DO NOT COPY 199 © Retail Banking Academy, 2014 RETAIL BANKING I RETAIL BANKING ACADEMY Course Code 109 - Credit Loss Management 109. Credit Loss Management
DO NOT COPY
199© Retail Banking Academy, 2014
RETAIL BANKING I
RETAIL BANKINGACADEMYCourse Code 109 - Credit Loss Management
109.Credit Loss Management
DO NOT COPY
RETAIL BANKING I
200
RETAIL BANKINGACADEMY
Course Code 109Credit Loss Management
Introduction
A retail bank, by virtue of being a financial intermediary in that it takes in short-term deposits and makes long-term loans, takes on risk and hence can incur loss. The bank can account for such loss through the pricing of its products. However, when the actual loss at the end of the contractual period is higher than the level that was expected at the start of the transaction, then profitability suffers. This simple illustration of the risks involved in financial intermediation brings into focus the importance of loss management. Indeed, recent innovations in behaviour economics show that people are more concerned with loss avoidance than previously thought. This is the essence of prospect theory,* which states that people view losses with greater regret than the pleasure they experience with an equal amount of gains.
Losses Loom Larger
This is no different for a bank since people make decisions and they will be duly affected by loss aversion. More than this, unexpected loss has important consequences for profits and capital.
Before we discuss the details regarding credit loss management in retail banks, it is important to review the fundamental concepts of risk and uncertainty and the risks that a retail bank faces as identified by Basel III. This will set the stage for an analysis of credit risk management, the complex issue of provisioning and, finally, delinquencies and collections. We also consider sources of non-financial risk such as customer fraud that can lead to significant reputation risk. Legal risk can be harmful in lending since contracts between the bank (lender) and the customer (borrower) may fail to comply with charges on collateral, or may contain deficient documentation.
Risk, Uncertainty and Large Numbers in Consumer Banking
What is the meaning of risk?
First, consider the case of a retail bank making a loan to a customer. Before making a loan, the bank professional will try to estimate the likelihood that the borrower will not be able to make the loan payments on time and in full. Already we see that there are two issues here. The first is
* Prospect theory was developed by Daniel Kahneman and Amos Tyversky and published under the title, “An Analysis of Decision under Risk”. Econometrica, Vol 47, No 2., pp 263-292 (March 1979).
Course Code 109 - Credit Loss Management
DO NOT COPY
201© Retail Banking Academy, 2014
RETAIL BANKING I
RETAIL BANKINGACADEMY
how likely it is that the customer will not pay on time and in full and the second is how much money will be recovered should the customer fail to make payments.
To put this as simply as possible, there are two dimensions to consider.
• The first dimension is the knowledge of the probability of an outcome – the chance.
• The second dimension is the knowledge of the outcome should it occur – the consequences.
An event may or may not happen – but if it does happen it can create a loss, the true value of which will be known only at the end of the time period.
Is there a difference between risk and uncertainty? The answer can be obtained from an address made by Guy Debelle, Assistant Governor (Financial Markets).* (Parts of the address that are not central to understanding the difference between risk and uncertainty have been omitted here. Notice the difference between what is quantifiable and what is not.)
Risk was misassessed by financial institutions, risk managers, investors and regulators. There was a false comfort taken from a misplaced belief that risk was being accurately and appropriately measured. To some extent, the technology provided risk managers with a false sense of security. Risk may well have been accurately measured for the particular regime that the economy and financial markets were operating in. But the risk assessment was not robust to a regime change that took the models out of their historical comfort zone. Not enough account was taken of uncertainty.
One of the messages I want to leave you with is that risk measurement based on historical models can only take you so far. Judgment must play an important role. Ultimately, the future is uncertain, in the sense that it cannot be quantified. The goal should be to design systems that are as robust as possible to this uncertainty. A system with less leverage is one obvious means of enhancing robustness.
Risk versus Uncertainty
In discussing risk, I would like to highlight the key distinction between risk and uncertainty: risk is quantifiable, uncertainty is not.
This is a distinction with a long tradition. Keynes made much of it, as did Frank Knight who lends his name to ‘Knightian uncertainty’ on which he elaborated in Risk, Uncertainty and Profit in 1921. Knightian uncertainty arises when you don’t know the underlying probability distribution, which makes quantifying the risks impossible.
More recently, this distinction has been given prominence by Donald Rumsfeld with his knowns and unknowns, as well as by Satyajit Das in Traders, Guns and Money and Nassim Taleb in The Black Swan.
Conclusion
The argument I have been seeking to make today is that the misassessment of risk has been a key element of the financial crisis. One of the contributing factors to this misassessment was an over-reliance on a model-based approach to risk management, which focused too much on measurable risk without taking full enough account of immeasurable uncertainty
* Address to Risk Australia Conference, Sydney – 31 August 2010.
Course Code 109 - Credit Loss Management
DO NOT COPY
RETAIL BANKING I
202
RETAIL BANKINGACADEMY
Open Question #1
What is the difference between risk and uncertainty?
We end this introduction by reviewing the main categories of risk that a retail bank faces.
The Basel Committee on Banking Supervision (BCBS) categorises banking risks as follows.
Credit Risk
When a bank enters into a loan agreement, the bank lends the principal loan amount. The transfer of the funds from the bank to the customer occurs at the time the contract is signed. But the terms and conditions of the contract are defined for repayment over a future period of time. This creates a risk for the bank as lender. The terms and conditions of repayment may not be fulfilled as previously agreed. For example, the borrower may miss a scheduled periodic payment or may default altogether. This is a description of credit risk.
Credit risk also includes country risk. Country risk is the uncertainty created by unexpected political and economic events that makes it more difficult for the borrower to repay the bank. An important part of country risk is transfer risk. One example of transfer risk is when the government has imposed prohibitive exchange restrictions or capital controls that may make it impossible for borrowers to transfer payments to the lender.*
Associated with credit risk, is an important concept that is normally introduced in statistics – the law of large numbers.
The Law of Large Numbers†
This law has important applications in retail banking where there are a lot of transactions that occur over a short time period. In simple terms, it states that the larger the pool of loans, the more predictable the amount of losses will be in a given period. Alternatively, the larger the number of similar credit exposures that a retail bank takes, the more predictable will be the credit risk for the bank’s portfolio of loans. Since retail banking is a high volume business, the law of large numbers will play an important role in credit risk management.
How does this apply to the management of credit risk?
The bank can increase the predictability of loss on its loan portfolio by spreading the risk over many borrowers. This requires that the bank makes many small-sized loans over many customers – a typical situation in retail banking. The alternative is large-sized loans over a small number of customers – a typical situation in wholesale banking. The law of large numbers will predict lower variability in loan loss rates and so the actually experienced default rates will be close to the expected default rates for this type of business. For this law to apply in retail banking, we require borrowers to have similar characteristics (this shows the value of customer segmentation) and that loan loss rates among customers are independent.
While this chapter deals only with credit loss management and with credit risk, we briefly describe the other two classes of Basel risks.
