Now you see it, now you don't: risk in the small and in the large
Keynote address by Mr Malcolm D Knight, General Manager of the BIS, at the Eighth Annual Risk Management Convention of the Global Association of Risk Professionals, 27-28 February 2007.
The evolving nature of financial risk reflects the interaction of structural trends in the financial system with the current global macroeconomic environment. This has raised several challenges from both the "micro" perspective of risk management within an individual firm and the "macro" perspective of risk in the system as a whole. In particular, one key question is whether the prevailing benign market signals indicating that risk is very low might not understate the potential range of future outcomes. This entails important lessons for risk managers and policymakers.
It is a pleasure and an honour to address this distinguished global association of risk managers. Your profession has made enormous strides over the last two decades and has moved to the very heart of the functioning of the financial system. It is an interesting observation that your talents and expertise, and indeed all the sophisticated risk management structures that financial institutions have put in place in recent years, are currently operating in a global financial environment that has been - using market-based measures of risk - remarkably benign and stable.
Ever since the dotcom equity bust at the turn of this new century, the financial system has been going through an extraordinarily tranquil phase. Financial firms have made strong and steady profits, the few episodes of stress have been easily absorbed, interest rate spreads have been very thin across a broad set of risk asset classes, and implied volatilities have been unusually low, particularly in the most recent period. Based on these market signals, risk is hard to detect, or at least appears quite low. Do these market prices provide an accurate reading of the potential risks - the range of possible outcomes that lies ahead - or do they offer an overly sanguine picture of risks?
Today, I would like to take this question as the starting point for some broader personal reflections on the evolution of risk in the global financial system over the last decade or so. Admittedly, the answer to my question - do current market prices give accurate signals of potential risks - is intrinsically unknowable. Indeed, I see good arguments supporting each of the two alternative answers. You will excuse me, however, if, wearing the hat of a prudent central banker, I consider the less sanguine possibility.
The hypothesis I would like to explore is that we may be witnessing an increase in what one might call "option-like" payoff patterns in the financial system. By that I mean payoffs that yield steady streams of income in good times (akin to the collection of the option premium), but may actually imply large discontinuities - sudden price movements and unexpected payouts - under sufficiently stressful conditions (akin to an option's "exercise"). This evolution towards instruments with option-like payment structures could potentially raise "tail risks", while at the same time giving the impression that the financial system is stable and that risks are low.
First, let me highlight some key characteristics of the recent evolution of financial systems, and of the macroeconomic backdrop. Next I will draw out their implications for the nature of risk. Then I will focus on some key implications for risk managers and policymakers.
In exploring the evolution of financial risk over the last ten years or so, I would like to highlight five features of the unfolding economic environment. The first four refer to the financial system itself, the fifth to the current macroeconomic backdrop.
The first feature is what one might call the "atomisation" of risk. Major advances in financial know-how and information technology have resulted in a quantum leap in the ability to create new financial products. Payoffs are unbundled and rebundled, and then priced accordingly. In the process, risk is deconstructed into its constituent elements and recombined in a variety of ways. The potential embedded leverage achieved through such financial engineering is much higher than in the past. Concomitantly, new markets have emerged: for example, the extraordinary expansion of credit transfer products.
The second feature has been the emergence of new financial players, outside traditional intermediation channels. One example is hedge funds. They now dominate trading in a wide array of financial markets, represent a major source of income for more traditional financial firms such as investment banks and commercial banks, and have become an established investment outlet for both retail and institutional investors. Another example is private equity firms. Their rapid growth is currently fuelling a major wave of leveraged buyouts, mergers and acquisitions.
The third feature has been the growing symbiosis between markets and financial institutions. Markets now rely on both "old" and "new" financial firms for the supply of securities, market-making services and backup liquidity lines. Conversely, financial firms increasingly rely on markets for their profits and, above all, their risk management activities.
The fourth feature has been the unprecedented surge in the volume of transactions. This naturally follows from the previous developments. While it is normally taken for granted, the smooth functioning of the payment and settlement infrastructure has become even more critical for the workings of the financial system.
