Chapter one
1.1 Overview of RBS
The Royal Bank of Scotland (headquartered in Edinburgh) was founded in 1727, making it one of the oldest UK banks. Over the years RBS has grown significantly and now serves over 30 million customers across four continents. Namely: Europe (Including United Kingdom), Middle East, Americas and Asia. Growth was driven mainly through an aggressive acquisition strategy. This resulted in RBS becoming a holding company for many smaller household names such as NatWest, Coutts, Ulster Bank. By 2007 RBS had become one of the largest banks in the world, acquiring Dutch Bank ABN Amro as the lead consortium member in 2008.
Following the Acquisition of ABN the Bank was hit hard by the financial crises due to high leverage, risk concentration and business stretch. Consequently the bank was bailed out by the UK government in October 2008. The government paid £20bn in return for an 84.4% stake of the Bank.
In response to the 2007/08 financial crises RBS drastically changed its strategy from ‘becoming the world’s biggest bank’ to downsizing to become a trusted and sustainable UK based bank. The latest reports released by the Bank (Annual Report 2015) state that RBS has begun regaining government shares, with the state owned percentage falling to 72%.
The Bank is currently split into three main operating divisions: Personal and Business Banking (PBB), Commercial and Private Banking (CPB) and Corporate and Institutional Banking (CIB). For the purpose of this research project CIB is the only relevant business area. CIB is the wholesale banking arm of the Dutch branch. Following the retraction of RBS from many European countries only coverage teams will remain in the Netherlands, with back office services such as Credit and Capital being migrated to London CIB. Products offered by the Dutch coverage teams are as follows:
• Lending: this includes standard loan facilities, both committed and uncommitted.
• Trade Products: Guarantees and Letters of Credit
• Derivatives: Instruments used to hedge risk (Interest and Currency)
1.2 Statement of the Problem
Credit exposure is the risk of borrowers defaulting on payments required by the bank in return for the credit extended to them. There is an element of credit exposure in all banking products. Clients are required to pay a rate of interest (return) for the risk the bank takes on in extending credit to them. In line with the Risk-Return trade-off principle, loans or assets with high levels of risk associated are expected to generate high levels of return. In banking terms higher interest is charged on loans with lower credit ratings.
Following the acquisition of ABN Amro RBS had taken on £6bn worth of CDO (Collateralised Debt Obligations) of which £4.5bn consisted of subprime mortgages. These investments put RBS in a less than desirable position with enormous concentration in leveraged finance and commercial real estate lending. The lack of sector diversification in the books opened the Bank up to concentration risk. This is a risk posed to a Bank when a group of exposures, in this case RBS’s CDO’s, has the potential to produce losses that have the ability to threaten the going concern of the Bank.
The acquisition of the ABN’s Investment Banking arm doubled RBS assets from 870bn GBP to 1.2 trillion GBP. The CDO’s were attractive as they offered high rates of return. As mentioned previously this high return was offset by the low credit quality of the underlying assets. Funding such an enormous balance sheet became increasingly difficult and with a leverage of 100 times and large sector concentrations banks and investors began cutting risk appetites towards RBS. When the subprime mortgage market collapsed RBS was left holding virtually worthless assets backed by enormous levels of debt. With funding drying up and the Bank suffering losses the UK government had no choice but to step in using tax payer’s money to prop up the ailing bank.
In the absence of the extortionately high concentration in leveraged finance and commercial real estate lending the Bank may have been able to survive the economic downturn. When a lending portfolio is sufficiently diversified the risks are spread across different sectors/industries/assets and therefore significantly decrease systemic risk.
Lack of diversification was not the sole contributor to the near collapse of RBS, a lack of capital meant that the Bank was unable to absorb the level of losses being realised from the concentrated loans. This threatened the liquidity of the Bank and if it had not been for the UK government stepping in the Bank would not have survived. This research paper will focus solely on the Credit Risk Concentration element of this problem due to time constraints.