* It seems that Basel II restricts country risk to being transfer risk. The Basel Committee explicitly includes transfer risk in the list of criteria on risk assessment of a borrower by stating that the risk characteristics of the country (the borrower is operating in), and the impact on the borrower’s ability to repay, (including transfer risk) where the borrower is located in another country and may not be able to obtain foreign currency to service its debt obligations (page 50 of Basel II Accord).
† Due to the eminent Swiss mathematician, Jakob Bernoulli (1655-1705).
Course Code 109 - Credit Loss Management
DO NOT COPY
203© Retail Banking Academy, 2014
RETAIL BANKING I
RETAIL BANKINGACADEMY
Market Risk
As stated above, retail banks may create assets by making loans and mortgages to consumers and companies and hence incur credit risk and associated country risk and transfer risk. But the bank may also invest in straight bonds, listed equities, mortgage-backed securities, covered bonds, etc., as well as conduct trading activities. In this way, the bank incurs market risk.
Simply put, market risk is defined as the risk that the value of the bank’s investment portfolio or the income stream will change due to changes in an underlying market factor such as prices of shares, interest rates, prices of commodities or foreign-currency exchange rates. This immediately implies that the principal sources of market risk are equity risk, interest-rate risk, commodity-price risk and foreign-currency risk.
Note that there are factors that may create both market risk and credit risk. For example, a change in interest rate may affect the default probability of an obligor as well affect the current market price of a bond.
Operational Risk
The risk of direct or indirect loss resulting from inadequate or failed internal processes, people and systems or from external events. It includes legal risk but excludes reputational and business/strategic risk.
Operational risk arises from banking operations that are banking processes managed by people on IT systems (including information databases) in order to deliver a financial service for customers. This definition identifies the sources of operational risk – people risk, process risk and systems risk.
Sources of Process Risk
1) Lack of internal controls – internal auditing is seen as a third line of defence after the initial checks by front-line managers and middle office risk and compliance professionals.
2) Inaccurate or incomplete documentation – leads to sloppy risk evaluation, leading to potential adverse selection of borrowers or unacceptable clients, e.g., money-laundering risk, as depositors.
Sources of People Risk
People risk arises from processing and reporting errors, i.e., human error and fraud. Human errors are inadvertent and different with respect to intent; fraud is a deliberate criminal activity that operates against the interest of the bank’s stakeholders.
Sources of Systems Risk
Systems risk arises from:
• Virus invasion of database or breach of security. These are system security problems (see the MasterCard International reference made above).
• There could be normal problems associated with fire or electrical lapses that lead to data corruption, service-delivery interruption.
• Software may be inflexible and be insufficiently tailored to the bank’s needs. This could lead to lack of scalability or improper functional support for the people and processes in the bank.
Course Code 109 - Credit Loss Management
DO NOT COPY
RETAIL BANKING I
204
RETAIL BANKINGACADEMY
The rest of this module is organised as follows:
Chapter 1 places credit risk management in a model of financial and non-financial risks where the latter category includes the important case of fraud risk.
Chapter 2 considers credit loss management along with an emphasis on delinquencies and collections.
Chapter 3 deals with the thorny issue of provisioning for credit losses and the debate surrounding the conflicting views of accountants and regulators.
This module concludes with a summary and review questions.
Course Code 109 - Credit Loss Management
DO NOT COPY
205© Retail Banking Academy, 2014
RETAIL BANKING I
RETAIL BANKINGACADEMY
Chapter 1: Credit Risk Management
The Basel Committee on Banking Standards (BCBS) defines credit risk as the potential that a borrower or counterparty with the bank fails to meet its obligations in accordance with agreed terms. There are important implications of this definition.
Clearly, there is always a chance that the bank’s customer or counterparty will not meet its obligations. This is the meaning of credit risk. Hence, there is always a chance that the bank will incur a ‘credit’ loss. But when does the loss occur? There are two ways that retail banks look at this question:
The first way is to assume that a credit loss has occurred only when the counterparty has actually defaulted on its payment. This is called the Default Mode Model (DMM). We present its fundamental details below and there is a fuller discussion in the Risk Management module of Retail Banking II.
The second way is to assume that a deterioration of the counterparty’s ability to make outstanding future payments is a signal of potential loss. This approach is embodied in the Mark to Market (MtM) approach. We present some details below and there is a larger discussion in the Risk Management module. At this point, we summarise the main features of our approach.
Course Code 109 - Credit Loss Management
DO NOT COPY
RETAIL BANKING I
206
RETAIL BANKINGACADEMY
Credit
Market
Operational
Financial
Non-Financial
Risk
s
DMM
MtM
Fraud/Legal Risk
109.1: Summary of Risk Management
Comment
We have only considered, to date, financial risks (credit, market and operational). But credit risks can also give rise to special underwriting risks such as incorrect choice of customers (this is called ‘adverse selection’) that can lead to instances of customer fraud and legal risk. These are labelled ‘non-financial risks’ and are considered below.
Non-Financial Risks
For the purposes of credit loss management, these risks include legal risk and consumer fraud. Legal risk occurs when there is violation of statutory and compliance requirements of a jurisdiction (e.g., a country). Examples are deficient documentation that makes the loan contract between the bank and the customer unenforceable. For example, current charges against collateral were missed in the risk evaluation phase.
The Institute of Internal Auditors’ International Professional Practices Framework (IPPF) defines fraud as:
“… Any illegal act characterised by deceit, concealment, or violation of trust. These acts are not dependent upon the threat of violence or physical force. Frauds are perpetrated by parties and organisations to obtain money, property, or services; to avoid payment or loss of services; or to secure personal or business advantage.
“Clearly the deliberate falsification of information on a loan contract in an effort to obtain money (through a loan) from a lender constitutes fraud. This is a very important source of non-financial risk that lenders face. One reason is that consumers can hide or modify their financial and demographic information so that banks incur information risk. Indeed, this could lead to the bank getting a too optimistic assessment of the customer’s creditworthiness.
“The Association of Certified Fraud Examiners’ 2010 Global Fraud Study found that the banking and financial services industry had the most cases across all industries – accounting for more than 16 percent of frauds.”
Models to Estimate Credit Risk
We now consider each of the credit risk approaches and discuss their respective implications for the calculation of credit loss.
Course Code 109 - Credit Loss Management
DO NOT COPY
207© Retail Banking Academy, 2014
RETAIL BANKING I
RETAIL BANKINGACADEMY
Default*
† Mode Model
This model assumes that there are two states – either the customer defaults or the customer does not default. Accordingly, the retail bank recognises a credit loss only when the customer defaults. This model does not recognise a potential deterioration of the customer’s creditworthiness probably due to financial distress. The expected credit loss (EL) is measured as the difference between the current value of the exposure that is recorded in the bank’s books and the present value of expected future net losses (i.e., recovery value less costs of recovery that include legal costs, expert’s fees in disposal of collateral etc.).