The fifth feature is that the current economic expansion has been taking place against the backdrop of remarkably quiescent inflation, unusually low nominal and real interest rates, rapid credit growth and buoyant asset prices. We may be witnessing a progressive change in the characteristics of business fluctuations, in a new environment defined by liberalised financial markets, globalisation and central bank anti-inflation credentials. Indeed, market participants often refer quite aptly to the loose notion of "ample liquidity", in the sense of low financing costs and easy access to credit or funding liquidity.
First, complexity has grown tremendously. This complexity is now characteristic of individual financial products, of the business of financial firms and of the linkages that tie together the various elements of the financial system.
Second, market liquidity is playing an increasingly important role. On the one hand, there is now a much greater ability to trade assets and risks in markets of increasing breadth and depth. On the other hand, it is critical for agents to be able to trade even under stressful market conditions.
Third, we have been moving into new territory. And we should be cautious in applying old rules of thumb to understand how macroeconomic developments might unfold and interact with the financial system.
On balance, these profound changes have undoubtedly been for the better: they have yielded long-term benefits and can result in a more resilient financial system. They have gone hand in hand with major improvements in the risk measurement and risk management culture. They have made it easier to shift risk towards those economic agents who are more willing and better placed to manage it. The resilience exhibited by the financial system in recent years to adverse events like the equity bust of the early 2000s, 9/11 and the failures of the likes of Enron and Amaranth is indicative of these benefits.
Even so, securing the full benefits of these structural shifts does put a premium on effective risk management. At this juncture, it is particularly important to understand the implications of the option-like payoff patterns that are arguably embedded in the recent evolution of the financial system. I will discuss this first "in the small" - that is, taking financial asset prices and the macroeconomy as given . Then I will discuss it "in the large", taking into account the feedback of changes in the financial system on the behaviour of asset prices and the macroeconomy. These might be termed the "micro" and "macro" levels, respectively.
As regards financial products, several examples come to mind. We have seen an increasing use of debt-type contracts and financial guarantees. Thanks to the atomisation of risk, complex new financial products that promise capital preservation with upside potential have become ever more popular among investors. Similarly, credit derivatives have evolved towards more complex structures: by combining embedded leverage with collateralisation, they yield higher expected returns, but at the cost of higher potential losses if risks materialise. As a recent report by the BIS-based Committee on the Global Financial System has emphasised, these are the conditions under which credit ratings can be inaccurate or even misleading indicators of the risk profiles of financial instruments.
As regards trading strategies, what I have in mind, in particular, are trades predicated on the sustainability of a positive carry - strategies that yield steady returns under "normal" conditions, but at the risk of losing part or all of the cumulative gains in the event of sudden changes in the relevant prices. These strategies have been followed for a wide variety of asset classes, including foreign exchange and fixed income.
"In the large", the recent evolution of the financial system has also led to the greater prevalence of "option-like" payoff patterns as market participants collectively issue their new financial products and pursue their evolving trading strategies. As a result, there may be feedback effects that could make sudden price changes more likely. Prices, implied volatilities and spreads may be moved away from sustainable levels. Leverage can be increased in an effort to maintain returns in ever more "crowded trades", raising the vulnerability of the financial system as a whole.
As regards market functioning, these processes can lead to the sudden evaporation of market liquidity. Recall how liquidity abruptly disappeared in autumn 1998, as spreads were pushed to unsustainably low levels and positions had to be unwound following unexpected losses. In such circumstances, concerns with counterparty risk and the need to meet sudden surges in demand for funding liquidity combine to exacerbate the withdrawal from markets. Paradoxically, these mechanisms are akin to those behind old-fashioned bank runs. Indeed, it has transpired that markets, like institutions, are vulnerable to such runs. And the fundamental changes in the financial system have arguably made it more, not less, dependent on continuous access to funding liquidity.
As regards the macroeconomy, similar self-reinforcing processes can amplify the fluctuations of the economy as a whole. Easier financing conditions during expansions, on the back of rising asset prices, credit expansion and muted perceptions of risk, can lead to balance sheet overextension, increasing the vulnerability of the economy and sowing the seeds of the subsequent downturn. And monetary policy itself may at times unwittingly accommodate such conditions, especially if inflationary expectations and associated inflation pressures remain subdued, paradoxically underpinned by central bank credibility itself. If so, economic expansions these days may be longer than they were in the past. But, by the same token, in the downturn the unwinding of financial imbalances may play a bigger role, potentially leading to option-like discontinuities - large and rapid jumps in asset prices and the disappearance of market liquidity.