1.3 Objective of the study
• Understand RBS risk measures used to create sector risk appetites
• Compare theories underlying RBS portfolio diversification with MPT
• Analyse if Corporate Coverage teams are fulfilling RBS models for portfolio diversification
• Recommend actions for CIB Amsterdam to reduce credit risk exposure on WE portfolio
1.4 Methodology
The source data for this research paper was collected from The Royal Bank of Scotland, Amsterdam Headquarters. The data consists of confidential information detailing client names, facility types and exposures including risk measures, appetites and policies. For the purpose of this study only high level data (sector overviews) will be shown as not to breach confidentiality. RBS internal credit grading models were used for analysis alongside excel models.
Qualitative information has been collected from many different sources, such as academic articles, books, interviews and other online sources.
Add to at end
1.5 Justification
Many banks have acknowledged that a sever lack of risk management within their credit departments was a major contributor to the 2008/09 financial crises. This awareness has paved the way to tighter regulation and closer monitoring of credit risk management. RBS has a comprehensive set of risk management software’s it uses to monitor and report credit risks. These models have been tested and approved by financial regulators for the banking group as a whole. However the sector diversification for the Western European lending portfolio has not been analysed in particular detail as the group credit standards are within limits and therefore do not pose a going concern risk on the Bank. This paper gives RBS deeper insight into the level of exposures they currently have in specific sectors ……….results…….
1.6 Limitations
• Not looking at capital modelling
• For the purpose of this study I will analyse credit risk under “default mode”. This is restrictive as it does not take into account any migrations of credit risk over the life of a facility.
• Sectors have been restricted
• The RBS Wholesale Credit Risk Appetite Framework accounts for three forms of concentration: Single Name Concentration (SNC), Product and Asset Class Concentration and finally Sector Concentration. For the purpose of this research paper I will only be looking at the Sector Concentration framework.
• Each legal entity in RBS lending book is assigned with a specific SIC code, this is used to identify the sector cluster, sector, and subsector. I will use the Sector cluster classifications to perform my research on the sector optimization of RBS Western European lending portfolio.
• RBS risk appetite frameworks are still evolving, therefore there are still legacy positions which pre-date current appetites and therefore fall outside these parameters.
1.7 Organisation of Thesis
Chapter one of this Thesis outlines the background of the study, an overview of RBS as an organisation and the limitations to this research.
Chapter two is focussed on secondary research. It covers some of the regulations relevant to credit risk and gives definitions for banking terms used throughout the paper. Relevant theories on portfolio modelling and predicting defaults are discussed.
Chapter three applies the theoretical aspects of this paper to the current operations within RBS regarding managing and monitoring credit risks.
Chapter four outlines the current lending portfolio for the Western European.
Chapter five is my analysis of the current portfolio, where it differs from what is defined as being an optimal mix and any reasoning as to why the portfolio may differ.
The final chapter is a conclusion of this study and includes my recommendations to the management of RBS.
Chapter two
2.0 Bank for International Settlements
In 1930, 60 central banks came together to form the Bank for International Settlements (BIS). The purpose of BIS is to ‘serve central banks in their pursuit of monetary and financial stability, to foster international cooperation in those areas and to act as a bank for central banks’ (Bisorg, 2016). Within BIS there are many committees and associations formed to set guidance and regulation on a wide spectrum of banking issues. One of these committees is the Basel Committee; the purpose of the Basel Committee is to set prudential regulation for banks and ‘provide a forum for banking supervisory matters’ (Bisorg, 2016). One of the main concerns for the Basel Committee is capital adequacy within financial institutions. Capital is an extremely important function in banking as it is essentially a buffer, protecting debt holders if the bank were to experience losses larger than the bank had provisioned for. In 1988 the Committee released its first accord (Basel I) the accord was predominantly focussed on credit risk. The accord introduced the concept of having a two tiered capital structure. The first tier is the ‘core’ capital, it’s formed of reserves held aside to cushion against potential future losses or income variations. The second tier, known as ‘supplementary capital’ is all other forms of capital such as: gains on investment assets or long-term debts. Credit Risk was quantified as Risk Weighted Assets (RWA) with total capital mandated at being at least 8% of the banks RWA’s.
In 2004 the second Basel accord was released, with the purpose of amending the standards controlling how much capital banks are required to hold against financial and operational risks.
