Associate PRM Exam Reading Material - adapted excerpts from the PRM Handbook and relevant articles

This document contains reading material that is part of the syllabus of the Professional Risk Manager’s International Association’s (PRMIA) Associate Professional Risk Manager (APRM) exam.

The material is made up of adapted excerpts from the Professional Risk Manager’s Handbook, the definition of PRMIA’s standards of risk management, and other material. This document contains: a. An adapted excerpt from the PRMIA Handbook, Edition 1, 2004 on Credit Risk Management b. A paper on Counterparty Credit Risk Management The complete syllabus consists of this document, the Essentials of Risk Management (Crouhy, Galai and Mark - McGraw-Hill, 2006), excerpts from the PRMIA Guide to Financial Markets (McGraw-Hill, © 2008) – in the old PRMIA Handbook (PRMIA Publications © 2004), these are Chapters I.C.1 to I.C.8 in Volume I - available on the PRMIA website as a separate document, the PRMIA Standards of Best Practice, Conduct and Ethics, PRMIA Governance Principles and the Case Studies – all these except the Essentials of Risk Management are available on the PRMIA website.

Copyright © 2008 The Professional Risk Managers’ International Association.

1

counterparty ratings. Copyright © 2004 J.1 Introduction After a decade of success in the field of market risk modelling mainly driven by advances in derivative pricing and computer technology. credit modelling team.From the PRM Handbook – an adapted excerpt from Volume III. 2. credit risk policy and reporting teams. 2 We will introduce terminology and show components of a detailed credit risk report below. 1 Partner. discussion about the extent to which these developments actually find their way into daily credit risk management seems to be absent. that one is tempted to ask to what degree the wealth of new information has reached the ‘central figure’ in this context: the credit risk manager! While the business of publishing the latest developments in glamorous magazines and discussing it at conferences has reached a high point. These two worlds. Global Financial Services Risk Management. exposures. etc. Ernst & Young. Although a significant part of this shift is attributable to regulatory initiatives such as Basel II. pricing on the one hand and risk management on the other. The portfolio group supports the credit officer by complementing a typically rather transaction-focused view with monitoring expected and unexpected losses of the credit book. Jörg Behrens1 III.1. This includes responsibility for the management of credit-related work groups such as the credit portfolio group. etc.B. 2 . the past years have seen a clear shift towards credit risk modelling. including information on limit excesses. let us browse through his job description in order to better understand his responsibilities: 1. are now converging as credit risk is actively managed using credit derivatives. concentrations. Compilation and responsibility for credit risk reporting. an important contribution comes from the demand for more sophisticated credit products and from better management of credit risk.B. Before looking at a typical day in the working life of a credit risk officer.1 Credit Risk Management Dr. Behrens and The Professional Risk Managers’ International Association. In fact so many methods have been developed and so many papers are being published. reviewing provisions. As a result a wealth of new models for pricing and risk management has been developed. Credit Risk Management III.B.2 Supervision of credit data quality process and delivery of all critical credit risk information to various stakeholders and board level. The credit modelling team reviews the bank’s credit risk methodologies and needs to be independent of the business. Ownership of credit risk function and framework.1.

but how does it translate into practice? III.1. including limit setting. Table III. Behrens and The Professional Risk Managers’ International Association.1). This also allows a deputy to smoothly take over the work whenever he is not around. Such a list looks good on paper. Credit Risk Management 3.1: To-do list • Review strategic credit positions: o Any changes to largest exposures (net of collateral)? o How about changes to counterparty ratings? o Any significant credits to be approved by chief credit officer or board? Credit limits and provisions: o Any limit excesses? o Limits to be reviewed? o Provisions still up to date? o All concentrations within limits? Credit exposure: o All exposures covered and correctly mapped? o Any ‘wrong-way’ positions? Credit reporting: o All significant risks covered in credit report? o Report distributed to all relevant parties? o Any significant credits that must be discussed at top management/board level? Stress and scenario analysis: o Any surprises from stress and scenario analysis at portfolio or global level? o Anything not covered by current set of scenarios? Provisions: o Any past or anticipated changes in general loss provisions? o Any changes to specific provisions? Documentation: o Full documentation in place for all transactions? o Break clauses and rating triggers fully recognised? Credit protection: o Credit protection utilised and understood? o Any further possibility to exploit credit protection? • • • • • • • Copyright © 2004 J.1.1. In particular. 3 . 5. Ownership of credit processes.B. 4. Representative for credit risk.2 A Credit To-Do List In order not to get sidetracked by too many ‘other’ issues such as administration.B. our credit risk officer has produced a ‘to-do’ list to remind him of the key tasks he wants to focus on (Table III.B.1. credit stress and scenario testing and calculating capital requirements.From the PRM Handbook – an adapted excerpt from Volume III. dealing with external credit bodies such as rating agencies and regulators. provisioning. ownership of all credit risk-related aspects of Basel II. Benchmarking of performance of credit risk functions within business units according to group credit policy.B.

