Credit risk transfer: developments and policy implications

Speech by Lucas Papademos, Vice-President of the ECB,
at a Workshop on “Risk management and regulation in banking” jointly organised by Basel Committee on Banking Supervision, the Centre for Economic Policy Research (CEPR) and the Journal of Financial Intermediation
Basel, 29 June 2006

Ladies and Gentlemen,

I. Introduction

“A lot of people approach risk as if it is the enemy, when it’s really fortune’s accomplice”, British rock singer Sting once remarked. You, as participants in this workshop, certainly approach risk with an inquisitive mind; risk is one of your main objects of enquiry – whether it has become fortune’s accomplice for you I cannot judge! I am delighted to be here in Basel and to address such a distinguished audience of experts on risk management and banking regulation. At the same time, I feel that I have also taken a risk by agreeing to stand between you and dinner after a long day of stimulating presentations and discussions. My initial reaction to the invitation was to consider a risk transfer. But, finally, I concluded that, considering the importance of this workshop, it would be preferable and appropriate to talk about credit risk transfer.

In my remarks, I will first recall some of the facts on what we know about the development of the market for credit risk transfer (CRT) instruments, clarifying some of the specific aspects that are especially important for central banks’ financial stability monitoring and assessment. Since the ECB has devoted considerable effort in trying to understand the scale of the CRT activity in the euro area, I will partly draw upon that work in seeking to pinpoint the main policy challenges related to the CRT market. I will then conclude with some ideas as to the direction of future research.

II. Credit Risk Transfer – broad characteristics and systemic consequences

Over the past few years, growth in the trade of credit derivatives has been unrelenting and exponential: in December 2005, the market for credit default swaps (CDS), measured by the notional principal outstanding, was almost ten times larger than it was in June 2002. This multiple admittedly reflects a small base, but growth rates have remained impressive: the BIS reported that, in the first half of 2005, the notional amount of credit derivatives outstanding (CDS only) grew by 60%. Moreover, in the following six months, that is, until December 2005, the International Swaps and Derivatives Association reported a growth rate in the notional amount outstanding of credit default swaps of about 40%.[1]

The extraordinary growth of this market also reflects the involvement of parts of the financial sector that extend beyond banks. Indeed, figures provided by the British Bankers’ Association earlier this year identified growing participation in this market of a variety of non-bank financial institutions which were involved in both the provision and purchase of credit risk protection. Only two years ago, credit risk transfer was essentially a bank-to-bank business, at least in the EU (as highlighted in a report of the Banking Supervision Committee of the ESCB), with banks seeking to better manage their credit portfolios, or to act as facilitators of CRT among other institutions.[2] However, the share of banks’ participation in the CRT market may gradually decline, as investment funds – including hedge, pension and mutual funds – and insurance companies increase their shares in this market.

The rapid pace of growth and widespread participation in the credit derivatives market is transforming the financial landscape. I should first note that the development and growth of the market for credit derivatives has positively contributed to a more effective management of credit risks by banks, for instance by allowing the largest among them to reduce the degree of concentration of loan book exposures to single corporations or industries. It is also reassuring to observe that in parallel with the growth of the market, the management of credit risk within the financial system has improved, as suggested by market-based measures of credit risk (in the financial system). Although other factors have been at work over the past few years, including significant balance sheet restructuring in the non-financial corporate sector, CRT markets seem also to have played a supportive role in facilitating a better spread of risk across the financial system, both within the banking sector and beyond. Furthermore, credit derivatives appear to have had favourable effects on the corporate bond market too, as the ability to hedge risks via the CRT markets contributed to improving liquidity in the underlying markets. In addition, investors continue to effectively discriminate risk across sectors in periods of greater stress – as was manifested in the case of the GM downgrade or in the more recent period of financial market volatility.