The latest accord released was Basel III. The new accord comprises of two capital adequacy measures. The first of these being capital to cover potential future losses from defaults. The second is capital adequacy measures for potential future defaults due to changes in counterparty credit quality. The capital structure remains the same, formed of two pillars; changes have been made to the purpose of each pillar. The first pillar should be sufficient to satisfy default levels at any given moment, whereas the second pillar is should provide a large enough buffer to cover any future capital requirements.
2.1 Credit Risk
Commercial banks generate returns, through lending funds in return for interest payments over a specific time horizon. The funds lent are credit, this is defined as “A contractual agreement in which a borrower receives something of value now and agrees to repay the lender at some date in the future” (Root, 2003). When the Bank enters into this agreement they take on a certain level of credit risk, this is the risk that the counterparty will default on the repayment of the loan.
Default situations can be classified by any of the following:
1. Delaying payments (temporarily or indefinitely)
2. Deterioration of credit rating forcing the restructuring of debt
3. Bankruptcy
Basel II legislation states that a firm is said to be in default if either or both of the following situations are met:
I – The credit institution considers that the obligor is unlikely to pay
its credit obligations to the credit institution in full, without recourse
by the credit institution to actions such as realizing security (if held).
II – The obligor is past due more than 90 days on any material
credit obligation to the banking group. Overdrafts will be considered
as being past due once the customer has breached an advised limit or
been advised of a limit smaller than current outstanding’s.
(Bisorg, 2004)
Traditionally Credit Risk has been, and still is, the largest risk facing Banks. Although many studies have been performed on this risk there is no proven way to mitigate it completely. Therefore managing the level of risk a Bank takes on is an essential part in ensuring the continuity of its lending activities. When borrowers apply to take out loans at a Bank they will first have to go through credit checks. These are essential to assess the ability of the borrower to repay the both the principle and interest of the loan. This is otherwise known as an assessment of the borrower’s creditworthiness. Creditworthiness is not fixed indefinitely; there are many factors that can cause a decline in a customer’s creditworthiness. Examples of these include poor management or external factors such as competition, rising inflation and volatility in asset values. Consequently even clients that have solid credit ratings impose credit risk on Banks.
Credit Risk can be defined as:
Credit Risk = max (Actual loss – Expected loss)
To quantify Expected Loss Banks must rely on a series of probability measures. There are three commonly used measures; these are Probability of default (PD), Exposure at Default (EAD) and Loss Given Default (LGD).
Expected Loss, under default mode, for a single loan facility can be defined as:
Expected Loss = LGD x EAD x PD
Probability of default is the likelihood that a borrower becomes unable to repay their loan and will therefore fall into default. The probability of default is a percentage; it is created following an analysis into the client’s financial health and non-financial aspects such as quality of management, industry and geographic location.
Exposure is the current outstanding amount a client is indebted to the Bank. Exposure at Default is an estimate of the total amount the Bank is exposed to at the time a default occurs. Loss Given Default is the amount the Bank can expect to lose at the point of default. The LGD is based on a number of estimations and therefore the Basel Committee on Banking Supervision has singled out specific lending facilities that have fixed LGD percentages. For example: “Senior claims on corporates, sovereigns and banks not secured by recognized collateral are assigned at 45% LGD” (Bessis, 2015).
Rating agencies such as Moody’s produce credit ratings for publicly traded companies, however these do not cover small and medium-sized companies, thereby creating the necessity for Banks to create processes for rating the credit worthiness of all of their corporate and retail clients. Obtaining credit ratings is an essential step in being able to calculate the probability a client will default on its debt obligations. A client with a low credit rating signifies a lack of free cash flow indicating it will encounter more difficulties in repaying and of its liabilities such as a loan. Therefore a Bank with a client bearing a low credit rating will predict a higher probability of default.
The PD, EAD and LGD are measures used on every facility taken out by a client, even if the same client has multiple facilities these risk measure must be calculated per facility. Once the risk is known on a facility basis it is possible to calculate the total exposures and risk of loss for an entire portfolio. “Sound practices now involve measuring the credit risk at a facility level and portfolio level” (Bessis, 2015).
There are ways that a Bank can reduce the level of risk it exposes itself to on a facility. One way to do this is to ask the counterparty to post collateral against the value of the loan. Collateral is usually made up of assets that the Bank is able to liquidate for example low risk bonds. The presence of collateral reduces the level of credit risk. Another way for Banks to reduce their credit risk exposure is to require the guarantee from a third party that if the Bank’s client were to default at any point of time the third party will cover the remaining exposure.
