lender. The first is that banks are required by their regulators to also hold capital against available but undrawn co mmitments. The rate of this regulatory capital requirement is, roughly, 50% that applicable to drawn loan advances. For example, if the loan is subject to 8% capital requirement on loan advances, it will also be subject to 4% capital requirement on undrawn commitment amounts. Again this capital has a carrying costs, but in this case it can be quantified as the difference between the lender’s cost of equity and the return it could receive by holding the capital in risk free investments that are not, themselves, subject to a capital charge. The second cost the commitment fee has to cover is opportunity cost. W here the bank has lending limits that force it to ration the investments it can enter into then providing a loan facility, even if it is not utilized by the borrower, may preclude it from entering into other investments. As described in Chapter 5, an undrawn facility commitment, which may be drawn at the borrower’s option, uses up scarce lending limits in the same way as loan advances. The commitment fee helps to compensate the lender for this opportunity cost. In Financial Institutions Lending Unit Workbook 71 Queensland University of Technology School of Economics & Finance some cases where the facility is not expected to be utilized, for example a standby facility, some lenders might charge an additional, periodic facility fee on the entire facility amount, whether it is utilized or not. This is not typical, however, and is usually resisted by borrowers. Financial Institutions Lending Unit Workbook 72 Queensland University of Technology School of Economics & Finance 6D. ADDITIONAL COSTS There are a number of additional lender’s costs contemplated in the Facility Agreement. The first of these provides that the borrower shall make all payments to the lender free and clear of taxes. If the borrower or lender is required by law to deduct taxes from any payments then the borrower must increase the amount of, or “grossup”, the payments so that the lender receives the amount they would have received had tax not been deducted. This is called the grossup provision. In section 6B it was observed that the cost of complying with regulatory capital requirements, the capital charge, must be recouped by the lender through the interest margin. There is a major exception to this, however. This provides that where there is a change in laws or regulations or their interpretation after the facility is entered into, and this has the effect of reducing the return to the lender or increasing its costs, then the borrower shall pay the amount of this increased cost to the lender. An unanticipated increase in the regulatory capital requirement on the loan, with a proportional increase in the capital charge, would be an example of an increased cost that would fall within this provision. The Facility Agreement also sets out a general catchall of lender’s costs that must be indemnified by the lender if they are realized. The most significant of these include its costs, losses and liabilities if the borrower defaults under the facility. Default is the subject of Chapter 9 of this workbook, but simply it means that the borrower has failed to meet any of its obligations under the facility. Other costs that fall within this provision include a lender’s costs if it arranges funding for a loan advance pursuant to a drawdown notice and that loan advance is not then drawn, or if the borrower gives a notice to repay a loan advance and then does not do so. Similarly where the borrower prepays a loan advance prior to the end of an interest period then it is also liable to pay the lender’s breakcosts, which are defined in Clause 1.1 as its costs of reinvesting the funds for the remainder of the interest period. Financial Institutions Lending Unit Workbook 73 Queensland University of Technology School of Economics & Finance Chapter 7 Relationship Compensation Financial Institutions Lending Unit Workbook 74 Queensland University of Technology School of Economics & Finance A major focus of this unit is the distinguishing of banks from investors in order to explain the fundamental differences between their debt products, loan facilities and securities. This chapter continues to do so, but this time as the entry point into a discussion of the additional, noncontractual revenue or compensation that a bank might generate pursuant to the provision of a loan to its client. Banks are firms that generate revenue from various sources, only one of which is investment in the form of loans, and which might include the sale of other financial products and services to their clients. By contrast, an investor in a debt security might be assumed to be limited to the revenue directly related to the securities in which it invests, either the contractual payments under them and/or the price it receives from selling them. Following is a nonexhaustive list of loan and nonloan products and services typically provided by banks to their corporate clients. 