Some reflections on the future of the originate-to-distribute model in the context of the current financial turmoil
Speech by Mr Malcolm D Knight, General Manager of the BIS, at the Euro 50 Group Roundtable on "The future of the originate and distribute model", London, 21 April 2008.
The financial markets are facing serious turmoil. The disturbances which started last summer have deepened and widened, often in ways that have been very hard to predict. While it is true that the originate-to-distribute model has, together with the easy global credit conditions that existed over many years, contributed to the current financial turmoil, it would be wrong to discredit the model itself. What needs to be done is to strengthen the underpinnings of the originate-to-distribute model, in particular by enhancing the information available to market participants, incentives structures, risk management practices and the use of credit ratings. Both public authorities and private market participants have to work together on these issues so as to make the originate-to-distribute model more resilient and thereby more able to contribute to efficient and sound financial markets.
Good evening, ladies and gentlemen. It is a pleasure to be here tonight and to have the opportunity to speak to such a distinguished group of participants. In my remarks, I will touch on three themes: first, I will say a few words about the current global financial turmoil; second, I will focus on the originate-to-distribute model and the role it has played in the current turmoil; and finally, I will touch on the future of that model.
1. The current financial turmoil
The global financial turmoil that has persisted for the past nine months is certainly severe. When it started last summer, many observers initially thought that it could be confined to the market for subprime mortgages, particularly in the United States, and to related structured products. But then the disturbances both deepened and widened, often in unexpected ways. One example is the remarkable speed with which the uncollateralised interbank term credit markets became dysfunctional.
In trying to assess the underlying causes of what has been happening, there is a significant risk of oversimplification. There are two main schools of thought on the causes of the current turmoil. The first school sees the current disruption of financial markets as the long-run consequence of the easy global money and credit conditions that existed, particularly from early 2002. Accommodative monetary policies, advances in financial technology, and a persistent rise in risk-taking and financial leverage contributed to the current troubled conditions. This is the general macroeconomic, or "macrofinancial", explanation.
A second school of thought, which in a sense complements the first, focuses on the shortcomings of specific types of financial products introduced in recent years. Concern has focused on structured products and bank-related structured investment vehicles (SIVs) and conduits, particularly those based on pools of subprime mortgages. The rising role of these instruments and entities, taken together, has made the global financial system much more opaque, with respect to the size and location of risk exposures, the degree of leverage, and the nature of the ultimate holders of risk. This second explanation emphasises more the structure of the financial markets, and less the macroeconomic environment.
Obviously, the recent turmoil has highlighted specific weaknesses in the current structure of the financial markets, in particular in structured finance. This, indeed, was the focus of the recent report of the Financial Stability Forum Working Group on Enhancing Market Functioning and Resilience, of which I was a member and which presented its conclusions to the G7 central bank Governors and finance ministers in Washington DC on 11 April.
And, of course, the central banking community and the BIS, in particular, have also stressed the importance of the "macroeconomic view" when analysing the current unwinding of global credit excesses. [Many academics have theorised about the underlying causes of financial crises. Hyman Minsky's work, done in the 1970s, is of particular relevance. He warned that a continuous deterioration of credit underwriting and monitoring standards over a period of years would eventually culminate in a moment of recognition and recoil from risk-taking (what has since been dubbed "a Minsky moment"). For Minsky, the resulting liquidity crisis was a symptom of the underlying credit problem.]
It is a fact that, prior to the onset of the current global financial turbulence, the prices of many financial assets were unusually high for an extended period. Not only was the yield on US Treasuries historically low; the risk spreads on emerging market sovereign debt, high-yield corporates and other risky assets also fell to record low levels. And during the middle years of the present decade, equity prices in the industrial countries continued to be highly valued, while equity values in many emerging markets rose spectacularly. Residential property hit record highs in virtually all countries, with only limited exceptions in Japan and a number of countries in Europe. The cost of insurance against market price movements - approximated by implicit volatility - also stood at persistently low levels.