2.2 Predicting Defaults
In 1968 Edward Altman pioneered a statistical formula derived from five accounting ratio’s that can be used to predict default rates. The original formula was designed for use on publicly traded manufacturing companies, however the model has since been updated for use on companies from all industries (provided they are publicly traded). The theory is known as “Z-Score” (Altman, 1968).
The equation used to calculate the Z score is as follows:
Z = 1.2 X1 + 1.4 X2 + 3.3 X3 + 0.6 X4 + 0.999 X5
Where
X1: Working Capital/ Total Assets
X2: Retained Earnings/ Total Assets
X3: Earnings before interest and taxes / Total Assets
X4: Market value of equity / Book value of total liabilities
X5: Sales/ Total Assets
The outcome of the formula is a weighted score, the Z score. If this score is above 3.0 then the company is considered stable and financially sound, therefore unlikely to default. If the score is between 2.7 and 3.0 then the company should be put on ‘watch’ for any adverse changes in its operating environment, and below 2.7 was categorized as high probability of default. The model was initially tested on a sample of 66 companies, of which 33 went into bankruptcy. The model was able to predict this fairly accurately. Since its initial state the Z score theory has been extended and is now used in Basel II frameworks.
2.3 Portfolio Optimization
When an investor is looking at investing in a specific asset or stock they will want to know the returns they can expect to receive for investing their money. Every investor knows there is an element of risk with each investment; therefore they look to invest in the stocks with the highest return for the lowest level of related risk. In 1952 Harry M. Markowitz published an essay titled ‘Portfolio Selection’ in the Journal of Finance. It was that essay that formed the basis of what is now known as Modern Portfolio Theory (MPT). The theory is hinged on the idea that by diversifying the assets you hold, so that they have little or no covariance, you can reduce the level of risk for the portfolio as a whole.
One of the main factors investors will look at first is the expected return. Expected Return is not the actual return you will receive when investing in an asset. It is impossible to predict the actual return, therefore ‘statisticians define the expected value of a variable as its average (or mean) value’ (Hull, 2015). The result is therefore a weighted average of a number of different scenarios’s where the investment could produce a specific return. The weights given to each scenario’s outcome are a reflection of the probability. This is represented as:
E(WiRi) = Wi E (Ri)
Where
E = Expected Return
Wi = The weight of the individual asset
Ri = Anticipated return on the individual asset
This equation is valid for a single asset. When we look at a portfolio of assets the expected return will be the sum of the weighted average expected returns.
E(W1R1 + W2R2 +…+ WnRn) = W1E(R1) + W2E(R2) + … + WnE(Rn)
If you were simply interested in maximizing possible returns you would pick the stock with the highest return and invest all your funds in that. However if there were to be an adverse market movement and the value of that particular stock plummeted you would lose your entire investment. This paves the way to portfolio diversification. By separating your investment over stocks that are completely unrelated it is likely that if one fails another will outperform expectations thereby balancing out the portfolio return.
Risk of an asset is measured by the variance; this is a measure of dispersion around the mean. It is all well and good to measure the risk of an asset individually, however in order to successfully diversify your portfolio it is essential to know how the assets move together. The mathematical term for this is the covariance. Knowing the covariance is essential, as if you buy a set of assets that have a strong covariance your portfolio will not be diversified as it is highly likely that if one fails, the others will aswell.