This is a forward-looking concept since it clearly depends on future recovery value that is uncertain at the time the calculation is made. The actual credit loss is ex-post in that it is calculated after the default has occurred – i.e., incurred cost. Chapter 3 deals with the debate between the incurred cost and the expected cost models of provisioning.
The default mode model proposes that expected credit loss be calculated in terms of the probability of default by the customer, the loss that the bank will suffer should the customer default and the amount of exposure that the bank has at default. Specifically,
expected credit loss is given by:
EL = PD * LGD *EAD where:
EL = expected credit loss
PD = probability of default (expressed in percentage terms)
LGD = loss given default (expressed in percentage terms)
EAD = exposure at default (expressed in monetary values)
We present further explanations as follows:
Recovery Rate: This is the percentage of the EAD that is recovered during a default. A related concept is loss given default (LGD) where LGD = 100 percent - Recovery Rate. So if the bank can recover 90 percent of its exposure at default, then the loss given default is 10 percent.
Probability of Default (PD): This is the average likelihood of default over the remaining period of the loan.
Example
Suppose that current loan exposure is £2.8 million; the probability of default (PD) is estimated by the risk management team to be 5 percent and the estimated recovery rate is 80 percent and hence the loss given default (LGD) = 20 percent.
Hence the expected loss = 0.05 x 0.20 x £2.8 million = £28,000
This means that the bank expects to lose £28,000 if the customer defaults and the estimated values for PD and LGD are correct. However, as is obvious, the expected credit loss can be different from the actual credit loss. Hence, the bank could experience unexpected credit loss. This is one of the main roles of capital in a bank. Capital provides a cushion to absorb unexpected credit loss.
† The Basel Committee (2004, page 92) states: “A default is considered to have occurred with regard to a particular obligor when either or both of the two following events have taken place: (a)The bank considers that the obligor is unlikely to pay its credit obligations to the banking group in full, without recourse by the bank to actions such as realising security (if held). (b) The obligor is past due more than 90 days on any material credit obligation to the banking group.”
Course Code 109 - Credit Loss Management
DO NOT COPY
RETAIL BANKING I
208
RETAIL BANKINGACADEMY
Comment
This example shows that:
The DM Model being a two-state paradigm – default and no default – the bank loses zero GBP if there is no default. If the customer defaults, then the bank expects to lose £560,000
While banks provision for expected credit loss, they provide capital for unexpected credit loss – that is, risk. Of course, expected loss can be priced into the cost of loans to consumers as would be the case in ‘risk-based pricing’ presented in Risk Management, a module in Retail Banking II.
We delve further into provision for expected loss versus incurred loss in Chapter 3. The Balance Sheet Management Module in Retail Banking II covers aspects of capital allocation to cover unexpected loss.
Comment
For practical purposes, the default mode model is appropriate for periods up to one year. The reason is that over longer periods, there is likely to be change in the creditworthiness of the customer or counterparty. In other words, credit risk (and risks in general) change over time. Since this model only recognises credit losses upon default, it is sensible not to apply it over a long period of time.
It is also for this reason that retail banks routinely employ risk professionals to monitor credit risk changes over time and thereby predict which loans are likely to become delinquent. This is a critical activity that is central to credit loss management.
Open Question #2
“Under the Default Mode model, if the planning period is one year, then credit losses from exposures beyond one year are ignored.”
If your bank utilises the DM model, how would you deal with this ‘exclusion’ problem?
The MtM Model
In this paradigm, the Mark to Market (MtM) assumes that the bank experiences a credit loss in two situations. First (as in the DM model) if the customer defaults, or, additionally, if there is a deterioration of the customer’s credit quality. This is called credit risk migration and it recognises the possibility that changes in the economic environment (such as recession, unemployment and inflation rates) can create changes in the customer’s creditworthiness.
In the next chapter, we analyse the behaviour of delinquent accounts over time using a net flow matrix.
Course Code 109 - Credit Loss Management
DO NOT COPY
209© Retail Banking Academy, 2014
RETAIL BANKING I
RETAIL BANKINGACADEMY
Chapter 2: Credit Loss Management
We begin with an important question regarding credit loss management. We stated earlier that the law of large numbers that is relevant to lending in retail banking states that current average credit losses are a good predictor of future average credit losses. Since the law of large numbers offers a good prediction of average portfolio credit losses, what are the sources of average portfolio credit loss for the bank?
This is an important question to answer since retail banks do incur, at times, relatively large credit losses arising from legal risk and consumer fraud.
There are several reasons for a departure from the predictions of the law of large numbers.
a) In the banking world, uncertainty arises because of our lack of knowledge in things over which we have limited control. If we lived in a laboratory and the conditions in which we worked remained fixed, then the behavioural relationship discussed above would indeed be unchanging – uncertainty would disappear. But obviously we don’t live in a laboratory and things change all the time. Sometimes we have an understanding of how these changes may affect our customers, sometimes not. Sometimes we are aware that changes have taken place; sometimes they happen without our knowledge.
This then is the biggest source of our uncertainty. We are simply not sure if the explanations that we have made of different behaviours are still valid. We know, for example, when the credit score was developed that the relationship between socio-economic profiles and default existed. However, over even the shortest period, that relationship can change and the model then doesn’t work. Industrial sectors become higher risk, economies go into recession or into boom, cultural habits change. They all have a measurable effect on the performance of our portfolio, and we have to try and identify and quantify that effect as soon as possible.
As stated in the introduction, the Assistant Governor of the Reserve Bank of Australia identifies quantifiable and non-quantifiable sources of risk.
We now consider the main elements of Credit Loss Management – dealing with delinquencies and collections and, importantly, post-contract risk monitoring in line with the MtM approach.
Course Code 109 - Credit Loss Management
DO NOT COPY
RETAIL BANKING I
210
RETAIL BANKINGACADEMY
First we introduce some key concepts.
a) Contractual Ageing
In consumer banking, the models are discussed with reference to a contractual ageing system. It is felt that statistically managed credit products require systematic accounting practices which reflect the risk inherent in an account. In the contractual ageing system the accounts flow through a series of monthly buckets until they reach a level of delinquency, at which point the account is written off.
A missed payment after 30 days causes the account to move into the 30-day bucket. If a partial payment had been received, the partial outstanding moves into the 30-day bucket. If an account is already in the 30-day bucket and the customer resumes paying a full instalment each month but always one month late, the account stays in the 30-day bucket. In the case of the matched-payments system, this account would become increasingly delinquent without necessarily becoming a greater risk. Accounts in the 30-day bucket that miss a second payment have the whole outstanding amount transferred to the 60-day bucket. The 60-day bucket is a measure of a higher level of risk, and it is reasonable to suppose that although only one payment is 60 days past due (DPD), the whole outstanding amount is at risk from default.