1) Treasury a) Research b) Foreign Exchange c) Futures d) Derivatives e) Money market f) Precious metals 2) Securities underwriting 3) Mergers & acquisitions 4) Corporate and project advisory 5) Payments and cash management a) Trade services b) Letters of credit c) Guarantees d) Documentation 6) Custodial services 7) Asset management Banks realize that the provision of a loan facility can increase the probability of selling other products and services to the borrower client, and vice versa. For this reason, banks usually assess and base their decisionmaking on the overall revenue and profitability of a relationship with a client, the relationship return, and not the revenue or profitability of individual products or services on a standalone basis. This raises the possibility that a loan facility might be used as a lossleader, in the sense that it might not being profitable on a standalone basis but will increase the probability of the bank securing lucrative opportunities to sell other products or services to the client, either simultaneously or in the future. Financial Institutions Lending Unit Workbook 75 Queensland University of Technology School of Economics & Finance To take a simple example, recall from Chapter 6 that a bank loan has a floating or variable interest rate, but that the borrower can effectively fix the interest rate by entering into an interest rate swap. If the bank can simultaneously sell the borrower an interest rate swap as well as providing it with a loan then it makes sense for the bank to set the pricing and assess the profitability of both products together. The model now used by banks almost universally in assessing the return from their relationships is RAROC, or Risk Adjusted Return on Capital. In its general form this model captures all the cost and revenue elements from all the products that comprise a bank’s relationship with a customer. The major exception is that it does not capture opportunity costs. As such the RAROC model can be placed in the general class of economic value added, or EVA, models. EVA models are characterized as such because they quantify the return to a firm’s equityholders which is theoretically in excess of their required return for the risk that their equity is exposed to; in other words, the theoretical value that a product or relationship adds to the firm for its equityholders. The specific form of the RAROC models is slightly different from that of usual EVA models, however. RAROC models express their output as a % return on the risk adjusted equity (capital) used by the products or relationship, which is then compared with the firm’s required return on capital. If RAROC is in excess of required return, then the product or relationship generates a positive EVA and so increases the value of the firm to its equityholders. RAROC less than the required return is negative EVA and destructive of firm value. Using the cost and revenue elements identified in Part 2, a RAROC model for the first period of loan facility might be expressed as follows. Financial Institutions Lending Unit Workbook 76 Queensland University of Technology School of Economics & Finance Typical RAROC Model RAROC % = Amounts Advanced $ x [ Margin % Credit Loss Provision % (Funding Premium % x (1 – Capital on Advances %)) + (Base rate * Capital on Advances %) ] + Undrawn Commitments $ x (Commitment fee %) + Facility Amount$ * UpFront Fee% ] Allocated administrative costs $ All divided by Advance Amount $ * Capital on Advances % + Undrawn Commitment $ * Capital on Undrawn Commitments% RAROC Example Assume the following in relation to the first period of a loan facility Facility Amount Loan Advances UpFront Fee Margin Credit Loss Provision Funding Premium Base Rate Commitment Fee Riskadjusted Capital on Advances Riskadjusted Capital on Undrawn Commitments First period RAROC = [$50*(0.50%0.30%0.05%(18%)+5%(8%)]+[$50*0.20%]+[$100*1%] / $50*8% = 34.425% $100 $50 1% 0.50%p.a. 0.30%p.a. 0.05%p.a. 5%p.a. 0.20%p.a. 8% 0% Financial Institutions Lending Unit Workbook 77 Queensland University of Technology School of Economics & Finance This loan has positive EVA if required return on capital is less than 34.425%. BUT what if the loan has more than one period? W ithout the upfront fee, RAROC is only 9.425%. Now let’s say that, if the bank provides the loan facility in the example above, it will also receive a mandate from the firm to advise on the way it should