Come back to this… Stats are an issue (speak to Mikel, how to solve)
Chapter three
3.1 RBS definition of Credit Risk
The definition of credit risk is one that is universally accepted, it is the risk that ‘a bank borrower or counterparty will fail to meet its obligations in accordance with agreed terms’ (Bisorg, 2000) RBS defines three scenario’s that constitute a counterparty failing to meet its obligations as:
Default: whereby no money is recovered
Partial Default: Some of the funds are recoverable
Late payments: Payments that are past 90 days due
Within RBS the basic default model incorporates three measures to predict the expected loss and the unexpected loss on any facility that is approved. The expected loss, denoted as:
Expected Loss = PD x EAD x LGD
Where :
PD: Probability of Default
EAD: Exposure at Default
LGD: Loss Given Default
Is the loss that RBS expects to bear over a certain period (usually one year). Expected loss is seen as a consequence of doing business; therefore these losses are budgeted and accounted for (as impairments). Expected Loss is only calculated for portfolios as for a single loan the value of the loan will either be its book value, or the Loss Given Default. Whenever RBS enters into an agreement with counterparty there is also an element of unexpected loss. Unexpected loss is the risk of loss that exceeds the expected loss within the same time period. Unexpected losses are estimated on a 99% confidence level, the formula is:
Unexpected Loss (99%) = VaR99% – EL
Where:
VaR99%: The Value at Risk, 99% confidence
EL: Expected Loss
Unexpected losses influence the amount of capital the Bank has to hold as there must be enough to protect the bank from larger than expected losses. To calculate the level of capital needed to cover unexpected losses RBS uses Risk Weighted Assets (RWA’s). The RWA’s are calculated by adjusting the EAD by an appropriate risk weight.
3.2 Master Grading Scale
Master Grading Scale (MGS) is a measure of credit worthiness. RBS rates all clients internally with a scale of 1-27, with one being the best. Grades 1-12 are viewed as investment grade and can be mapped directly to Moody’s external rating as BBB and above.
Many companies in Western Europe are privately held, making them smaller organizations. The MGS grading model will downgrade these types of companies as they are seen as being less financially stable. However in the majority of cases this is not an issue, therefore smaller private companies with MGS grading ~14 are still viewed as investment grade within RBS.
MGS scores are derived from the following information needs:
1.) Financial information. Line items from Income Statement and Balance Sheets are required to measure the financial health of a company.
2.) Non-Financial Information: Credit teams are required to fill out a number of questions that are designed to assess the ability of the management of the company, the quality of information the company is able to supply to RBS, the demand and cyclicality of the market the company operates within, the market position they hold and the companies ability access to alternative funding.
The MGS model assigns specific weightings to each information point and will return an MGS score for the company. This framework has been tested and approved by regulators, however if the credit team feel the MGS is too low they are able to override the score with sufficient reasoning.
There are exceptions from this system, RBS legacy clients that are not heavily active and have secured facilities, such as guarantees that have been cash backed, will be automatically assigned MGS 18, unless there has been significant reason to downgrade. The credit grades are then cascaded across any subsidiaries. Subsidiaries backed by parent companies (topco) guarantees are graded equally. Having a lower MGS indicates the company is more credit worthy and therefore the bank is required to report less RWA’s which affect Capital level requirements.
3.3 Probability of Default
Probability of default is calculated in accordance to the MGS score the company is assigned. For each score there are three PD percentages given for three separate scenarios: Low, Medium and High.
** Need info on when each is used**
3.4 Loss Given Default
Loss given Default is a widely used metric for measuring the fraction of exposure on a facility that is effectively unrecoverable when a counterparty enters into default. RBS official definition of LGD is: An estimated percentage of exposure value that is unrecoverable if a client defaults within one year. The LGD is calculated by an internal application called CRADLE (Credit Risk and Default Loss Estimation). CRADLE calculates the LGD per facility per counterparty, however if LGD is needed for all facilities owned by a particular counterparty weightings are assigned to each facility and an aggregate LGD is given. The application derives LGD’s from the information inputted by RBS credit teams. The application first calculates a ‘Core LGD’ based on:
• Industry. Some Industries are by nature riskier than others. Many Industries may also have cyclical aspects which will require additional information to gauge which period within the cycle the market is currently in.
• Seniority of the debt. If the debt is senior it must be repaid before all other forms of subordinate debt, therefore the LGD will be lower.
• Collateral. By posting collateral against a facility the risk of loss will decrease directly with the value of collateral.
Once the core LGD is calculated it is adjusted further by the subsequent calculation steps:
• Country of Origin/Operation. This is an important aspect as it represents the timeliness and enforceability of the legal system in which RBS would seek recovery.
• Type of entity. If the entity requiring the facility is an operational unit it is more likely they will fall into default than if it was a parent company. In these situations parent companies may guarantee the facility, therefore the operating unit will be given the same level of creditworthiness as the parent. If there is no guarantee present the creditworthiness may be downgraded by one or two levels.