There are a number of issues that arise in association with this, and for which there is no right or wrong solution. The first relates to partial payments. This is when, of the given amount due, only a part is paid. If the system demands that the full instalment be paid then a payment of 99 percent of the instalment will cause an account to become delinquent. The overdue balance on this account however will be small. This may mean that an account gets routed to collections for the collection of a few cents. This may be considered a waste of time and might lead to inefficiencies in the collections area, because they have to process these accounts. One solution is to allow a tolerance, which means that if a sufficiently large percentage of the balance is paid then the account does not age. This will normally be based on no more than five to 10 percent of the instalment. Clearly, if smaller payments are made and are treated as a full payment there is a danger of understating the risk attached to the account.
b) Restructuring
For certain customers, depending on their situation, it may be considered necessary to offer some kind of forbearance to their credit programmes. Forbearance programmes have many different elements, depending on both product and customer. They may be payment ‘holidays’; payment of capital only; reduced interest rates or a number of other debt-relief options. The most common approach to debt relief, however, is to restructure the debt. This usually means that the tenor of the loan is extended so that the instalment amount is reduced.
The treatment of this process is crucial in providing an honest representation of the quality of a lender’s portfolio, and is often abused by lenders to disguise their problems. This means, in reality, that the possibility of restructuring accounts is made to customers who are already well behind with their contractual payments and represent a high level of risk. As a result of the restructuring process the delinquent account is closed and a new account is opened which, at the date of opening, is current. The result is that for a short time at least the lender has made the problem go away. The situation is exacerbated when collectors are offered the possibility of initiating the restructuring process, because it takes away the need to collect and it provides them with an easy alternative.
(c) Write-Off
In order for a business to state accurately its profits and assets, it needs to establish a rule under which accounts are considered to be either written off or fully provisioned. Rules and practices vary dramatically from country to country – and are variously influenced by international
Course Code 109 - Credit Loss Management
DO NOT COPY
211© Retail Banking Academy, 2014
RETAIL BANKING I
RETAIL BANKINGACADEMY
accounting standards (IFRS), tax rules and the requirements of banking regulators. In some countries, the standard rule for writing off an account that is not paying may be as long as seven years of non-payment. It may well have been made clear from the behaviour of the customer that the loan would have only a slight chance of being collected as early as six months after it went into default.
Fundamentally, bank accounting should be no different from the accounting applied in any industry. It should provide management and shareholders alike with useful information to judge performance, and should follow IFRS. Accounts prepared for regulatory and tax purposes should be entirely separate.*
Open Question #3
“Financial institutions can avoid high incidence of non performing loans, and costly loan write-offs when a thorough background investigation is conducted at the start of the loan review process*.”
Do you agree?
(d) Risk Dynamics
Risk changes over time. There are different reasons for this and it is important to understand and manage those different reasons. The first part to understand is that the risk represented by an individual customer changes the longer the customer has a relationship with the lender. For new customers the inability to recognise the potential for default means that there is inevitably a number of accounts that, because of uncertainty in the underwriting process, default at the start of their relationship. As the customers default, those that remain are relatively less risky and the default level falls.
The evolution of the defaults follows the shape of the graph underneath:
109.2: Evolution of defaults
Default continues among those customers that remain for a number of reasons: economic difficulties caused by an economy in recession; an individual employer who goes out of business or personal problems – health, family or other. The potential for a customer to default is always present.
However, if the shape of this default curve is universally true, which thus far seems to be the * I. Ajah, “Loan Fraud Detection And IT-Based Combat Strategies”, Journal of Internet Banking and Commerce, Vol 16, no 2 (August 2011).
Cumulative Default Rate
Time on Books
Course Code 109 - Credit Loss Management
DO NOT COPY
RETAIL BANKING I
212
RETAIL BANKINGACADEMY
case, then it has important implications both for periods of rapid growth in a portfolio or the launch of a portfolio, when there is a higher than average concentration of new accounts on the books. This gives rise to higher than average losses and has nothing to do with the quality of management of the portfolio.
A portfolio that stays the same size over the years – as the number of new accounts is offset by the loss of existing accounts – is said to be in a steady state.
To have a better understanding of the impact of new accounts or the effect of any defects in the credit-management process, we need to look at the way the level of delinquency in the whole portfolio evolves over time. There are a number of ways to do this but the simplest and the easiest to understand is the trend analysis of delinquency via the net flow matrix.
Understanding Delinquency Movements
It is common to report the status of a portfolio broken down into the different levels of delinquency as measured by contractual ageing. That is the number of days delinquent as measured against the schedule established when the agreement was signed. The amount reported is the ‘balance at risk’ (net of any unearned income).
A report of this kind will look like the one shown here. The numbers used are realistic for a mature portfolio. In the example, we have drawn a line across the table at 180 days past due which is the point at which the accounts are written off.
Write-off is often mistakenly understood to mean that there will be no future revenue from the account. In fact, in most markets there is still significant value in an account that has been written off. The debt for the customer still stands and there is still a legal obligation on the part of the customer to settle it.
The extent to which debts are collected depends very much on the local legal environment or the efforts of the collector. In countries such as Germany, where it is usual for banks to apply to the courts for a portion of the customer’s wages to be
assigned directly to the bank, the rate of recovery can be very high.
Where this legal support is not available, the proportion collected will depend on the skill of the collections department and usually varies between zero and 33 percent.
In most countries there is an active market for the sale of written-off debt, with prices varying up to a maximum of around eight percent of the face value of the debt. The logic for this maximum is that if the debt is recoverable at 33 percent then the cost of recovery will typically be about 25 percent.
In our example the delinquency may not be important in terms of the reported results of a business but for the credit-risk manager, understanding the evolution of delinquency beyond write-off is critical.
Changes between Months
If the delinquency figures for two consecutive months are written side by side, a direct comparison can be made between the two months. In the table shown you can see that not only are the receivables growing but so are the delinquent balances at each level of delinquency.
Unfortunately, the table doesn’t say anything about the movement of accounts between the two
109.3: Example of a risk portfolio
Jan 11DelinquencyCurrent 1,234,5671-29 dpd 49,38330-59 dpd 11,35860-89 dpd 4,99890-119 dpd 3,798120-149 dpd 3,342150-179 dpd 2,975180-209 dpd 2,796210-239 dpd 2,712240-269 dpd 2,658270-299 dpd 2,631300-329 dpd 2,605330-359 dpd 2,579
720 + dpd 2,566
Course Code 109 - Credit Loss Management
DO NOT COPY
213© Retail Banking Academy, 2014
RETAIL BANKING I
RETAIL BANKINGACADEMY
months, so accounts that are, for example, 1-29 dpd in February 2011, may have been in any non-written-off bucket in January.
The maximum an account can age between two months is 30 days but of course the customer has the option to pay off as much of the overdue balance as they can and become less delinquent – all the way back to current or anywhere in between.
However, experience tells us that the evolution of delinquency for accounts beyond 60 dpd is likely to be determined primarily by most customers making no payment and simply ageing to the next bucket.
It suits our purpose to make the simplifying assumption that all accounts arriving in a given delinquency bucket come from the previous (one lower) stage of delinquency. This assumed movement from month to month is called the ‘net flow’ of delinquent balances.
From the above table we will be saying that the flow from the 30-59 dpd bucket in January to the 60-89 dpd bucket in February is 11,358 > 5,111 or a flow of 45 percent.
