Acknowledgement

I would like to express my special thanks to my supervisor Mrs. Margaret Woods, who has given me strong support and encouragement during the whole research, and I am very appreciate of the expert guidance and inspiration she brought me.

I am very grateful to my parents for their love and encouragement during my whole education period. The academic suggestions my father has given help me a lot in designing the dissertation.

University of Nottingham

Last but not least, I would like to thank all my friends especially those in Melton Hall. I will never forget the help they have offered, which raises my confidence in completing this dissertation.

Credit Risk Management in Major British Banks

By Xiuzhu Zhao 2007

A Dissertation presented in part consideration for the degree of MA Finance and Investment

Abstract

Credit risk is always treated as the major risk inherent in a banks banking and trading activities. And if not well managed, this kind of risk may drag a bank into great trouble or even bankruptcy, which can be proved by various bank failure cases. For banks, managing credit risk is not a simple task since comprehensive considerations and practices are needed for identifying, measuring, controlling and minimizing credit risk.

In this dissertation, the credit risk management practices of major British banks are examined through the quantitative research on all Major British Banking Group members and qualitative analysis on the four sample banks. The key areas in the generalization and comparison of techniques and practices of sample banks are chosen according to the Basel (1999a, 2000) requirements, which are also adopted as benchmarks in the evaluation of banks credit risk management. The robustness and weakness of larger and smaller major British banks in managing credit risk are identified respectively, which indicates the areas for further improvements as well.

List of Contents

Acknowledgement............................................................................................................................................i Abstract...........................................................................................................................................................ii Chapter 1 Introduction....................................................................................................................................1 1.1 Introduction to Credit Risk................................................................................................................1 1.2 Motivation and Objectives of Study..................................................................................................1 1.3 Structure of the Contents...................................................................................................................2 Chapter 2 Literature Review...........................................................................................................................3 2.1 The Credit Risk of Banks..................................................................................................................3 2.1.1 Another Definition..................................................................................................................3 2.1.2 Categories of Credit Risk.......................................................................................................3 2.1.3 Identifying Credit Risk Exposures in Banks..........................................................................6 2.2 General Principles of Sound Credit Risk Management in Banking................................................10 2.2.1 The Goal of Credit Risk Management..................................................................................11 2.2.2 The Principles of Credit Risk Management.........................................................................11 2.3 Credit Risk Measurement................................................................................................................14 2.3.1 Fundamentals of Credit Risk Measurement.........................................................................14 2.3.2 Credit Risk Rating................................................................................................................15 2.3.3 Credit Scoring Systems........................................................................................................16 2.3.4 Credit Risk Modeling...........................................................................................................17 2.4 Credit Risk Mitigation and Transfer................................................................................................21 2.4.1 Traditional Methods for Controlling Credit Risk.................................................................22 2.4.2 Newer Methods for Credit Risk Transfer.............................................................................24 2.4.3 Banks Other Defenses against Credit Risk..........................................................................28 2.5 Issues in Credit Risk Management..................................................................................................29 2.5.1 Adverse Selection.................................................................................................................29 2.5.2 Moral Hazard........................................................................................................................29 2.5.3 Credit Risk Concentrations..................................................................................................31 Chapter 3 Methodology.................................................................................................................................32 3.1 Research Aims and Objectives........................................................................................................32 3.2 Research Design..............................................................................................................................32 3.3 Research Sample and Data Description..........................................................................................34 3.3.1 A Brief Introduction on the Structure of UK Banking..........................................................34 3.3.2 Research Sample Selection..................................................................................................35 3.3.3 Research Data Source and Description................................................................................36 3.4 Limitations of Methodology............................................................................................................40 3.4.1 Limitations on Sampling......................................................................................................40 3.4.2 Limitations on the Singleness of Research Approaches.......................................................41 3.4.3 Limitations on Ratio Selection and Availability..................................................................42 Chapter 4 Findings........................................................................................................................................43 4.1 The Overall Level of Major UK Banks Credit Exposure and Quality...........................................43 4.2 Credit Risk Management Techniques and Practices at RBS and Barclays......................................47

4.2.1 An Overview on the Royal Bank of Scotland Group and Barclays PLC.............................47

4.2.2 Credit Risk Management Techniques and Practices at RBS and Barclays...........................48 4.3 Generalization and Comparison with Basel Regulations................................................................64 4.3.1 The Establishment of Credit Risk Environment...................................................................64 4.3.2 Credit Granting Process........................................................................................................64 4.3.3 Credit Measurement and Monitor Process...........................................................................65 4.3.4 Controls over Credit Risk.....................................................................................................67 4.3.5 Credit Risk Mitigation..........................................................................................................67 4.4 Credit Risk Management Practices at Bradford & Bingley and Northern Rock.............................68 4.4.1 An Overview on B & B PLC and Northern Rock................................................................68 4.4.2 Credit Risk Management Practices at B & B and Northern Rock........................................69 4.5 Generalization and Comparison with Both Basel Regulations and Other Sample Banks...............73 4.5.1 The Establishment of Credit Risk Environment...................................................................73 4.5.2 Credit Granting Process........................................................................................................74 4.5.3 Credit Measurement and Monitor Process...........................................................................75 4.5.4 Controls over Credit Risk.....................................................................................................77 4.5.5 Credit Risk Mitigation..........................................................................................................77 4.5.6 Credit Disclosure..................................................................................................................78 Chapter 5 Conclusion....................................................................................................................................81 References.....................................................................................................................................................84 Appendix.......................................................................................................................................................92

List of Figures

2.1: CreditRisk CreditMetrics framework: The four building blocks..........................................................19 Figure 2.2: risk measurement framework........................................................................20 Figure 3.1: Number of banks and building societies 1985-2005 ..........................................................34 Figure 3.2: MBBG sterling lending to UK residents 1985-2004...........................................................35 Figure 4.1: Credit risk management organizations at RBS...................................................................49 Figure 4.2: Barclays Credit Risk Management Organization................................................................50 Figure 4.3: Distribution of credit risk assets by industry sector............................................................55 Figure 4.4: Distribution of credit assets by geography..........................................................................55 Figure 4.5: Distribution of credit risk assets by product and customer type.........................................56 Figure 4.6: RBS Distribution of credit risk assets by asset quality.......................................................58 Figure 4.7: B & B Credit Risk Management Organization...................................................................69 Figure 4.8: Northern Rock Credit Risk Management Organization......................................................70

List of Tables

Chapter 1 Introduction
1.1 Introduction to Credit Risk Table 2.1: Bands according to ratings from eligible rating agencies.....................................................16
Table 2.2: CreditPortfolioView-data input............................................................................................21 Table 4.1: Ratio analysis results (%) of 9 major UK listed banks in the last three years......................43 According to Duffie and Singleton (2003), credit risk can be defined as the risk of default or of Table 4.2: The six ratios (%) of the sample banks in the last three years..............................................45 reductions in market value caused by changes in the credit quality of issuers or counterparties. Table 4.3: An expression of RBS group level business. asset quality grades.......................................................54 Generally speaking, it is common in every Whenever a payment or performance to a Table 4.4: Maturity by analysis of loans and advances to customers.........................................................57 contractual agreement counterparty is expected, this risk exists. Conventionally, credit risk arises through lending, investing as well as credit granting activities and concerns the return of borrowed Table 4.5: RBS loans that are classified as REIL and PPL....................................................................59 money or the payment for sold goods. Besides, it also appears through the performance of Table 4.6: Total customer provisions balance........................................................................................60 counterparties in contractual agreements such as derivatives (Horcher 2005). Undoubtedly, when Table 4.7: Provision coverage ratios.....................................................................................................61 the obligation is not discharged completely, a loss occurs. Horcher (2005) suggests that when an Table 4.8: Impairment coverage of losses, PCRL.............................................................................................61 organization has accumulated large owes many other counterparties or when its creditors or Table 4.9: Summaries of credit risk management techniques and failure practices RBS and Barclays.......63 counterparties have financial difficulties or have failed, credit is at more likely. Table 4.10: Derivatives Use at RBS in 2006.........................................................................................67 Table 4.11: Carrying amount of securitized assets at Barclays in 2006................................................68 Table 4.12: Four risk categories of residential and commercial lending...............................................71 Table 4.13: Summaries of credit risk management techniques and practices at B & B and Northern Rock..................................................................................................................................73 Table 4.14: Comparisons and recommendations between the two groups of banks against Basel requirements......................................................................................................................80

1.2 Motivation and Objectives of Study Credit risk is one of the oldest and most important forms of risk faced by banks as financial intermediaries (Broll, Pausch and Welzel 2002). Since this risk carries the potential of wiping out enough of a banks capital to force it into bankruptcy, managing this kind of risk has always been one of the predominant challenges in running a bank (Broll, Pausch and Welzel, 2002). The goal of credit risk management is to achieve the maximum risk adjusted rate of return by identifying credit risk inherent in individual bank transactions as well as portfolios

and controlling the credit risk exposure to an acceptable level. And the effective management of credit risk is a critical component of a comprehensive approach to risk management and essential to the long-term success of any banking organization (Basel 1999a). Until now, some work has already been made to contribute to the literature on bank credit risk management, such as that of Nishiguchi, Kawai and Sazaki (1998), Hoggarth and Pain (2002) and Arsov and Gizycki (2003). However, all such work is focusing on part of the credit risk management in banks, like provisioning, credit derivatives use or credit risk measurement, instead of taking a view of the full picture. Besides, through searching the literature, it is found that most of the work is related to the performance of the US banks, which may be due to the better availability of data, while the practices of UK banks are seldom the focus. Therefore, this dissertation will focus on the credit risk management of major British banks, and an entire review of the management techniques and practices will be made.

1.3 Structure of the Contents

The structure of the following contents in this dissertation will be organized as follows. In Chapter 2, a comprehensive literature review will be made, which intends to cover most of the background information that will be needed for understanding banks credit risk management practices. The research aims, samples and methodologies will be introduced in Chapter 3. While Chapter 4 is for all the findings to be explained, including the quantitative analysis results and detailed review on how major British banks manage their credit risk. Eventually, a conclusion about the whole dissertation will be presented in Chapter 5.

Chapter 2 Literature Review

2.1 The Credit Risk of Banks

2.1.1 Another Definition

According to the Basel (1999a), credit risk is defined as the potential that a bank borrower or counterparty will fail to meet its obligations in accordance with agreed term. And the Monetary Authority of Singapore (2006) has defined it to be the risk arising from the uncertainty of an obligors ability to perform its contractual obligations, where the term obligor refers to any party that has either direct or indirect obligations under the contract. Regarding the importance of this kind of financial risk, Kaminsky and Reinhart, as cited by Jackson and Perraudin (1999), think of it to be the largest element of risk in the books of most banks and if not managed in a proper way, can weaken individual banks or even cause many episodes of financial instability by impacting the whole banking system. Thus to the banking sector, credit risk is definitely an inherent and crucial part.

2.1.2 Categories of Credit Risk

To gain a better understanding on the nature of credit risk, it is necessary to introduce the types of credit risk involved in financial activities before any further discussion. Concerning the categorizing of credit risk, different authors have expressed various criteria. For example, Hennie (2003) points out in his book that the three main types of credit risk are consumer risk, corporate risk and sovereign or country risk, while Culp and Neves (1998) consider realized default risk and resale risk to be the two types of credit risk. What is adopted here is part of

the views from Horcher (2005), who defines six types of credit risk, including default risk, counterparty pre-settlement risk, counterparty settlement risk, legal risk, country or sovereign risk and concentration risk. However, since legal risk is more likely to be considered as independent or belonging to operational risk nowadays (see HSBC 2006 annual report, Casu, Girardone and Molyneux 2006, etc) and concentration risk, together with adverse selection as well as moral hazard, is more reasonably to be thought of as an important issue in managing credit risk rather than a type of the risk itself (see Duffie and Singleton 2003), in the following illustration, only the rest four kinds of credit risk mentioned by Horcher (2005) will be touched upon.

Pre-settlement risk arises from the possibility that the counterparty will default once a contract has been entered into but a settlement still does not occur. During this period, a contract has unrealized gains, which indicates the risk. The potential loss to the organization depends on how market rates have changed since the establishment of the original contract, which can be evaluated in terms of current and potential exposure to the organization (Horcher 2005). 1

a. Default Risk

According to Horcher (2005), traditional credit risk relates to the default on a payment, especially lending or sales. And a likelihood of the default is called the probability of default. When a default occurs, the amount at risk may be as much as the whole liability, which can be recovered later, depending on factors like the creditors legal status. However, later collections are generally difficult or even impossible in that huge outstanding obligations or losses are usually the reasons why organizations fail.

b. Counterparty Pre-Settlement Risk

c. Counterparty Settlement Risk

According to Casu, Girardone and Molyneux (2006), settlement risk is a risk typically faced in the interbank market and it refers to the situation where one party to a contract fails to pay money or deliver assets to another party at the settlement time, which can be associated with any timing differences in settlement. Horcher (2005) points out that the risk is often related with foreign exchange trading, where payments in different money centers are not made simultaneously and volumes are huge. The case of the small German bank Bankhaus Herstatt, which received payments from its foreign exchange counterparties but had yet to make payments to counterparty financial institutions on the shutting down date, can serve as a typical example for the failure caused by settlement risk (Heffernan 1996).

d. Country or Sovereign Risk

Country risk arises due to the impact of deteriorating foreign economic, social and political conditions on overseas transactions and sovereign risk refers to the possibility that governments may enforce their authority to declare debt to external lenders void or modify the movements of profits, interest and capital under some economic or political pressure (Casu, Girardone and Molyneux 2006). Then as Horcher (2005) has concluded, since

evidence shows that countries and governments have temporarily or permanently imposed controls on capital, prevented cross-border payments and suspended debt repayments etc, problems arise for issuers to fulfill obligations in such environment. Also financial crisis may precipitate sometimes.

As explained by Horcher (2005), current exposure is the organizations exposure if the counterparty defaulted on its obligation at current market rates and potential exposure is an estimate of losses if a counterparty were to default under different rate scenarios.

Definition generated from http://www.standardchartered.com.hk/cb/invest//invest_debt.html

This section is based the consultative paper Principles for the Management of Credit Risk issued by Basel in 1999a.

According to Jackson (2001), there are two possible approaches for slotting interbank exposures. Loans to banks will be slotted according to the rating of their sovereign; and under option 2, according to the banks own rating. Exposures to borrowers without a credit rating will carry a 100% weight under unrated band.

