The Practice of Lending: A Guide to Credit Analysis and Credit Risk

Part IThe Criteria-Based Approach to Credit Assessment and Credit Rating

© The Author(s) 2020

T. M. Yhip, B. M. D. AlaghebandThe Practice of Lendinghttps://doi.org/10.1007/978-3-030-32197-0_1

1. Credit Analysis and Credit Management

Terence M. Yhip¹   and Bijan M. D. Alagheband²

(1)

University of the West Indies, Mississauga, ON, Canada

(2)

McMaster University and Hydro One Networks Inc., Toronto, ON, Canada

Terence M. Yhip

Electronic Supplementary Material:

The online version of this chapter (https://​doi.​org/​10.​1007/​978-3-030-32197-0_​1) contains supplementary material, which is available to authorized users.

Keywords

Credit originationCredit managementTwo-dimensional rating systemBorrower risk ratingFacility risk ratingInformation asymmetry

Learning Objectives

1.

Provide a general understanding of the various processes in lending and credit analysis

2.

Review the Five Cs method of credit analysis

3.

Introduce the Criteria-Based Approach to risk rating

4.

Examine the reasons that asymmetrical information poses serious problems for lending and risk rating

5.

Discuss credit culture and its influence on credit analysis and lending

6.

Look at banking regulations and their relevance to credit analysis and lending

1.1 Introduction

Historical and Philosophical Bases of Credit Analysis

This chapter provides a broad discussion on lending, which consists of two connected activities: credit origination and credit administration. A basic understanding of the issues around them allows a deeper appreciation of the work of a credit professional and the philosophical approach the credit analyst brings to the task of assessing creditworthiness. Credit and the credit market are vital to the efficient functioning of a modern economy that is intrinsically characterised by complexity across production, consumption, and financial intermediation. This third activity featuring a developed credit market is intrinsic to a modern economy. For example, we instinctively think of loans when we see a commercial bank, a credit union, or an investment bank, some of the principal institutions that deliver the full array of credit, including sophisticated financing products such as securitisations. Regardless of the complexity of the loan facility, lending in its simplest form is as old as civilisation going back thousands of years to agrarian societies. The oldest surviving codes governing lending date back to 2000 BC in Mesopotamia.¹ Not surprisingly, the fundamentals of lending have not changed over millennia. A lender still asks the same old question: Should I lend to this person or firm? So, in terms of the modus operandi, there is no essential difference between a moneylender in 2000 BC and the manager of your local bank, where you go to for a personal loan or a business loan.

The word credit comes from the Latin word creditum, something entrusted to another; or some would say, credere meaning to trust or believe. In the Western world, credit involves a contract, articulated in a legal document called the Loan Agreement or Credit Agreement, in which a borrower receives a sum of money, or anything of value, now, and promises to repay the lender, usually with interest, at some future date. By definition, a credit agreement is enforceable by law. In Islamic countries where Sharia law² rules, banks lend not on the basis of interest but on the principle of profit-and-loss sharing, through which the parties in the transaction agree to distribute profits according to an agreed ratio, and to absorb the loss in proportion to the capital invested by each partner. Regardless of the cultural difference in the approach to lending or, broadly speaking, investing, the universal and fundamental principle embodies two concepts: willingness and ability to repay. Islamic commercial lending practices do not eliminate risk and, thus, a financial assessment of creditworthiness is still necessary as in traditional Western banking. The models that we will be introducing in later chapters attempt to quantify directly the second concept. You will observe in later chapters that we are not ignoring willingness because we capture it in business risk and country/sovereign risk analysis.

Credit risk analysis is both art and science; how much of each depend on the problem in question. It is part art because it involves experience, practice, skill, and imagination; but it is also science in the narrow sense of credit risk analysis employing the essential methodology of natural science, which consists of the procedures and the practices of theorising, testing, and revising with new information. All this goes on with the awareness that even the reigning theory is always provisional.

