Opening remarks at the press briefing on the occasion of the publication of the December 2007 ECB Financial Stability Review
Speech by Lucas Papademos, Vice President of the ECBFrankfurt am Main, 12 December 2007
My colleagues and I would like to welcome you to today’s press briefing on the occasion of the publication of the December 2007 edition of the ECB Financial Stability Review. The financial stability assessment contained in the Review has been prepared with the close involvement of the ESCB Banking Supervision Committee. It is for the most part based on information that was available up until 9 November 2007, the “cut-off” date for this Review.
The primary objective of the FSR is to identify and evaluate the main sources of risk and vulnerability to financial stability and to provide a comprehensive assessment of the capacity of the euro area financial system to absorb adverse disturbances. The December 2007 edition of the Review places special emphasis on the financial market turmoil, which is also the focus of the overview chapter. This edition of the Review addresses several topical issues in 19 boxes and also contains five Special Feature articles. The choice of the content of some of these boxes and articles has been motivated by the financial market developments over the past few months.
In my presentation, I will first devote some time to the ongoing financial market turmoil. Second, I will review the risks and vulnerabilities in the external environment as well as in the euro area financial and non-financial sectors, which are partially affected by the re-pricing of financial risk. I will conclude with an overall assessment of the risks to euro area financial stability, as we currently see them.
2. The financial market turmoil
Starting with the market turmoil, the principal and underlying cause for the market tensions and volatility that erupted in August is that, over the past few years, several intertwined vulnerabilities were created by a mutually reinforcing interplay between three developments: (1) a strengthening investor risk appetite and a search for yield; (2) a sizeable expansion of financial market liquidity from mid-2003 onwards; and (3) the increasingly widespread adoption by banks of the “originate and distribute” business model.
More specifically, in an environment of historically low risk-free asset yields, the search for yield by investors made it necessary to seek and acquire alternative assets offering higher returns but associated with greater risks. These assets included complex structured finance products that are relatively illiquid. The “search-for-yield” phenomenon is evidenced by the compression of credit spreads and the decline in market volatility over the past four years (see chart on the left of slide 3).
Against this background, banks pursuing the “originate and distribute” business model met a growing base of highly receptive investors willing to include in their portfolios riskier loans re-packaged into increasingly complex structured credit products. While the “originate and distribute” business model facilitated a wider distribution of credit risk across the financial systems, there were growing concerns among the global central banking community that this model was reducing the degree of transparency concerning the ultimate location of risks in the financial system. The amount of loans securitised in euro denominated markets has grown over the past three years (see chart on the right of slide 3). The tendency of European banks to resort to securitisation of loans has allowed them to dispose of the credit risk and at the same time it has fuelled banks’ loan growth on the supply side.
Before elaborating on the risks emanating from the interaction of the three aforementioned factors, it might be worth recalling how the ECB saw the situation six months ago, before the financial market turmoil gathered pace. In this context, let me refer to some pertinent quotes and conclusions included in the June 2007 FSR (see slide 4).
First, we had highlighted the potential of the US sub-prime mortgage market crisis deepening and spreading to other markets, to which euro area financial institutions could have exposures.
Second, we had stressed the prospect of financial market liquidity abruptly drying up when investor risk aversion rises and had warned that an abrupt and sharp decline in market liquidity could reveal vulnerabilities in the financial system.
Third, concerns had been expressed about the implications of the growing reliance on rating agencies in credit risk assessment.
Fourth, we had discussed the possible loosening of standards and compromising of counterparty risk management at the margin by some banks in an environment of intensifying competition about lucrative fee income.
Fifth, we pointed out that reduced financial market liquidity had the potential to hamper banks’ ability to sell securitised risky loans and thus adversely affect the credit cycle.
2.2. Propagation channels
Let me now elaborate on the various stages and the propagation of the turbulence as this is useful for the assessment of the prospects for the financial markets and for deriving some lessons. At the early stages of the crisis, when it was still contained to the US sub-prime mortgage markets, delinquency rates on US sub-prime mortgages increased markedly after mid-2005, for two main reasons. First, the interest rate charged on sub-prime mortgages is initially much lower than for a standard mortgage, but it is typically reset to a much higher rate, usually after a two to three year period. Second, high rates of house price inflation since the early 1990s had encouraged some households to borrow for house purchases with the intention of refinancing, or repaying the mortgage, before the reset date. Once interest rates on these mortgages started to rise, and house prices began to moderate and subsequently to fall, many borrowers were unable to service their loans, sometimes even before the reset date if they had been aggressively betting on further house price increases. The developments in delinquency rates and house price inflation are shown in the charts on slide 5.