First, system-wide exposures have become more opaque due to the growing complexity of risk profiles, to the interconnections across different market segments and to the scarcer information about the exposures of some of the new key players, including hedge funds and private equity funds. This opaqueness can make it harder to assess the actual exposures to risk, and that can lull participants into a false sense of security.
Third, the role of historically low interest rates and easy credit terms deserves to be highlighted. Lower interest rates can encourage risk-taking and higher leverage, because asset holders' target rates of return have not been adjusted downwards accordingly. To take just one example, this can result from established contractual arrangements. Think of insurance companies that still have outstanding policies offering guaranteed rates of return which originated in a period when the yields these companies could earn on their assets were higher. Likewise, easy access to credit can induce market players to mistake the indulgence in refinancing that many creditors are willing to show at the current juncture for the emergence of a fundamentally new environment of low funding risk.
In these conditions, some of the "alpha" performance that is currently observed in markets may be more apparent than real. It may reflect the option-like strategies that I have described, in which tail risk is incurred without being mirrored in the returns that are being earned. And it may also reflect the taking-on of illiquid positions, where values could easily decline or disappear as markets come under stress.
First, we should avoid being lulled into a false sense of security and continue to instil caution in behaviour. Experience has repeatedly shown that it is precisely in good times that the seeds of financial distress are sown. Policymakers and risk managers need to act to temper the natural exuberance of market participants. Encouraging a "long view" is critical, because it can help to develop a better appreciation of the likely impact of the tail risks to which the financial system is subject.
Second, we must work to better understand and assess risk, both in the small and in the large. The art of stress testing plays a key role here, since it is essential to capture non-linearities in the positive or negative payoffs from various configurations of assets and liabilities embedded in derivative instruments. Much progress has been made by risk managers in developing stress tests for market risk and, increasingly, credit risk for the portfolios of their own institutions; that is, at the micro level. For their part, policymakers are also developing analogous stress tests for the financial system as a whole - at the macro level. Even so, further progress is urgently needed: at the micro level, to incorporate liquidity risks in a satisfactory way; at the macro level, to better incorporate feedback effects between the financial sector and the real economy; and at both levels, to translate the results of stress tests of "tail events" into effective risk management decisions.
Third, we should improve transparency in the financial system. This means encouraging better and more in-depth reporting of individual institutions' risk profiles to counterparties and the broader public and having better information about the distribution of risk across the financial system. Examples of such efforts include those made by risk managers to strengthen the reporting of information by hedge funds to counterparties, and those made by supervisory authorities to shed light on the transfer of credit risk across the financial system.
Fourth, we should continue to improve the trading and settlement infrastructure. The improvements made recently in the settlement of over-the-counter credit derivatives trades are a case in point. They represent an excellent example of what cooperation between market participants and policymakers can achieve.
Fifth, and finally, do not assume that the public sector will always be there to pick up the pieces! Tail risk or catastrophe insurance is not just for the public sector to address. Indeed, the overblown perception that public authorities will always bail out financial firms could lead market participants to pay even less attention to tail risks.
Let me conclude. There is no doubt that the financial system is now more resilient than it was in the past. But, at the same time, its ability to absorb recent episodes of financial distress and the recent benign conditions in financial markets should not mislead us into a false sense of security. In fact, there is a certain tension between the potential risks facing the world economy - burdened as it is by high household debt levels, elevated house prices and global current account imbalances - and the low level of risk suggested by yield spreads and measures of implied volatility. The greater incidence of what I have called "option-like" payoff patterns in the financial system is an issue that justifies continued vigilance.
Let me stress that my remarks this morning have been of an exploratory nature: they are intended to raise questions, rather than to provide answers. Only the future will tell. Even so, as a prudent policymaker, I cannot help but note that the most insidious risks may well arise just at the time when market participants begin to believe that risk, like the business cycle itself, has finally disappeared from the world economic and financial system.