The predictions of an economic model (of which financial and accounting models are a subset) will never be precise for many reasons, which is a subject all of its own. In physics, the king of all sciences, 100% model accuracy is an unrealistic expectation. The common expression this or that method is not an exact science is based on the fallacy that science is absolutely exact or accurate—a belief that was debunked long years ago by the nineteenth-century English economist and logician, Will Stanley Jevons.³ Consider gravity. Isaac Newton described this force by the Inverse Square law, which explains and predicts the paths of planetary bodies accurately for all practical purposes, but more than one and a quarter century later Albert Einstein’s General Theory of Relativity (1915) came along and proved more accurate. Many people have the false notion that science is exact and certain without stopping to think that science is essentially an endless series of refinements, and that the answers are never 100% certain. Consider the laws of motion in physics: the certainty or the determinism of the laws in classical or Newtonian mechanics has been supplanted by uncertainty in quantum mechanics, which teaches that the laws of physics would only allow one to calculate relative probabilities of various future outcomes.⁴

A related fallacy is the belief that the mathematisation of credit analysis makes it more accurate and objective. At best, mathematics and statistics are just tools, albeit indispensable, to detect, test, and quantify patterns in large datasets. Consider asset valuation models that we will examine in later chapters. For one thing, the inputs or assumptions used for valuation leave plenty of room for personal bias, but that does not mean valuation models are worthless. Used properly, they are powerful and useful tools for making informed investment decisions. Mathematical models, no matter how sophisticated or carefully constructed, will still be a limited although important tool in credit analysis because much of the information inputs is qualitative.

There is also a common fallacy to equate subjectivity with everything from guesses to personal prejudices. In Chaps. 4 and 5, we will be discussing the criteria approach to credit risk analyses, a hybrid model that combines features of qualitative/heuristic models and mathematical/causal models, which includes statistical models. In criteria-based modelling, we allocate weights to the criteria and the predictors judgementally. The weights come from expert judgement rather than from a statistical procedure. That said it would be a mistake to equate judgement with guesswork or bias.⁵ Subjectivity ought to be grounded in observation, meaning both empirical data and interpretation borne out of experience (which, by definition, is not impersonal). As more information emerges, overall understanding of the issues improves, thus compelling the credit practitioner to review the thinking, the current model, the methodology, and the methods.

Updating the priors as in Bayesian analysis is a natural and essential way to approximate objectivity and to make more accurate decisions, which implies there is no escaping the fact that the model will still fall short of the ideal.⁶ Expert judgement is both unavoidable and necessary. The good thing about the criteria-based approach is that the subjectivity is not buried deep within the analysis but is laid bare in the weights one assigns to the risk criteria and sub-factors, and the choice of predictors. In this way, one can go back and do over the analysis with revised weights in the light of better information. Other adjustments, such as adding or replacing variables, may also be made.

Process

You would be seeing the word process used numerous times in this book. The word comes up regularly in all activities that require a series of steps to transform inputs to outputs. Credit underwriting, credit structuring, credit analysis, loan monitoring, credit review, problem loans management, Watch List, and so on are quintessentially processes with defined aims (outputs). In the case of loan underwriting, the ultimate aim is to get the loan on the books with all the features of an appropriately structured loan. Related to process is process mapping, an important tool for not only operations but also risk management.

1.2 A Framework for Credit Operations

We may view the lending process as part of a much larger system characterised by linkages between decision-making units of an organisation, its internal credit policy and ERM (Enterprise Risk Management) framework, and the overarching economic and legal environment in which the organisation operates. Figure 1.1 explains the system. The ellipse in the centre is the credit process of originating and managing the exposure of individual borrowers and loan portfolios. The centre is where loan underwriting and approval processes occur, and where loans are booked, serviced, and managed. The rectangle nesting the ellipse depicts the organisation’s lending policy, the ERM framework, and the related oversight by the board of directors. Credit policy determines single name limits, portfolio limits, and sector limits. The ERM framework of a bank (a) establishes risk tolerance and self-imposed risk limits by business type, (b) manages and monitors the risk categories, limits, and targets, and (c) prepares regular reports that at least address and document developments pertaining to (a) and (b). To give an example, the business unit (e.g., mortgage finance) might want to put more residential real estate loans on the books (because the demand is high), but due to risk appetite the total for the whole organisation cannot breach the limits, otherwise a formal approval process has to be followed to exceed the targets.

Fig. 1.1

A framework for credit operations

The external environment encompasses the economy, law of the land, and financial regulations. The economy includes business conditions, industry risks, and the business cycle. These limit the potential risk ratings of borrowers. In any given economic cycle, some industries will be highly risky and lenders would be unwise to ignore the trends. In a recession, lenders exercise more caution in lending. Law of the land is self-explanatory: a country’s laws set the parameters for what a financial institution can and cannot do. Financial regulations come under various acts or laws governing domestic and international banking. In relatively developed financial markets, many regulators are responsible for supervising financial organisations. Table 1.1 gives a list of the independent regulatory bodies for the US, the UK, and Canada.