What are the prospects for the sub-prime mortgage markets? Recall that the typical fixed-rate period for adjustable-rate mortgages is two to three years, after which they are reset and tied to one of a number of common indices, which are closely related to market interest rates. Notwithstanding a slight decrease in these indices as of recently, an estimated USD 400 billion of sub-prime mortgages will face a reset in interest rates over the next two years.
The surge in delinquency rates quickly translated into falling values of residential mortgage-backed securities (RMBSs), especially those backed by the most recently originated vintages of loans. The values of such securities, and the indices of credit default swaps (CDSs) on such securities, have mostly continued to deteriorate until today (see chart on the left of slide 6); Box 2 of the Review examines these issues in detail.
The fact that claims on the cash-flows generated by sub-prime mortgages were embedded in a broad array of structured credit products, often layered on top of one another, explains why the unexpectedly severe credit losses on the underlying US sub-prime mortgages and the risk of potential future credit losses had such widespread effects on the financial system.
The announcement by some rating agencies in early July that they would downgrade many asset-backed securities (ABSs) and collateralised debt obligations (CDOs) backed by pools of sub-prime mortgages and that they would be revising their methodologies for assigning new ratings resulted in a loss of confidence in ratings and caused investors to fundamentally reassess the expected distribution of returns on a wide variety of structured finance products and the risks embedded in these products. These developments contributed to an increase in risk aversion and a substantial decline in market liquidity (as illustrated in the chart on the right of slide 6) and resulted in a sharp drop in issuance volumes in structured credit and leveraged loan markets.
One of the main channels of propagation and amplification of the financial market turmoil proved to be the off-balance sheet vehicles created by banks for clients or for themselves to invest in structured finance products. Money-market fund managers and other investors apparently had underestimated the exposure they were incurring towards the risks in US sub-prime mortgages by purchasing asset-backed commercial paper (ABCP) issued by these vehicles. Box 8 of the Review provides a detailed explanation of the various ABCP structures and how they were affected by the market turmoil. Ultimately, a liquidity squeeze was triggered in the unsecured interbank money markets because of the larger-than-expected funding liquidity needs, or perceived future needs, of some banks with liquidity commitments to off-balance sheet vehicles, which had become unable to roll over a large share of their ABCP in the market. At the same time, other financial institutions with liquidity balances tended to hoard them.
Box 9 of the Review discusses the various reasons for this hoarding behaviour which has persisted until the present. The first is an adverse selection problem: counterparties with liquid balances have found themselves unable to distinguish between financial institutions that have sizeable exposures to structured credit products and those who do not. Second, banks have hoarded liquidity for precautionary reasons because of heightened uncertainty about their own potential funding liquidity needs. Third, money-market funds have also built up precautionary liquid balances to cover the risk of potential redemption requests by investors.
I will not review here in detail the central bank operations that aimed to ensure the orderly functioning of the money market as they have been discussed on several occasions, and are summarised in the Review. The ECB acted swiftly and decisively, with several refinancing operations, including initially operations of very short maturities and then of three-month maturity to address tensions in the term money market. It has also provided additional liquidity, above the benchmark allotment amount, in the context of the regular main refinancing operations (MROs) but has continued to aim at balanced liquidity conditions at the end of the maintenance period. These operations have had stabilising effects on money market rates, by reducing the volatility of the very short-term interbank rate around the key policy rate of the ECB and by containing pressures in the interbank term money market. They have thus contributed to mitigating the effects of the turmoil on the financial system and the economy.
However, tensions have persisted in the term money markets, as shown by the spreads between euro deposit rates of different maturities and EONIA swap rates which have remained at elevated levels (see the chart on the left of slide 7). Continuing pressures have characterised the term money markets globally, including the US dollar, euro and pound sterling money markets (as shown by the chart on the right of slide 7). A main reason for this is the persistence of uncertainties surrounding the financial positions and liquidity needs of financial institutions.