CreditPortfolioView is first developed by Wilson (1987).

In practice, the seller bank usually provides no recourse to the buyer to show that the risk is totally shifted to the buyer, which allows the bank to legally remove the loan from the balance sheet.

Tier_1 capital includes the book value of common stock, non-cumulative perpetual preferred stock and public reserves from post-tax retained earnings, according to Basel (1988).
2

The discussion on both adverse selection and moral hazard are based on the relative parts in the book of Duffie and Singleton (2003).

According to Heffernan (1996), Toyo Sogo Bank ran into problems because of excessive exposure to a local shipbuilder; Johnson Matthey Bankers got into trouble because of a huge amount of bad loans to the third world countries; and the US commercial bank failure is largely because of a concentration of lending in the energy sector.

Calculated from the total assets of MBBG members, the average total assets are $906,800.6 million. Barclays, HBOS, HSBC and RBS have higher than average assets, while the rest five banks total assets are all lower than that. Therefore Barclays (total assets 996,787.0) and RBS (871,432.0) are chosen as representatives of larger banks and B & B (45,354.2) and Northern Rock (101,010.6) are chosen as smaller banks.
2

The description of the first four ratios is based on Abdus (2004) work.

Calculated as capital/(credit risk + market risk + operational risk) as required by Basel II.

All the summaries on the recommendations are based on the paper issued by Basel (2000).

The NPLGL ratio of Northern Rock cannot be found for the last three years, as shown in Appendix B.

1 2

This section is based on the information generated from company websites at http://www.rbs.com and http://www.aboutbarclays.com. Information regarding the Top 1000 World Banks 2007 is generated online at http://www.thebanker.com/news/fullstory.php/aid/5050/TOP_1000_World_Banks_07. html.

This section bases on RBS and Barclays 2006 annual report.

represents a probability of default at a point in time.

According to the notes in the annual report, non-accrual loans include all loans against which an impairment provision is held and PPL refers to loans for which an impairment event has occurred but no impairment provision is necessary, which are used for fully collateralized advances.

The RBS ratings are compared with the S & P equivalent while the Barclays one is said to have no more direct link with rating agency ratings.

Base on website information at http://www.bbg.co.uk and http://www.northernrock.co.uk.

Base on Bradford & Bingley and Northern Rock 2006 annual report.

2.1.3 Identifying Credit Risk Exposures in Banks

Generally, credit risk is related to the traditional bank lending activities, while it also comes from holding bonds and other securities. Basel (1999a) reports that for most banks, loans are the largest and most obvious source of credit risk; however, throughout the activities of a bank, which include in the banking book as well as in the trading book, and both on and off the balance sheet, there are also other sources of credit risk. Various financial instruments including acceptances, interbank transactions, financial futures, guarantees, etc increase banks credit risk. Therefore, it is indispensable to identify all the credit exposures-- the possible sources of credit risk for most banks, which can also serve as a starting point for the following parts of this work.

a. On-Balance Sheet Exposures

Loans

According to Saunders and Cornett (2006), the major types of bank loans are commercial and industrial (C&I), real estate, consumer and others. Commercial and industrial loans can be made for periods from a few weeks to several years for financing firms working capital needs or credit needs respectively. Real estate loans are primarily mortgage loans whose size, price

and maturity differ widely from C&I loans. Consumer loans refer to those such as personal and auto loans while the so called other loans include a wide variety of borrowers such as other banks, nonblank financial institutions and so on.

Credit risk is the predominant risk in bank loans. Over the decades the credit quality of many banks lending has attracted a large amount of attention. The only change is on the focus of the problems from bank loans to less developed countries and commercial real estate loans to auto loans as well as credit cards, which is an American example. Since the default risk is usually present to some degrees in all loans (Saunders and Cornett 2006), the individual loan and loan portfolio management is undoubtedly crucial in banks credit risk management.

Nonperforming Loan Portfolio

According to Hennie (2003), nonperforming loans are those not generating income, and loans are often treated as nonperforming when principal or interest is due and left unpaid for 90 days or more. Thus the nonperforming loan portfolio is a very important indication of the banks credit risk exposure and lending decisions quality.

Debt Securities

Besides lending, credit risk also exists in banks traditional area of debt securities investing. Debt securities are debt instruments in the form of bonds, notes, certificates of deposits, etc, which are issued by governments, quasi-government bodies or large corporations to raise capital.

1 In general, the issuer promises to pay coupon on regular basis through the life of the

instrument and the stated principal will be repaid at maturity time. However, the likelihood that the issuer will default always exists, resulting in the loss of interest or even the principal to banks, which can be a damaging impact.

b. Off-Balance Sheet Exposures

Since the 1980s, off-balance sheet commitments have grown rapidly in major banks, among which there are swaps, forward rate agreements, bankers acceptances, revolving underwriting facilities, etc. (Hull 1989). Those commitments give rise to new types of credit risk from the possibility of default by the counterparty. In this section, some of the off-balance sheet credit exposures will be introduced, among which the first one is related to derivative contracts.

Derivatives Contracts

According to Saunders and Cornett (2006), banks can be dealers of derivatives that act as counterparties in trades with customers for a fee. Contingent credit risk is quite likely to be present when banks expand their positions in derivative contracts. Since the counterparty may default on payment obligations to truncate current and future losses, risk will arise, which leaves the banks unhedged and having to substitute the contract at todays interest rates and prices. This is also more likely to happen when the banks are in the money and the counterparty is losing heavily on the contract. Comparatively, the type of credit (default) risk is more serious for forward contracts and swap contracts, which are nonstandard ones entered into bilaterally by negotiating parties. While trading in options, futures or other similar contracts may expose banks to lower credit risk since contracts are held directly with the

exchange and there are margining requirements. However, the credit risk is also not negligible.

Guarantees and Acceptances

Bank Guarantee is an undertaking from the bank which ensures that the liabilities of a debtor will be met, while a bankers acceptance is an obligation by a bank to pay the face value of a bill of exchange on maturity (Basel 1986). It is mentioned by Basel (1986) that since guarantees and acceptances are obligations to stand behind a third party, they should be treated as direct credit substitutes, whose credit risk is equivalent to that of a loan to the ultimate borrower or to the drawer of the instrument. In this sense, it is clear that there is a full risk exposure in these off balance sheet activities.

Interbank Transactions

Banks send the bulk of the wholesale dollar payments through wire transfer systems such as the Clearing House InterBank Payments System (CHIPS). The funds or payments messages sent on the CHIPS network within the day are provisional, which are only settled at the end of the day. Therefore, when a major fraud is discovered in a banks book during the day, which may cause an immediate shutting down, its counterparty bank will not receive the promised payments and may not be able to meet the payment commitments to other banks, leaving a serious plight. As pointed out by Saunders and Cornett (2006), the essential feature of the above kind of settlement risk in interbank transactions is that, banks are exposed to a within-day, or intraday, credit risk that does not appear on its balance sheet, which needs to

be carefully dealt with.

Loan Commitments

A loan commitment is a formal offer by a lending bank with the explicit terms under which it agrees to lend to a firm a certain maximum amount at given interest rate over a certain period of time. In this activity, contingent credit risk exists in setting the interest or formula rate on a loan commitment. According to Saunders and Cornett (2006), banks often add a risk premium based on its current assessment of the creditworthiness of the borrower, and then in the case that the borrowing firm gets into difficulty during the commitment period, the bank will be exposed to dramatic declines in borrower creditworthiness, since the premium is preset before the downgrade.

2.2 General Principles of Sound Credit Risk Management in Banking Hennie (2003) states that despite innovations in the financial services sector over the years, credit risk is still the major single cause of bank failures, for the reason that more than 80 percent of a banks balance sheet generally relates to this aspect of risk management. The consultative paper issued by Basel (1999a) also points out that the major cause of serious banking problems continues to be directly due to the loose credit standards for borrowers and counterparties, poor portfolio risk management and so on. All such evidence proves the extremely vital role credit risk management plays in the whole banking risk management approach as well as the sustainable success of the organization. In this section, the goal and principles of banking credit risk management will be summarized briefly, which together with

the above part on the identification of the existing credit risk in banking activities, will provide a basic framework for the understanding and discussion of banks

credit risk management practices.

2.2.1 The Goal of Credit Risk Management

The goal of credit risk management, as presented by the Basel (1999a), is to maximize a banks risk adjusted rate of return by maintaining credit risk exposure within acceptable parameters. Consistent with principles of managing portfolio, it is requested that both the credit risk arising from individual creditors or transactions and the risk of the entire portfolio should be managed, and the relationship between credit risk and others must be considered as well.

2.2.2 The Principles of Credit Risk Management1

To review the general principles of credit risk management can provide a clearer picture on how banks carry out their credit risk management, despite of the specific approaches that may differ among banks. According to Basel (1999a), the sound practices of bank credit risk management should cover the following four areas:

a. Establishing an Appropriate Credit Risk Environment

To establish an appropriate credit risk environment mainly depends on a clear identification of credit risk and the development of a comprehensive credit risk strategy as well as policies.

To banks, the identification of existing and potential credit risk inherent in the products they offer and the activities they engage in is a basis for an effective credit risk management, which requires a careful understanding of both the credit risk characteristics and their credit-granting activities, especially the complicated or newly developed ones. Besides, the design of objective credit risk strategies and policies that guide all credit-granting activities is also the cornerstone in bank credit risk management process. It is stated that a credit risk strategy should clarify the types of credit the bank is willing to grant and its target markets as well as the required characteristics of its credit portfolio. While credit policies express the bank

s credit risk management philosophy as well as the parameters within which credit risk is to be controlled, covering topics such as portfolio mix, price terms, rules on asset classification, etc (Hennie 2003). Both the strategies and policies should be designed and implemented well in conducting credit-granting activities, and they help to establish a beneficial credit environment. Moreover, establishing an appropriate credit environment also indicates the establishment of a good credit culture inside the bank, which is the implicit understanding among personnel about the lending environment and behaviour that are acceptable to the bank (Strischek 2002).

b. Operating under a Sound Credit Granting Process

A sound credit granting process requires the establishment of well-defined credit granting criteria as well as credit exposure limits in order to assess the creditworthiness of the obligors and to screen out the preferred ones. A banks credit criteria are designed to shape the types and characteristics of its preferred obligors, and they should set out who are eligible for the

credit, the amount of the credit and the relative terms and conditions (Monetary Authority of Singapore 2006). These criteria, together with the credit exposure limits on single and groups of counterparties that usually base on internal credit rating, should help banks to generate sufficient information on credit risk profiles and instruct the safe credit approval process, which are applicable to credit extension activities as well.

c. Maintaining an Appropriate Credit Administration, Measurement and Monitoring Process

Credit administration, as emphasized by Wesley (1993), can play a vital role in the success of a bank, since it is influential in building and maintaining a safe credit environment and usually saves the institution from lending sins. Therefore, banks should never neglect the effectiveness of their credit administration operations. Then talking about credit risk measurement in banks, it is required that banks should adopt effective methodologies for assessing the credit risk inherent both in the exposures to individual borrowers and credit portfolios, and this will be explained in details later. The last focus in this area of principles is related to credit risk monitoring, which is definitely a must in banks risk management procedure. Banks should keep track on the borrowers current financial conditions and ensure their compliance with the covenants. Both cash flows and collateral adequacy should be ensured and the potential problem credits should be considered. In this way, banks are well in control of their credit qualities as well as all the related situations, and can react to any future changes timely and readily.

d. Ensuring Adequate Controls over Credit Risk

The means for guaranteeing adequate controls over credit risk in banks lay in the establishment of different kinds of credit reviews. Regular credit reviews can verify the accordance between granted credits and the credit policies, and an independent judgement can be provided on the asset qualities.

2.3 Credit Risk Measurement

Measuring risk is always a crucial part in risk management process, and as suggested by Fabozzi (2006), quantifying credit risk can be complicated due to the lack of sufficient historical data, the diversity of involved borrowers and the variety in default causes. With the dramatic development of technology, credit risk measurement evolves greatly during the last 20 years. In the following, the fundamentals of credit risk measuring and the three categories of methods for bank credit risk measurement---credit rating, credit scoring and credit modeling will be explained.

2.3.1 Fundamentals of Credit Risk Measurement

Generally speaking, measuring risk is about trying to obtain some measures of the dispersion of possible future outcomes, and in practice, the focus is usually on the downside outcomes (Lowe 2002). The credit risk in banks should be measured by size as well as scope of the exposure, and as pointed out by Lowe (2002), all kinds of credit risk measuring approaches comprise of four common building blocks, including the probabilities of borrowers defaulting (PDs), the correlation of PDs across borrowers, the possible loss in the event of default (LGD)

and the correlation between PDs and LGD. Based on these elements, the approaches may differ in assumptions and modeling methodologies.

2.3.2 Credit Risk Rating

A credit rating is for assessing the creditworthiness of an individual or corporation to predict the probability of default, which is based on the financial history and current assets and liabilities of the subject. As mentioned by the Federal Reserve (1998), credit risk ratings may reflect not only the likelihood or severity of loss but also the variability of loss over time. For banks, both the internal credit rating and the external one are involved in their credit risk assessment.

a. Internal Credit Ratings

The internal credit ratings of banks, as suggested by Jacobson, Linde and Roszbach (2003), are the summary of the risk properties of the bank loan portfolio. They can be treated as monotonic transforms of the probability of default and shape the nature of credit decisions that banks make daily (Treacy and Carey 1998). A consistent and meaningful internal risk rating system can be a useful means for differentiating the degree of credit risk in loans and other sources of exposure (Basel 1999a). The internal credit ratings of banks are becoming increasingly important since the recommendations, as per the latest Basel II accord (Basel 2006), emphasize the adoption of robust internal credit rating system for risk assessment and buffer capital calculation, which will certainly encourage and lead banks to further development in this method.

b. External Credit Ratings

The external credit ratings are provided by credit rating agencies such as Moody, Standard & Poor and Fitch, and are a measure of the long-term fundamental strength of companies (Gonzalez et al. 2004). One noticeable issue is that credit rating agencies usually take a long-term perspective, which implies a lower sensitivity of their ratings to short-term fluctuations in credit quality, and rating migrations are triggered only by significant credit quality change (Altman and Rijken 2004). Despite of this, those ratings still play a key role in pricing credit risk. Since in the standardized approach to credit risk of Basel II accord (Basel 2006), banks are allowed to slot assets into weighting bands (see Table 2.1) according to ratings from eligible external agencies (Jackson 2001), it is quite possible that the future role of external ratings will keep on expanding.