Table 1.1

National regulators of the financial system of the United States, the United Kingdom, and Canada

But in emerging financial markets, the responsibility for the day-to-day supervision is usually executed by the central bank that has the regulatory authority. Regardless of whether the economy is developed or developing, regulators are expected to be independent bodies because of their authorities. For example, they monitor compliance and enforce prudential regulations and lending practices. Financial institutions face heavy fines and penalties for not complying. The regulatory environment helps to shape a bank’s lending policy and rating system.

1.3 The Credit Underwriting Process

For an individual borrower, the main source of financing is a loan from a financial institution (banks and credit unions). Businesses or commercial enterprises raise financing through various channels:

1.

Bilateral borrowing from a bank or a syndicate of banks

2.

Leasing (capital and operating leases)

3.

Issuing secured and unsecured bonds (notes and bonds, respectively) directly in the fixed-income capital market

4.

Issuing common shares/private placement (equity market)

5.

Issuing hybrid instruments with bond and equity features

6.

Securitising and factoring/forfaiting

In this book, we focus on credit or, more narrowly defined, balance sheet debt and off-balance sheet debt. This type of debt in the Western banking tradition is called funded debt because the debt is funded by the interest payments. The credit rating system that we will be looking at is a two-dimensional system proposed under Basel II and supervised by the regulators. The essential requirement of such a system is that borrower risk is separate and distinct from facility risk. Together, however, they determine the expected credit loss of a loan that is classified impaired.

Difference Between Funded and Unfunded Debt

Liabilities consist of funded and unfunded obligations. A funded debt is a credit facility that is funded with actual cash and bears interest. It includes overdraft, loans, and bonds, so whether it gets reported on or off the balance sheet is irrelevant in credit analysis. The debt is said to be funded by the interest payments to the lender. In contrast, unfunded debt are contractual obligations for future lending, such as guarantees and documentary letters of credit. For such lending, the financial institution (FI) does not advance actual cash but only assumes the risk of non-payment. In return, the FI charges a commission (but not interest). These are contingent or potential liabilities that depend on an uncertain future event.

Determining the BRR (Borrower Risk Rating) and the Facility Risk Rating (FRR) are essentially the work involved in loan underwriting. In particular, loan underwriting is the process of a lender determining whether an obligor’s loan application is a safe risk based on the entity’s capacity to repay the loan primarily through cash generated from the business and secondarily through collateral. The loan can be the funds to buy a car for personal use and not commercial use, which makes it a personal credit; or the loan can be a term loan to acquire plant and equipment for the enterprise; hence, a business loan. For business or commercial loans, underwriting includes in this general definition the evaluation of the business owner and the business. Consequently, the underwriting process for business and corporate loans will be more complex and lengthier than for personal loans and residential mortgages.

A business usually needs more than one credit facility (or a loan product), the amounts would be larger than for personal loans, and altogether there is more documentation to complete. The customer provides some of the documents, but the lender does most of the paper work to support the loan application. They are the legal documents to close the transaction. One important document is the contract that binds the obligor and the lender, called the Loan Agreement or the Credit Agreement. Other important documentation would include business licences and registration, Articles of Incorporation, and various security agreements for secured loans. Regardless of the type of credit, credit underwriting involves the same process flow from start to finish and includes the following steps:

1.

Credit Initiation

2.

Credit Analysis

3.

Loan Structuring

4.

Credit Submission and Adjudication

5.

Loan Documentation

6.

Loan Closing and Disbursement

7.

Loan Monitoring

8.

Problem Loans Management

Although the focus of this book is credit analysis and credit risk rating, it is important for credit analysts to have a fundamental understanding of the lender’s credit policies and the procedures. They are the controls that make the lending process systematic and methodical, but most of all minimise credit losses (e.g., due to poorly executed documentation) and enhance the likelihood of full repayment. Lending is a very risky business and creditors face the statistical certainty that a fraction of their loan portfolios will go bad, just like firms that sell products expect their accounts receivables will not all be paid. An essential part of the credit process is identifying and managing problem loans in order to prevent and recover losses. We provide an overview of each of the underwriting components listed above.