• Carve Out Non-Financial Collateral. Haircuts are applied to certain types of non-financial collateral posted against the facility to account for any deterioration over a specific time frame.
• Carve Out Financial Collateral. Haircuts are applied to financial collaterals to account for any deterioration over a specific time frame.
• Cost adjusted LGD. 1% add-on is applied to cover any costs incurred in recovery of the debt.
• Minimum adjusted LGD. If facilities are 100% cash backed, or the collateral is sufficient to cover the entire facility amount a minimum LGD of 1% is applied.
For every facility two forms of LGD are calculated. For reporting purposes RBS will give a Down Turn LGD figure, this is usually 4% higher than the Through the Cycle LGD which is used for internal purposes.
3.5 Risk Appetite
Following the financial crises of 2007/08 banks have stepped up their governance and management frameworks in a bid to prevent mistakes of the past repeating themselves. Historically RBS has suffered due to highly concentrated exposures in a specific sector; therefore a crucial aspect of this tightened governance at RBS is the Credit Risk Appetite and Concentration Risk Framework. As explained in the ‘Limitations’ chapter I will only be focusing on one aspect of this framework: Sector Concentration. The purpose of the framework is to provide a clear understanding of the level of risk the Bank is willing to accept on a sector basis per business unit, in order to ensure the Banks strategy remains consistent, avoids over exposure and maintains capital adequacy. This framework is supported by numerous policies that employees are expected to adhere to on a daily basis.
The sector credit risk appetite is the level of credit risk that RBS is willing to accept in a sector. Appetites are defined for all sectors RBS is active in. The process of setting the appetite limits per sector begins with the business units (CIB,CPB, PBB) in each country. The units are required to submit ‘business franchise papers’ to the Risk Appetite & Policy department in London. The papers need to detail the following:
• Current market share and what is the target state?
• What are the characteristics if customers we propose to support to meet the target state?
• Where do we not want to lend?
• What products do we intend to use?
• Are there any barriers to achieving the target state?
• What do potential customers think of us?
• Why does this strategy make sense from a capital and returns perspective?
Credit will evaluate the proposal from a credit risk point of view, using both historical data and market outlooks to form an opinion on if the proposed strategy is viable or will leave the Bank open to an unacceptable level of credit risk.
Once the business units have completed their proposed sector risk appetite requests the information is sent to the ‘Sector Risk Appetite’ team. This team will then perform their own credit risk analysis on each of the business units proposed limits on an individual basis and collated for the entire group. The outcomes of this analysis and proposed group sector risk appetites are then presented to the Group Credit Risk Committee (CRC) and Group Chief Credit Officer (GCCO). It is down to the discretion of this committee to decide if the proposed limits are acceptable. Under the circumstances the proposal is not acceptable as limits are too high the Sector Risk Appetite team will be required to make amendments to ensure the limits fall in line with RBS strategy. Once the limits have been realigned and accepted by the committee they are cascaded back across the business units. This formal review of all sectors takes place annually.
Each Sector is given a level of oversight. The CRC and GCCO can assign one of four classes to each sector; High, Medium, Low-Medium and Low. The level of oversight assigned to each sector is classified based on the relative size and riskiness. Size is measured as the percentage of RBS Total Committed Exposure in comparison to the portfolio average. Riskiness is measured using the Economic Capital Total loss percentage, using a 96% confidence interval. This is then compared to the portfolio average. The sectors are reviewed on an individual basis according to their oversight classification. Sectors that are high oversight require reviews on, at least, an annual basis. However for low oversight sectors no regular reviews are required.
On a monthly basis Sector Concentration Reports are released by the Enterprise Credit Risk Team, London. These reports are group wide and are used to update business units on the group’s current exposures. Sectors where TCE is within 10% of the limit are highlighted in orange. Further in the report notes are given on all sectors that have been highlighted, these notes give explanation as to why the limit is almost reached and an advice to teams if there are actions that need to be taken.
The final section of the sector report is a ‘triggers’ section. For every sector there are a number of predefined triggers that if breached signal that the sector should be watched closely as there are movements that are not favorable for RBS. For example the decrease in Oil price led to an increase in the % of defaults in the Oil & Gas sector. The level of defaults surpassed the allowable change and caused the sector to be put on watch.