Table 2.1: Bands according to ratings from eligible rating agencies (source: Jackson 2001)1

2.3.3 Credit Scoring Systems

Credit-scoring approaches, as stated by Reto (2003), can be found in virtually all types of

credit analysis and share the same concept with credit ratings. A credit scoring system determines points for each preidentified factor, which are combined to predict the loss probability and the recovery rate. According to Altman and Saunders (1998), there are two types of accounting based credit-scoring system in banks---univariate and multivariate. The first one can be used to compare various key accounting ratios of potential borrowers with industry or group norms while in the latter one, key accounting variables are combined and weighted for producing a credit risk score or a probability of default measure, which if higher that a benchmark, indicates a rejection to the loan applicant or a further scrutiny.

2.3.4 Credit Risk Modeling

According to Basel (1999b), credit risk models attempt to aid banks in quantifying, aggregating and managing credit risk across geographical and product lines, and the outputs can be very important to banks risk management as well as economic capital assignment. Those models, despite of the possible differences in assumptions, share the common purpose to forecast the probability distribution function of losses that may arise from a banks credit portfolio (Lopez and Saidenberg 1999). And regarding the potential benefits from the application of credit risk models in banking sectors, Basel (1999b) has concluded that they are responsive and informative tools offering banks a framework for examining credit risk in a timely manner, centralizing data on global exposures and analyzing marginal and absolute contributions to risk. According to Jackson, Nickell and Perraudin (1999), four types of models that are better known.

a. Merton-based Models

Merton-based models are also referred to as structural models. The basic principle of this category of models, as suggested by Merton (1974) first, is that a firm is considered to be in default when the value of its assets falls below that of its liabilities. Merton has modeled a firms asset value as lognormal process, with the equity modeled as a call option on the underlying assets, and the default is allowed at only a future time t (Arora, Bohn and Zhu 2005). The current value and the volatility of the firms assets, the outstanding debt and its maturity are required as inputs, from which the borrowers default probability can be determined (Hull, Nelken and White 2004). Based on the Merton model, Moodys KMV model has been developed for providing a term structure of default risk probabilities. The term Distance-to-default is proposed in the KMV model, which is calculated from the market assets value of the firm, the volatility as well as the default point term structure, and the model derives the actual probability of default---the Expected Default Frequency for each obligor instead of relying on the average historical transition frequencies produced by the rating agencies (Crouhy, Galai and Mark 2000).

b. Ratings-based Models

One of the most widely used ratings-base models is the CreditMetrics from JP Morgan. It is a tool for assessing portfolio risk that arises from changes in debt value caused by changes in obligor credit quality, and causes of the changes in debt value include possible default events and upgrades as well as downgrades in credit quality (JP Morgan 1997). According to Jackson, Nickell and Perraudin (1999), the obligor credit quality change probability can be expressed

as the probability of a standard normal variable falling between various critical values that are calculated from the borrower

s current credit rating and historical data of credit rating migrations. The measurement framework of CreditMetrics is illustrated in Figure 2.1. CreditMetrics suffers from a major problem, as pointed out by Crouhy, Galai and Mark (2000), that this approach relies on transition probabilities based on average historical frequencies of defaults and credit migration, which affects the accuracy of calculations.

Figure 2.1: CreditMetrics framework: The four building blocks (source: JP Morgan 1997)

c. Actuarial Models

The example of actuarial models is CreditRisk (see Figure 2.2 for the model framework), which predicts the loan loss distribution with the help of statistical techniques. In the model, borrowers are allocated among country-industry segments and default of individual loans is

assumed to follow a Poisson process. The development of a loan is understood as that either the obligor pays the amount due or the loan defaults, and no profits or losses from rating migrations have to be considered. As concluded by Gundlach and Lehrbass (2004), the attractiveness of CreditRisk are that it is more intuitive and better suited to practical needs. There is no requirement to provide the complete cash flow vector for each loan or to update spread curves for each rating class on regular basis.

Figure 2.2: CreditRisk risk measurement framework (source: Credit Suisse 1997)

d. Macroeconomic Models

Among macroeconomic models, CreditPortfolioView is most widely used, which is proposed by McKinsey.
1 based on the casual observation that default and migration probabilities are linked to the economy (Crouhy, Galai and It is a multi-factor model Mark 2000). The model measures only default risk, and the times series of default rates per industry sector (see Table 2.2 for an example) are required as the most important data input for simulating macroeconomics scenarios, which are used for estimating the conditional distribution of

default probabilities for individual credits (Kern and Rudolph 2001).

Table 2.2: CreditPortfolioView-data input: country-industry-sectors and the time series of sector-specific default rates (source: Kern and Rudolph 2001)

2.4 Credit Risk Mitigation and Transfer

The last step for any kind of risk management is to mitigate and transfer the risk in order to avoid or reduce losses. In this area, a variety of approaches are available and new methods keep on emerging. Generally speaking, the traditional methods for reducing credit risk focus on loan underwriting process and diversification, while the new means refer to asset securitization and hedging with credit derivatives (Neal 1996). In this section, all those essential ways for minimizing credit risk will be covered.

2.4.1 Traditional Methods for Controlling Credit Risk

a. Accurate Loan Pricing

One of the most obvious ways for minimizing credit risk, as mentioned by Heffernan (1996), is that banks should ensure the price of a loan exceed a risk adjusted rate, and include any loan administration costs. Basically, the risk premium is higher for riskier borrowers and the loan rate should keep changing with the alteration of the loan risk profile. However, adverse selection is a potential problem, in which case the higher loan rate actually implies higher default probability, and a later section will cover that topic.

b. Credit Rationing

The tenet of credit rationing, as concluded by Horcher (2005), is that credit granting favors the most attractive risk-to-return tradeoff. According to Mishkin (2004), credit rationing takes two forms in a bank. The bank may either refuse to make a loan to a borrower regardless of his/her willingness to pay a higher interest rate, or restrict the quantity of the credit, which can be much less than what the borrower want.

c. Credit Limits

The use of credit limits on both single and group obligors assists a careful counterparty selection in the risk management process, and the limits should be prepared for different products, activities as well as industries respectively. For example, they can be applied to the asset management in the form of exposure size limits while to trading activities in the form of

exposure position limits.

d. Collateral

The use of collateral to support various lending agreements for reducing credit risk has been adopted by banks for a long time already. It is applied not only to loans but also in other transactions such as derivative trading, where it works as an initial margin. However, the usefulness and importance of collateral use has actually been doubted. Wesley (1993) used to comment that collateral seldom provides a way out for a loan because when it matters most it has the least value. Since the value of collateral diminishes very quickly, it may not be able to generate enough cash flow for the debt as planned. Therefore, what really matters is actually whether the borrower has continuing access to sufficient cash flow when needed, not merely an existence of the collateral.

e. Diversification

Diversification is a very common concept in the area of risk management. For minimizing credit risk, diversification can be used to offset the additional volatility created from an increase in the number of risky loans. Seeking out assets whose yield returns are negatively correlated, banks can combine different types of loans into a portfolio and diversify away the non-systematic risk. Since the loss from defaulted loans will be offset by the earnings from other loans, the total probability for the bank to suffer a loss can be well reduced.

f. Netting Agreements

A netting agreement nets the amounts to be exchanged between counterparties, which reduce the credit exposure. For banks, netting agreements are mostly applied to interbank transactions, including bilateral payments netting, multilateral payment systems with net settlement and master derivative agreements (Emmons 1995). In a bilateral netting agreement, for instance, all payments for the given day are totaled and what the participants need to make is only the net payments. Emmons (1995) has commented that interbank netting agreements reduce interbank credit exposures and shift the default risk to bank creditors whose claims are not included in such netting agreements. Therefore, they are also useful tools for bank credit risk control.

2.4.2 Newer Methods for Credit Risk Transfer

The last two decades have witnessed the development of more efficient and less costly financial instruments, which make credit risk more manageable. Banks begin to pool assets with credit risk and sell parts of the pool; loan sales come into play; and credit derivatives are also gaining importance (Broll, Pausch and Welzel 2002). In the following part, those methods will be described individually.

a. Asset Securitization

Asset securitization is about turning traditional, non-marketed balance sheet assets into marketable ones and moving them off balance sheet (Twinn 1994). For a bank, securitization requires it to set aside a bunch of incoming-earning assets and to sell securities against those

assets in the open market, which means transforming loans into public traded securities in effect (Rose and Hudgins 2008). Many kinds of assets can be securitized including residential mortgages, commercial mortgages, credit card loans, trade receivables. Asset securitizations can improve credit risk management because they help to diversify a bank

s credit risk exposure and reduce the need for monitoring each individual loan

s condition. For instance, if a bank finds its lending too concentrated in a given sector, it can securitize some of lending to reduce exposures.

However, there are also two important problems with asset securitization. The first one involves the Special Purpose Entities (SPE), which are set up by banks for separating and securitizing loans. As pointed out by Tavakoli (2003), the SPEs are off-balance sheet and bankruptcy remote, which facilitate the risk management of banking community, but can also be used for illegitimate uses, such as embezzlement and mischaracterizing revenues and losses. Just as he mentioned, the SPEs are ideal for both securitizing assets and hiding assets, since any legitimate means can be exploited for illegitimate gain. Good news is that ever after the Enron case, much more attention is already given to this issue and it should always be treated as a focus of the regulators. The second problem about asset securitization is that banks which are securitizing their assets for transferring credit risk may also purchase the assets-backed securities issued by other banks. Due to the fact that those purchased securities may be riskier than the issued ones, the extent to which credit exposures can be reduced by asset securitization can actually be quite ambiguous for banks to tell.

b. Loan Sales

Rather than being collateral in the securitization, bank loans themselves can be sold in entirety to a new owner, and such a bank loan sale occurs when a bank originates a loan and sells it either with or without recourse to an outside buyer (Saunders and Cornett 2006). 1 According to Haubrich and Thomson (1993), there are two types of loan sales, the participation and the assignment. In the participation, the original contract between the borrower and the bank remains in place, and the bank still keeps on collecting payments, overseeing the collateral as well as examining the books. While in the assignment, which is a less common type, the debtor-creditor relationship is transferred to the buyer by allowing the buyer to take direct action against the borrower.

Loan sales have existed for many years already and their use has been increasingly recognized as a valuable tool in a bank managers portfolio of credit risk management techniques. Neal (1996) thinks the strategy of loan sales is attractive to banks since a fee from the loan origination is earned but the credit risk is assumed by the new investor. And in the occasion where the banks lend large amounts in a single takeover, loan sales are especially important for controlling credit risk.

c. Credit Derivatives

The continuous development of credit risk transfer techniques brought credit derivatives several years ago, which recently have gained importance rapidly in situations where the

diversity of loans and credit risk makes it difficult to carry out securitizations or sell loans individually (Broll, Pausch and Welzel 2002). They are contractual agreements based on credit performance that mainly refers to predetermined events such as default, nonpayment of loan obligations, downgrading and insolvency. According to Neal (1996), credit derivatives can help banks to manage the credit risk by insuring against adverse movements in the credit quality of the borrowers, and the major types of credit derivatives are credit default swaps, credit options and credit-linked notes.

Credit Default Swaps

Credit default swaps (CDS) are the most common and liquid credit derivatives, which together with the related products, account for roughly half of the credit derivative market today (Duffee and Zhou 2001). A CDS is a bilateral contract that offers protection against the risk from a particular credit event, under which the buyer makes periodic payments to the seller until the occurrence of a credit event or the maturity while the seller makes a contingent payment to compensate the buyer if the defined event occurs. Blanco, Brennan and Marsh (2003) point out that the economic effect of a CDS is like that of an insurance contract, but the difference is that it is not necessary to hold an insured asset for claiming compensation under a CDS.

Credit Options

Credit options are credit-spread derivatives based on the interest rate differences between the debts of different types of issuers (Horcher 2005). For example, in a credit spread call option,

the buyer pays premium to the seller in exchange for a contingent payment when the credit spread widens past a certain level, thus a bank that worries about the increasing risk on a loan can purchase such an option for hedging purposes.

Credit-linked Notes

A credit-linked note is a kind of debt instrument with an embedded credit derivative, which can therefore be treated as a combination of a bond and a credit option. It is the credit derivative suitable for debt issuers to hedge against credit risk, which reduced payments for the issuer when a specified credit event happens.

The reason for the wide acceptance of credit derivatives in controlling credit risk is that these instruments provide a mechanism permitting the transfer of unwanted risk from organizations with too much or wrong types of credit risk to the willing counterparty, as concluded by Horcher (2005). Besides, credit derivatives facilitate a portfolio approach to credit risk management and diversification since banks can enter into a derivative contract to transfer some credit risk in order to reduce the risk in their portfolios.

2.4.3 Banks Other Defenses against Credit Risk

When all the approaches for managing credit risk fail in a bank, its capital will form the ultimate defense against risk. According to Rose and Hudgins (2008), the capital acts as the last line of defense against failure and absorbs losses from bad loans to keep the operating of the bank. Regarding this, the Basel (2006) has defined comprehensive formula for banks to calculate minimum capital requirements against credit risk, which safeguard the whole

banking sector. It is required that the total capital ratio of a bank should be no less than 8% with the tier_1 ratio no less that 4% to cover the potential credit, market and operational risk, in which credit risk is the primary focus.

1 adequate level of capital, which is important for risk management purposes And by meeting the Basel requirements, banks can maintain an and the achievement of financial stability.

2.5 Issues in Credit Risk Management

2.5.1 Adverse Selection2

It is often the case that in the lending process, a borrower knows more than the bank about his/her own credit risk and the bank, being at an information disadvantage, may attempt to increase a borrowers interest rate for compensating the unknown credit risk. The problem is that the higher interest rate does not prevent riskier borrowers but those with less probability to default, and it is suggested that the more effective way be used to limit access to credit instead. However, what is likewise noticeable is that the quantitative credit exposure limits also deserves careful consideration. Since a smaller limit reduces lending volume as well as profits while a larger one encourages borrowers with low credit quality, the choice of an optimal limit is an important task for banks.