finance a new project. This advisory role will not require the bank to take any additional credit exposure to the firm, but will generate income (revenue less costs) of $1 per period. The relationship would now generate RAROC of 59.425% in the first period and 34.425% in subsequent periods of the loan. The net revenue from the advisory role increases only the numerator of the RAROC equation, leaving the denominator unaffected. This explains the attraction to banks of products that do not generate exposure and so do not require capital. Now further assume that the loan and the advisory role both have terms of 3 periods, the bank has a required return on equity of 20%, and that whether the bank receives the advisory mandate is not certain. If the bank makes the loan then the probability of receiving the advisory mandate is 75%. If it does not make the loan, then the probability of receiving the advisory mandate falls to 10%. W hat should the bank do? There is, in fact, no correct answer. To this we can add difficulties with the RAROC measure itself, which does not reflect reality. It assumes the bank actually raises the new capital required to support the new loan, but in reality the capital raised by banks is very lumpy, comprising new issues of equity plus profits less dividends. No bank ever raises new capital on a loan by loan basis, and so a new loan actually shares in the existing capital of the bank, making the RAROC measure more of a disciplinary tool rather than a real reflection of economic value added. Financial Institutions Lending Unit Workbook 78 Queensland University of Technology School of Economics & Finance Section 4 Loan Terms and Conditions Financial Institutions Lending Unit Workbook 79 Queensland University of Technology School of Economics & Finance Section 4 of this workbook explores the nonoperative terms and conditions of the loan facility. The operative terms and conditions are those that describe the facility and the mechanics of its utilization; these have been covered in earlier sections of this workbook. The nonoperative terms and conditions are those concerned with protecting the position of the lender. The next chapter, Chapter 8, considers the conditions precedent, representations and undertakings by which the lender confirms the circumstances under which it enters the loan and attempts to control the way these circumstances change throughout the life of the loan. The following chapter, Chapter 9, describes the remedies available to the lender in the event that circumstances change unfavourably and the lender is in danger of suffering credit loss. Financial Institutions Lending Unit Workbook 80 Queensland University of Technology School of Economics & Finance Chapter 8 Conditions Precedent, Representations and Undertakings Financial Institutions Lending Unit Workbook 81 Queensland University of Technology School of Economics & Finance The nonoperative terms and conditions are the heart of any loan facility. They are the means by which the lender confirms the circumstances of the borrower, induces it to disclose further information, and then induces the borrower to act, or cause others to act, in a manner which protects the position of the lender. Loans are often distinguished from bonds on the basis that they contain more onerous terms and conditions, and it is true that these nonoperative terms and conditions often significantly constrain the activities of the borrower. Borrowers often resist them as much as possible, and this is where the majority of the difficult negotiation of any loan facility relates takes place. This tends to miss the point of these terms and conditions. They are concerned, ultimately, with the credit risk of the borrower and it is the ability of banks to negotiate and enforce such terms and conditions that can enable them to either accept credit risk that would be unacceptable in the capital markets or mitigate credit risk and thus reduce the compensation they require to provide the facility. In fact it has been observed that a firm’s bank loan facilities may even have a higher external credit rating than the same firm’s equal ranking bond issues. This is attributed to the presence in the loan facilities of such terms and conditions and the ability of the lender to monitor and enforce them. The nonoperative terms and conditions also play an important role in mitigating the conflicts between the firm’s lenders, shareholders and management. Debt is a fixed claim, with no share of upside profits, but in return has priority over the firm’s assets. Equity is a residual claim, subsidiary to the debt claim, but has the benefit of unlimited upside potential. Shareholders may, therefore, have very different views of the way they want to firm to