With this simplifying assumption in place it becomes possible to analyse in detail what is happening to risk within a portfolio.
The Net Flow Matrix
Following this section there are two tables. The first is a delinquency matrix which shows the balances delinquent for a series of 12 months; the second is a net flow matrix which shows the transition to a state of delinquency in a given month from the state of lower delinquency in the previous month, under the assumption that accounts become only one step more delinquent every month or cure entirely (a simplification but adequate especially for the later buckets).
The net flows tell quite a detailed story about how the portfolio is being managed. First of all the evolution of an account through the different stages of delinquency gives the monthly gross loss rate (before recoveries). The simplest way of stating this is to say that the losses that arrive at 180 dpd in August 2011 were current in January of that year. So the percentage that become write-offs from that month is 3,785 / 1,234,567 or 0.306 percent. If this happens on each of 12 months of the year, then the annualised loss rate is 3.68 percent.
Another way of calculating the same number is to take the product of the monthly flows:
5% × 25% × 45% × 75% × 85% × 90% × 95% = 0.306%
The advantage of this way of looking at things is that it breaks the losses down into different stages of their evolution, and if there is a deterioration in any part of the process it can be quickly identified and remedial action taken.
If the losses were to deteriorate by 20 percent from 0.306 percent per month to 0.366 percent per month, the reason may not be obvious immediately. However if it can be seen that the flow rates have changed to those below:
5% × 25% × 54% × 45% × 75% × 85%× 90% × 95% = 0.366%
109.4: Example of a two-month risk portfoliio
Jan 11 Feb 11DelinquencyCurrent 1,234,567 1,296,2951-29 dpd 49,383 61,72830-59 dpd 11,358 12,34660-89 dpd 4,998 5,11190-119 dpd 3,798 3,748120-149 dpd 3,342 3,228150-179 dpd 2,975 3,008180-209 dpd 2,796 2,826210-239 dpd 2,712 2,740240-269 dpd 2,658 2,658270-299 dpd 2,631 2,605300-329 dpd 2,605 2,579330-359 dpd 2,579 2,553
720+ dpd 2,566 2,566
Course Code 109 - Credit Loss Management
DO NOT COPY
RETAIL BANKING I
214
RETAIL BANKINGACADEMY
It can be seen that the flow from the 30-day bucket to the 60-day bucket has increased from 45 percent to 54 percent, which accounts for the 20 percent increase in the gross losses. The problem lies in a specific part of the collections area. It could be due to staff shortages or some other such change in collections strategy but at least the problem has been located and can be treated.
The other major advantage of this kind of analysis is that it can be used to understand the true state of the portfolio without being misled by the impact of the effects of growth.
In our example we see the accountant’s view of the percentage of losses would simply be the balances written off divided by the sum of the non-written-off balances. This would be calculated for the month of August as follows: 3,785 /(1,737,160-3,785) = 0.002184 = 0.2184 percent or almost a third lower than the lagged value.
This is because the receivable that is used as the denominator refers to the month of the write-off. As the portfolio has been growing since the time when the newly written-off account was current, the denominator is higher and the ratio of losses is lower. Of course, if the situation is reversed then the denominator shrinks and the loss ratio becomes higher.
Given this discussion of using a net flow matrix to forecast delinquencies, collections become quite important.
Course Code 109 - Credit Loss Management
DO NOT COPY
215© Retail Banking Academy, 2014
RETAIL BANKING I
RETAIL BANKINGACADEMY
109.5: Example of a year-long risk portfolio
Jan
11Fe
b 11
M
ar 1
1Ap
r 11
May
11
June
11
July
11
Aug
11Se
p 11
Oct
11
Nov
11
Dec
11
Del
inqu
ency
Curr
ent
1,23
4,56
71,
296,
295
1,36
1,11
01,
429,
166
1,50
0,62
41,
575,
655
1,65
4,43
81,
737,
160
1,82
4,01
81,
915,
219
2,01
0,98
02,
111,
529
1-29
dpd
49,3
8361
,728
64,8
1568
,056
71,4
5875
,031
78,7
8382
,722
86,8
5891
,201
95,7
6110
0,54
930
-59
dpd
11,3
5312
,346
15,4
3216
,204
17,0
1417
,865
18,7
5819
,696
20,6
8021
,714
22,8
0023
,940
60-8
9 dp
d4,
998
5,11
15,
556
6,94
47,
292
7,65
68,
039
8,44
18,
863
9,30
69,
772
10,2
6090
-119
dpd
3,79
83,
748
3,83
34,
167
5,20
85,
469
5,74
26,
029
6,33
16,
647
6,98
07,
329
120-
149
dpd
3,34
23,
228
3,18
63,
258
3,54
24,
427
4,64
84,
881
5,12
55,
381
5,65
05,
933
150-
179
dpd
2,97
53,
008
2,90
62,
867
2,93
23,
187
3,98
44,
184
4,39
34,
612
4,84
35,
085
180-
209
dpd
2,79
62,
826
2,85
82,
760
2,72
42,
786
3,02
83,
785
3,97
44,
173
4,38
24,
601
210-
239
dpd
2,71
22,
740
2,76
92,
801
2,70
52,
669
2,73
02,
968
3,70
93,
895
4,09
04,
294
240-
269
dpd
2,65
82,
658
2,68
52,
714
2,74
52,
651
2,61
62,
676
2,90
83,
635
3,81
74,
008
270-
299
dpd
2,63
12,
605
2,60
52,
632
2,66
02,
690
2,59
82,
564
2,62
22,
850
3,56
33,
741
300-
329
dpd
2,60
52,
579
2,55
32,
553
2,57
92,
607
2,63
62,
546
2,51
22,
570
2,79
33,
491
330-
359
dpd
2,57
92,
553
2,52
72,
502
2,50
22,
528
2,55
42,
583
2,49
52,
462
2,51
82,
737
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
720
+ dp
d2,
566
2,56
62,
566
2,56
62,
566
2,56
62,
566
2,56
62,
566
2,56
62,
566
2,56
6
Jan
11Fe
b 11
M
ar 1
1Ap
r 11
May
11
June
11
July
11
Aug
11Se
p 11
Oct
11
Nov
11
Dec
11
1-29
dpd
5%5%
5%5%
5%5%
5%5%
5%5%
5%30
-59
dpd
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
60-8
9 dp
d45
%45
%45
%45
%45
%45
%45
%45
%45
%45
%45
%90
-119
dpd
75%
75%
75%
75%
75%
75%
75%
75%
75%
75%
75%
120-
149
dpd
85%
85%
85%
85%
85%
85%
85%
85%
85%
85%
85%
150-
179
dpd
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
90%
180-
209
dpd
95%
95%
95%
95%
95%
95%
95%
95%
95%
95%
95%
210-
239
dpd
98%
98%
98%
98%
98%
98%
98%
98%
98%
98%
98%
240-
269
dpd
98%
98%
98%
98%
98%
98%
98%
98%
98%
98%
98%
270-
299
dpd
98%
98%
98%
98%
98%
98%
98%
98%
98%
98%
98%
300-
329
dpd
98%
98%
98%
98%
98%
98%
98%
98%
98%
98%
98%
330-
359
dpd
98%
98%
98%
98%
98%
98%
98%
98%
98%
98%
98%
0
%
0
%
0
%
0
%
0
%
0
%
0
%
0
%
0
%
0
%
0
%
-
-
-
-
-
-
-
-
-
-
-
-72
0 +
dpd
2,56
62,
566
2,56
62,
566
2,56
62,
566
2,56
62,
566
2,56
62,
566
2,56
62,
566
Course Code 109 - Credit Loss Management
DO NOT COPY
RETAIL BANKING I
216
RETAIL BANKINGACADEMY
Collections
Many consumer banks have origins as commercial and savings banks, where collections don’t really exist. When a corporate client defaults this usually means the closure of the enterprise and the only actions are to realise the collateral. An individual, on the other hand, doesn’t close down. They continue to work, (or not) to earn and to spend money. There is often the possibility of recovering any debt owed.
