The mid-summer blues are not quite over yet: with subprime default rates still on the rise, 3-months interbank rates stay abnormally high, credit conditions remain tight, gross issues of mortgage-backed bonds and commercial paper are all but dried up, and banks lick their wounds and attempt to set up emergency vehicles to dispose of the backlog of illiquid assets left in their books. The system remains vulnerable. Still, as the worst fears for financial stability have subsided, the debate now shifts from the central banks’ ex post emergency reactions to the preventative reforms needed for the future. Unfortunately, given the nature of this crisis, there is no quick fix this time.
The textbook paradigm
In its unfolding, this crisis conforms to the textbook paradigm. In a financial system where intermediaries hold illiquid assets against liquid liabilities, there are two possible equilibria. When only those agents subject to liquidity shocks require the service from intermediaries, the latter are able to carry out maturity transformation and allow society to earn superior returns. When instead, as a result of a shock, all agents, simultaneously but independently, seek liquidity, the intermediaries’ balance sheets go under stress, there is no demand for less liquid assets and disruptive liquidations may threaten financial stability: a succinct description of what has happened between July and September.
As noted by Mervyn King,1 the “most unusual nature” of this crisis was the disproportion between the shock (“a relatively small size of…bad loans compared with the total assets of the banks”) and its widespread systemic consequences. Echoing Mervyn King, Ben Bernanke wondered how the impact could be so large, comparing the US subprime mortgage market with “the enormous scale of global financial markets”.2 True, also in the textbook model, “crisis“ equilibria may be triggered by potentially insignificant events. But according to the textbook prescriptions, undesirable outcomes can be avoided through informed supervisory action: supervisors possessing the relevant information regarding potential exposures to shocks are better able to prevent a crisis, thereby reassuring all market participants that threats to financial instabilities can be contained. When, on the other hand, market participants not only do not know how serious and widespread the impact of a dislocation is, but also become aware that the supervisory authorities are no less ignorant, they rationally cut their risk positions by more than would be warranted if they possessed greater information and could rely on the presence of a better-informed coordinating agent. The surge in volatility and the drying-up of liquidity make the worst scenario self-fulfilling.
This is, in our view, what has happened this time. A generalised lack of information multiplied the effects of the initial shock.
The information gap
The information gap was wide and deep. Mortgage brokers had an incentive to provide the raw material by quantity, regardless of quality. The valuation of the structured and complicated financial instruments pooling credit risks rested on rating agencies’ models, biased by observations limited to a relatively short span of very benign history. Those products were issued and (rarely) traded over the counter: marked to model, as there was no proper market assessing their liquidity. By the very nature of the CRT, nobody had a clue as to where the credit risks had ended up.
This would have mattered less if the ultimate risk recipients had been only the usual suspects: hedge funds, pension funds and insurance and re-insurance companies. The systemic consequences of the collapse of a few of those would be confined to the counterparty risks assumed by some intermediaries in their lending or broker-dealer activities. But it turned out that there were many banks amongst those more heavily exposed to the direct risk of credit products, through off-balance sheet liquidity commitments granted to vehicles investing in those illiquid assets, equity tranches in the CDOs, own portfolio investment, and reputational commitment to proprietary mutual funds engaged in ABSs. The authorities in charge of stability supervision were seemingly unaware of this exposure: certainly they appeared to be caught by surprise by the consequences of the subprime insolvencies on the banking system, ignorant as to where the losses were located and therefore unable to deal selectively with the problem. The consequence was widespread mutual mistrust causing the hoarding of banks’ liquidity and the hike of interbank rates.
Filling the multi-dimensional information gap that was responsible for transforming a spate of subprime defaults into a full-fledged crisis should be a priority of any reform effort. Unlike in earlier crises, however, there are no obvious solutions to this problem. We confine ourselves to drawing a list, in order of importance, of what we believe to be the more relevant issues.
Filling some gaps
At the origin there is a purely American problem: a crowd of unlicensed non-bank brokers, governed by wrong incentives, offering mortgage loans to all and sundry, irrespective of any assessment of the debtor’s potential solvency. Though the party is over by now, the problem remains and will have to be addressed by Congress.