1.3.1 Credit Initiation

Either a borrower approaches a lender for a loan or a relationship manager (or account manager) finds a prospective loan customer as part of his sales effort. The credit initiation and analysis process ensures that loans made by a lending institution adhere to the lender’s enterprise-wide credit policy, guidelines, credit procedures, and credit standards. The credit policy lays out the types of loans that are acceptable, the loan purposes, tenor, collateral, structure, and acceptable guarantees. Furthermore, the credit policy establishes that policy exceptions must have the explicit approval of the loan approval authority, which adjudicates credit transaction requests.

In banks, the authority rests in Risk Management that provides independent risk oversight across the enterprise. The credit policy sets the criteria that a prospective borrower must meet. Let us suppose that the preliminary client screening is completed, the customer meets the threshold requirements, the relationship manger has identified the customer’s credit needs, based on which, the customer submits a loan application. What follows next in the underwriting process is a thorough analysis of the borrower’s creditworthiness, which is the capacity and the willingness to repay the loan.

1.3.2 Credit Analysis

Loan underwriting is the process of determining if a loan application is an acceptable risk. An important objective of the process is to evaluate the credit risk, which has two components: the borrower’s ability to repay the loan and collateral supporting the loan. The traditional Five Cs method is a good jumping-off point for a deeper dive into credit risk rating methodology. The credit analysis process begins with the collection, analysis, and evaluation of the information pertaining to the Five Cs:

1.

Capacity

2.

Capital

3.

Collateral

4.

Condition

5.

Character

Capacity to Repay

The capacity to repay a loan is the borrower’s financial ability to repay a loan with interest on schedule. In order to assess capacity, reliable and timely financial statements are required. In later chapters, you will learn how to interpret and use financial statements to calculate financial ratios and perform a forecast of cash flow.

Capital⁷

Capital refers to shareholders’ equity in the enterprise. It is what the owners of the firm have invested and what they have at risk if the business were to fail. In addition, we will be looking at two other types of capital in credit analysis: regulatory capital and economic capital. Regulatory capital, as the name suggests, is capital that banking authorities define for regulatory purposes, and the items in the definitions are balance sheet items. The economic capital of a bank is derived from an economic capital model that does not require assets as inputs. It is the difference between some given percentile of a loss distribution and the expected loss derived from combining the probability of default (PD) and the loss given default (LGD). Capital, as a factor of production, refers to physical assets that a business uses to produce goods and services.

Collateral

The primary source of repayment is cash flow. The secondary source of repayment is collateral in case the borrower cannot repay the loan through cash flow. Personal assets, paper assets, physical assets, and even future income (such as customers’ orders) are all considered collateral. A borrower would pledge these assets to secure a loan; hence, the phrase secured loan. In the event of default, the central concern of the lender of a secured loan is whether the realisable value of the collateral is sufficient to cover the expected loss. Since the quality of the asset is important, lenders usually keep current and accurate record of the realisable value of the collateral. Part of the exercise of updating collateral value is obtaining current evaluations of the assets such as property, inventory, and accounts receivable. An important function of loan administration is to ensure that collateral documentation held by the lender is in (legal) order to facilitate orderly sale of assets and orderly recovery of losses.

Condition

The desire to grant a loan depends on certain conditions that are external to the lender. The relationship manager and the credit analyst must be aware of recent and emerging trends in the borrower’s line of work, the borrower’s industry, business conditions affecting sales, and economic forces such as inflation and interest rates that might affect the loan. Additionally, certain conditions internal to the lender are important to lending. Therefore, credit officers need to know the institution’s lending policy and the guidelines. Consider, for example, the purpose of a loan for three different transactions: buy a house purchase, buy a car, and personal needs. The interest rate on a mortgage would be the lowest because the lender has a better chance of recovering the loan if the borrower were to default. In the case of personal needs, which are not specific, there is more risk or uncertainty of repayment and, thus, the interest rate would be the highest of the three.

Character

An assessment of an obligor’s credit history is necessary to determine willingness to repay a loan. A poor credit history is a good predictor of future repayment problems. Therefore, reputation for repaying debt is an essential part of character assessment. There is also name lending, which is lending based on a borrower’s social standing rather than the capacity to repay. The practice exists in mature markets but is prevalent in emerging markets. This is not to deny the fact that in emerging markets, borrowers of high social status usually have stronger capacity to repay. For a corporate borrower, the character assessment will include an assessment of the competence, capability, and the integrity or honesty of management. The analysis of leadership and reputation is qualitative and subjective, but the qualitative nature of the analysis makes it no less reliable or rigorous than the more quantitative and objective financial analysis.