invest in its assets; typically shareholders desire more risk and debtholders less. Shareholders appoint management, however, and in the absence of some form of control by debtholders, they would prevail. The non operative terms and conditions are the mechanism by which the bank attempts to constrain the way that the firm’s management deal with its assets on behalf of shareholders or in their own interests, where they do not coincide. It is often the case in practice, however, that lenders demand nonoperative terms and conditions which are badly conceived, do not address specific risks of the borrower and/or are difficult or impossible for the borrower to comply with. In such cases borrowers do have legitimate objections to them. In this chapter I provide an example of the way these conditions might be structured in a specific case. It takes the form of a hypothetical transaction which is described by a credit application memorandum, the standard document by which a proposed loan facility is documented within a bank. This is followed by a draft term sheet which sets out the terms and conditions of the facility agreement on which the bank would be willing to proceed. Financial Institutions Lending Unit Workbook 82 Queensland University of Technology School of Economics & Finance You should also refer to the Facility Agreement, where the form of the traditional major nonoperative conditions are set out, but recall that the entire purpose of nonoperative conditions is to address the specific risks of the particular borrower and facilit y and must be tailored to each. Financial Institutions Lending Unit Workbook 83 Queensland University of Technology School of Economics & Finance 8A. EXAMPLE Short Form Credit Application Memorandum Borrower: ACME WIDGET COMPANY Pty Ltd (“AWC”) Borrower description: AW C is a special purpose vehicle which would acquire all the assets and rights of the W idget division of Megacorp under a proposed management buy out. W idgets are used as a key input in the construction industry. The proposed acquisition price is $100,000,000 (comprising $95,000,000 purchase price and $5,000,000 capitalized acquisition expenses) which will be funded as follows: $75,000,000 senior bank loan $20,000,000 mezzanine (subordinated) loan Investmentfund (final terms yet to be agreed) $5,000,000 management equity Proposed new credit exposure $75,000,000 7 year term acquisition facility. Recommended risk rating: BB Primary Source of Repayment: Cash flow from operations (Moderate) Megacorp’s W idget division generates strong and consistent cashflows based on profitability of its underlying operations, but this will moderated by the relatively high debt service costs under the proposed acquisition structure. Secondary Source of Repayment: Sale of business or disposal of assets (Moderate/weak, Strongly related to primary source). A number of other groups also bid for to acquire the W idget business, suggesting possible purchasers for the business on a going concern basis at reduced price. The major asssets are specific to the business. But these sources are likely to realize significantly lower value in a default situation. Major Credit Risks · · · · Recession in building industry activity (Probability: medium/high. Impact: moderate/high) Significant interest rate increase (Probability: medium. Impact: moderate) Competitive pressure on operating margins (Probability: low/medium. Impact: moderate) Unsuccessful transition to standalone entity (Probability: medium. Impact: moderate/high) to be provided by Financial Institutions Lending Unit Workbook 84 Queensland University of Technology School of Economics & Finance (Note that the summary of the major credit risks above would be more comprehensive in a real credit application memorandum.) Competitive Position Strong Megacorp currently holds 30% of national market share in W idgets. Its major competitors are ABC Corp with 35% market share and a collection of smaller competitors in each regional market. Megacorp and ABC are the only national operations. The W idget market has significant barriers to entry due to the significant investment in capital equipment and access to the raw material, Qwertinium. There are no perfect substitutes for the use of W idgets in the construction industry. Balance Sheet Position W eak Due to the proposed acquisition strructure, AW C will have a high level of senior leverage and total leverage. The company will have limited ability to withstand significant financial shocks. Management Strength Moderate Management is very experienced in the W idget industry, with the proposed CEO and CFO having combined 60 years experience. The proposed CEO has been director of the W idget division of Megacorp for 10 years. It is intended