2.5.2 Moral Hazard

It is said that if you owe your bank $100 million that you dont have, your bank is in big

trouble

. For banks, large loan are considered riskier than small ones because they provide incentives for borrowers to undertake riskier behaviors. Also, large borrowers who default have stronger bargaining power, which puts lending bank in a worse position. Therefore banks always have their own limits on credit for certain counterparties, to deal with moral hazard.

Another interesting aspect with moral hazard in the banking system involves central banking, since it is argued that central banks help will encourage banks to engage in riskier behaviors. The most recent example may be the one about the crisis with American subprime mortgage-backed securities in the last few weeks. As mortgage lending institutions, hedge funds and many banks are seriously affected, worldwide central banks made quick reaction for maintaining the financial stability and public confidence, among which European Central Bank provided $ 131 billion of extra funds to the money market (Anonymous 2007a) and the Federal Reserve cut the rate for emergency lending to banks from 6.25% to 5.75%, with a lengthening term to 30 days (Anonymous 2007b). However, whether such central bank bailout is correct is quite arguable. It is mentioned by Grauwe (2007) that central banks should provide liquidity only against good assets and let the banks which loaned massive amounts of money to hedge funds get their punishment or even collapse. If the irresponsible banks can get away with it cheaply, they will be incited to take such activities again. Therefore, moral hazard at this level may be a dilemma, in which banks need to make decisions considering both current and future financial stability.

2.5.3 Credit Risk Concentrations

Concentrations, as pointed out by Basel (1999a), are probably the single most important cause of major credit problems. They are regarded as any exposure where the potential losses are large relative to the banks capital and are quite common in the banking sector. The reason is that banks usually cannot avoid specializing in certain industries or geographical areas due to the convenience for collecting information and the benefits of being more knowledgeable as well as better able to predict defaults of the targets they are familiar with. However, by doing this, banks should also bear the cost of charge-offs, nonperforming asset and strict reserve requirements. In fact, concentration is the major cause of bank failures due to credit risk management problems, as concluded by Heffernan (1996), which is proved by many examples such as the cases of Japanese Toyo Sogo Bank, Johnson Matthey Bankers in the UK the US commercial bank failures. 1

Chapter 3 Methodology

3.1 Research Aims and Objectives

Generally, the research aims for mapping a picture of the current credit risk management practices of major British listed banks, and the focuses can be divided into five research questions, including:

Have credit exposure and lending decision quality changed significantly among major British banks in the last three years?

What techniques do major British banks adopt to manage their credit risk currently?

What are the differences between the practices of larger and smaller banks?

Have major British banks current practices satisfied the requirements of Basel (1999a, 2000) on credit risk management and disclosure?

What are the recommendations for improvements in major British banks credit risk management?

3.2 Research Design Due to the aims and objectives of this research, a combined methodology of both quantitative and qualitative approaches is applied. The quantitative approach in this research is mainly for answering the first research question. From the research aims it can be seen that the focus is on reviewing the credit risk management techniques and practices among major UK banks,

therefore this quantitative part of research is actually only serving as a preliminary stage work for making the whole dissertation a little more comprehensive. In order to get a general understanding of major British banks

credit risk management level through comparison of their credit exposures and lending decision quality performance in the last three years, ratio analysis will be adopted, together with the method of Oneway Anova. No very complicated analysis will be used in this part, and the major hypothesis for testing is whether there exists significant difference among the three years

credit risk exposure and quality level of major British banks.

As already mentioned, qualitative study part will be given more attention in this research, which provides an easy understanding of the analysis of words and images rather than numbers (Hammerskey 1992). This part can actually be viewed as a collective case study, which, as defined by Silverman (2005), means a number of cases are studied in order to investigate some general phenomenon. As to the techniques chosen for this research, textual analysis is employed and a comparison method is also adopted. It is believed that these methods are suitable for the research because through analyzing the relative bank performance reports, access to all the available information needed in the research can be gained and assessment can be made, which provides both a separate evaluation and generalization on the sample banks credit risk management practices. Those outcomes will be the findings about the second research question and the basis of the rest of questions. While comparisons, as already implied in the research questions, are obviously necessary means for answering the last three research questions, from which the possible suggestions and recommendations can also be drawn. In fact, although it is not the focus of this research, comparison also exists in

descriptions of the two banks in each group, so that redundant explanations about their credit risk management can be avoided, and additional findings may be generated.

3.3 Research Sample and Data Description

3.3.1 A Brief Introduction on the Structure of UK Banking

Compared with many other countries, the UK has a smaller number of banks which is attributable to the large scale of consolidation in the domestic market. The change of the numbers of banks in the UK during the last 20 years is shown in Figure 3.1.

Figure 3.1: Number of banks and building societies 1985-2005 (Source: Casu, Girardone and Molyneux, 2006)

In the UK, the investment banking industry, according to Casu, Girardone and Molyneux (2006), is dominated by major US and European banks which focus on wholesale foreign currency activities. While the major players in retail banking industry are the Major British Banking Groups (MBBG), together with the mutual building societies. The following diagram shows the make up of MBBG lending business, from which great growth in domestic

mortgage lending and low level lending to manufactory sector are indicated.

Figure 3.2: MBBG sterling lending to UK residents 1985-2004 (Source: Casu, Girardone and Molyneux, 2006)

3.3.2 Research Sample Selection

MBBG, according to Casu, Girardone and Molyneux (2006), refers to Abbey National, Alliance & Leicester, Barclays, Bradford & Bingley, HBOS, HSBC Bank, Lloyds TSB, Northern Rock and the Royal Bank of Scotland (RBS). Those listed retail banks dominate the sterling-denominated banking business and are the focuses of this research and will all be included in the quantitative research part. While by reading through the relative information about their credit risk management approaches, it is found that some of the banks dont differ much in the practices and can present a whole picture on the standard practices, which indicates the possibility and reasonableness of conducting the qualitative research only on part of the MBBG members. Regarding this, two groups of samples are chosen for the qualitative research. By using average total assets of the nine banks as the grouping criteria, RBS and Barclays, which have higher than average total assets, have been put into Group One (larger

banks) while Bradford & Bingley and Northern Rock with lower than average total assets are selected as Group Two (smaller banks).

Judging from the amount of their total assets, it is confirmed 1 that the two groups have presented a relatively large gap in size. Besides, one issue that is worth noticing is that among MBBG members, four were mutual building societies which converted to banks, including Abbey National (converted in 1989), Alliance & Leicester (1997), Bradford & Bingley (2000) and Northern Rock (1997). Taking this into consideration, since Group Two is just composed of two those former building societies, it is hoped that the formation of the two groups will be more reasonable and representative.

3.3.3 Research Data Source and Description

a. Quantitative Data

The quantitative data that will be needed are six financial ratios relating mainly to banks credit exposure and lending decision quality. The choice of these ratios is according to Zhangs (2006) work, in which the author has based the method on four ratios Abdus (2004) has used and two Basel (1998) has recommended. The description of the six ratios is as follows. 2

Non-performing loans to gross loan ratio (NPLGL) = impaired loans/gross loans

This ratio is for assessing the quality of banks loans because it can measure the percentage of

doubtful loans in banks

credit portfolio. The credit performance of the banks is moving adversely with this ratio.

Loan Loss Reserve to impaired loan ratio (LSRIL) = loan loss reserve/total impaired loans

This ratio measures the proportion of loan loss reserve held against banks non-performing loans. Obviously, the higher this ratio, the better is the quality of banks loans and the more confident the banks can be about their assets.

Net charge-off to gross loans ratio (NCOGL) = charge off excluding recoveries/gross loans

Since this ratio measures the percentage of the amount written off from the gross loans, a lower ratio indicates that s smaller part of todays gross loans are finally written off the book.

Loan loss reserve to gross loan ratio (LSRGL) = loan loss reserve/gross loans

The percentage of total loan portfolio that has been set aside but not charged off is indicated from this ratio, and a lower ratio is the symbol of better loan portfolio quality.

Total Capital Ratio (TCR) = Total capital/Risk-adjusted assets 1

This ratio measures the quality of a banks capital and is required by Basel to be no less than 8%. The higher the ratio, the more adequate the banks capital and the better the asset quality will be.

Tier 1 capital ratio (Tier_1) = Tier 1 capital/risk adjusted assets

This ratio is for measuring the capability for tier 1 capital to absorb bank losses, and 4% is the minimum level required by Basel.

Those ratios are obtained from Bank Scope, which is on-line database offering financial information for 27600 banks around the world. For the quantitative research aim, three years (2004- 2006) ratios of the nine MBBG banks are used.

b. Qualitative Data

The qualitative data needed in this research mainly comes from the four banks 2006 annual reports and other published results. After examining all the reports, it is confirmed that they all have independent review sections on credit risk management. Those reports outline items such as the banks credit risk management policies, structures, qualitative and quantitative measures for risk control together with all the relevant figures, thus they are the basis of the whole research. Besides, the consultative papers issued by Basel (1999a, 2000) also play an important role due to their updated and detailed requirements on banks credit risk management, which will serve as the benchmarks in achieving the research objectives. Between those papers, the one issued in 1999 has shown the principles on credit risk management, which is already treated of above. As regards with the paper in 2000, a brief explanation will be given below.

Best Practices for Credit Risk Disclosure

The Basel paper issued in 2000 on best provides sufficient guidance for banks to reveal information that can facilitate meaningful assessments of their credit risk profiles. And that is

the reason why the contents in this paper are important and meaningful to this research. As to the main recommendations on credit risk disclosure, there are three aspects that will be involved in the following chapters.

Recommendations on Credit Exposures1

The recommendations on Credit exposures put emphasis on three areas---segment exposure information, credit concentration and risk mitigation techniques. Among them, segment exposure information requires credit exposures in banks be categorized according to business line, types of counterparties and geographical areas respectively; information about significant credit exposure concentrations should be revealed; while the disclosure about credit mitigation techniques includes the use and effect of collateral, guarantees, credit insurance, legally enforceable netting agreements, credit derivatives, securitization and recourse transactions.

Recommendations on Credit Quality

Concerning credit quality, it is suggested that information about a banks internal rating process and certain problematic assets conditions be shown. Amounts and changes in allowances should be revealed and the credit exposures on which the contractual cash flows have ceased are also subject to disclosure.

Recommendations on Credit Risk Management

The recommendations on credit risk management base on and further the above two aspects. They point out all essential parts in a bank

s credit risk management. The suggested disclosures include those on the nature and sources of credit risk in banking activities, the structure of credit risk management function, control policies as well as practices, techniques for managing past due and impaired assets and the use of credit scoring as well as portfolio credit risk measurement models. By following all of those recommendations, a picture can be drawn on a bank

s credit risk management process, which allows both clear understanding and evaluation.

3.4 Limitations of Methodology

3.4.1 Limitations on Sampling

Sampling in qualitative research can be a hard issue, and it represents the major limitation in this research. Due to the time and information availability restrictions, only four UK banks are chosen, which is really a small sample size, despite of the fact that the focus of this research is only on nine major British banks in total. It is reasonable that choosing some of the banks from MBBG can already help to draw a general picture of British retail banks credit risk management practices, since the well developed levels of western banks on risk management indicates no big variance in their individual practices from the standard systems. However, the representativeness of the samples is unavoidably a problem.

Besides, the grouping strategy of the sample banks also suffers from flaws. According to Seale et al. (2004), the sample procedure adopted in this dissertation can be categorized as a

purposive one, which means sampling according to certain characters or extreme cases in order to have all the possible situations. And the samples in this research are chosen according to size difference, combining with the special feature among MBBG members that some of them are converted from mutual building societies to enhance the reasonableness, however, problems arise. It is verified that both Bradford & Bingley and Northern Rock hold assets that are much less that the other two sample banks, so comparatively they do represent the smaller sized banks. Nevertheless, if another criteria--the FTSE 100 and FTSE 250 firm constitutes is adopted, it will be found that only Bradford & Bingley is listed in the FTSE 250 index as a mid-capitalized company while the rest are all in FTSE 100 as most highly capitalized companies, including Northern Rock. Therefore, the size consideration cannot be treated as totally convincing, which also has weakened the representativeness of the samples.

3.4.2 Limitations on the Singleness of Research Approaches

This dissertation, as mentioned above, will adopt mainly qualitative approaches for achieving the research objectives. But it should also be pointed out that the quantitative approaches on measuring banks credit performance, by comparing certain ratios, are also acceptable and meaningful, which can be shown from works of Abdus (2004), Peura and Jolivuolle (2004), etc. In this research, quantitative approach is adopted only for answering the first research question, but not used much in the analysis of the two groups of British banks. The reason why such techniques are not chosen lies behind the fact that the comparing of ratios wont be quite significant or representative, considering the situation in this research. On the one hand, the comparison of ratios between group members is not necessary, since the aim is on generalizing

common practices; on the other hand, the comparison of ratios between groups also doesn

t make much sense, because the sample size is too small.

3.4.3 Limitations on Ratio Selection and Availability

Even the six ratios selected for the quantitative research are widely accepted as related to banks credit exposure and quality level and are already used in other literatures, their representativeness cannot be guaranteed. After all, it is difficult to find out that how many ratios will be enough for providing a comprehensive measure, or whether the ratios chosen in this research are the most important ones related to the management of credit risk. Besides, not all the six ratios are available for the target banks, which is also a problem of insufficient data.

Chapter 4 Findings

4.1 The Overall

Sig

NPLGL LSRIL NCOGL LSRGL TCR Tier_1 2004 1.5167 65.1400 0.5350 0.9271 12.0125 7.8625 2005 1.5413 53.8625 0.4500 0.7600 11.9778 8.0778 Level of Major UK Banks Credit Exposure and Quality 2006 1.4825 55.4750 0.3200 0.7400 11.9111 8.1000 0.975 0.706 0.73 0.746 0.979 0.762

In order to understand the level of credit risk management among the nine major British listed banks, the comparison of six relevant financial ratios is made among three years, and the major results from the oneway anova are shown in Table 4.1, including the means and statistical significance level (see Appendix A for the full results). The null hypothesis is that the mean values for different ratios do not show significant difference respectively during the last three years.

Table 4.1: Ratio analysis results (%) of 9 major UK listed banks in the last three years

So far from the results, it can generally be concluded that each of the six ratios doesnt vary significantly during the last three years, which is drawn from the fact that all the significance value is greater than 0.005 and the null hypothesis cannot be rejected. Therefore, this result partly indicates a steady credit risk exposure and quality level of the nine MBBG banks. Despite of this stability, certain changing trends can still be inferred by comparing the means of the ratios in each year, which will be explained below.