Next, when credit risks are pooled and re-packaged, comes the role of the rating agencies whose decisions affect the allocation of risks in different investors’ portfolios. Apart from their conflicts of interest from their semi-monopolistic, officially sanctioned status,3 a major information problem arises from the suitability of the statistical models used to provide the ratings on which many investors rely blindly. The spate of downgradings affecting them in recent months is evidence of serious flaws. Some propose that the rating agencies should be treated as underwriters, with the attendant responsibilities; at the very least their models should be subjected to an independent enquiry and, as it were, be themselves rated.4
A deep and wide secondary market insuring at least post-trade transparency is an essential provider of information; over-the-counter transactions instead remain opaque and known only to the parties concerned. The heterogeneity of structured products (each with idiosyncratic features) is an obstacle to the supply of a public good. An agreement prompted by industry associations in consultation with the supervisors to standardise the most diffuse classes of instruments, as was done for some derivative contracts, would be a step towards the creation of a market.
There is then the question of the bank-sponsored investment vehicles (SIVs) and of the treatment of the liquidity facilities provided to them by banks, which under Basel I are exempt from capital requirements (and hence from disclosure) as long as the commitment is for less than 365 days. The somewhat more stringent prescriptions of Basel II are still short of achieving adequate transparency. This however is only a part of a more general issue: that of designing an efficient structure of information flows in order to fill those gaps that are prejudicial to stability.
A wider problem: Basel II?
Ideally the authorities in charge of stability should be empowered to acquire all the information needed to assess the system's (and not only an individual agent’s) vulnerabilities from all financial entities whose actions may have systemic effects. They would thus be better equipped to prevent the eruption of dislocations as well as to provide guidance to market participants on the risks present in the system. In the view of the Basel Committee on Banking Supervision, the implementation of the Basel II capital framework, by improving “the robustness of valuation practices and market transparency for complex and less liquid products”, “would have gone some distance” to alleviate the present crisis. We believe that the distance would have been very short, as the Basel II framework represents only a small approximation to a satisfactory solution.
First, disclosure belongs to the third pillar of the accord (market discipline), which is recognised as by far the weakest, in terms of both prescriptions and enforcement.5 Second, Basel II disclosure is required in order to assess an individual bank’s capital adequacy. That is not enough: a strong bank capital base, while essential to avoid the collapse of any major financial institution, was not sufficient to prevent the systemic effects of the subprime crisis. Third, any disclosure obligation imposed by the accord only concerns banks. But all the entities having liquidity mismatches between assets and liabilities may produce systemic effects, either directly with counterparties or through the structure of their balance sheets: not only traditional intermediaries, but also broker-dealers, non-thrift financial institutions borrowing wholesale in the market, any kind of vehicle with the same characteristics, as well as hedge funds.
Finally, designing an efficient structure of information flows meets with institutional obstacles. In a closely connected financial world, where cross-border entities prevail, information on global stability is the more valuable the less its gathering and processing are fragmented. There are natural limits to this and cooperation between bank supervisors helps. But there are obvious steps to be be taken to improve the situation. At the national level, the single regulator model, whereby banking supervision is not a responsibility of the lender of last resort, has shown important flaws, at least in Germany and the UK. More importantly, similar faults are present in the euro area, where the ECB, which, though not a lender of last resort, is responsible for providing liquidity, has no supervisory competence and must rely on the information voluntarily provided by the national central banks.
Adequate, reliable and timely information is essential to ensure financial stability. Filling the information gap however has so far been a slow and hesitant process. Do we need more crises to move forward at a faster pace?
1 Speech at the Northern Ireland Chamber of Commerce and Industry, Belfast, 9 October 2007.
2 “The Recent Financial Turmoil and its Economic and Policy Consequences”, Economic Club of New York, October 15, 2007.
3 On the role and the shortcomings of rating agencies see J.R. Mason and J. Rossner, “Where Did the Risk Go? How Misapplied Bond Ratings Cause Mortgage Backed Securities and Collateralized Debt Obligation Market Disruptions”, mimeo, May 2007.
4 Credit rating agencies have themselves “acknowledged the need to review the information they receive from originators and they provide to investors in structured credit products” (Financial Stability Forum, Working Group on Market and Institutional Resilience, 15 October 2007),
5 It has been alleged that Basel II has a “bias in favor of nondisclosure”, strengthened by the opposition of the banks (Institutional Risk Analytics, Comments to the proposals, October 2006).