How Important Is Character in Risk Rating?

Character assessment has been under closer scrutiny in the light of some recent financial scandals involving companies like Enron Corporation, WorldCom, Carillion Plc, and Nortel that made world headlines.⁸ The list of reported accounting scandals in the 2000s is the longest, making this period the worst for the notoriety in the last century.

The ease of obtaining character-related information depends partly on the type of borrower. For a personal borrower, the lender has more readily available information from a variety of sources that include credit reports, collection agencies, and media reports. For a public corporation or private company, character valuation requires assessing management conduct and ethics. If the borrower is a current customer or is new, the bank’s credit analyst obtains the information partly from call reports, face-to-face interviews with the principals of the company, meetings with lower level staff, and site visits. But in the absence of these means of gathering information, the analyst must rely on media reports and local market knowledge. For public companies listed on stock exchanges, character-type information is available from the Securities Commission of the country in question. In the United States, for example, the SEC (Securities Exchange Commission) Office of the Whistle-blower rewards eligible individuals for sharing original information that will lead to successful law enforcement actions.

Management can be a nebulous activity, with much of the information limited, qualitative, and hard to quantify, making Management Risk the most difficult of all Risk Criteria in a BRR (Borrower Risk Rating) scorecard to evaluate. A common remark that one hears is that evaluating management is a thankless job. Yet Management Risk cannot be swept under the rug because it subsequently surfaces in Financial Risk of the scorecard, but then it is too late because the lender had been making credit decisions based on an inflated Financial Risk rating and hence, an inflated BRR. The common thread in both risks is the problem of inadequate information, specifically the information asymmetry that occurs when the borrower possesses better information than the lender in a credit relationship. Thus, the fact that management can be vague is not a reason to ignore it in the scorecard. This book attempts to include information asymmetry explicitly in the BRR. The lender faces an information asymmetry problem stemming from the difficulty of performing an accurate assessment of the borrower’s creditworthiness. If the bank has the ability to obtain full information when loan applications are accepted, the assigned BRR will be a more accurate measure of the default risk and the lender will be able to minimise default risk. We will examine information asymmetry and the resulting problems later in this chapter.

Credit Risk

What do we mean by credit risk? It is the potential for a borrower or counterparty to fail to make full and timely payments of interest and principal. Credit risk consists of these two components.

1.

Default risk: measured by assessing the borrower’s capacity and willingness to service the debt under the terms of the loan agreement; and

2.

Loanrecovery prospects: The lender determines the expected loss based on the default risk (as a separate factor) and the loan structure and the value of the collateral held (as a separate factor).

The expected credit loss is the probability of default multiplied by the exposure at default, times the fraction of loss given default. The expected loss has a direct bearing on economic capital, which we will examine in Chap. 7. To convert loss expectation in percentile to a dollar amount, an estimate of the exposure at risk is needed; hence, the following equation.

where:

EAD = Exposure at default

LGD = Fraction of the loan amount lost given default (%)

PD = Probability of default (%)

The EAD is an estimate of the drawn amount, or the amount expected to be owed by a borrower at the time of default; stated another way, the EAD estimate is based on past utilisation of the undrawn credit and the possible future changes in that exposure due to the nature of the credit commitment before default. The Facility Risk Rating (FRR) reflects information that is specific to the transaction, so it includes collateral, guarantees, seniority, and maturity for each transaction or credit facility, such as a short-term revolving loan, a term loan, and a lease. Each FRR has a LGD rate calibrated against each other. The BRR reflects the credit creditworthiness of the borrower and takes into account the attributes that are specific to the obligor (and sometimes the guarantor of a borrower). The combination of PD risk and LGD risk gives the loan quality risk. Figure 1.2 illustrates low PD risk combining with low LGD risk translates into acceptable loan quality risk, and vice versa in quadrant III. Quadrant II and quadrant IV involve trade-off between PD and LGD. When PD Risk > LGD Risk as in quadrant II, which often happens as the BRR deteriorates over time, prudent banks ensure that the collateral remains solid with low LGD and legal enforceability. Lenders encounter this situation for borrowers that are on the Watch List or are non-performing. When LGD Risk > PD Risk as in quadrant IV, prudent lenders will ensure that the BRR is accurate so that the implied PD is also accurate. Banks cannot avoid quadrant III and they have a specialised group that manages only non-performing loans.