To sum up, credit derivatives have effectively helped to enhance the efficiency of the financial system by (i) providing to both bank and non-bank financial institutions access to a broader range of risk-return combinations and a wider pool of underlying risks, and (ii) enhancing the liquidity of corporate bond markets. Finally, the rapid growth of CRT instruments in Europe also points to a more integrated market for credit risk, and may reflect the market response to the persistent segmentation of the underlying cash market due to national regulatory barriers, legal difficulties to transfer loads, etc. In other words, the CRT market is a good example of an effective private-sector impetus to deepen financial integration in Europe.

Since I am not the elusive “one-armed economist”, this overall positive assessment, on the one hand, is complemented by some other considerations of caution and concern on the other. The benefits of CRT markets may have come at the expense of creating some new types of risk for financial stability. Why should we be concerned? First, there are legitimate questions regarding the ability of all risk-takers to assess the risks they assume through this market as effectively as the banking sector has been able to. In addition, we know little about how this market would function in a period when the credit cycle deteriorates substantially and the frequency of idiosyncratic defaults increases, especially because of the risk – which was highlighted in the June 2006 issue of the ECB’s Financial Stability Review – that all investors may want to “run for the door at the same time” because of a crowding of trades. There are also other sources of risk, notably (i) concerns that a few large global banks dominate a large part of the market; (ii) delays encountered in the clearing and settlement of transactions; and (iii) the potential risk of a double default where the credit protector (maybe a bank or hedge fund) defaults at the same time as the so-called reference entity defaults. Such a concern, if not fear, has recently been fuelled by the observation of market behaviour in a period of stress, for example during the Delphi episode where market dynamics behaved rather differently from what was expected.

Often, fear has its origin in a lack of knowledge, and I am convinced that the uncertainty regarding the functioning of the credit derivatives market stems from the added degree of complexity that credit risk transfer instruments, and their trade, bring to the financial landscape. To begin with, the growth in the use of these instruments epitomises the complex interaction between cash and derivative transactions, which is dynamic by definition. In addition, the pricing of these financial instruments by market participants is highly model-dependent, and we know that a model is usually as good as its assumptions – of which there are plenty and (some) are not generally open to scrutiny. In this context, a 2004 study by the Joint Forum showed that, reassuringly, there was no evidence of hidden concentrations of credit risk in the financial system as a whole and that there was a broad awareness of the risks associated with CRT. However, somewhat less reassuringly, even the most sophisticated market participants faced challenges in fully understanding the credit risk profile of what were then the more complex instruments.[3] This type of challenge became manifest, for instance, in the assessment of default correlation across different reference entities, the handling of collateralised debt obligation (CDO) tranches, and the excessive reliance on rating agency assessment. In addition, it remains true to date that – owing to an almost complete lack of relevant information about the time-varying characteristics of the correlation structures among the returns of assets included in the underlying portfolios – market participants are still unable to accurately and fully assess the information content of the prices of complex instruments or to properly characterise the size and nature of the relationship of structured products with other financial markets.[4]

We also perceive that some form of regulatory arbitrage is taking place: credit risk is being transferred to financial intermediaries with lighter, or no regulatory burden and, as some believe, which are less effective in their management, or is being transferred to the less-regulated trading book within banks. One noteworthy aspect regarding the former is the growing role of counterparty risk arising from the more extensive transfer of credit risk, particularly when highly leveraged and concentrated hedge funds are counterparts to banks that have similar characteristics, as seems to be increasingly the case. A better understanding of the interlinkages between key players and strategies and more information on the potential concentration or unwinding of risks is needed in order to be in a better position to assess the magnitude of this risk. Another aspect, related to the “within the bank transfer of risk”, is the rapid evolution of the trading book, from a tool to hedge credit risks to an actively managed portfolio. It is becoming less and less obvious how the complex structured products that are now part of the trading book fit with the notions underlying the trading book’s capital requirements – such as the liquidity requirements that would need to apply to equity tranches of CDOs. Furthermore, because of the nature of structured products, the trading book’s concentration risk and/or correlation risk pertaining to underlying names remains difficult to ascertain, thus possibly obviating potential benefits from the diversification the structured products offer. Indeed, the blurring of the lines between the market risk in banks’ trading books and the credit risk in banks’ portfolios is one of the most formidable challenges stemming from the expanding transfer of credit risk.