Firstly, the non-performing loan to gross loan ratio is decreasing from 2005 to 2006 for major

British banks, which is a good sign of loan quality improvement in those banks.

1 By analyzing the ratios of each bank individually (see Appendix B), it is found that most banks have presented smaller NPLGL ratio in 2006 than the other two years, while Bradford & Bingley has in contrast a deteriorating NPLGL ratio in the last three years from 0.72% in 2004 to 1.53% in 2006, which indicates the weakness for B & B in controlling bad loans.

Secondly, regarding net charge-off to gross loan ratio, the nine banks generally have shown a satisfactory decreasing trend from 0.535% in 2004 to 0.32% in 2006, which means the loans that are finally written off the book are possessing smaller proportions of the gross loans year by year. Thus, an indication of banks better asset quality can be made in terms of NCOGL ratio.

Thirdly, considering the two capital adequacy ratios of the last three years, all the nine major banks perform steadily well. Although the average total capital ratio is experiencing a slight decrease from 12.0125% in 2004 to 11.9111% in 2006, it is still nearly 50% higher than the Basel requirements. While for the tier_1 capital ratio, a pleasant increase in the last three years can be found, which shows the ratio is maintained about 100% higher than the Basel requirements. Both of the two ratio analysis is evidence of major British banks good performance in keeping adequate capital, which indicates their strong ability to absorb losses and the confidence in assets quality.

Finally, coming to the loan loss reserve to impaired loans and gross loans respectively, a generally decreasing trend is found by comparing the means of the ratios in last three years.

Although the difference is not significant when P value is considered, it still means the nine banks on average are keeping less loan loss reserves to impaired and gross loans now. By studying the nine banks one by one, it is found that seven of them are following this trend (Abbey National are generally keeping an increasing amount of reserves in 2005 and 2006 and there is no relevant data about Northern Rock). Conclusion cannot be simply made since the phenomenon of banks setting aside less money to cover bad loans may be a sign for weakening credit risk management but may also indicate an improvement in default control. Judging from the fact that all the nine banks are writing more loans year by year from 2004 to 2006 (see Appendix C), the explanation that decreasing LSRIL and LSRGL ratios for major British banks indicate improvements in credit exposure control make more sense in this situation.

Bank LSRGL Group 1 RBS 0.84 LSRIL 62.06 NPLGL 1.35 NCOGL .

TCR 11.7

Tier_1 7.5

Year 2006

Group 2

RBS 0.92 65.42 RBS 1.08 70.31 Barcla 1.17 74.96 ys Barcla 1.26 75.74 ys Barcla 1.01 76.02 ys Bradfor 0.14 8.96 d& Bingley

1.41 1.53 1.56 1.67 1.33 1.53

0.46 . . . 0.50 0.02

11.7 11.7 11.7 11.3 11.8 13.2

7.6 7.0 7.7 7.0 7.1 7.6

2005 2004 2006 2005 2004 2006

Bradfor 0.16 11.52 1.35 0.00 13.4 7.8 d& Bingley Bradfor 0.20 27.83 0.72 0.00 13.2 7.5 d& Bingley Northe 0.15 . . . 11.6 8.5 rn Rock Considering the target sample banks in this research, a little more analysis should be applied to the 0.18 . . 12.3 7.7 four banks in particular, besides theNorthe above overall analysis. And .the six ratios of these four banks in the last three years are shown in thern following table. Rock Northe 0.23 . . . 13.5 8.0 rn Rock

2005 2004 2006 2005 2004

Table 4.2: The six ratios (%) of the sample banks in the last three years (source: Bankscope)

From the above table, it can be found that the four banks have followed the similar decreasing trend in the LSRGL (last two years) and LSRIL ratios performance, on which the indications are already inferred in the above paragraph. The decreasing NPLGL ratios of RBS and Barclays (Group 1) imply improved control of nonperforming loans, while B & B obviously haven

t done well enough in the last three years, with deteriorating NPLGL ratios. It is confirmed that the four sample banks all meet the Basel II requirements on capital adequacy ratios. The last point that is worth mentioning is about the missing ratios among these four banks. There are no LSRIL, NPLGL and NCOGL ratios for Northern Rock, and the NCOGL ratios of the Group 1 banks are not complete. This may be due to the problems of Bankscope, however, it is also a possible explanation that those banks haven

t disclosed relevant information for such ratios calculation, which then indicate the area for improvement in credit disclosure.

To sum up, through the quantitative research on the nine major British banks, the conclusion that those banks have all maintained a steady level of credit exposure and quality can be drawn. Besides, the ratios have in general implied slight but satisfactory improvement in banks credit quality in the last three years despite of certain fluctuations. The extra analysis on the four sample banks ratios have indicated that they generally accord with the results gained from overall analysis, while one obvious problem lies on B & B, which have done a comparatively bad job in managing nonperforming loans. Also, it should not be ignored that some of the ratios are not included for certain banks in the analysis, which may have impact on the convincingness of the results.

4.2 Credit Risk Management Techniques and Practices at RBS and Barclays

After getting a full but rather general picture of the credit risk management performance in the major British banks from 2004 to 2006, it moves to the stage for analyzing the current practices of certain sample banks, which is particularly important to get an understanding on how British banks manage their credit risk nowadays.

4.2.1 An Overview on the Royal Bank of Scotland Group 1 and Barclays PLC

The Royal Bank of Scotland Group, founded in 1727, is one of the leading banking groups in the UK and is among the top 25 world banks by total assets, according to the Bankers reports. 2 The total assets of the bank amount to 871,432 million among which loans and advances to customers account for 466,893 million, as shown in the balance sheet at 31 December, 2006. The constituents of the group include the Royal Bank of Scotland, National Westminster Bank, Adam & Company, Child & Co, Citizens, Coutts & Co, Direct Line, Drummonds, Isle of Man Bank and Ulster Bank. The business has been extended from the UK to Europe, the US and Asia Pacific. In the UK, RBS is a market leader in the provision of retail and corporate banking services.

Barclays PLC, with 27 million customers and clients in more than 50 countries, is also among the Bankers 2007 top 25 world banks by total assets and market capitalization. This banking

group offers a full range of retail and wholesale services to customers, including retail, business and private banking, credit cards, investment banking, investment management and wealth management. The business unites embedded inside Barclays are UK Banking, Barclaycard, International Retail and Commercial Banking, Absa, Barclays Capital, Barclays Global Investors and Barclays Wealth. Nowadays, about 50% of the group

s profit comes from outside the UK. The total assets of the group are

996,787 million,

282,300 million of which are loans and advances to customers, at 31 December, 2006.

4.2.2 Credit Risk Management Techniques and Practices at RBS and Barclays 1

a. Credit Risk Management Governance

Both at RBS and Barclays, the risk appetite and philosophy, which indicate the strategic directions, plans and limits on controlling risk, are set by the Board, and there exist various sub-committees carrying different tasks. At RBS, the Group Credit Committee is responsible for credit proposals approval and the delegation of authority to divisional credit committees, and it is supported by Group Internal Audit and two dedicated functions, one of which is Group Risk Management (GRM). This function deals with all kinds of risks including credit risk, and is responsible for setting credit risk management standards, which are a combination of governance structures, credit risk policies, control processes and credit systems and are known as the groups Credit Risk Management Framework (CRMF). At divisional level, the management of credit risk is conducted by divisional credit departments, which are in charge of the establishment and maintenance of individual CRMF that complies with the group one.

Those departments also have the task to control and monitor the asset quality, and are subject to the regular assessment undertaken by the Group Risk Management function. Joint reporting lines are enforced in the credit risk management process---the divisional head of the credit reports to both the divisional CEO through the divisional Chief Risk Officer (CRO) and to the Head of Group Credit Risk.

While at Barclays, the Board Risk Committee monitors the risk profile. Different responsibilities are held by different levels of officers and the implementation is always carried out by each business risk departments. For managing credit risk, business credit risk management teams take the direct responsibilities and are led by the Credit Risk Function of the group, which is established for assisting the Group Risk. CEO and credit risk directors at different levels are the ones who are in charge of the risk management duty and to whom the credit risk management teams should report. In managing credit risk, a five-step risk management process and internal control framework, including direct, assess, control, report and manage as well as challenge, is strictly followed within Barclays, which clarify the responsibilities as well as procedures. Better illustration about the risk organization of each group is made in the following figures respectively.

Figure 4.1: Credit risk management organizations at RBS (based upon RBS 2006 annual report)

Figure 4.2: Barclays Credit Risk Management Organization (based upon Barclays 2006 annual report)

b. Credit Granting Process

Prior to the approval of the credit exposures, an assessment of the customer credit risk and credit facilities is undertaken at RBS, which include a review of the purpose of the credit and sources of repayment. The affordability tests will be carried out, and the repayment history, repayment capacity, sensitivity to economic and market developments as well as the risk-adjusted return are all examined.

Within the group, different customer types are subject to different credit approval processes, which will ensure that the appropriate skills are employed and the specialties are well dealt with. For instance, the financial market counterparties of the group are assessed by dedicated credit function that specializes in trade market product risk.

Considering different features of different types of customers, the techniques applied also vary

at RBS. The Assessments of corporate borrowers, transaction risk and financial markets counterparties are carried out by using credit risk models, while the evaluation of consumer lending and personal businesses mainly employ credit scoring techniques, which combines scores to management judgement for guaranteeing an effective approval process.

c. Quantitative Credit Risk Measurement

Credit Risk Models

Credit Risk models are adopted throughout the two groups for supporting the quantitative part of the risk assessments, which are needed for credit granting, monitoring and the portfolio level analysis and reporting. At RBS, those models are developed by divisional credit risk departments, tailoring different types of borrowers and credit facilities, and the validity should all be ensured by GRM before any implementations. Generally, there are four categories of credit models:

Probability of default (PD)

Obviously, the PD models are designed for calculating the probability that a customer will fail to make a full and timely repayment of credit obligations, and the time horizon applied is one year. Different such models are developed for different types of customers, considering the characteristics in every credit portfolio. And according to the outcome of the PD models, each customer will be assigned an internal credit grade for further use.

The inputs to those models include both quantitative ones, including recent financial

performance and customer behaviour, and qualitative ones which are company management performance, sector outlook and so on.

Exposure at Default (EAD)

EAD models are for the estimation of expected level of utilization of a credit facility in the credit portfolios at the time of a borrowers default. It is recognized by the methodologies in EAD modeling that customers may make more use of the credit facilities in the run to a default.

Loss Given Default (LGD)

By using LGD models, the economic loss the group will suffer on a credit facility in the event of default will be estimated, which means the amount of debt that will not be recovered. The factors that are taken into account by the LGD models include the type of borrower, facility as well as risk mitigations such as collateral, the industry sector of the borrower and general economic conditions.

Credit Risk Exposure Measurement

The major application targets of the credit risk exposure measurement models are derivatives and other traded instruments, where the exposure is not 100% of the gross nominal amount of the credit obligation. In these instruments, the credit exposure amounts may depend on external variables such as interest and foreign exchange rate, which will be calculated by the models carefully.

While at Barclays, the three building blocks in the measurement system are also PD, EAD and LGD models, which share and follow the same principle with models adopted by RBS, so no further description will be given. What is worth mentioning is a self-developed measurement of expected loss used at Barclays---Risk Tendency (RT).

Risk Tendency

RT is adopted in the group for estimating the average loss for the loan portfolio on a 1-year basis, by considering the current size and risk characteristics of the portfolio. The calculation of RT for both retail and corporate loans is:

By calculating and interpreting RT, the average loss that is inherent in the groups credit exposures can be grasped by each business unit, so that credit risk management can be performed with clearer focus. Besides the models developed internally, external models and rating tools also help with the estimation of RT, which are checked for appropriateness in the model approval process on a regular basis.

Internal Credit Rating system

Besides those credit risk models introduced above, the internal ratings are also used inside RBS group for assessing borrower credit quality, which base on the results from the PD models. All the customers are assigned credit ratings respectively, which then are mapped into

a group level rating conclusion as the reflection of asset quality scales. For illustration, the upper and lower boundaries and the midpoint for each asset quality scale are shown as follows, which are expressed as an annual probability of defaults.

Table 4.3: An expression of RBS group level asset quality grades (source: RBS 2006 annual report)

Besides the information on asset quality grading, the distribution of assets according to this grading is also provided by RBS, as shown in Figure 4.6.

At Barclays, an internal credit rating scale ranging from grade 1 (lowest PD) to grade 21 (highest PD) has been developed for the wholesale credits, which represents the best estimation can be achieved on the level of credit risk for each counterparty. Other methodologies such as external models and rating agency ratings also help with the internal assessment. For smaller credits, results from a single rating model may already be sufficient while for individual large credits, multiples approaches are used.

d. Credit Risk Exposure Management

RBS has a very clear definition about the credit risk it will manage and has adopted credit risk assets as an internal measure of the groups exposure to its customers, which consist of loans and advances, instalment credit, finance lease receivables, debt securities and other

traded instruments.

Regarding the credit risk assets distribution, RBS has very comprehensive understanding and provides a detailed report according to industry sector, geographic areas and product as well as customer type respectively, which are shown in the following three figures.

Figure 4.3: Distribution of credit risk assets by industry sector (source: RBS 2006 annual report)

Figure 4.4: Distribution of credit assets by geography (source: RBS 2006 annual report)

Figure 4.5: Distribution of credit risk assets by product and customer type (source: RBS 2006 annual report)

To assess the potential concentration risk that arises from within the loan portfolio, industry, geography and product as well as customer type analysis play a very important part since the riskier areas can be identified, which can prepare the group for the future potential volatility. From the figures above, some facts and numbers about the groups credit exposure distribution can be generated. For instance, considering geographical areas, the three largest exposures are in the UK, North America and Europe, with the rest of world only comprises 6.4% of the credit risk assets. And through differentiating credit risk assets by products and customer type, it is shown that lending to corporate customers are definitely the largest category, representing 35% of the total; the second place is taken by mortgages lending to individuals, which is about 20% of the total amount; while debt securities issued by banks, sovereign and quasi government account for 19%.