This can be an expensive process involving repeated contacts, letters and visits before the negotiation results in either a settlement or a final credit loss.
A rule-of-thumb for some regulators is that on a written-off debt it is usually possible to collect about a third of the amount due. However, the cost of collecting it is about a quarter of the amount due making a net return of about eight percent on write-offs. The return on early collections is obviously much greater but unfortunately, the number of lenders willing to make the investment in collections is rare.
Lesson
An evaluation of the bank’s portfolio of written-off accounts can yield not only returns but also rehabilitated customers over a number of years.
Collections strategy and process is always a balancing act. On one side of the equation is the need to preserve a good customer relationship, with the possibility of increasing revenues and retaining customer loyalty and preventing attrition. On the other side are the costs incurred by the collection process through use of the bank’s resources, as well as its costs in incurring credit losses and having to hold provisions.
Having dealt with delinquencies we now come to underwriting issues. Underwriting risk is extremely important in credit loss management. Increasing the cost of borrowing to non-creditworthy customers is not a substitute for the proper selection of potential borrowers. We know that in investment theory the higher the risk the investor takes, the higher the expected return. This is true of consumer lending as well.
But the higher the expected return, the higher the chance of a large actual loss. In credit loss management, there is no substitute for creditworthy customers.
Open Question #4
Bank decision makers may decide to lower the cost of borrowing in order to attract credit- worthy customers. What is the nature (if any) of the underwriting risk that the bank will likely face?
Course Code 109 - Credit Loss Management
DO NOT COPY
217© Retail Banking Academy, 2014
RETAIL BANKING I
RETAIL BANKINGACADEMY
Open Question #5
Underwriting Risk: Failure to Recognise Fraud
“The first – and perhaps the one that gives rise to the most stories – is fraud. It may exist in the form of organised crime or it may simply be opportunistic individuals exploiting a weakness in the bank’s processes. Certainly, if the weaknesses are there, they will be found, and their existence will be communicated widely and quickly.
“Of course, recognising fraud is the first step to identifying its cause. While it is true that the standard credit-evaluation procedures will not typically identify a fraudulent application, it is the verification and collections processes that detect the fraud, either before the credit is granted or once it is issued.”
Do you agree?
Booking New Customers
It is important that each part of the credit process is adapted to handle the target market, the sales channels and the product characteristics.
Generally speaking, the higher the level of credit risk expected, the more detailed and exhaustive the credit analysis.
If a customer is already known to the bank and has a long history of performance, with an array of savings and credit products and their monthly salary is paid into the bank, the credit analysis for additional credit products such as a second credit card will not require much investigation or verification.
However, if a customer is unknown to the bank and comes simply to borrow money with no prior relationship with the bank and no salary lodgements, then it is likely that a more thorough credit assessment will have to be made.
In all cases, there are three elements to the credit decision. The first is the decision about whether the applicant can be offered credit. The second part is how much credit they can be offered, and the final part consists of verification of the customer to ensure that they are who they say they are, and they live and work at the places claimed, and that they can be contacted if need be.
A good deal of commercial banking concentrates on how much money can be lent to a company, however, the assessment of the character of management is also an essential part of the credit risk evaluation process. The existence and performance of companies is usually provided by accounts registered with a government authority, and the legal documentation is normally fairly comprehensive. Analysis of the accounts, balance and the balance sheet and the forecasted results of companies is the usual method of determining whether or not they can be granted a loan, and its size.
Individuals are different. They are not legally incorporated and do not provide annual accounts, indeed, they may well not keep an accurate record of either their expenses or their earnings, and so there is no possibility of carrying out the same kind of analysis as for a company. They also differ from companies in that they normally do not cease to exist because of poor economic performance.
Whereas companies that fail economically default and are forced to close by their creditors, individuals who get into difficulty continue to live on, and so does their obligation to repay their debts, unless of course they have declared themselves bankrupt, which is an increasingly available option in many countries. This can mean a recovery stream lasting many years. Typically,
Course Code 109 - Credit Loss Management
DO NOT COPY
RETAIL BANKING I
218
RETAIL BANKINGACADEMY
in the consumer-lending environment we are mostly considering unsecured loans.
For the most part this makes the underwriting ‘style’ of consumer lending very different from corporate lending. The emphasis is on trying to ascertain the stability and character of the individual borrower.
Indeed, any attempt to carry out an analysis of the balance sheet of the individual is fraught with difficulty. Income may only be known by reference to pay but many people have secondary income, which they are loath to disclose, and self-employed people will rarely give their total income for fear that it might be passed on to the tax authorities. Added to that, there can be added occasional income or regular bonuses, or even inherited income and other gifts.
If income is difficult to estimate then expenditure is even worse, since one person’s necessity is another’s luxury.
The analysis of character usually proceeds via the demographic information that is available from the application form. This will typically include not only personal information such as age, address, education, residential status etc. but also information about employment and income.
Normally this information can be used to describe credit policy, either in terms of rules, or used as data in order to make a credit score. The rules are often simple and exclusive, taking the form of; ‘do not accept applicants self-employed for less than two years’, ‘do not accept professions, a, b, c’. While such rules may be useful at times, they can be a disadvantage because they are rules about people to be rejected. They cannot be verified.
The alternative to a rule-based system is to use a credit score. Credit scores are developed by looking at the experience from previous clients that have been booked in the business. It gathers the information available to the lender at the time the decision to accept or reject the customer was made, and then looks at the performance at a later date (typically 12 to 18 months). Some of the customers will have defaulted and others will be within contract. The analysts then try to find if there was a difference in the application information available between those who have defaulted and those who have not. They use a range of statistical techniques to do this. The result is usually presented as a points system in which more points are associated with a lower probability of default, and each item of data on the application form can contribute positive or negative points. The points that they contribute are added together to make a total.
In this sense, the credit-scoring approach follows the way in which an individual underwriter may evaluate an application, taking everything into account but giving more positive weight to the details that they thought were important, or negative weight to things they thought were associated with risk.