Let me now turn to the lessons that have been learned so far from the market turmoil. Although it is too early to draw definite conclusions, a number of weaknesses that have been revealed by the turmoil could be generally considered as a consequence of inadequate risk management by some institutions and insufficient market discipline. But let me be more specific and point to four broad areas of weakness in risk management which partly relate to the originate-and-distribute banking model and concern (i) the monitoring, assessment and management of credit risk; (ii) the management and pricing of funding liquidity risk ; (iii) the assessment of counterparty risk and (iv) the role of the so-called conduits and structured investment vehicles in such a model and in the transmission of tensions from the credit and structured finance product markets to the money markets. Many of these weaknesses had been previously identified. But they combined in such a way that few had anticipated the potential severity of their joint impact on the core money markets.
Several issues related to these risks and weaknesses require further examination, which is currently being undertaken by various fora, namely the Financial Stability Forum and the Basel Committee for Banking Supervision.
One important issue concerns the valuation of complex structured finance products. Investors who had assumed that the model-based ratings for structured credit products had similar statistical properties to those for corporate bonds discovered that these products were very different. For example, the so-called bespoke Collateralised Debt Obligations (CDOs) have became impossible to value in an acceptably precise manner and, at the current juncture, efforts are made to either mark-to-market the products using some comparable tradable asset prices or the latest available price quotes.
Another problem – closely related to the valuation issue – is the role the rating agencies have played in the development of structure finance products and the propagation of the turmoil as well as the excessive reliance of investors on ratings in their credit risk assessments. Questions have also been raised about the potential conflicts of interest in the activities of rating agencies.
Finally, with regard to funding liquidity risk, an important ingredient in the turmoil which contributed to the propagation of tensions from the credit markets to the money market, was the maturity mismatch on the balance sheets of ABCP conduits. The risks associated with these funding mismatches were further aggravated when market liquidity evaporated from the markets for complex structured credit products. In addition, it became clear that banks often did not have adequate contingency plans in place to deal with unexpected funding liquidity needs arising from the contingent liquidity facilities they had provided to conduits or to deal with the risk that they would face difficulty in syndicating the bridge loans they had extended to finance leveraged buyouts.
Looking forward, at the current juncture the outlook for the subsequent evolution of the sub-prime originated turmoil is highly uncertain. In order to obtain some indication on the possible propagation of market tensions, I would like to refer to Box 5 in the Review which focuses on the interactions between credit and market de-leveraging cycles. It concludes that in an environment in which bank loans are widely used as collateral for asset-backed securities, a sudden increase in borrower default rates could have implications on the financial performance of banks not only via credit risks, but also through market and income risks. From this perspective, an important issue for the euro area and global financial stability outlook is the way in which the ongoing re-pricing of credit risk feeds into the evolution of the credit cycle.
A crucial channel that will determine the impact of money market tensions and the re-pricing of credit risk on the credit cycle and the real economy is the extent to which they will affect banks’ lending behaviour and, consequently, the financing conditions of households and the corporate sector. Indications of a considerable tightening of lending standards have already emerged on both sides of the Atlantic (see chart on the left of slide 9). In addition, there are signs that corporate default rates – which reached historically low levels in recent years – are forecast to increase both in the US and in the euro area (as shown in the chart on the right of slide 9). In the event that continued or increased market tensions and volatility would constrain banks’ ability to extend credit to the non-financial sectors, the turn in the broader credit cycle could be accelerated and the downside risks to economic activity could increase. So far, however, bank credit expansion to the private sector in the euro area has remained robust.
3. The external macro-financial environment
Let me now turn to the other risks and vulnerabilities that have been identified in the FSR which pertain to the external environment and the euro area financial and non-financial sectors. Most of these risks have been previously stressed, but recent developments have a bearing on the likelihood that they materialise or on the expected impact on the euro area financial system and economy.
The significant appreciation of the euro vis-à-vis the US dollar over the past months has had some implications for the outlook for the adjustment of global imbalances, notably for the financing of the US current account deficit (see the chart on the left of slide 10). The financial market re-pricing seems to have triggered a noticeable change in the pattern of capital flows to the US. In July 2007, foreign net acquisitions of US long-term securities declined sharply and investors rebalanced their portfolios towards short-term instruments. In the following month, the US recorded net outflows from long-term securities (particularly equities), as well as a decrease in short-term liabilities. Although some of this capital appeared to flow back into the US capital markets in the months that followed, these events underlined the potential of a disorderly unwinding of imbalances in the case of financing challenges for the US external deficit.