While at Barclays, credit risk, especially together with the sub-types of country risk and settlement risk is well recognized within the group. The risk management covers a wide range including customer loans and advances, loan commitments, contingent liabilities, debt securities and other exposures in trading activities, and clear explanation as well as differentiation is given to the credit risk nature of these exposures respectively in the report. Geographical and industry analysis on the loans and advances to customers are also carried out by the group in the similar as what RBS has done. Besides, Barclays also include maturity analysis on its credit exposure, as shown in the following table.

Table 4.4: Maturity analysis of loans and advances to customers (source: Barclays 2006 annual report)

Credit concentration issue is given particular attention at Barclays, where relevant information can be found in an independent part of the report named Credit Concentration. A framework called mandate and scale is adopted in the group, setting limits such as the caps on UK commercial investment property lending and proportion of lending with maturity in excess of seven years. Credit exposure limits are also given to some sub-investment grade countries for reducing potential risk.

e. Credit Quality Management

Quality Statistics by Internal Rating

As mentioned in the risk measurement part, RBS uses internal credit ratings for assessing the quality of borrowers. To give general information on its credit risk assets, the statistics about the distribution of those assets by quality is provided by the group.

Figure 4.6: RBS Distribution of credit risk assets by asset quality (source: RBS 2006 annual report)

According to this figure, 97% of the credit exposures are to counterparties rated AQ4 or higher, which means an S & P equivalent of no lower than B. Exposure to investment grade counterparties account for nearly half (46%) of the total credit risk assets. As for Barclays, no such information can be found.

Loan Impairment

Impaired assets at RBS have been classified into two categories---Risk Elements in Lending (REIL) and Potential Problem Loans (PPL). Non-accrual loans, loans that are accruing but are

past due 90 days and restructured loans are referred to as REIL, while PPL represents impaired assets that are not included in REIL but upon which management has serious doubts due to the known information about possible credit problems. The loans classified as REIL and PPL in the group are shown in Table 4.5.

Table 4.5: RBS loans that are classified as REIL and PPL (source: RBS 2006 annual report) 1

Regarding this issue, different categorizing method is adopted at Barclays, where potential credit risk loans are identified as either NPL or PPL at Barclays, between which NPL are defined as loans where the repayments are failed in part or full and PPL are defined as performing loans upon which serious doubts exist for their further repayment. Within the group, statistics on NPL and PPL are made both by location and as a percentage of loans and advances.

Provisions/ Allowances

Different provision analysis methods are tailored for different credit portfolios at RBS. Largely automated processes are used for detecting problem credits in the consumer portfolios,

which are comprised of small value, high volume credits. While an approach of case by case analysis is adopted by specialists for identifying problems in corporate portfolios, which consist of higher value lower volume credits. Both of them help the group to ensure an early management of problem exposures and possible minimization of the losses. Three kinds of provision assessments are used in credit risk management practices:

Individually assessed provisions are assessed on a case by case basis by considering the financial condition of the counterparty, which means the estimation of the discounted value of any recoveries and realization of security or collateral, and are used for individually significant impaired assets.

Collectively assessed provisions are determined from a quantitative review of the portfolio, which are applied to impaired credits below a certain limits.

Latent loss provisions are those held against the estimated impairment in the performing loan, which are still waiting for the identification at the balance sheet date.

According to the three categories of impairment provisions, a summary at the group level can be drawn in the following table, together with the provisions coverage condition.

Table 4.6: Total customer provisions balance (source: RBS 2006 annual report)

Table 4.7: Provision coverage ratios (source: RBS 2006 annual report)

At Barclays, no such detailed information on the allowance assessment methodologies is provided. And what is known is that the impairment is measured either individually for significant assets or collectively for portfolio of homogeneous assets, which seems share some similarity with the RBS approaches. The coverage of these impairment allowances is calculated and provided in the report (see Table 4.8), from which the adequacy can be judged, but only for different geographical areas.

Table 4.8: Impairment coverage of PCRL (source: Barclays 2006 annual report)

f. Credit Risk Mitigation Methods

Unfortunately, there is no independent section in RBS 2006 annual report that can provide any information on the methods adopted for minimizing credit risk within the group. By searching through the whole report and other published information, it can be found that traditional ways for controlling credit risk are used by the group, such as collateral and predetermined credit limits. Besides, asset securitization is carried out against portfolios of

mortgage and credit card receivables totaling

18,589 million in the UK and Ireland by the special purpose entities. However, there is no detailed description about the diversification strategy or practices of loan portfolios. Credit and other derivatives transactions are also undertaken by the group.

Credit risk management governance

RBS Group Risk Management sets the Credit Risk Management Framework

Barc Credit Risk Func responsibilities an management proc

Different credit approval N/A processes are applied to different customer types. Therefore it seems Barclays has done a better job this time since it has been shown that various approaches have been adopted for managing credit risk. Assessment of the borrowers repayment Credit risk measurement Self developed PD, EAD, LGD PD, EAD and LG ability is made, and collateral as well as security is required certain funds advanced. Netting andfor Credit Risk Exposure used and a self-de agreements and covenants are also used for minimizing credit exposures. of Measurement modelsDiversification and internal named Risk Tend portfolio is well managed inside the group, by distinguishing different types ofon counterparties and credit ratings based the PD adopted for estim setting maximum exposure limits to avoid unwanted concentration. Besides, other methods such losses. as model are adopted. A clear A 21-grad loan sales, asset securitization and credit derivatives are also adopted, a clear purpose for comparison of with internal ratings system, together managing credit risk. with the external equivalent is ratings, is used fo made. risk. Credit exposure management Assets with credit risk are all identified. Credit assets distribution is controlled according to industry sector, geographical area and customer type. Awareness of credit risk concentration has been raised.

Credit granting

Analysis on credi made according t area, industry and concentration is g and limits to subcountries are clar

Credit management Quality Statistics below, based on internal Potential credit ri To sum up, quality based on the above illustration, a consolidated table is provided which includes ratings has been made. Impaired identified as eithe the key techniques and practices in credit risk management of the Group 1 banks. A clear general assets are classified as either Risk loans or potential view can be drawn regarding credit risk management at RBS and Barclays. And certain potential Lending is me areas for the two banks to learn from each other can also be Elements indicated,in although it or is Potential not specified Impairment in Problem Loans. Three kinds of individually or co the research questions. provision assessments are used.

Credit risk mitigation techniques

Traditional methods such as collateral and credit limits are adopted. Newer measures---asset securitization and credit derivatives are used as well.

All kinds of tradi are used, includin agreements and d Loan sales, asset and credit derivat adopted for mana

Table 4.9: Summaries of credit risk management techniques and practices at RBS and Barclays

4.3 Generalization and Comparison with Basel Regulations

From the above section, the current credit risk management practices in the two large banking groups RBS and Barclays are already shown from several major aspects, with RBS given more detailed descriptions. On the basis of this, a generalization will be drawn from their practices in this section, and those practices will also be compared with major Basel (1999a, 2000) regulations.

4.3.1 The Establishment of Credit Risk Environment

The first common feature of the credit risk management at RBS and Barclays is that they both have built up a favorable atmosphere for sound credit risk management. Credit risk is given clear definition and the nature is understood completely. Credit risk organizations have been established for the management and monitor purpose and responsibilities are reasonably designated at different levels, which are already shown in Figure 4.1 and 4.2. Credit policies and strategies are set by the board and implemented by every division or business unit, which work as the guidance for the whole risk management process, such as the Credit Risk Management Framework at RBS and the predetermined five-step process at Barclays. All of the above proves the central role credit risk management has played in the whole risk management of these two banking groups. And it also indicates that they do follow the first area of Basel (1999a) principles well.

4.3.2 Credit Granting Process

Regarding this area, it should be admitted that not much generalization can be concluded from

since Barclays hasn

t given any particular description on its credit granting process in the report. As for RBS, it is confirmed that careful assessment is made to different types of counterparties, however, no information is given about its credit-granting criteria, as mentioned in Principle 4 of Basel (1999a)

s requirements. Credit limits for individual borrowers and counterparties are set and followed in the risk management process of both banks, but processes about extension of existing credits are not revealed much. Since the Basel (2000) paper on credit risk management disclosure doesn

t include particular recommendations on credit granting process, no absolute comment can be made in this area about the two banks.

4.3.3 Credit Measurement and Monitor Process

a. Credit Risk Measurement

Both RBS and Barclays apply credit risk modeling and credit rating approaches to the quantification of credit risk exposure of individual borrowers and portfolios. Similar basic models---the PD, EAD and LGD models are used, which are also the building blocks mentioned in Lowes (2002) paper. Besides, each bank has developed more measurements on the exposure and potential loss related to credit risk, which are the credit exposure measurement models of RBS and the Risk Tendency of Barclays. So it can be concluded that multiple modeling methodologies are employed in both banks, targeting different conditions.

The application of internal rating systems is also found in both banks, which base on the PD model especially. The difference is that the rating system of RBS is a 5-grade one while that

of Barclays is a newly improved 21-grade one. However, what still needs improvement is that none of them are given detailed explanations in the reports.

1 each rating means and the degree of risk Since it is mentioned in the Recommendation 22 of Basel (2000) paper that information like what being distinguished should be disclosed, such descriptions in the two banks annual reports are obviously not sufficient. Besides, the specification made by Basel II

on the use of internal rating for the calculation of risk-weighted capital also indicates that this methodology deserves more attention.

b. Credit Monitor

The monitoring on the composition and quality of credit portfolios is carried out well both at RBS and Barclays, which also accords with the requirements of Basel (1999a). Credit exposure is analyzed by different categories such as business lines and geographical areas, as shown in Figure 4.3, 4.4 and 4.5 and concentration problems are identified by these banks.

The quality of credit risk assets is monitored by revealing necessary information about potential problem loans and the relevant coverage ratio. Although judged from the recommendation made by Basel (2000), the disclosure may not be detailed enough, which provides inadequate information on whether the banks have strong monitor on the impaired loans, it can be found that basically, the two banks do give attention to the impairment as well as allowances issue.

4.3.4 Controls over Credit Risk

Although it is already mentioned that clear definition and designation of responsibilities are found in these two banks and according to their reports, different functions and committees will give regular assessments to the implementation of credit risk management, it cannot be proved that a valid and efficient credit review system is already established in both banks, due to the lack of relevant information.

4.3.5 Credit Risk Mitigation

Traditional credit risk mitigation approaches such as collateral and diversification are applied in both banks, although RBS doesnt give much description on its diversification strategy. Newer approaches are also applied, such as the use of securitization and credit derivatives, together with the relevant quantitative disclosure. For instance, RBS has clear explanation about the amounts of credit derivatives, as shown in Table 4.10. And Barclays has revealed the carrying amount of its securitized assets, as shown in Table 4.11.

Table 4.10: Derivatives Use at RBS in 2006 (source: RBS 2006 annual report)

Table 4.11: Carrying amount of securitized assets at Barclays in 2006 (source: Barclays 2006 annual report)

4.4 Credit Risk Management Practices at Bradford & Bingley and Northern Rock

4.4.1 An Overview on B & B PLC and Northern Rock1

Bradford & Bingley plc is a UK-based bank, converted from Bradford & Bingley Building Society in 2000. The business activities of the group are focused on providing specialist mortgages and saving products. The group also participates in commercial property lending and lending to Housing Associations throughout the UK. The company has total assets 45,354.3 million, among which loans and advances to customers that account for 36,131.7 million take an absolutely large portion.

Similar to Bradford & Bingley, Northern Rock is also a bank that is converted from building society (conversion is in 1997), and its core business lies in the provision of UK residential mortgages and secured commercial lending and personal unsecured lending products. The total assets of the company are 101,010.6 million, with 86,681.5 million loans and advances to customers, at 31 December, 2006.

4.4.2 Credit Risk Management Practices at B & B and Northern Rock

a. Credit Risk Governance

Generally, the credit risk policies and strategies, the risk management structure and responsibilities are determined by the Board of Directors at B & B. Under those frameworks, Credit Risk Committee supports the Group Risk Committee by advising, reviewing and monitoring all credit matters, while line managers are responsible for identifying, measuring and managing credit risk within the area of responsibility (see Figure 4.7).

Figure 4.7: B & B Credit Risk Management Organization (based upon B & B 2006 annual report)

At Northern Rock, the Board is in charge of setting general risk appetite and management strategies while the articulation of such appetite is delegated to Management Board Asset and Liability Committee. Under its instruction, the Risk Committee, on behalf of the board, reviews and reports the credit risk management system and guarantees the internal control. While the identification and implementation jobs for the credit risk management, is carried out by each business area respectively, as shown in Figure 4.8.

Figure 4.8: Northern Rock Credit Risk Management Organization (based on Northern Rock 2006 annual report)

b. Credit Approval Process

The credit approval process at B & B is guided by the lending policies issued by the board, which is supported by credit scoring techniques for sound credit decision making. Certain limits have been set for screening the accepted types of borrowers or counterparties. For instance, within the treasury trading book, counterparties are required by the group to have a minimum long term credit rating of Moodys A2 or equivalent, while for short term credit rating, Moodys P2 or equivalent is the minimum level.

Comparatively, a more rigorous approval process is found in Northern Rocks practices. Within the company, the approval process is undertaken by two functions---the Retail Credit Risk Approval Group and Treasury Credit Risk Approval Group. They exist for different customer types and ensure the appropriate allocation of the skills and resources. For retail credit approval process, what is worth mentioning is that certain factors are reviewed for assuring the prudent lending, among which there are loan size, geographical spread, size of loan relative to the value of the property and so on. Through the statistics results and analysis of those factors, the lending policies and limits can gain a better match.

c. Credit Risk Measurement

Loans with low risk of default and such high levels of security that even in markedly adverse economy no loss Internal credit rating system has been established within B & B, would which be is incurred. a four-category one as shown in the following table,Low based on the assessment of relative risk ofin default and of the ofadverse risk Even the event a strength markedly the underlying security. The assignment of loans to different categories is achieved by projecting economy, only modest reductions in the assessment results forward according to planned advances and redemptions. Besides, credit profitability would be recorded. scoring techniques and external ratings are also applied to measuring the credit risk. Medium/low risk In markedly adverse economy, substantial reductions in profitability may be caused, but profits would still be positive. Medium risk In markedly adverse economy, high levels of bad debts would result in negligible profits or losses being incurred.