that all the current management team of Megacorp W idget division will transfer to AMC. This experience is moderated by the lack of experience of the management team in operating a standalone entity, as external capital raising and financial management is centralized in Megacorp. Financial Institutions Lending Unit Workbook 85 Queensland University of Technology School of Economics & Finance Summary Base Case Acme Widget Company Amounts in $’million Income Statement Sales rev enue Cost of goods sold Gross Income S&A expenses EBITDA 6 Interest Expense Depreciation/amortisation Taxes Net Income after tax 0 1 2 3 4 5 6 7 110 70 40 12 28 9 5 1 13 0 112 70 42 13 29 9 5 1 14 5 114 70 44 13 31 8 5 3 15 5 116 72 44 14 30 7 5 3 15 5 120 74 46 15 31 6 5 5 15 5 122 74 48 16 32 4 5 7 16 5 125 76 49 18 31 3 5 7 16 5 Proposed Dividend Balance Sheet ASSETS Cash and equiv alents Net other current asset s Fixed Assets Total Assets CAPITAL Senior Debt Subordinated Debt Common Equity Retained Earnings Total capital Cash Flow Beginning cash CF f rom operations CF f rom investment CF f rom financing End cash Ratio Analysis Interest Cov er DSCR Senior Leverage Net Senior Debt/EBITDA Gross Margin Net Margin 6 11 5 95 111 17 6 90 113 19 7 85 111 17 7 80 104 13 9 75 97 12 9 70 91 12 8 65 85 75 18 5 13 111 70 16 5 22 113 60 14 5 32 111 45 12 5 42 104 30 10 5 52 97 15 8 5 63 91 0 6 5 74 85 0 0 100 100 0 0 18 5 2 11 11 19 1 12 17 17 20 1 17 19 19 20 0 22 17 17 20 2 22 13 13 21 0 22 12 12 21 1 22 12 3.02 2.49 0.68 2.14 0.36 0.12 3.22 1.81 0.62 1.82 0.38 0.13 3.88 1.55 0.54 1.32 0.39 0.13 4.29 1.25 0.43 0.93 0.38 0.13 5.17 1.35 0.31 0.38 0.13 8.00 1.52 0.17 0.39 0.13 10.33 1.55 0.00 0.39 0.13 Interest assumed at 7%p.a. senior debt (floating rate), 20% fixed rate for subordinated debt. Financial Institutions Lending Unit Workbook 86 Queensland University of Technology School of Economics & Finance Acme Widget Corporation $75 million Term Loan Facility Summary of Proposed Terms and Conditions Borrower Facility Lender Purpose Final Maturity Availability : Acme W idget Corporation (“AW C”). : Term Loan Facility. : Bank of QUT. : To finance the acquisition of the assets and rights of the Widget division of Megacorp. : 7 years from the drawdown date. : Subject to the satisfaction of conditions precedent, the loan may be drawn in one lump su m during the Availability Period. Any amount not drawn will be automatically cancelled. : 6 months from the date of signing of the facility agreement, but in no event later than (date). : Subject to any Mandatory Prepayment, the loan will be repaid in 6 annual instalments commencing two years from drawdown date as follows: Instalment 1 $5 million Instalment 2 $10 million Instalment 3 $15 million Instalment 4 $15 million Instalment 5 $15 million Instalment 6 $15 million Availability Period Amortisation Mandatory Prepayment : On the date of each scheduled repayment, the Borrower will also prepay 50% of the Excess Cashflow for the year preceding that date, as defined below. Prepayments will be applied to reduce scheduled repayments in inverse order of maturity (i.e. last ones are reduced first). Excess Cashflow for any period is equal to the beginning cash plus cashflow from operations for that period, less the sum of senior debt service for that period and scheduled senior debt service for the following period. Optional Prepayment : At its option and subject to 3 day’s written notice, the Borrower may prepay additional amounts on any interest payment date, in minimum amount of $5 million and Financial Institutions Lending Unit Workbook 87 Queensland University of Technology School of Economics & Finance integral multiples of $1 million. Amounts prepaid will be applied to reduce instalments in inverse order of maturity. Pricing UpFront Fee Interest Rate : 3.50% flat on the Facility Amount, payable on loan signing date. : Libor + 4.00%, payable in arrears at the end of each interest period. Interest periods shall be threemonths duration. : 2.00% on the daily unutilized and uncancelled commitment amount, payable quarterly in arrears. : Conditions precedent to first drawdown will include, but not be limited to, the following: 1. Executed agreement of sale of the assets and rights of the W idget division of Megacorp to the Borrower, in form satisfactory to the Lender, under which completion is conditional only upon the receipt of funds drawn under this facility. 2. Executed indemnity from Megacorp in favour of the Borrower against misrepresentation of the condition of the assets or the nature of rights transferred to AW C under the agreement of sale. 3. Executed subordinated debt facility in form satisfactory to the Lender, and under which all funds have been drawn by the Borrower. 