So, for example, an applicant who is young, without permanent employment, without qualification, living in rented accommodation etc. would be considered high risk almost everywhere in the world. On the other hand, an applicant aged 40, who is qualified as a doctor and had been in the same job for the last 15 years, who is married with two children and owns their own house in an affluent part of town would almost certainly be considered low risk. Those two examples would be easy to explain but the scoring system does as well as possible with all the in-between cases.
The use of scoring systems is explained later in these notes but there are a couple of points that are worth remembering. The first is that the system depends on the past being the same as the future. Clearly in the real world this is never the case, but there has to be sufficient similarity for us to make reasonable inferences from the past. Secondly, any model that is developed is only as good as the data used to develop it. The old maxim applies.
Predictive Modelling – Scoring
It is generally believed that credit scoring had its origins in the Midwest of the United States
Course Code 109 - Credit Loss Management
DO NOT COPY
219© Retail Banking Academy, 2014
RETAIL BANKING I
RETAIL BANKINGACADEMY
during the Second World War. The recruitment of soldiers to fight in Europe left lenders without qualified staff to underwrite loans. The solution was to implement a credit-scoring system.
The pioneer credit-score company, FICO, was founded in 1956 as Fair, Isaac and Company by engineer Bill Fair and mathematician Earl Isaac. FICO was first headquartered in San Rafael, California. It sold its first credit-scoring system two years after the company’s creation, and sales of similar systems soon followed. In 1987, FICO went public. That year also saw the introduction of the first general-purpose FICO score when BEACON debuted at Equifax.
Originally, the models were mostly deployed in America, where Fair Isaacs developed its first application credit-scoring models in the early 1970s, but since the start of the 1980s they have been widely used in Europe. Now many of the major organisations accept or reject their new customers on the basis of an application credit score. There remains a wide divergence in the levels of development. Some businesses are still unaware of the importance of systematically recording and analysing the data concerning their decisions. They have been content to make the same mistakes repeatedly. Undoubtedly, under the strictures of increasing competition, they will be forced to reconsider.
The area in which statistical decision models have been mostly used is in the development of application credit scores. A credit score in this sense is simply a way of assessing the response that a customer has given to a series of questions when applying for a loan or banking product. Typically, a number of points are associated with each answer. The points are added up and if a required threshold, called the cut-off, is reached then the loan is booked. If not, the loan is refused. Their use has been widespread for the last 15 years. The major reason for their use has been the ability of institutions to reduce costs and undertake the underwriting process centrally, thereby cutting down on staffing levels, and at the same time satisfying the need to make consistent and accurate decisions.
We now consider the accounting treatment of credit losses through the concept of provisioning.
Course Code 109 - Credit Loss Management
DO NOT COPY
RETAIL BANKING I
220
RETAIL BANKINGACADEMY
Chapter 3: Accounting for Credit Loss
As we have seen, credit loss is a normal part of a bank’s lending business. Let us first outline some definitions:
1. Provisions are amounts set aside to cover known losses and liabilities at the balance sheet date. They are balance sheet items and are deducted from loans, which are assets. They are never reserves, unless overstated – when the amount of the overstatement might be described as a hidden reserve.
2. Reserves are part of shareholders’ capital. 3. Charges in the P & L (and the reverse) result from increases (or decreases) in the amount of
provisions.4. Counter-cyclical provisioning is a highly controversial matter in bank accounting, similar
to profit-smoothing and secret reserves. To accountants and company law specialists, it is designed to understate profits in good times and understate losses in bad times – so that banking regulators get a bit of added comfort. However, bank accounts, like those of all companies, are not prepared for banking regulators – but for shareholders and the public at large.
Critics argue that counter-cyclical provisioning will take us back to the dark ages of accounting (pre-1980) and will allow bank management to play games with profit calculation – making the industry less and less competitive. (They also say that banks already deliberately under-provide for bad debts so as to flatter their reported results).
To manage the effect of loan losses, banks create a provision for bad and doubtful debts on the asset side of the balance sheet. It also has an income statement effect that was illustrated in Module 102 (Retail Banking Overview) and it is suggested that many companies have a natural tendency to under-provision since this increases after-tax profit and so has a positive impact on dividend payout.
Course Code 109 - Credit Loss Management
DO NOT COPY
221© Retail Banking Academy, 2014
RETAIL BANKING I
RETAIL BANKINGACADEMY
Consider an example:
Assume that a bank makes a loan of €5 million to a business enterprise. The bank, after non-payment of an instalment, wants to make an additional specific provision for expected loan losses. It decides that a provision of €100,000 is appropriate based on the estimates provided by the risk-management group. Other things being equal, two accounting entries will occur. First, this loan loss provision of €100,000 will reduce net income by €100,000 – so debit the P&L account. Second, there will be an increase in the loan loss provision (LLP) on the balance sheet that will reduce the gross value of loans.
The Debate between Accountants and Regulators
Let us first consider when credit loss is incurred as prescribed by International Accounting Standards (IAS 39); IAS 39 states that a “loan [is] impaired if, on the basis of objective evidence, it is partly or wholly uncollectible, so that its carrying amount is greater than its estimated recoverable amount”. In other words, a bank can make a provision only if the loss is known to have occurred at the date of the balance sheet.
This is not as difficult as it seems but bank and accounting regulators are still arguing about the matter. The accounting standards bodies say that banks often deliberately underprovide so as to boost management bonuses – while banking regulators claim that accounting standards are allowing banks to overstate profits. The regulators’ response is to demand a new provisioning model called counter-cyclical provisioning. (The interesting question is, where are the auditors? )
What evidence will support provisioning according to IAS39?
The answer is default or probable default.
Comment
A provision can only be made when there is reasonable evidence of impairment of the loan – this is the accounting view.
As already mentioned, there are broadly two approaches to bad debt provisioning that are now being argued about by accountants and banking regulators. The first is the traditional incurred cost model (which is written into company law in many countries) while the second is based on expected losses. They reflect a tension between the accounting view and the bank regulatory view – but note that bank accounts are not prepared primarily for bank regulators. The accountant wants to provide a true and fair view of the financial condition of the bank at a point in time (the balance sheet date) while the regulator is ultimately concerned about the bank`s solvency. The obvious solution is separate regulatory accounting, as applies in the insurance industry.
Let us examine the implications of the incurred cost model. During an economic upturn, the frequency of default and consequently, the level of credit losses should decline. Hence provisioning for loan losses will be low and profits should rise. During a downturn in the economy, borrowers will default at a higher rate and the provisioning for credit losses will also rise. Incurred cost accounting reflects this reality, but critics say it makes a bad economic situation for the bank even worse.
This creates a wider swing in bank’s profits over the economic cycle and it is sometimes called ‘pro-cyclical provisioning’. Note that the cyclicality of bank lending (higher in upswings and lower in downturns) is also amplified. This is due to the fact that during economic up-cycles, collaterals that support loans typically increase in value and so lending increases. The opposite is the case during a downturn when the value of collateral assets will most likely decline.