Negligible risk

Table 4.12: Four risk categories of residential and commercial lending (based upon B & B 2006 annual report)

Two basic credit risk models are used throughout Northern Rock for the quantitative assessment part of credit approval, monitoring and portfolio management, which are the PD and LGD models. Both of the two categories of models are governed by the two credit approval groups mentioned above.

d. Credit Exposure and Quality Management

Bradford & Bingley Northern Rock Credit risk governance Credit Risk Committee Risk Committee on behalf supports the policies and of the board is in charge of Both Bradford & Bingley and Northern Rock have clearly identified credit risk exposure, both strategies set bytheir the Board. credit risk management on and off balance sheet, although the credit risk arising from activities suchmonitor. as trading and settlement activities process actually only takes a small proportion. No analysis according to Credit approval Lending policies issued bymade Retail Credit Risk geographical areas or maturity types has been found, the nor board has the impairment isinformation guiding the about loan Approval Group and as well as allowances allocation. process and credit limits Treasury Credit Risk are applied to different Approval Group are set up types of borrowers. for assessing different customer types. Credit risk measurement A four-category internal PD and LGD models are rating system is adopted, adopted, which are subject together with credit scoring to the governance of the and external ratings. two credit risk approval groups. e. Credit Risk Mitigation Credit exposure and quality management Clear identification on Clear identification on credit risk exposure. credit risk exposure. At B & B, traditional methods for minimizing credit risk, such as collateral and diversifications through limits around individual counterparties, types of counterparties, industry and geographical Credit risk mitigation techniques Traditional methods for Traditional methods, regions. Similar techniques are also found in Northern Rocks practices, besides, the company also minimizing credit risk are together with adopts methods such as margining, master netting agreement and credit derivatives periodically credit for derivatives are used. minimizing credit risk in particular transactions. used. However, no relevant information about the 4.13: Summaries used of credit management and at B &report. B and Northern Rock amount Table of credit derivatives forrisk managing risk techniques is found in itspractices 2006 annual

4.5 Generalization and Comparison with Both Basel Regulations and Other Sample Banks

4.5.1 The Establishment of Credit Risk Environment

Similar to what has done to the Group banks, a consolidated is also provided The establishment of been an appropriate credit risk1 environment is the basis table in sound credit risk for a brief conclusion on the credit risk management techniques and practices at B & B and Northern management, as emphasized by Basel (1999a), and through the analysis, there is no doubt that Rock, as shown below. credit risk has been defined and identified comprehensively at both B & B and Northern Rock, just like what is found about RBS and Barclays, since all the four banks have expressed their understanding about the maximum credit exposure in the business, and the framework for the management has been set. The responsibilities for different levels of function are explained

and designated, among which the only difference is that the larger banks RBS and Barclays may have more complicated structure for managing credit risk, due to their size and business range. All the above findings provide good evidence that a satisfactory average level of credit culture and credit risk management environment has already been build among major British banks.

4.5.2 Credit Granting Process

Strictly speaking, this second area of the Basel (1999a) requirements on sound credit risk management is not that comparable among the four banks, due to the lack of sufficient information at Barclays and B & B. From what is gained at RBS and Northern Rock, it can be concluded to some extent that those two banks, different in size, both present acceptable level of credit granting management, because factors such as differentiation on types of counterparties, relationship between group borrowers and certain credit limits are all taken into consideration in the credit decision making. However, due to the problem mentioned at first, this result cannot be treated as representative, which causes the difficulty to make a conclusion that will apply to all the four banks and further to major British banks. What can be confirmed is that all the sample banks do have systematic procedure for assessing and granting credit, and they do put credit granting in a very important place because after all a good practice at this stage can save much efforts in the future risk control process. But whether their credit granting process is comprehensive and forwardlooking enough as Basel (1999a) has suggested, and whether all the relevant self-issued guidelines are followed strictly during each of the banks implementation are still quite arguable.

4.5.3 Credit Measurement and Monitor Process

a. Credit Risk Measurement

It is quite obvious from the illustration of the four banks credit risk management practices that a wide variation exists in the depth of quantitative credit analysis since measuring techniques such as credit modeling are better developed and used at RBS and Barclays than the Group 2 banks. This is shown by the categories of models and measures those four banks have mentioned in each report. Both RBS and Barclays have sophisticated systems including multiple choices of building block models and their own measures; however B & B gives little description on its credit modeling practices while Northern Rock only adopts PD and LGD models. Since Basel (1999b) already has emphasized the importance of credit risk models in bank risk management and performance measurement process, the two smaller banks, despite of the fact that the credit risk they face may not be that complicated as what larger banks face, should still make better use of credit risk modeling technique.

Internal rating system is a also a common measuring system at B & B and Northern Rock, which bases on similar principles, but may differ in the scale of the grades and explanation. This, together with the earlier findings, has shown that the major British banks have already given attention to internal rating systems by developing the one that will mostly tailor to their business characteristics. It is found that the basic methodology of the internal ratings doesnt vary much in the four banks since they all have mentioned the use of PD model in the calculation of the ratings, which implies the role of internal rating as a default indicator at major British banks. However, common fault also exist in the insufficient disclosure about the

development, meaning and use of the system, which as mentioned in last section, definitely is an area that needs further improvement for both large and small banks.

External ratings dont seem to play a very important role at the four banks, while comparatively the smaller banks have presented more heavily reliance on them because they both emphasize the reference of rating agency ratings to the credit analysis process.

Besides, what has not been mentioned in the earlier section is that both of the Group 1 banks have adopted stress testing on the portfolio credit risk, which will measure the potential vulnerability to exceptional economic and geographical events in order to estimate the potential loss. However, this test has not been adopted either at B & B or Northern Rock, which may indicate a general lack of awareness of the sensitivities in counterparty credit risk portfolios at smaller major British banks (FSA 2004).

b. Credit Monitor

No doubt that credit monitor is a dispensable responsibility at both B & B and Northern Rock, however, compared with the other sample banks, there is no evidence proving that they have done that as well as RBS and Barclays or what Basel (1999a) has required. For instance, credit exposure analysis, which is quite common at the two larger sample banks, has not been shown by any form, leading the doubt on the banks credit exposure management performance. Very little information can be generated about their practices related to impaired loans or allowances. Although it is true that the two smaller banks, due to their core business nature, may have smaller scale of loan exposures, it is not reasonable to think that such

analysis or effort is unnecessary.

4.5.4 Controls over Credit Risk

Similar to the other two banks, credit risk management practices at B & B and Northern Rock are assessed by different levels of functions, and it is certain that some credit reviews are undertaken, according to the description in the annual reports. However, it can never be affirmed that those controls are really made appropriately and timely.

4.5.5 Credit Risk Mitigation

The similarity between B & B and Northern Rock, regarding their credit risk mitigation techniques, is that they both make good use of traditional approaches for minimizing credit risk, among which, credit limits and collateral as well as security take a leading role. For instance, B & B has stated that all its residential and commercial lending is secured. Diversification is also a common means in these two banks, since they both express the effort to try to allocate mortgage lending to the different regions and different types of managers. However, compared with the other two larger banks, B & B and Northern Rock havent applied much of the newer approaches such as securitization, loan sales or credit derivatives for minimizing and transferring credit risk, so it can be concluded that their practices for credit risk mitigation is still limited. This situation can be explained partly by the reason that most of their credit exposures are coming from the mortgage and commercial lending, for which the use of ways such as credit limits and collateral may already be sufficient to reduce the risk to an acceptable level.

4.5.6 Credit Disclosure

Credit disclosure has great impact on the whole research and it is the major area where recommendations for MBBG banks credit risk management arises. Comparatively, B & B and Northern Rock have done obviously much worse than the other larger sample banks in credit disclosure, and they are far from following the recommendations of Basel (2000). What can be found most in the Group 2 banks annual report is the general description of the management structures, policies and certain credit limits. While no detailed information regarding credit exposure and quality, no analysis explanation or statistic results, and no introduction about the rating or modeling systems have been disclosed. In general, RBS and Barclays have better credit risk disclosure performance (take the detailed analysis shown in Figure 4.3, 4.4 and 4.5 as an example), which can provide the outsiders more comprehensive credit risk profiles of the group, and make the assessment of the groups credit risk management easier. However, they also have presented weakness in certain aspects. For instance, the lack of relevant information on credit granting process in Barclays annual report has caused great difficulty in the review. In fact, all the weaknesses that are concluded above are related with credit disclosure problems, since it must be made clear that whether the sample banks are really doing badly in the credit risk management process or they just havent revealed enough information to the public. For example, only words such as sophisticated credit rating system cannot provide any information for self exploration and judgement, and will certainly lead outsiders to doubt the soundness of banks credit risk management, which may actually be quite wrong and misleading. The last issue that deserves mentioning is that although there are only three aspects of disclosure mentioned in section 3.3.3, upon which the

evaluation and recommendations are based, accounting policies are also important in the credit disclosure. Due to the lack of time and knowledge of the researcher, analysis on relevant accounting policies among sample banks is not made, but the impact of accounting rules on the calculation of credit risk exposures and allowances can never be ignored, which is also emphasized by Basel (2000).

Group 1 Banks Credit risk environment -comprehensive identification of credit risk -clear designation of credit risk management responsibilities

Group 2 Banks -clear understanding on maximum credit exposure in the business -straightforward responsibilities at different management levels

Analysis and Recommendations

-generally, the sample banks have met the Basel requirements on establishing appropriate credit environment -indicate satisfactory level of practices in Based on all the above arguments, a table is made for summarizing the comparison results and this area among emphasizing certain recommendations, which is displayed as follows. MBBG members Credit granting process -systematic procedure for credit granting -at RBS, different assessments are designed for different customer types -Barclays lacks relevant information -credit granting is put in an important place -Northern Rock follows systematic procedure for granting credit -the information B & B has revealed is too general -both large and small banks should make efforts to make their credit granting process more comprehensive and tailor to different customer types -relevant information should be disclosed

Credit risk measurement

-self-developed credit risk models, including PD, EAD and LGD -new measures are developed and adopted, such as Risk Tendency of Barclays -internal ratings and external ratings are used, with the former one playing the key role -stress testing is carried out

-lack of selfdeveloped credit risk models -internal ratings are adopted and external ratings are also relied on in credit analysis process -lack of stress testing

-smaller banks should give more attention to the use of credit risk models -the standards and principles of internal rating systems and the use of external ratings deserve more emphasis and information should be disclosed to a larger extent -stress testing should be taken by all banks

Credit monitor and control

-analysis on overall credit composition and quality is made -credit concentration is given attention to -credit reviews are undertaken

-credit reviews are undertaken

-smaller banks should pay more attention to monitoring credit exposure and quality -larger banks accord better to Basel requirements than smaller ones -no absolute recommendations can be made on the adoption of credit risk mitigation techniques since banks should choose the techniques they are best at and that tailor most to the nature of their credit exposures -both large and small banks should make improvement to meet the Basel requirements on credit disclosure -accounting policies related to credit risk assessment deserve attention

Credit risk mitigation techniques

-combination of both traditional and newer methods for minimizing credit risk

-heavier reliance on traditional techniques than on the newer ones

Credit disclosure

-most credit risk management information is recorded and reported, in different depth

-limited information available in annual reports

Table 4.14: Comparisons and recommendations between the two groups of banks against Basel requirements

Chapter 5 Conclusion

Almost in all the books and articles about bank management, credit risk is declared to be the major risk the banking sector is facing. Wesley (1993) used to conclude that few banks have been left untouched by the consequence of poor lending and other credit risk management practices. Although it should be understood that perhaps banks can never know their borrowers or counterparties as well as they should, an appropriate credit risk management can help to minimize the loss, while on the contrary, a poor one can be extremely damaging (see Heffernans book for various cases of bank failure caused by credit problems).

In this dissertation, the credit risk management of major British banks is analyzed, from which various findings have been achieved. It has been shown by the ratio analysis that the UK banks are generally maintaining the same level of credit exposure and quality which also means a steady performance on controlling credit risk from 2004 to 2006. However there is no doubt that future improvement is still to be made; even when the six ratios contained in the analysis are all showing pleasant trends for major British banks. As for how on earth the major British banks perform their credit risk management tasks, two groups of larger and smaller British listed banks are compared from various aspects of credit risk management techniques and practices, in order to draw the similarities as well as differences. It has been found that generally the major British banks with larger size have managed credit risk inherent in their banking and trading activities more comprehensively than the smaller ones. They follow the Basel guidelines better, and have adopted multiple choices of means for assessing, granting and mitigating credit risk. While for the major British banks with

comparatively smaller size, it seems that they put more focus on the credit granting phase, and actually may have stricter limits for the credit risk they can accept due to the low risk tolerance. In this way, they may control and limit the credit risk to some extent at the very beginning of any credit inherent behaviors, but at the same time, the further control and risk mitigation as well as transfer practices are not that well carried out as larger banks. However, when compared with the major Basel regulations regarding credit risk management, it should be admitted that none of the large or small major British banks have already met all the requirements. It is observed that all the sample banks have followed the general principles to some extent, but the individual practices may vary and suffer from all kinds of weakness and flaws. The problem with credit disclosure also plays an important role in the research and findings. Because no banks can be treated as exactly following the recommendations for the best practice of credit disclosure issued by Basel (2000), certain information about banks

practices cannot be found in the published reports, which may lead to the misjudge of banks

credit risk management profile or incomprehensive findings. This problem with credit disclosure is thus also treated as an area all the nine major British banks need to give more attention in the future, even though the larger banks actually have done an acceptable job compared with smaller ones.

As mentioned before in the chapter of methodology, this research suffers from several bias and limitations, which is hoped to be improved in future works. Also, there are several important issues about bank credit risk management, such as the adoption of asset securitization, the abuses of special purpose entities, moral hazard inside banking system, the use of stress testing, the impact of accounting rules and the efforts of supervisors that are not

given much focus in this dissertation, and they may deserve more discussion in the future work as well.

References

Abdus, S., 2004. Bahrains Commercial Bank Performances During 1994-2001. Management Review. Available at: http://www.findarticles.com/p/articles/mi_qa3857/is_200401/ai_n13454903

Credit & Finance

Altman, E. and Rijken, H. A., 2004. How Rating Agencies Achieve Rating Stability. Available at: http://dse.univr.it/safe/Workshops/MCMR/2004/rijken.pdf.

Altman, E. I. and Saunders, A., 1998. Credit Risk Measurement: Developments over the Last 20 Years. Journal of Banking & Finance, 21, pp. 1721-1742.

Anonymous, 2007a. Bankers Mistrust. Economist, 384(8542), pp.64.