Fig. 1.2

Loan quality risk from combining PD and LGD

There are numerous studies on the negative relationship between credit ratings and probability of default, also known as the PD curve.⁹ The findings show a reasonably close correlation between the BRR and PD. Following the Basel II accord, banking institutions that implement the AIRB (Advanced Internal Rating-Based) approach, calibrate the BRR and the PD against each other—a statistical procedure that entails as much art as it does science.¹⁰ In general, the mapping depicts a common pattern: default rates start out low for the least risky grades and then rise rapidly as the grade worsens. This means that the PD increases non-linearly and monotonically as rating grades deteriorate from highest to lowest. For example, the average one-year default rate of say AAA (investment grade) may be 0%, compared to 20% for CC (non-investment grade).

Many banks report their PD-BRR mapping in their annual reports. Table 1.2 gives the PD-BRR calibration of the 22-point rating scale of Canada’s largest Bank, Royal Bank of Canada. It is a detailed mapping and, for this reason, serves as a perfect illustration of the non-linearity and the monotonicity. The better is the rating, the lower the PD; the worse the rating, the higher the PD. The PD values range from 0% for the highest grade to 100% for the lowest grade and the range subdivides into discrete bands, as they should for a one-on-one correspondence. Banks align their scale to the ratings used by the ratings agencies. The RBC Internal Rating Map is one such example where the bank uses S&P and Moody’s. The main reason they compare their internal ratings to external ratings is to determine whether they are too conservative or too liberal in their lending policy. If a bank is too conservative, it risks losing safe customers; if it is too liberal, it risks making loans to weak borrowers that are more likely to default.

Table 1.2

RBC internal ratings map

Source: 2018 Annual Report

An illustration of the EL equation for a given transaction is the following. Consider a weak BRR (e.g., CC), a PD of 75%, and an LGD of 40%, the latter representing the average percentage loss rate over time. The product of PD and LGD is the statistical or expected loss of 30%, which translates to 30 cents/1 dollar EAD. Expected loss is an increasing function of each of the three variables. As you can see in this example, the LGD or the LIED (Loss in the Event of Default) is always transaction specific, whereas the PD is borrower specific and independent of the LGD rate. This is the essence of a two-dimensional rating system of the AIRB Approach that grades the borrower and the credit facility separately rather than mixing them together.

In this book, we are concerned with the PD component of credit risk, so our focus is on the methods that credit analysts use to assign BRRs. We look at four types of borrowers:

1.

Business/Commercial: Commercial or business credit is used to fund capital expenditures and a firm’s day-to-day operations. Commercial borrowers are companies operating in the non-financial sector, such as agriculture, fishing, mining, mineral exploration, manufacturing, and services.

2.

Banks: Simply put, banks borrow money to make money in the form of loans. Thus, bank risk assessment adopts a different approach than the analysis of non-financial companies, though the basic mechanics of deriving a rating are the same for all borrowers. Spend a few minutes on a bank’s financial statements and you will see how different the line items are from those of non-financial firms even though the common language is accounting and the terms are the same, such as revenue, expense, assets, liabilities, and capital. The risks are different. Banks receive deposits (essentially borrowing from depositors) or borrow in the wholesale capital market, and recycle money to make money (loans and advances) on which the bank expects to make a profit margin over its funding cost. This operation involves numerous risks (which we will discuss in detail later).

3.

Commercial real estate (CRE): Commercial real estate comprises income-producing real estate (IPRE) assets used solely for business purposes, such as shopping centres, offices, apartments, motels, and hotels. The financing of these assets is CRE mortgages, which are loans secured by liens on the property. Thus, both the repayment and recovery of the loan depend primarily on the IPRE’s cash flows and secondarily on the owner of the asset or guarantors of the mortgage.

4.

Sovereign states: Cross-border lending has been a major activity by banks. Direct foreign investments by companies are essentially cross-border lending. Country risk analysis is used to assess the creditworthiness of sovereign or country debtors, and to obtain a more accurate BRR for private entities operating in the same jurisdiction. The country equivalents of a firm’s audited financial statements are national accounts statistics (gross domestic product, gross national income, prices, production, balance of payments, and government finance).