Credit Risk Management Clearly this list is not exhaustive. for example to better integrate new businesses into the existing credit framework. The credit risk officer must ensure that model calibration is done seriously so that credit granting is based on the best information available and provisions are computed with realistic default probabilities and loss given default values.1. • Review of credit process: Another topic that is usually triggered by ‘events’ – for example. Copyright © 2004 J. Clearly the frequency of such meetings varies with the volatility of the business. This is a good opportunity to ‘clean up’ legacy positions.From the PRM Handbook – an adapted excerpt from Volume III. a large unforeseen loss. • Review of credit strategy: The review of strategic positions clearly is a frequent key task for any credit officer.B. Clearly there are other tasks with an inherently long time horizon. it does not address topics with an inherently long time horizon.1. Some of these were listed in the job description at the beginning of this chapter. III. These sometimes start with a complex ‘spreadsheet trade’ on a remote desk rather than with a formal approval of the ‘new business committee’ and might require a dedicated infrastructure including appropriate management information systems. • Review of stress testing: Credit risk officers will also from time to time meet with economists to hear about changes in the ‘global economical outlook’ and adapt their stress scenarios accordingly. 4 . Behrens and The Professional Risk Managers’ International Association.B. or when a new chief credit officer assumes his role – is the review of the credit process itself.3 Other Tasks We have mainly focused on the credit officer’s more frequent tasks. In the following we highlight a few tasks that are usually addressed only annually but then sometimes lead to very controversial discussions. In particular. However. Backtesting should be performed by an independent modelling team that also approves any new models before they go into ‘production’. perhaps even unasked: ‘Do we really want to be in that business? Are we fooling ourselves when we believe we can exploit what is called “arbitrage” today?’ • Review of models: Backtesting of ratings and models for ‘loss given default’ is not only required by Basel II. a change in regulatory requirements. a more fundamental question often remains unanswered.

is an essential qualitative requirement.1. Behrens and The Professional Risk Managers’ International Association. while ‘good risk management’ links risks to returns. 5 . under the Basel regulations. Credit officers have begun actively using these instruments to manage credit risk. and some firms have started to significantly expand the credit officer’s role and in fact integrate the two functions.1.From the PRM Handbook – an adapted excerpt from Volume III. Credit Risk Management III.B. ‘Good business’ links returns to risks. So it seems that the new credit risk officer. Copyright © 2004 J. unlike his traditional predecessor. we nowadays see joint teams exploiting opportunities to integrate both aspects. identifies opportunities and provides insights? Can he proactively support the business to better generate value for the firm? The answer is a clear yes. the head of the business unit has the job of optimising profits. the traditional role of the credit risk officer is to control the business units and avoid losses.4 Conclusions In the simplest terms. Another development is the formation of portfolio teams comprising business people and credit risk managers who jointly explore opportunities to do ‘the right trades’ – trades that optimize return versus risk. at least for banks. By contrast. In addition to joining the business team and providing it with a better perspective on risk. Whereas in earlier times traders would have done ‘their business’ and a credit risk officer would have either ‘approved’ or ‘blocked’ the business.B. What used to be a two-stage process has been integrated into one. But is a credit risk officer necessarily restricted to a defensive role – ‘to protect’ – or can he also be someone who actively makes suggestions. can directly benefit from many advances in credit risk management. This is best highlighted by the use of credit derivatives. the new credit officer has the role of independent oversight of credit risk which.