Anonymous, 2007b. Paper Losses. Economist, 384(8543), pp.40-41.

Arora, N., Bohn, J.R. and Zhu, F. L., 2005. Reduced Form vs. Structural Models of Credit Risk: A Case Study of Three Models. Available at: http://www.moodyskmv.com/research/files/wp/Aroara_Bohn_Zhu_reduced_structural_20050217 .pdf

Arsov, I. and Gizycki, M., 2003. New Measures of Credit Risk. Reserve Bank of Australia Bulletin , June, pp. 10-14.

BANKSCOPE: http://bankscope.bvdep.com/version-200793/cgi/template.dll

Barclays: http://www.barclays.com

Basel Committee, 1986. The Management of Banks Off-Balance Sheet Exposures. Basel Committee on Banking Supervision, March.

Basel Committee, 1988. International Convergence of Capital Measurement and Capital Standards. Basel Committee on Banking Supervision, July.

Basel Committee, 1998. International Convergence of Capital Measurement and Capital Standards. Basel Committee on Banking Supervision.

Basel Committee, 1999a. Principles for the Management of Credit Risk. Basel Committee on Banking Supervision, July.

Basel Committee, 1999b. Credit Risk Modelling: Current Practices and Applications. Basel Committee on Banking Supervision, April.

Basel Committee, 2000. Best Practices for Credit Risk Disclosure. Basel Committee on Banking Supervision, September.

Basel Committee, 2006. International Convergence of Capital Measurement and Capital Standards. Basel Committee on Banking Supervision, June.

Bradford & Bingley: http://www.bbg.co.uk

Blanco, R., Brennan, S. and Marsh, I. W., 2003. An Empirical Analysis of the Dynamic Relationship between Investment-Grade Bonds and Credit Default Swaps.

Bank of England Working Paper No. 211.

Available at: http://papers.ssrn.com/so13/papers.cfm?abstract_id=478825

Broll, U., Pausch, T. and Welzel, P., 2002. Credit Risk and Credit Derivatives in Banking. Saarland University and University of Augsburg Discussion Paper . Available at: http://ideas.repec.org/p/aug/augsbe/0028.html

Casu, B., Girardone, C. and Molyneux, P., 2006. Introduction to Banking. Harlow: Pearson.

Credit Suisse, 1997. CreditRisk +: A Credit Risk Management Framework. Credit Suisse Financial Products.

Crouhy, M., Galai, D. and Mark R., 2000. A Comparative Analysis of Current Credit Risk Model. Journal of Banking & Finance, 24, pp. 59-117.

Culp, C. L. and Neves, A. M. P., 1998. Credit and Interest Rate Risk in the Business of Banking. Derivatives Quarterly, 4(4), pp. 19-35.

Duffee, G. R. and Zhou, C., 2001. Credit Derivatives in Banking: Useful Tools for Managing Risk? Journal of Monetary Economics, 28, pp. 25-54.

Duffie, D. and Singleton, K. J., 2003. Credit Risk: Pricing, Measurement and Management. Oxford: Princeton University Press.

Emmons, W. R., 1995. Interbank Netting Agreements and the Distribution of Bank Default Risk. The Federal Reserve Bank of ST. Louis Working Paper.

Fabozzi, F.J., 2006.

Bond markets, analysis and strategies

. Harlow: Prentice Hall.

Federal Reserve, 1998. Sound Credit Risk Management and the Use of Internal Credit Ratings at Large Banking Organizations. Federal Reserve Working Paper.

FSA, 2004. Management of Credit Risk within a Trading Environment. Financial Services Authority Report.

Gonzalez, F. et al., 2004. Market Dynamics Associated with Credit Ratings. European Central Bank Occasional Paper 16, June.

Grauwe, P., 2007. Banking Bail-Out Sows Seeds of Future Crises.

Available at: http://ftchinese.com/sc/story_english.jsp?id=001013490.

Gundlach, M. and Lehrbass, F., 2004. CreditRiskin the Banking Industry. New York: Springer.

Hammerskey, M., 1992. Whats Wrong with Ethnography? Methodological Explorations . London: Routledge.

Haubrich, J. G. and Thomson, J. B., 1993. Loan Sales, Implicit Contracts and Bank Structure. Federal Reserve Bank of Cleveland Working Paper , October.

Heffernan, S., 1996. Modern Banking in Theory and Practice. Chichester: John Wiley & Sons Ltd.

Hennie, V. G., 2003.

Analyzing nd and Managing Banking Risk: A Framework for Assessing Corporate Governance and , 2 edition. Financial RiskWashington DC: World Bank Publications.

Horcher, K. A., 2005. Essentials of Financial Risk Management. Hoboken: John Wiley & Sons, Incorporated.

HSBC 2006 Annual Report: http://www.hsbc.com

Hull, J., 1989. Assessing Credit Risk in a Financial Institutions Off-Balance Sheet Commitments. Journal of Financial and Quantitative Analysis, 24(4), pp. 489-501.

Hull, J., Nelken, I. and White, A., 2004. Mertons Model, Credit Risk, and Volatility Skews. Journal of Credit Risk, 1(1), pp. 3-28.

Hoggarth, G. and Pain, D., 2002. Bank Provisioning: the UK Experience. Financial Stability Review, June.

Jackson, P., 2001. Bank Capital Standards: The New Basel Accord. Bank of England Quaterly Bulletin, Spring, pp. 55-63.

Jackson, P. and Perraudin, W., 1999. The Nature of Credit Risk: The Effect of Maturity, Types of Obligor and Country of Domicile. Financial Stability Review, November.

Jackson, P., Nickell, P. and Perraudin, W., 1999. Credit Risk Modelling. Financial Stability Review, June.

Jacobson, T., Linde, J. and Roszbach, K., 2003. Internal Ratings Systems, Implied Credit Risk and the Consistency of Banks

Risk Classification Policies.

Sveriges Riskbank Working Paper 155,

December.

J.P. Morgan, 1997. CreditMetrics---Technical Document. Available at:

http://212.59.24.64/Kredito%20rizika/Portfelio%20modeliai/Portf4.pdf.

Kaminsky, G. and Reinhart, C., 1999. The Twin Crises: The Causes of Banking and Balanceof-Payments Problems. American Economic Review, June.

Kern, M. and Rudolph, B., 2001. Comparative Analysis of Alternative Credit Risk Models---an Application on German Middle Market Loan Portfolios. Center for Financial Studies Working Paper, March.

Lopez, J. A. and Saidenberg, M. R., 1999. Evaluating Credit Risk Models.

Available at: http://www.frbsf.org/econrsrch/workingp/wp99-06.pdf

Lowe, P., 2002. Credit Risk Measurement and Procyclicality. BIS Working Paper 116.

Monetary Authority of Singapore, 2006. Credit Risk. Monetary Authority of Singapore, February.

Merton, R., 1974. On the Pricing of Corporate Debt: The Risk Structure of Interest Rates. Journal of Finance, 29, pp. 449-470.

Mishkin, F. S., 2004.

The Economics of Money, Banking and Financial Markets

. London: Addison Wesley.

Neal, R. S., 1996. Credit Derivatives: New Financial Instruments for Controlling Credit Risk. nd Economic Review, 2 quarter.

Nishiguichi, K., Kawai, H. and Sazaki, T., 1998. Management Based on the Quantification of Credit Risk. FRBNY Economic Policy Review, October.

Northern Rock: http://www.northernrock.co.uk

Peura, S. and Jokivuolle, E., 2004. Similation Based Stress Tests of Banks Regulatory Capital Adequacy. Journal of Banking and Finance, 28, pp. 1801-1824.

Reto, G., 2003. Risk Management and Capital Adequacy. Blacklick: McGraw-Hill Companies.

Rose, P. S., and Hudgins, S. C., 2008. Bank Management & Financial Services. London: McGraw-Hill Education.

Saunders, A. and Cornett, M. M., 2006. Financial Institutions Management: A Risk Management Approach. London: McGraw Hill.

Seale, C. et al., 2004. Qualitative Research Practice. London: SAGE.

Silverman, D., 2005. Doing Qualitative Research, 2 Publications.

nd

edition. London: Sage

Standard Chartered: http://www.standardchartered.com.hk/cb/invest//invest_debt.html

Strischek, D., 2002. Credit Culture. RMA Journal, November.

Tavakoli, J. M., 2003. Structured Finance: Uses (and Abuses) of Special Purpose Entities. Proceedings, May, pp. 444-453.

The Banker Top 1000 World Banks 07: http://www.the banker.com/news/fullstory.php/aid/5050/Top_1000_World_Banks_07.html

The Royal Bank of Scotland Group: http://www.rbs.com

Treacy, W. and Carey, M. S., 1998. Credit Rating at Large US Banks. Financial Stability Review, June.

Twinn, C. I., 1994. Asset-backed Securitisation in the United Kingdom. Bank of England, May. Available at: http://www.bankofengland.co.uk/publications/quarterlybulletin/qb940204.pdf

Wesley, D. H., 1993. Credit Risk Management: Lessons for Success. Journal of Commercial Lending, 75, pp. 32-38.

Wilson, T., 1987. Portfolio Credit Risk I. Risk, 10(9).

Zhang, L., 2006. Credit Risk Management in Chinese Banking Industry. Available at: http://edissertation.nottingham.ac.uk/ archieve/00000375

Appendix

Descriptives N Mean

Std. Deviation 0.48140 0.57477 0.50762 0.50185 0.57982 0.49719 0.49046 0.50328 0.54519 0.41102 0.25495 0.38925 25.20519 26.49199 26.80831 25.45673 1.05348 0.94443 1.07523 0.98385 0.63906 0.89551 0.57879 0.69973

Std. Error

NPLGL A SPSS Output of Ratio Analysis 2004 Appendix

6 2005 2006 Total 2005 2006 Total 8 8 22 9 9 25 4 2005 2006 Total 7 5 16 6 2005 2006 Total 8 8 22 8 2005 2006 Total 9 9 26 8 2005 2006 Total 9 9 26 7

1.5167 1.5413 1.4825 1.5132 0.9271 0.7600 0.7400 0.7996 0.5350 0.4500 0.3200 0.4306 65.1400 53.8625 55.4750 57.5245 12.0125 11.9778 11.9111 11.9654 7.8625 8.0778 8.1000 8.0192

0.19653 0.20321 0.17947 0.10699 0.21915 0.16573 0.16349 0.10066 0.27260 0.15535 0.11402 0.09731 10.28997 9.36633 9.47817 5.42739 0.37246 0.31481 0.35841 0.19295 0.22594 0.29850 0.19293 0.13723

LSRGL

One-way ANOVA: UK Major Banks Financial Ratio Comparison 2004-2006

2004

NCOGL

2004

LSRIL

2004

TCR

2004

Tier_1

2004

ANOVA Sum of Squares NPLGL Between Groups Within Groups Total Between Groups Within Groups Total Between Groups Within Groups Total 0.014 5.275 5.289 0.160 5.919 6.079 0.107 2.165 2.273

df 2 19 21 2 22 24 2 13 15 2 19 21 2 23 25 2 23 25

Mean Square 0.007 0.278

F 0.025

Sig. 0.975

LSRGL

0.080 0.269

0.297

0.746

NCOGL

0.054 0.167

0.322

0.730

LSRIL

TCR

Between 488.861 Groups Within 13120.08 Groups 5 Total 13608.94 6 Between 0.046 Groups Within 24.153 Groups Total 24.199 Between Groups Within Groups Total 0.286 11.954 12.240

244.430 690.531

0.354

0.706

0.023 1.050

0.022

0.979

Tier_1

0.143 0.520

0.275

0.762

Appendix B

Financial Ratios of Nine Major British Banks 2004-2006 (%)

Bank LSRGL Abbey National Abbey National Abbey National Alliance & Leicester Alliance & Leicester Alliance & Leicester Barclays Barclays Barclays Bradford & Bingley Bradford & Bingley Bradford & Bingley HBOS HBOS HBOS HSBC HSBC HSBC Lloyds TSB Lloyds TSB Lloyds TSB Northern Rock Northern Rock Northern Rock RBS RBS RBS 0.52 0.41 0.34 0.38 1.17 1.26 1.01 0.14 0.16 0.20 0.81 0.85 0.91 1.54 1.51 1.88 1.15 1.17 1.18 0.15 0.18 0.23 0.84 0.92 1.08

LSRIL 64.89 44.67 41.98 48.07 74.96 75.74 76.02 8.96 11.52 27.83 37.66 35.99 58.46 98.55 99.22 103.65 54.74 50.27 54.57

NPLGL NCOGL 0.80 0.92 0.80 0.78 1.56 1.67 1.33 1.53 1.35 0.72 2.16 2.35 1.56 1.56 1.52 1.81 2.10 2.33 2.15 0.25 0.30 0.23 0.15

TCR 12.6 12.5 10.2 10.5 11.5 11.7 11.3 11.8 13.2 13.4 13.2 12.0 12.4 12.3 13.5 12.8 12.0 10.7 10.9 10.1 11.6 12.3 13.5 11.7 11.7 11.7

Tier_1 8.0 10.0 7.9 7.6 8.3 7.7 7.0 7.1 7.6 7.8 7.5 8.1 8.1 7.9 9.4 9.0 8.9 8.2 7.9 8.2 8.5 7.7 8.0 7.5 7.6 7.0

Year 2006 2005 2004 2006 2005 2004 2006 2005 2004 2006 2005 2004 2006 2005 2004 2006 2005 2004 2006 2005 2004 2006 2005 2004 2006 2005 2004

0.50 0.02 0.00 0.00 0.39 0.35 0.35 1.26 1.29 0.71 0.63

62.06 65.42 70.31

1.35 1.41 1.53

0.46

Data Source: Bankscope.

Appendix C

Loans Written in the Nine Major British Banks 2004-2006 (million GBP)

Abbey National Alliance & Leicester Barclays Bradford & Bingley HBOS HSBC Lloyds TSB Northern Rock RBS

2004 94,390.0 35,787.0 262,409.0 28,868.4 286,668.0 672,891.0 161,162.0 54,967.6 379,791.0

2005 95,467.0 42,240.1 268,896.0 31,127.1 343,768.0 740,002.0 174,944.0 70,076.1 417,226.0

20 103,146 48,277 282,300 36,131 376,808 868,133 188,285 86,685 466,893

Data Source: Bankscope.