Monotonicity

The Basel 2 Accord makes the internal ratings a critical building block. Under the AIRB (Advanced Internal Rating Approach), both internal ratings and scores should map to a master scale of the default probability attached to each rating or score. The mapping implies that default probabilities of the master scale are monotonic functions of ratings. In general, a monotonic function is one that is consistently increasing and never decreasing, or consistently decreasing and never increasing in value. In the risk rating business, we call a forecast of grade-level default rates a PD (probability of default) curve. The reliability of the forecast requires monotonicity, which gives the rating system rank-ordering capability. If not we have a system that will forecast lower PD values as the BRR improves and reversions with even higher ratings.

Once the credit analysis of a borrower is complete, the rating model assigns a BRR based on the credit rating system. A credit rating system¹¹ must be capable of differentiating default risk effectively, a result achieved by the granularity of the rating scale (see Table 1.2). Insufficient grades in a rating scale severely limit a lender’s ability to differentiate default risk, quantify the profitability of a loan, and determine the amount of capital for the exposure (see the applications of the BRR in Sect. 1.6). That said, there is no gain in striving for finer rating scale because the data would be unavailable and because of the increased likelihood that the relationship between the numerous rating grades and the default probabilities will fail the test of monotonicity (more on this in Chap. 3). The rating scale of S&P, Moody’s, Fitch Group, and DBRS consists of 20 or more grades. As we noted earlier, banks align their internal BRRs to the credit ratings of the rating agencies for the loan portfolio of the firms that are externally rated. Both Risk Management and the business units use the information to determine whether the bank is losing creditworthy borrowers or encouraging and retaining less creditworthy borrower, or taking on excessive risk. One major criticism of rating comparison is that the external ratings used as a standard of sorts can be very unreliable in an ex ante or forward-looking sense. The most glaring example is the 2008 financial crisis where borrowers that were at the high end of the rating scale (even AAA) were downgraded to non-investment grade in just a short period of time after actual results contradicted the expectations of the rating agencies.

1.3.3 Loan Structuring

We think of loan structuring as a process to achieve two main goals (outputs). The first is to preserve the borrower’s ability to service the debt from a profitable operation. Repayments derive primarily from the cash flow and secondarily from collateral security. The second is to position the lender closer to the borrower’s assets than other creditors in the event of a liquidation. Banks prefer to lend only in the first position so that they have first rights on the borrower’s assets. Banks ensure or attempt to ensure that various payments (accrued expenses) are subordinated to bank debt. Hence, a lender must know a borrower’s capital structure—referring to the asset and liability sides of the balance sheet—to determine where and how the company funds its operations, and the rights of other creditors to the company’s cash flow and assets. Loan structuring is best understood as a process that involves all the following starting with the crucial ones.

Determining the purpose of the loan

Determining the amount of the loan

Matching the repayment schedule to the cash flow of the borrower. (Note: profit is not cash flow, which is the primary source of loan repayment)

Matching the term of the loan to the asset being financed (e.g., short-term funding for working capital; long-term loan for plant and equipment)

Setting interest rates appropriate to the credit risk and setting fees.

Drafting the Loan Agreement, the principal loan document

Determine the covenants, which are intended to ensure a loan gets repaid and to protect the lender

Deciding the need for a guarantor serving as a secondary or even tertiary repayment source (which means the lender must also perform a credit risk assessment on the guarantor to assign a guarantor BRR)

Deciding whether to take collateral and if so, what collateral is appropriate. (Note: Collateral is a secondary source of loan repayment)

Determining structural subordination. It refers to the positions of various creditors to the group’s assets in an event of insolvency. In lending to a holding company (holdco) especially, understanding the holdco’s structure is vital. A lender would want to ensure that in the event of a default, it has access to the cash flow and assets of the operating subsidiary or subsidiaries through upstream guarantees. In the absence of such guarantees, the claims of a lender of a holdco are structurally subordinated to the claims of the creditors of the operating companies (opcos) in the group. This is because the claim of the holdco, as a shareholder of the subsidiary, is subordinated to the claims of the other creditors of the opcos. In a bankruptcy scenario, a lender prefers to be at the top of the heap, or at least on an equal footing (pari passu) with other creditors for repayment. To ensure that lenders of the holdco rank equally with lenders of the opcos, the latter guarantee the liabilities of the holdco. The upstream guarantee offsets the structural subordination. (Note: In a downstream guarantee, the holdco guarantees the debts of the opcos.) There are other subordination mitigants like intercompany loans; however, this subject is outside the scope of this book.