Furthermore. however.htm for an interesting summary. there was only limited understanding of the presettlement credit implications of term derivative transactions. 6 . Copyright © 2003 David M.ch/134710. which meant there was no settlement risk. Ignorance is bliss – pre 1985 There are many claimants to the title of first swap arranger. The primary form of credit risk associated with trading in those days was settlement risk. one cannot claim loss of future interest payments on a loan when a borrower defaults. 1974. Some people thought that since there was no exchange of principal on an interest rate swap there was no credit risk.The Evolution of Counterparty Credit Risk The Evolution of Counterparty Credit Risk Management David M. Looking beyond the settlement risk issue. but the date of the first swap appears to have been sometime in 1981. By now the truth is probably lost in the mist of time.ifci. Increased Materiality and Visibility 3 See http://newrisk. from the time payment instructions became irrevocable to the time the agreed exchange amount was received. Recognition of this had been driven home painfully by the failure of Bankhaus Herstatt during the trading day on June 26. Rowe and Risk Magazine. As swap dealers were quick to point out. Since then there has been much progress in measuring and managing the associated risks but much remains to be done. Such volumes gave rise to serious concern about the credit risk of such contracts. daily turnover in the foreign exchange (FX) spot and forward markets was in the hundreds of billions of dollars. At that stage. These factors combined to make pre-settlement credit exposure a latent and largely ignored risk aspect of derivative trading in the early 1980s. derivative volumes were small and counterparties were virtually all investment grade names.3 Even in the early 1980s. I have even heard of deal parties to celebrate each swap like those often held after the closing of large syndicated loans. these were truly arranged contracts. settlement on most of their transactions was on a net basis. After all. Rowe Pre-settlement counterparty credit exposure came to the fore in 1986 when the Bank for International Settlements (BIS) insisted that such credit exposure be recognized in the first Basel Capital Accord. The fact that most credit officers were thoroughly steeped in the historical cost accounting framework was an additional obstacle to a clear understanding of these mark-to-market instruments.

For swaps. The First Basel Accord The derivative credit exposure issue received added attention when the Basel Committee on Banking Supervision began to develop its initial rules for minimum regulatory capital requirements in 1986 and 1987. Initial market values only reflect the slightly off-market impact of a dealer’s bid/offer spread. swap market volumes had grown significantly and so had current mark-to-market exposures. First. This initial value. b. The committee realized that the current market values of a bank’s derivative contracts did not capture their full potential credit exposure. As a result. this is intended to limit the maximum exposure at default. It was neither proposed nor intended as a satisfactory approach to measuring individual counterparty exposures. however measured. however. historical cost accounting treats the principal amount as fixed. market fluctuations lead to constant revisions in the exposure. based on volatility in the underlying market data and the resulting impact on fair values of broad categories of trades. 4 In current terminology. When making a traditional loan it doesn’t require complex mathematical analysis to answer the question “how big is the loan?” The answer to that question is clearly stated in the proposed term sheet. 7 .The Evolution of Counterparty Credit Risk By 1986. As such. Gradually the realization spread that these positive market values represented material credit risky balance sheet assets similar to corporate and industrial bank loans. they are usually done near par. is not a realistic measure of the true potential loss from a downgrade or default on the part of the counterparty. The approach was intended to be simple to implement so that even small banks with a few hedge contracts could perform the calculation without difficulty. they set out to derive reasonably simple rules to calibrate potential increases in such credit exposure. Rowe and Risk Magazine. a. Several comments are in order regarding the structure and parameters of the initial Basel add-on calculation. is only a small fraction of the potential future exposure that may materialize as a deal ages and market conditions change. Such unstable exposure presented an entirely new dimension of uncertainty that many traditional credit control personnel found difficult to incorporate into their thinking. The result was the now well-known mark-to-market plus add-on approach to measuring potential future exposure. there are two problems. While some would point out that the fair value of a loan fluctuates just as the price of a bond does. Thus current exposure. counterparties to such trades clearly should be subjected to credit review and transactions should only be done within an approved credit exposure limit.4 But there was one nagging problem. especially current exposure on newly executed deals. Second. It was only intended to capture the increase in aggregate exposure. Copyright © 2003 David M.