4. Intercreditor agreement (agreement between creditors, in this case the senior lender and subordinated lender about the rights of the parties to the security and assets of the company in winding up) in form satisfactory to the lender. 5. Interest rate hedging for the full amount and life of the senior debt in form and rate satisfactory to the Lender. 6. Insurances over assets in form satisfactory to the Lender. 7. Usual conditions precedent for a facility of this type including copies of relevant company documents and authorities, legal opinions, valid security documents, consents and licenses, and certification of representations and warranties. Representation : Representations will include, but not be limited to the s following: Commitment Fee Conditions Precedent Financial Institutions Lending Unit Workbook 88 Queensland University of Technology School of Economics & Finance 1. No material adverse change in the financial condition or business of the Borrower, or the W idget market, which could affect the ability of the Borrower to repay the facility. 2. All agreements required as conditions precedent are in force and in form originally approved by the Lender. 3. Usual representation for a facility of this type including accuracy of information, pari passu ranking, no default, no resulting breach, status and capacity of the borrower, validity, no litigation, etc. All representations and warranties will be repeated upon facility signing, loan drawdown, and each interest payment date. General Undertakings : General undertakings will include, but not be limited to, the following: 1. Negative pledge: the Borrower will not enter into any additional borrowings or provide any financial accommodation (e.g. guarantees), excluding the subordinated debt facility, in excess of $2 million in total. The Borrower will not pledge or secure any of its assets or interests to any other party in excess of $250,000 in total. 2. No disposal of assets valued in excess of $100,000 without Lender’s consent, except for sales of inventory in the ordinary course of business. 3. Borrower to provide management updated management (unaudited) financial statements and three year forecasts within 14 days of the completion of each quarter, except that it shall provide audited financial statements at the end of each halfyear. 4. Payment of dividend and debt service on subordinated debt only to be made from Excess Cashflow and after Mandatory Prepayment. 5. Usual undertakings for this type of business, including provision of information, financial statements, and no change of business. Financial Undertakings : Financial undertakings will include, but not be limited to, the following: 1. Minimum Interest Cover Ratio for each halfyear period as follows: year 1, 2.75; year 2, 3.00; year 3, 3.50; year 4, 4.00, year 5, 4.50; year 6, 7.50. 89 Financial Institutions Lending Unit Workbook Queensland University of Technology School of Economics & Finance 2. Minimum Debt Service Cover for each halfyear period of 1.25:1. 3. Minimum Net W orth as follows: year 1, $15; year 2, $20; year 3 onward, $30. 4. Maxi mum Debt:EBITDA of 2.50 in year 1 and 1.75 thereafter. Security : First ranking fixed and floating charge over all the assets and interests of the Borrower. : Documentation in form usual for this type of facility and in form acceptable to all parties. It will include additional terms and conditions usual for this type of term loan facility including, but not limited to, standard conditions precedent, representations & warranties, undertakings, material adverse change, illegality, increased costs, and events of default. : Documentation shall include provision allowing the transfer or assignment of the Lender’s rights and obligations under the facility, subject to the consent of the Borrower, which may not be unreasonably withheld. : The Laws of Queensland. Documentation Transferability Governing Law Financial Institutions Lending Unit Workbook 90 Queensland University of Technology School of Economics & Finance Chapter 9 Events of Default, Remedies and Security Financial Institutions Lending Unit Workbook 91 Queensland University of Technology School of Economics & Finance You will be provided a handout in class of the following readings. It is difficult to improve on Adams’ concise writings on this subject, and so I will not try. If you wish to access these readings before class please see me or you can find the book in the library. Corporate Finance: Banking and Capital Markets David Adams. Jordan Publishing Limited. 2003. (“Adams”) Adams: pp.103111, 161167, 171179 Financial Institutions Lending Unit Workbook 92 ...
View Full Document
- Three '10
- Accounting, Institutions Lending Unit, Queensland University of Technology School of Economics & Finance