Course Code 109 - Credit Loss Management
DO NOT COPY
RETAIL BANKING I
222
RETAIL BANKINGACADEMY
Comment
Procyclical Effects:
Other things being equal, this means fewer loan provisions (and possibly even write-backs in the profit and loss account) in upswings and significantly higher provisions and P & L charges during downturns.
Open Question #6
“Regulators criticise the incurred loss model for recognising credit losses too late”.
Do you agree?
The economic downturn reflects credit losses that originated in the period of economic upturn.
Hence, as asserted by proponents of dynamic provisioning, there should be higher provisioning in the economic upturn to counteract the higher losses that will be uncovered in the second period. This will have a smoothing effect on reported bank profits and is in essence ‘counter-cyclical provisioning’. This is preferred by regulators but criticism has emerged from accountants who point out that company law and accounting standards require that earnings must present a true and fair view of the financial position and cannot be manipulated.
Recent actions by bank regulators are moving towards provisioning that will tend to reduce the effects of procyclical provisioning. For example:
• Basel III imposes a countercyclical capital buffer that augments new capital requirements – (full discussion in Balance Sheet Management in Retail Banking II). It is important to note that this is a capital buffer (merely an allocation of shareholders’ funds), and not a provision and does not affect profit calculation.
• “The European Commission acknowledges that regulatory capital (again, this does not affect the P & L account and causes accountants no problem) covers unexpected losses whereas dynamic provisioning is intended to address expected losses.” *
We close this chapter with a recent view on the proposal to move towards dynamic provisioning. The debate is heating up. Here is a recent report that captures the opposing views:
“Sir David [Tweedie] who was chairman of the International Accounting Standards Board for 10 years until 2011, has attacked the Bank’s Financial Policy Committee (FPC) for concluding that accounting rules have allowed banks to ‘overstate profits’.
“He told the accounting sub-committee of the Parliamentary Commission on Banking Standards that the FPC was wrong to think the International Financial Reporting Standards (IFRS) prevented banks from making proper provisions against bad loans. “When Lord Lawson, chairman of the sub-committee, pointed out that the FPC was a ‘respected authority’, Sir David retorted: ‘It doesn’t mean they are right.’ “Andy Haldane, the head of financial stability at the bank, has said that the International Financial Reporting Standards (IFRS) have distorted bank accounts so badly that regulators need to devise an entirely separate set of rules for lenders. Mr Haldane has said that getting ‘an accurate view of a bank’s assets is like trying to pin a tail on a boisterous donkey’. Sir David, who was described by Lord Lawson as the ‘father of IFRS’ in the UK, said he did not agree. He said that a separate
* Dynamic Provisioning, a Price Waterhouse Cooper Report (April 2012).
Course Code 109 - Credit Loss Management
DO NOT COPY
223© Retail Banking Academy, 2014
RETAIL BANKING I
RETAIL BANKINGACADEMY
standard for banks ‘would cause more problems than it would solve’. He admitted that in 2008, the IASB recognised that banks needed to make more provisions. However, he insisted that it wasn’t the rules at fault but banks and their efforts to get around them.”*
Summary
This module began with a discussion of the difference between risk and uncertainty, and demonstrated the relevance of the law of large numbers in retail banking. The speech by the Deputy Governor of the Reserve Bank of Australia is important in this context. There was also a discussion of the BCBS risks – credit, market and operational risk. Chapter 1 considered credit risk as an important source of financial risk in retail banking, but also non-financial risks such as fraud risk and legal risk.
We presented two paradigms (Default Mode and Mark to Market) to estimate credit risk and hence, credit loss. The application of the DM paradigm shows the calculation of expected credit loss, while the MtM paradigm considers risk dynamics, and places emphasis on delinquencies, write-offs and collections. Finally, Chapter 3 dealt with accounting for credit loss. We considered the current debate between the accountant’s view and the regulator’s view of provisioning and pro-cyclicality.
* The Daily Telegraph, 29 January, 2013.
Course Code 109 - Credit Loss Management
DO NOT COPY
RETAIL BANKING I
224
RETAIL BANKINGACADEMY
Multiple Choice Questions
1. Which one of the following statements is consistent with the MtM (mark to market) approach?
a) The MtM tracks the flow of delinquent accounts through an analysis of a net flow matrix. b) The MtM provides an estimate of expected loss that is determined by probability of default, loss given default and exposure at default. c) The MtM assumes two states - default and no-default.d) The MtM is constrained for a period that is less than one year.
2. Which is incorrect?
a) Due to uncertainty in the underwriting of loans, new customers have a higher probability ofdefault at the start of the relationship with the bank.b) Cumulative default rates increase over time but at an increasing rate.c) A high growth of new accounts in the loan book may lead to a higher than average degree ofloan loss.d) Risk and (Knightian) uncertainty are not equivalent.
3. Expected Loss (EL) for credit risk is derived from the formula: EL = PD * LGD * EAD where PD = probability of default, LGD = loss given default and EAD = exposure at default.This means that recovery rate (RR) is given by which one of the following?
a) RR = 1-PDb) RR = 1-LGDc) RR = 1-EADd) RR = 1-EL
4. Which is correct?
There is evidence which shows that the average rate of return on collecting ‘bad loans’ for a typical retail bank is approximately:
a) 15 percentb) 8 percentc) 0 percentd) -3 percent
5. Which one of the following is not a principal source of market risk?
a) Equity risk b) Foreign currency risk c) Process risk d) Commodity price risk
6. Which one of the following statements is most consistent with the default model (DM) approach?
a) Migration of credit risk is recognised in this model.b) It is a two-state model where the customer defaults or does not default.c) It is consistent with the net flow matrix analysis of delinquent loans.d) It recognises credit losses that are incurred outside of the planning period.
Course Code 109 - Credit Loss Management
DO NOT COPY
225© Retail Banking Academy, 2014
RETAIL BANKING I
RETAIL BANKINGACADEMY
7. When the value of balances at 90 dpd in February is divided by the value of balances at 60 dpd in January the result is called:
a) The Probability of Defaultb) The 60 dpd to 90 dpd Net Flowc) The Markov Transition probabilityd) The Credit Score Bad Rate
8. Which of the following is an example of non-financial risk?
a) Credit riskb) Operational riskc) Market riskd) Legal risk
9. A net flow matrix of delinquent accounts show that there were 4,998 accounts in the 60 to 89 dpd bucket in January and 3,250 accounts in the 90-119 dpd bucket for February of the same year. This shows a flow of:
a) 54% b) 51% c) 60% d) 65%
10. Which of the following is incorrect?
a) An implication of the incurred loss model is that during an economic upturn, the frequency of default and, consequently, the level of credit losses should rise.b) Dynamic provisioning tends to smooth out reported earnings.c) Accountants view dynamic provisioning as earnings manipulation.d) An implication of the incurred loss model is that during an economic upturn, provisioning for loan losses will be low and profits should rise.
Answers:1 2 3 4 5 6 7 8 9 10
a b b b c b b d d a
Course Code 109 - Credit Loss Management