It provided only very limited recognition of the tenor of exposure by offering different parameters for contracts with more than or less than one year to maturity. f. it reflected an inherent distrust among traditional credit officers of these fancy new instruments. d. will produce an unrealistic increase in measured exposure in both cases. Primarily. it was felt the volumes could be limited and the associated risks constrained. most institutions adopted some variation in the add-on method for tracking and setting limits for counterparty credit exposure. however. it completely ignored the degree of co-variation in value among multiple deals. the marginal change in potential exposure from adding or deleting a single deal was highly unreliable and could even be directionally incorrect. and is. market risk was viewed as the most serious issue requiring attention and resources. individual counterparty portfolios exhibit widely differing diversification characteristics. it was. It treated each transaction in isolation.The Evolution of Counterparty Credit Risk c. If the two trades are not legally nettable. The most obvious and striking example is an offsetting transaction where the market value moves in the opposite direction to that of an existing trade in response to changing market conditions. Despite these widely recognized shortcomings. g. This was driven by the comparative simplicity and modest cost of deploying such a system. e. 8 . While this greatly simplified the mechanics of the calculation. the second trade has almost zero impact on potential exposure. then the second trade significantly reduces potential exposure. This resulted in greatly inflated potential exposure Copyright © 2003 David M. While adequate for estimating aggregate potential exposure. In fact. Rowe and Risk Magazine. If they are legally nettable. bank credit departments invariably insisted on more conservative parameters internally than those employed in regulatory capital calculations. The essential problem is the implicit assumption of an average degree of diversification. The same add-ons were applied to both at-the-money and away-from-the-money contracts. Recognizing that portfolio effects had been ignored. despite the amortization toward zero in the value of many contracts as they approach maturity. By making the numbers larger. While adopting the add-on approach. the parameters of the method were calibrated to reflect an assumed average degree of portfolio diversification. A negative consequence of Basel’s add-on method was to provide apparent regulatory sanction for this approach. however. Building a more sophisticated system for controlling trading credit risk simply was not a serious consideration at most institutions. Partly this was driven by recognition of the fact that the purposes for the two calculations were different. The add-on approach. quite unsuitable when applied at the individual counterparty level or to evaluating the marginal exposure of a new deal. Given the huge trading losses experienced in the mid-1980s. While the potential increase in aggregate exposure was probably not unreasonable.

This introduces an added layer of complexity to the analysis. Traders realized that add-on based exposure estimates. Rowe and Risk Magazine. even in the U. Certain deals are entered into under terms of an enforceable netting agreement. Indeed.The Evolution of Counterparty Credit Risk estimates that often had only an infinitesimal chance of materializing as actual exposure-at-default in the future. They particularly recognized that marginal exposure implications were unreliable. cannot be offset against exposure in other nettable pools or against transactions done outside any enforceable netting agreement. unsophisticated credit exposure measurement systems tended to intensify the inherent cultural conflict between traders and credit officers. The International Swaps and Derivatives Association (then known as the International Swap Dealers Association) (ISDA) lead the way in most countries around the world. Organisational Implications Unfortunately.S. 9 . it is still far from universal. It also means that legal contract information must be captured and properly reflected in the calculations if they are to be robust in reflecting the impact of enforceable netting. netting is of questionable enforceability when dealing with certain types of institutions. if negative. ISDA formulated standard contract language and waged multiple campaigns to gain legislative recognition of the enforceability of netting under the terms of such contracts when one party declared bankruptcy. While we think of netting as the accepted norm today. They insisted that they were being constrained by a system that was inconsistent and arbitrary while credit staff felt obliged to defend the indefensible. Netting Comes into its Own Another trend in the late 1980s and early 1990s was the legal battle to make netting enforceable under the bankruptcy laws in various countries. This became painfully obvious when counterparties did offsetting trades to neutralize future changes in their net positions and the system indicated that these transactions increased credit exposure! All this reinforced traders’ cynicism about credit oversight in general. at least at the counterparty level. were inconsistent and inflated. These are nettable against each other but their combined net market value. The most frequent circumstance today is for netting to be partially enforceable across relevant sets of transactions. Copyright © 2003 David M. Examples are insurance companies and public utilities where bankruptcy claims are adjudicated in some type of state administrative process rather than under the US bankruptcy code or laws governing federally insured financial institutions.