When talking about regulatory arbitrage, we should not jump to premature conclusions: we first need to carefully and clearly identify the form of “market failure” arising from the transfer of credit risk (or being reinforced by it) that would require some form of regulatory oversight. Again, greater and more precise knowledge about this matter is essential to ease financial stability concerns – and I hope this workshop will be helpful to that end.

Finally – and this is especially interesting for a central banker – it has been argued that the rapid development of the market for the transfer of credit risk was to some extent driven by the low interest rate environment that prevailed until recently and which set in motion a “search for yield”. However, as we all know, correlation does not imply causality, and the precise mechanisms and incentives are still to be ascertained. In addition, to the extent that this environment stimulated innovation and financial development – and not excessive risk-taking per se – a change in the interest rate environment would not necessarily imply a generalised unwinding of positions taken, as the response to very recent movements in interest rates seems to suggest.

III. Regulatory and supervisory challenges

Although the ancient Romans were probably not aware of credit risk transfer, Seneca already warned us to “Be wary of the man who urges an action in which he himself incurs no risk.” Despite the fact that this warning does not fully fit, or entirely capture, the essence of today’s CRT transactions, we should nevertheless heed his advice and be wary and alert – which brings me to the challenges posed by the transfer of credit risk to the regulatory and supervisory framework. I would say that with the various recent initiatives in the areas of risk management and capital requirements, the necessary regulatory framework for coping with the complexity of CRT instruments is now in place. Nevertheless, substantial implementation work still remains to be done, both by supervisors and financial institutions. In this respect, let me mention a few issues that in my opinion require a dedicated follow-up.

The first issue is the need for close cross-sectoral cooperation and information exchange between banks, the insurance sectors and other financial entities. The BSC report on credit risk transfer I referred to earlier identified a growing involvement of hedge funds in the CRT market. Whereas the appetite for selling protection, i.e. for buying credit risk, on the part of insurance companies that had earlier entered this market appears to be declining, it remains to be assessed whether this is a structural shift and to what extent the slack has been taken up by other non-bank financial intermediaries.

It is therefore very important not to take a narrow sector approach to CRT instruments but to adopt a broader perspective encompassing all of the financial system. The BSC report I just mentioned provides a good illustration. Together with the input provided by the level 3 Lamfalussy committees, that is, the Committee of European Insurance and Occupational Pensions Supervisors (CEIOPS) and the Committee of European Securities Regulators (CESR), it contributed to the cross-sectoral risk discussion by the Economic and Financial Committee’s Financial Stability Table and also proved to be a useful input to the Joint Forum’s report on credit risk transfer.

This kind of joint approach should be pursued in the future. At the European level, cross-sectoral supervisory cooperation is formalised through the so-called “3L3” – the regular meetings between the chairpersons and secretaries of the three level 3 committees CEBS, CEIOPS and CESR. These committees should devote sufficient attention to such cross-sectoral risks.

A second area requiring close monitoring is the interaction between accounting rules and financial firms’ CRT business. International Financial Reporting Standards (IFRS) allow for a greater use of fair value accounting. In the past, items in the banking book were valued on an accrual basis and items in the trading book on a market value basis; under the IFRS, this distinction is not so straightforward. Credit derivatives, which as I mentioned blur the traditional division between banking and trading book, raise important supervisory issues, in particular regarding banks’ incentives to allocate transactions with the aim of profit steering or minimising capital requirements. The Basel Committee has already taken various initiatives in this regard, including the work of the Accounting Task Force, the Accord Implementation Group (AIG)’s trading book sub-group and, importantly for this audience, the Research Task Force’s working group on the interaction of market and credit risk.

