How to avoid further credit and liquidity confidence crises

Guillermo de la Dehesa

19 October 2007



The US banking authorities and the EU finance ministries, central banks and supervisory authorities are both trying to design a roadmap to strengthen financial stability and crisis prevention after experiencing the effects of present confidence crisis. In principle, the best way to try to avoid another credit confidence crisis is just to learn from what went wrong in the present one, to make the necessary changes and to develop new policies. By now, it seems clear that some market, regulatory and supervisory failures have taken place in the last few years of low interest rates and leveraging euphoria that need to be addressed.

Subprime mortgage lending is not new. It has existed for a long time in consumer finance both in the US and Europe although subprime mortgage finance is much more important in the US. The key to successful subprime lending is to develop a very good credit scoring based on data concerning the historical behaviour of borrowers, both collectively and individually. These are then applied interest rates for every type of borrower and marked high enough to more than compensate for their expected levels of non-performing loan losses.

The main problems with subprime mortgage lending in the US have been the following:

First, half of their originators are agents and brokers, which are not part of a banking group and thus fall outside federal banking regulation. Moreover, these agents and brokers get paid by commissions based on the number of mortgage loans that they are able to sell to households, so that their incentives have nothing to do with the default risk involved in the loan, but, on the contrary, the higher the risk of the borrower, the larger the commission.

Second, the other half is originated by banks, which sometime ago tended to hold the mortgage for some years in their books, so as to have an incentive to be careful about its non-performing risks. But today, both brokers and banks which originate these loans sell them very quickly either directly or through another financial intermediary than securitizes and sells them to investors, thus losing their traditional incentive to monitor their risk. The way these mortgages are securitized is based on pooling thousands of mortgages and other loans in an off-balance-sheet vehicle which issues marketable CDOs or CLOs representing shares in the pool.

Third, unlike in most countries in Europe, under the US legal system, subprime mortgage loans carry a much higher risk for the lender because there is no legally binding property register; the loan does not give the lender the right to repossess the property, regardless of who owns house and the repossessing system varies from one state to another.

Fourth, there are always risk-hungry investors who are ready to invest in higher risk-higher-yield financial products like the CDOs, but the problem this time is that these products are so complex that either they were not able to understand fully what they were buying or did not wanted to invest enough on disentangling their supporting models before to purchase them. The fact is that even the more sophisticated risk-hungry investors (as the hedge funds) did not really know well enough how to value these assets and eventually they had to trust the rating given by the independent rating agency involved in the securitization.

Fifth, although securitization is a great innovation which makes it possible for banks to extend affordable mortgages to many more households (mainly the low income ones) and to small and medium firms, such a complex financially structured products are extremely difficult to value and also to rate. In the old times, a triple or double “A” rating was usually given to security issued by a highly stable and solvent country or company which was quoted daily in an organized market. Today, one of these CDOs can achieve a triple or double ”A” rating, when they are composed of blocks of different ratings, from “senior” (double or triple A) and “mezzanine” (triple B) to “equity” (triple B-, triple C or less).

It looks like “alchemy” but sophisticated mathematical models were supporting this ratings based on the fact that, given the large number of loans pooled, their probability of default was much less correlated than in the case of one single or several loans, since, in principle, it is more unlikely that all default at the same time. Moreover, these structured products do not trade and are not quoted in organized markets. They are mostly customised to suit different investors, so that they are only sold over the counter. As such, their price transparency and market liquidity tend to be extremely low.

Sixth, the rating agencies have been classifying these products and their different tranches with their own models without any apparent problem. However, since last June, they have started to downgrade them quickly, given the accelerating rate of non-performing subprime loans and the progressive falling of average house prices in the US. This general and fast downgrading has had a detonating negative effect to the investor’s confidence on the real value of these products. This, in turn, has triggered the present situation of general uncertainty and lack of liquidity for these and other related products collateralized with mortgages and even of other medium and long term loans.

Credit agencies: charges and counter-charges

The rating agencies have come under attack for their role in all this. They have a conflict of interest (they are paid by the issuer of these products and not by the buyer) and their ratings do not seem to have ‘got it right’ at least according today’s market. Even if nowadays there are still few transactions, the ABX HE indexes (January = 100) show that, on average, some triple A rated asset backed mortgage structured securities are being sold with a loss of 6 percentage points, that double A show a loss of 20 percentage points, that single A sell at a loss of 50 percentage points, that triple B show a losses of 65 percentage points and that triple B minus trade a loss of 70 percentage points.