Rowe and Risk Magazine. outstanding loan balances and potential trading credit exposures were two different currencies with an uncertain exchange rate between them. Credit approvers grew more reluctant to approve increased trading credit limits in the face of already large measured exposures for good names and for new exposure to lower quality names. the impetus for change came at the point when credit availability began to constrain the ability to do more business. the prevailing attitude on the credit risk side often tended to be. Credit limits were generally available to cover even the inflated exposure estimates produced by a conservative add-on approach. volumes continued to grow while smaller and less creditworthy counterparties entered the market. it was impossible to generate consistent credit decisions across the organization. period. Failure to treat the dynamic interaction among trades in a counterparty’s portfolio was recognized as the most obvious shortcoming of the add-on approach. in effect: “I don’t care how inflated the numbers are. Eventually situations arose where credit officers would say. As long as volume was modest and counterparties were solid investment grade names this was not a serious problem. Discussions generally settled on the following essential features. The result was an unhealthy behavioral feedback loop. 10 . Moving into the 1990s. however. I’m not putting my name on an approval bigger than the current limit. This was particularly obvious relative to marginal exposure calculations. Gradually the old arguments about the exposures “not being real” and the estimates “being inflated” became less compelling. The Impetus for Change More often than not.” This resulted in a real business incentive for more realistic exposure measurement and a willingness to spend some money to make it happen.The Evolution of Counterparty Credit Risk Given the cost and complexity of implementing an improved exposure measurement system. In such a situation. In effect.” Unfortunately. however. this tended to undermine the credibility of the exposure estimates even among credit approvers. Essential Features for Consistent Exposure Estimation Given the incentives and resources to build an improved exposure estimation system. a) Proper recognition of offsets and diversification. “make the numbers bigger and that will make us safer. it was necessary to consider what features were important and in such a system. Justify more realistic estimates and then maybe we can talk. Each limit approver was required by circumstances to decide how to translate the one to the other. An addon based system would show exposure increasing when a new trade was added even when it Copyright © 2003 David M.

The Evolution of Counterparty Credit Risk actually reduced exposure by offsetting pre-existing imbalances in the counterparty’s portfolio. Likewise. a small number of similar deals.5 c) Exposure profiles rather than single loan equivalent amounts. The common thread running through these essential features was an attempt to incorporate all significant structural factors into the exposure estimation process. Specifically. Few things do more to undermine the credibility of a credit risk system than getting the sign of marginal exposure wrong! b) Incorporation of market data correlations. This was based on the portfolio ratio of current net exposure (to the extent netting is deemed enforceable) versus current gross exposure. but this ignores the fact that interest rate swap exposures peak in the middle of a contract’s life. Rowe and Risk Magazine. Again. Capturing the impact of enforceable netting in the current exposure is easy. Sometimes the potential exposures by deal would be “stacked” in the order of their maturity. A more rigorous approach was clearly necessary. since credit exposure must be simulated  many months and even years  into the future. By the 1990s. Another major flaw of the add-on approach was that it gave no insight into the timing of potential future exposure. A minor revision to the Basel capital rules in April of 1995 introduced a very crude way of reflecting the growth in the enforceability of netting. Credit exposure should be based on estimates of longer­term trend correlations  rather than daily change correlations. This made a reliable treatment of netting increasingly important for accurate exposure estimates. 11 . d) Rigorous treatment of netting. the probability of the projected potential exposure actually materializing on any relevant future date should be as consistent as is practically possible. Nevertheless. netting was becoming more widely accepted in law around the world. The goal was to produce results that were conceptually consistent across counterparties and throughout the projection period. most analysts agreed that incorporating reasonable estimates of such correlations was an important aspect of realistic exposure estimation. or a large and varied range of deals with two-way sensitivities and multiple market drivers.  Copyright © 2003 David M. 5 A less commonly recognized point was that correlations appropriate for market risk may not be appropriate  for credit exposure estimation. Whatever else it did. a revised trading credit methodology needed to produce insights into the timing of exposure. Any credit analyst will say that the timing of when exposure occurs is often central to whether the risk is acceptable. Clearly correlations between pairs of rates and/or prices are among the less stable parameters of market dynamics. The difficult issue is reflecting it accurately in potential future exposure. It should not matter whether a portfolio contains one deal. not at the end. the statistical properties of the exposure estimates should not change over the course of a projection as some deals mature and the complexity of the portfolio changes. while arguably appropriate for an aggregate exposure adjustment. this approach is much too crude to be reliable when applied to individual counterparty portfolios.