The last implementation issue I want to raise relates to sound risk management practices. Considerable efforts have been made both by the industry and the supervisory authorities to develop concrete guidance. The most prominent and recent example of such private sector initiative is last year’s report by the Counterparty Risk Management Policy Group, better known as the “CRMPG II”.[5] On the part of supervisors, relevant initiatives are the Joint Forum’s report on credit risk transfer I referred to earlier and the Basel Committee’s recommendations on sound practices [6] on highly leveraged institutions, which still remains a relevant benchmark although it may need to be revisited to ensure – or confirm – that it fully captures developments since its adoption in 1999.

Banks have been making substantial progress in adopting the various recommendations on sound risk management practices, as the survey conducted by the Basel Committee’s Risk Management Group very recently confirmed. One important area where progress was made is in terms of the substantial reduction in confirmation backlogs for credit derivatives, after the Federal Reserve Bank of New York and the UK’s Financial Services Authority publicly rang the alarm bell. But in other areas, such as exposure measurement, stress-testing and knowledge of counterparties, more can and should be done in order to better manage risks and address financial stability concerns. Interestingly, some of these concerns were already identified by the BSC in its 2004 survey and by the Basel Committee in its 1999 Sound Practices Paper.

IV. Concluding remarks

I would like to conclude by highlighting some areas for potential future research, as I am convinced that central bankers require a strong conceptual and empirical basis for policy-making related to all our tasks and functions.

First of all and related to my previous remarks this evening, we need to broaden our knowledge of the functioning and effectiveness of risk transfer mechanisms and their policy implications. Whereas we are all aware of the considerable benefits which stem from the growth in credit derivatives, we need a better understanding of, and an effective monitoring framework for, the current allocation of risk in the financial system.

Second, despite the Basel Committee’s “regulatory pause” to give banks sufficient time to absorb and implement the new capital framework, the research community can provide an important input to this process. One such input I would see concerns the continued blurring of the banking book/trading book boundary I mentioned earlier. As this development is likely to continue, a more harmonised regulatory treatment of credit risk and market risk is likely to be high on the regulatory agenda in the future. Another related issue is the possible recognition of a full portfolio approach to capital requirements, recognising credit correlations and diversification effects. The Basel Committee is of the view that, at this juncture, full portfolio models are not sufficiently robust to use for capital requirements, but there is hope that such models will be further developed.

Finally, and taking a wider perspective, I think that we require much more analysis of the linkages between monetary stability and financial stability and between the associated risks. Given that these two concepts are at the core of central banks’ mandates, we need to know more about how one affects the other and about the implications for policy-making.

So, plenty of work awaits us. But, “if politics is the art of the possible, research is surely the art of the soluble. Both are immensely practical-minded affairs.” These reassuring and appreciative words of a 20th century English author (Sir Peter Medawar) also express my encouragement and expectations for the future work in the areas discussed at this workshop. But for now, I suggest we turn to another “practical-minded affair”, which is our dinner.

Thank you very much for your attention.

[1] This figure is already netted out by about USD 4.5 trillion worth of bilateral trades (representing about one-third of the total amount outstanding). For a description of the TriOptima service that reduces the burden on banks’ back offices by terminating derivatives transactions on a multilateral basis, see Box 17 in the ECB’s Financial Stability Review, June 2005.

[2] ECB (2004), “Credit risk transfer by EU banks: activities, risks and risk management”, joint report of the Banking Supervision Committee of the European System of Central Banks, May.

[3] These developments are discussed in detail in the special feature article in the ECB’s Financial Stability Review, June 2005, entitled “Has the European collateralised debt obligations market matured?”.

[4] See Fitch (2005), Global Credit Derivatives Survey, November, which expresses the view that “existing standards of financial disclosure do not provide sufficient insight into firm level positions and exposures, particularly with reference to how credit derivatives and credit structured products are used to either mitigate, diversify or take on additional risks”.

[5] Counterparty Risk Management Policy Group (2005), “Towards greater financial stability: a private sector perspective”, July.

[6] Basel Committee on Banking Supervision (1999), “Sound practices for banks’ interactions with highly leveraged institutions”, January.

Speaking engagements

Media contacts