Admittedly, arguments about fundamentals could be made to support independently each of these asset price trends. But it is particularly notable that the simultaneous occurrence of all these patterns is also consistent with credit being freely available internationally, and having a low price. And, in fact, these unusually accommodative global credit conditions reflected the interaction of monetary policy, the choice of exchange rate regime in a number of countries (particularly developing countries with a large labour surplus), and important changes within the global financial system itself. Let me briefly discuss each of these three elements.
First, policy interest rates in the industrial countries were unusually low, particularly from 2001 to 2006. This seemed appropriate owing to the apparent absence of strong inflationary pressures. In part, the low inflation over this period certainly reflected the build-up of central bank credibility over many years. But it also reflected the beneficial effects of positive supply side shocks of various sorts (including large increases in the effective global labour supply).
Second, exchange rate regimes. The accommodative monetary policies in the developed world that I have just described might have been expected to cause a general depreciation of the exchange rates of the United States and other advanced countries, relative to those of emerging market countries. However, in many emerging economies, upward pressure on the value of the domestic currency was met with an easing of monetary policy and massive foreign exchange intervention. The former is likely to have contributed to higher asset prices and increased spending in the emerging markets. The latter further eased financial conditions in the industrial countries - via the investment of emerging market official foreign exchange reserves in developed country sovereign debt. In this way, the monetary stimulus to credit growth became increasingly global.
Third, structural changes in the global financial system also contributed to the easing of credit conditions. Changes in the regulatory environment as well as technological developments supported strong trends towards securitisation, globalisation and consolidation in the financial industry. And these elements, in turn, reinforced the easy credit conditions. Securitisation was a crucial underpinning of the development of the originate-to-distribute model. And globalisation made credit more freely available in each specific economy, and allowed problems generated in one part of the world to have effects virtually everywhere. Furthermore, consolidation in the financial industry led to economies of scale in providing credit, and these were in many cases passed on to borrowers in the form of easier financing conditions.
2. The originate-to-distribute model
Let me now turn to my second theme: the originate-to-distribute model and its role in the current turmoil.
It is obvious that the shortcomings of certain financial products introduced in recent years have played an important role in the current highly disturbed financial market conditions. The explanation I find most persuasive focuses on where the troubles began: the US subprime mortgage market and the associated markets for structured products. It highlights the key features of the recent turmoil: the lack of transparency in the originate-to-distribute model; the role played by credit rating agencies in the evaluation of structured products; and the covert reliance on special purpose vehicles to conduct off-balance sheet financial transactions on a large scale. The effect of all these influences was that when the "Minsky moment" came suddenly last summer, perceptions of risky exposures, both to credit losses and to liquidity shortages, rose sharply, as did uncertainty about where those exposures might materialise.
Nevertheless, I personally believe that the originate-to-distribute model basically remains a good one. When it functions correctly, it has the capacity to distribute risk widely and efficiently, while at the same time diversifying the revenue streams of the banks that still form the core of the global financial system. But a number of serious concerns emerged in recent years as this model became widely implemented. One source of trouble was the decline in "due diligence" in making loans. Those at the beginning of the subprime chain received fees to originate mortgages, and felt secure in the knowledge that someone else would buy them. Banks at the centre of the securitisation process focused on the profits associated with distributing these instruments, rather than on possible threats to their reputations and their capacity to provide liquidity. Those closer to the end of the securitisation chain probably placed too much trust in the due diligence of originators and packagers, the judgments of the credit rating agencies, and the capacity of modern technology and diversification to manage financial risks. As a result, when the quality of mortgage credit declined in the subprime area, much of this ended up being held in highly leveraged positions.
A second source of trouble was the very rapid and unexpected deterioration of credit quality and associated market liquidity. As far as structured products are concerned, many were highly rated by credit rating agencies. Not surprisingly, purchasers were shocked by the size of the subsequent downgrades. In retrospect, it is clear that the ratings were highly sensitive to even minor changes in assumptions about underlying fundamentals, as well as correlations among defaults, and recovery rates. Moreover, the data used to justify initial assumptions were themselves highly deficient, since they were based on incomplete or even erroneous information drawn from a time period too short to cover a full credit cycle. Furthermore, many investors apparently forgot that ratings only reflect credit risk, and high ratings thus provided no indication of the likelihood of possible major declines in market liquidity or market prices. For those holding such instruments, particularly in portfolios that effectively had to be marked to market, the size of the potential losses turned out to be much larger than was thought possible at the time of purchase.