Aggregate demand and inflation pressures globally, and in the euro area in particular, have been affected by the substantial and persistent rise in oil prices, although the simultaneous appreciation of the euro has dampened the impact on the euro area economy. Limited spare capacity amid robust demand especially from emerging economies, and continued geopolitical and supply risks are likely to keep oil prices at elevated levels as can be inferred from futures markets. However, there is considerable uncertainty which is skewed to the upside, as indicated by the implied distribution for future oil prices, extracted from options contracts (see chart on the right of slide 10).
4. Euro area corporate and household sectors
Turning to the vulnerabilities in the euro area economy, while the financial position of the euro area corporate sector has remained relatively sound at an aggregate level, the likelihood of a turn in the credit cycle, while still low, may have risen. This is because the amount of debt on firms’ balance sheets has continued to grow, as has the cost of rolling over short-term debt. Corporate profitability, although high, could be adversely affected by a less favourable macroeconomic outlook, tighter credit standards and the higher cost of debt.
One issue highlighted on several occasions throughout the current year is the risk associated with leverage buy-out (LBO) transactions. The financing packages granted to the largest LBO projects finalised in 2006 and 2007 were characterised by excesses that bear some similarities to the practices in the US sub-prime mortgage markets. A further deterioration of financing conditions could cause some of the firms involved to default, with adverse consequences to the collateralised loan obligation (CLO) market where LBO loans are frequently used in asset pools. In addition, a number of new large LBO deals have been delayed or restructured and some planned transactions have been withdrawn from the market due to the turmoil (see chart on the left of slide 11). All in all, most market-based indicators suggest that the current outlook for the euro area non-financial corporate credit quality, while deteriorating somewhat, still remains overall fairly solid. Default rates have remained low, but they started to rise in the second half of 2006 for the first time since 2003 and are expected to rise further in the period ahead. The balance between credit rating upgrades and downgrades of non-financial corporations remained at a level close to that observed over the previous years.
With regard to risks in the euro area household sector, despite some heterogeneity across euro area countries, the latest available data confirm a gradual moderation in the annual rate of house price inflation. In those countries where overvaluation appears to be most acute, house market price developments continue to represent a source of risk for household sector balance sheets, although income growth will typically be a more decisive determinant of the overall risk to household finances in euro area economies. At the same time, the growth rate of lending to households for house purchases has continued to decrease on average in the euro area, suggesting that the liability side of the households’ balance sheet has also grown more slowly (see chart on the right of slide 11).
However, the continuing rise, although at moderating pace, of household sector indebtedness and the signs of vulnerability in some housing markets add to the credit risk faced by euro area banks in the short to medium term. All in all, vulnerabilities could be exposed in those parts of the euro area where residential real estate valuations are particularly stretched, where households are highly leveraged, and where borrowing has taken place primarily at variable interest rates and where loan-to-value ratios are high. The analysis presented in Box 6 concludes that due to the on average relatively comfortable loan-to-value ratios, it would take a sizeable house price decline in addition to any adverse disturbance to the debt servicing capacity of mortgage borrowers before banks would incur large credit losses.
5. Euro area financial institutions
During the past few weeks, the focus on the implications of the financial market turmoil has concentrated on its potential effects on banks’ balance sheets and profitability. While the financial conditions of euro area large and complex banking groups (LCBGs) has further improved in the first half of 2007, the extent of the negative impact of the ongoing credit risk re-pricing on their financial condition will become clear gradually as they report their full year 2007 audited financial results.
Based on data obtained from publicly available sources, Box 11 in the Review presents an analysis of the outcome of a simple stress-test pertaining to 21 LCBGs’ funding needs and capital ratios. Among these euro area LCBGs, 18 have exposures to ABCP programmes and 9 to leveraged loan warehousing risks. In aggregate, the exposures correspond to an additional funding requirement for these banks which represent 5.2% of their total outstanding loans, or 10.4% of their deposit base. This additional funding requirement, should it materialise, is likely to adversely affect these institutions’ earnings prospects.
In a stress scenario where these exposures are fully taken back onto the balance sheets of the sponsoring banks, but where no potential second round effects are accounted for, some banks would see their capital ratios declining quite substantially, but even the worst hit institutions would still remain solvent (see chart on the left of slide 12). However, because many LCBGs may also target some particular capital ratio above the regulatory minima in the pursuit of higher credit ratings, deteriorating capital ratios could have adverse consequences for their credit quality and future funding costs.