Even if sophisticated exposure profiles are calculated periodically and fed back into the trading limit system. Nevertheless. especially grid computing. there is little excuse for any institution with significant derivative activity to operate with nothing but add-on style credit exposure estimates for internal measurement and control. the counterparty exposure profile or profiles affected by a new deal are updated in the background. and effective analytical shortcuts offer the possibility of updating such profiles intraday and potentially in real time as traders check limits and book new deals. This approach requires capturing new trades as they are made and transferring them to a central exposure evaluation system. such batch systems also can provide valuable portfolio concentration and risk source sensitivity information. daily trading activity is measured by fairly unsophisticated metrics. b. a. Properly configured. and the volume of trades being processed in this way. A key success factor is making such information available to relevant decision-makers in multiple locations. resets. deals done between updates are typically assessed in a much simpler fashion. the latency is likely to be from a few minutes to an hour or longer. Alternate Aspirations for Use of Simulation-based Exposure Given the advances of the past ten years. barrier breaches and any other structural event over the horizon of the simulation. Depending on the size of the portfolios affected. Given that. this limited capability can provide support for booking risk-reducing deals that Copyright © 2003 David M. it is possible to configure such a system to allow a small number of what-if simulations based on proposed new deals. 12 . In this approach. Such a system avoids most of the problem of unrealistically inflated exposures but it cannot offer meaningful insight into the marginal impact of a proposed trade. While not able to support a large volume of whatif inquiries. Initially this was so computationally burdensome that daily calculation of such exposure profiles was all that could be commercially justified. it is a giant step beyond having nothing but add-on based exposure estimates. Also. The advantage of this approach is that portfolio effects are reflected faster than with nothing but periodic batch updates of the simulations. there are three distinct levels of sophistication in the application of simulation-based exposure estimates. at least relative to newly booked transactions. Once received centrally. Rowe and Risk Magazine. with appropriate aging of the portfolio to recognize payments. More recent advances in computational power. Intra-day simulated exposure updates for new trades.The Evolution of Counterparty Credit Risk The Challenge of Dynamic Monte Carlo Simulation Gradually a consensus emerged that dynamic Monte Carlo simulation of the portfolio. if a counterparty’s exposure is close to the limit. As an overnight supplement to less sophisticated intraday measures. deal maturities. was the logical approach to effective credit exposure estimation.

reflecting all major effects of offsets and diversification. it would offer significant advantages. By providing an accurate estimate of the amount and timing of incremental exposure. this aspiration was commercially impractical. the cost of such a change is often viewed as excessive.The Evolution of Counterparty Credit Risk would otherwise be denied based on limit excesses. Rowe and Risk Magazine. Copyright © 2003 David M. there is a major difference. Traders’ performance could be measured on the basis of risk adjusted returns not gross trading profit. Sufficient trade details need to be captured to allow the simulations to run and exposure updates need to be made while new limit inquiries and transactions are taking place. the cost in hardware and support has heretofore made it commercially impractical. While this may seem like a small departure from the first configuration. such a system would encourage aggressive pricing for exposure reducing deals. Use of simulation-based exposure for what-if limit checking and pricing implications. 13 . To achieve intra-day updates requires moving from a batch to an event driven system architecture. This requires a significant revamping of the entire process when starting from a batch mode approach. In addition to creating a minimum spread for exposure increasing deals. this would be the basis for reasonable estimates of the cost of the associated incremental credit risk. Given the rather modest improvement in business functionality. c. Given recent advances. While such a system could have been built technically using Monte Carlo techniques. If accurate simulation-based exposure could be performed on a what-if basis in a matter of seconds. Until the recent improvements in computing power and advanced analytical approximation methods were available. such a system becomes far more commercially justifiable.