Structural Subordination: Don’t Forget the Opco in the Credit Analysis

Banks do not like the idea that opco creditors are paid before holdco creditors. Whist upstream guarantees from the opco to the holdco make for a tighter loan structure, the primary source of repayment on a loan is cash flow and collateral is only secondary, which comes into play only in a default situation. Moreover, if the opco files for bankruptcy protection, a bank cannot realise on the collateral. Credit analysts must always examine the holdco’s assets and its ownership structure to ensure they include the opco in the credit analysis regardless of the upstream guarantees. Advice: get an up-to-date org chart as part of due diligence.

The Right Credit Exposure

How does a lender know the size of the loan is too much or too little? Lenders consider many factors but the major one is whether the borrower has the repayment capacity. This entails analysing the borrower’s capital structure, leverage ratios such as debt/capital, and debt-service ratios. We leave the details of this discussion for later chapters.

Loan Pricing

The interest rate on a loan reflects creditworthiness. Consider a ceteris paribus scenario and two borrowers with the same amount of loan and the same maturity. In this example, the creditworthy borrower is less likely to default, so the interest rate on his loan is expected to be lower. Loan pricing involves establishing the interest rate on a loan, and the way a bank set a rate is far from being formulaic. For this reason, loan pricing is a mystery to the public. These are some of the questions usually asked: Why so many rates? Why are some rates higher for some customers and lower for others? Why are some rates higher for some types of credit and lower for others?

Banks use a combination of loan pricing models and practices to set loan rates. Whilst the macroeconomic environment sets the overall level and direction of interest rates, banks have hurdle rates or target RAROC (risk-adjusted return on capital)¹² as guidelines on a case-by-case basis. The target rates help decide which loans meet the threshold for consideration. That said, inter-bank competition often forces a lender to charge interest that is lower than the hurdle rate. In this case, the rate serves as a loss leader for higher future returns. Let us look at three pricing models to get a basic idea of the factors determining loan rates:

Cost-plus loan pricing

Price leadership

Risk-based loan pricing

But let us first consider some common factors that determine loan pricing:

Relationship lending: Banks look at the value of the total client relationship. Long-standing and valuable clients tend to get favourable loan rates and fees when they do most or all of their banking transaction with one. As mentioned earlier, a common practice by banks is to use the loan interest rate as a sweetener to secure long-term rents on the relationship business.

The borrower’s size: There is evidence that interest charged by commercial banks on small businesses varies inversely with the size of the loan and therefore the size of the borrower; however, the relationship may be capturing differential credit risk premium as much as differential costs of loan servicing.

Loan type and purpose: There are different types of credit facilities: term loan, demand loan, revolving credit, standby letter of credit (LC), lease, and so on. The interest rates on these facilities are different reflecting the options available to the bank and the borrower for drawn down, payment, prepayment, and redrawing. There are also many loan purposes, including working capital, acquisition, debt repayment, stock buyback, confirmed LC (letter of credit) to support commercial paper (CP), leveraged buyout, and so on. The differences in the rates reflect default risks. For example, a loan for leveraged buyout raises the debt/equity ratio and makes the borrower riskier, although only for a time until the ratio returns to normal. The LC, on the other hand, does not increase the borrower’s debt/equity ratio, until and unless the borrower draws on the line, which could occur if the company cannot issue new CP.

Geographical location of the borrower: The location in which a borrower is based or does most its business is an important factor. This risk is usually factored into credit risk (e.g., country and sovereign risk).

Clearly, the interplay of the numerous factors determines loan pricing. Borrowers who understand this reality are better informed to shop for the best rate. Let us examine these pricing models to see what the main determinants are.

Cost-Based Loan Pricing

Cost-plus pricing involves four steps. First, the bank determines the average cost to attract deposits and to borrow. The funds could be retail deposits (e.g., individual savings accounts and time deposits, chequing accounts) and wholesale deposits, which are very large deposits made by a large business, another bank, and institutional investors. Second, the bank calculates the average cost of servicing the loan, which covers the fixed, administrative costs we discussed earlier. The administrative cost is expressed as a percentage of the loan. Finally, after the costs are covered, the bank adds a percentage figure representing compensation for risk taking. Fourth, the bank applies a profit margin. The pricing formula in percentage terms is:

Price Leadership

Recall that banks maximise not only profit but also growth, and in order to grow they must compete for market dominance; hence, there is a trade-off between maximising current profitability and maximising future growth. Competition in banking has always been intense and more so with non-bank entities encroaching on traditional