All in all, it cannot be excluded that the market re-pricing process could become more disorderly, possibly revealing further and, so far hidden, risk exposures. Moreover, dividend policies could be adversely affected for several institutions. In addition, those LCBGs that rely on funding from non-deposit sources, and those that are particularly active in the securitisation businesses, could see their revenues decline significantly. Reflecting these concerns, forward-looking financial market indicators, such as banks’ CDS spreads and share prices, currently suggest that challenges pertaining to the banking sector are likely to remain in the near future (see the chart on the right of slide 12).
The market turbulence has also raised concerns about the extent and nature of euro area insurers’ exposures to structured credit products. Against that background, some insurers have chosen to disclose their exposures to the US sub-prime mortgage asset-backed securities at the end of the first half of 2007 as shown in box 17 in the Review. These data shows that the exposures in general are limited, on average less than 1% of overall invested assets, and that they mostly relate to high-rated investment segments. These rather limited exposures, especially when compared with the US monoline insurers, can be explained by the differences in business models of euro area insures and by the fact that they had limited appetite for high levels of credit risk. Nevertheless, while euro area insurers might have fairly limited direct exposure to US sub-prime mortgages or more generally to structured credit products, secondary effects of the ongoing credit market turmoil might be of greater concern.
6. Overall assessment of the financial stability outlook
Let me now provide an overall assessment of the risks to the outlook for the euro area financial stability, as we see it at the current juncture.
With financial systems undergoing a process of de-leveraging and re-intermediation, the uncertainty surrounding the financial stability outlook for the euro area has heightened. This uncertainty could persist for a considerable period of time until it becomes clearer how the total valuation and income losses, which could be sizeable, facing the euro area financial system will be spread across individual financial institutions. Clarity will also be needed on how liquidity providers intend to deal with their commitments to off-balance sheet investment vehicles, and how much risk will eventually flow back onto the balance sheets of banks. Moreover, until conditions in the US housing market show signs of improvement, the possibility of further tensions surfacing in structured credit markets cannot be excluded, especially if credit quality were to deteriorate in the broader US mortgage market.
As the adjustment process in the financial sector over the coming months is likely to prove challenging, the financial system could be more vulnerable than before to the crystallisation of other risks that have been identified in previous issues of the FSR and which remain relevant.
Specifically, these risks are related to the following facts and developments: First, within the euro area, the substantial increase in household sector indebtedness together with signs of vulnerability in some housing markets adds to the credit risk facing euro area banks in the short to medium term. Second, the surge of leverage in parts of the corporate sector, especially that associated with LBO activity, raises the possibility of an adverse turn in the credit cycle involving a rise in the default rates of the most heavily indebted firms. Third, although so far the hedge fund sector was relatively unaffected by the recent market turmoil, some uncertainties remain regarding hedge funds’ exposures to credit markets, as well as their leverage and liquidity risk. Finally, outside the euro area, persistently wide global imbalances continue to pose a risk that they will be unwound in a disorderly manner. This could bring about further tensions in global capital markets and, if this risk were to materialise, it could pose a challenging test for the risk management and loss-absorption capacities of key financial institutions.
All in all, the risks to euro area financial system stability have materially increased over the past 6 months.
There are, however, several mitigating factors:
The economic outlook remains broadly favourable and, although pockets of vulnerability can be identified, the balance sheets of households and firms are largely in good shape, supporting the overall creditworthiness of the non-financial sector. Moreover, the capital positions of core financial firms are also generally sound.
This overall positive assessment of shock-absorbing capacity should not provide any grounds for complacency given the heightened uncertainties. In an environment where balance sheet conditions could unexpectedly change, vigilance is of the essence and financial institutions in particular should step up their efforts to effectively manage the risks that may lie ahead.
In this respect, recently launched initiatives and measures that are being taken, both by policy-makers and by the financial industry, aimed at restoring confidence and addressing the weaknesses that have been revealed by the market turmoil should contribute to strengthening the resilience of the financial system. I would like to stress particularly the importance of the new Basel II Capital Accord, which is to be fully implemented by January 2008. Indeed, the new accord strongly relies on significantly improved risk management in determining the appropriate level of capital buffers for banks. Nevertheless, and despite these improvements, the degree of sophistication of financial markets also calls for constant investment in stress-testing methods and practices in order to ensure that banks have full understanding of the potential losses they might incur in highly unlikely but plausible scenarios and that they are adequately prepared to address the associated risks.