A third weakness in the originate-to-distribute model was its sheer uncertainty, the "unknown risks". Many of the structured products created in recent years bundled together traditional asset-backed securities and new products based on subprime mortgages. As a result, there was a pervasive uncertainty about where the risks were concentrated and how sensitive they might be to the economic cycle. In this environment, everyone suddenly became suspect when things turned wrong. In addition to this growing concern about counterparty risk, banks became increasingly concerned about liquidity risk. They had to fund their associated SIVs while they could no longer count on securitising the other loans they had made. And they started to reconsider their earlier commitments to provide liquidity, say as primary brokers to hedge funds.
In this climate of prevailing uncertainty, banks were reluctant to deal with each other in the interbank term market at anything but the shortest maturities, and other markets tied to the interbank market were also disrupted.
3. The future of the originate-to-distribute model
Looking forward, there are two main questions: is there a future for the originate-to-distribute model and, if so, how can the model be strengthened so that it will be more resilient when markets are subjected to stress?
On the first question, I do believe that the originate-to-distribute model provides the avenue for both increased diversification and financial innovation. Certainly, almost all market participants have displayed important weaknesses in risk management and a need to write down their structured product portfolios substantially. But some firms seem, at least so far, to have handled these challenges better than others. These differences in performance across firms were especially notable in collateralised debt obligations (CDOs), syndication of leveraged loans, and off-balance sheet vehicles. This suggests that it is not the originate-to-distribute model itself that is problematic.
So let me deal with the second question, ie how to fundamentally strengthen the originate-to-distribute model. In my view, there are four main avenues for action:
First, the steady increase in the complexity and opacity of securitised instruments has led to a lack of transparency about the risks in securitised products, in particular with respect to the quality and the potential correlations of the underlying assets. Clearly what is needed here is more transparency, by which I mean easier and better access for all market participants to all the information that is relevant.
Second, there is a problem with the incentives structure. The turmoil has revealed that many participants in the originate-to-distribute model had weakened incentives to make accurate risk assessments. The incentives of all parties - and this includes the originators, the arrangers, the managers, the distributors, the credit rating agencies and the investors - were misaligned. There is therefore a need to ensure that the incentives of all participants in the securitisation process are properly aligned.
Third, the turmoil showed that institutions using the originate-to-distribute model poorly managed the non-credit risks associated with the securitisation business, such as market risk, liquidity risk, concentration risk and, of course, pipeline risk. This means that market participants need to have adequate controls over their exposures, including effective scenario analyses and stress testing procedures. They also need to manage exposures that they retain on the balance sheet or that could return to it, and they need to avoid excessive risk concentration.
Fourth, there is the issue of credit ratings and their usefulness and transparency. Credit rating agencies play an important role in the originate-to-distribute model and should therefore make both the process of establishing the rating and the information underlying the rating more transparent. Only this will enable investors to make informed decisions and strengthen discipline amongst all participants.
Let me conclude. We continue to face serious financial turmoil. The disturbances which started last summer have deepened and widened, often in ways that were very hard to predict. It is true that the originate-to-distribute model has, amongst other factors, contributed to the current financial turmoil. But, in my view, it would be wrong to discredit the model itself. What needs to be done is to strengthen the underpinnings of the originate-to-distribute model, in particular by enhancing the information available to market participants, incentives structures, risk management practices and the use of credit ratings. Both public authorities and private market participants have to work together on these issues so as to make the originate-to-distribute model more resilient and thereby more able to contribute to efficient and sound financial markets. The recent report of the Financial Stability Forum on Enhancing market and institutional resilience provides concrete recommendations towards this goal.