The rating agencies have counterattacked by showing that, at the demand of the sellers, their ratings were made only on the default risks of these securities, which have been downgraded accordingly to the new information appearing in the marketplace, but not on their market and or liquidity risks, which are even more complex (and expensive) ratings. They argue that it is the present lack of liquidity what makes those securities loose value and not so much their probability of default which was rightly captured by their ratings.

The curious geographical transmission of the crisis

Another problem is how is it possible that a relatively minor and expected issue (with present losses of about $200 billion) arising in the subprime mortgage market in the US, has been able to contaminate so many American and European banks and markets. The answer is: because of the large proliferation of conduits and SIVs created by them off-balance-sheet, in order to avoiding regulatory capital consumption, to invest in long-term assets, financing them by issuing commercial paper backed by these assets.

Their basic aim was to borrow short and invest long (as banks always do) in a way that was more profitable since it allowed them to lend without “consuming” their regulatory capital, i.e. without having the investment ‘on balance sheet’ and therefore counting in their loans-to-capital requirements. The volume of conduits created is large (around $600 billion in the US banks and around $500 billion in European banks).These banking conduits did invest in CDOs and CLOs issued by American and other European banks which had subprime loans among other better rated corporate and mortgage securities.

Nevertheless, the main problem with banks in the US and Europe is not only that their conduits invested in subprime and other low quality credit structured products, (when their assets were meant to be of higher grades) but that, when their asset backed commercial paper market financing dried up, the borrow-short-lend-long wheel stopped. The conduits have to pay off their short-borrowing positions, but have problems selling off their long lending positions.

This left the banks with two options: take them into their balance sheets, provoking a credit crunch, or get enough temporary liquidity from a central bank to refinance them – to keep the wheel turning, as it were. The credit crunch in the case of the Euro Area banks would not be very large but substantial. The average ratio of regulatory capital to total loans is 8% in Euro-Area banks. The total volume of conduits needed to be taken into their balance-sheets would absorb only 0.7 percentage points of that ratio, that is, on average, they would have to reduce total lending by 8,75% to absorb these conduits. But for some banks with lower capital levels the impact would be fairly large.

Avoiding future crises

Regulators, supervisors and central banks should try to solve these perverse incentives and conflict of interest problems that lead to the crisis. Here are some of the measures they should take besides continue to inject liquidity until some confidence is regained:

First, the American banking authorities should regulate all American agents and brokers which are originating these mortgage loans in order to avoid their perverse incentives when dealing with their potential borrowers and to try to standardise their property registration and collateral execution systems across states.

Second, all banking supervisors should oblige all banks, which originate and sell loans and mortgages, to retain their “equity” or first loss risk block, as it happens today in some European countries, in order to make them share part of the risk when they sell them to intermediaries or final investors and, therefore, to be much more careful when monitoring their credit risks and when choosing the mortgages to be pooled for sale.

Third, the banks and financial institutions, which structure and securitize these loans, should be extremely transparent about their package processes, their supporting models and their associated risks. Moreover, they should try to increase the standardization of these products up to making them suitable to be traded in an organized and transparent market.

Fourth, the rating agencies should try to regain credibility by showing that they are truly independent and that their rating process is fully transparent and reliable, mainly for these complex structured products.

Fifth, in the case of conduit proliferation there has been a major supervisory coordination failure, at least in Europe, given that some central banks (as in Spain) have not allowed their supervised banks to create these conduits while other supervisors have done so at large. It is clear that these conduits have been created mainly by sophisticated wholesale banks and not so much by more traditional retail banks, but it is even clearer that in those countries where the banking supervision is not done by the central bank, but by another government agency or institution, the problem created by conduits has been much larger in size and risk involved. The main examples are Germany, the UK and the US (with the Netherlands the main exception to this rule). The case in point is probably the UK, where the tripartite division of responsibility between the Treasury, the Bank of England and the FSA has complicated to the extreme an, in principle rather easy, sale of Northern Rock to another bank, which eventually has ended in an expensive bank run and a bail out.

This issue is extremely important for two reasons. First, some supervisors in the Euro Area without coordinating with their other Euro-Area colleagues or even with their central banks, have allowed their supervised banks to develop a competitive advantage versus other competing Euro-Area banks, in the same single market, by allowing them to create large and highly profitable (but risky) conduits. Second, now, when such decisions have proved to be wrong and the conduits are on the verge of producing a credit crunch unless they are refinanced by the ECB System, all the rest of banks without conduits in the Euro-Area are also suffering the consequences of that decision. Something needs to be done about these supervising structures to avoid this lack of coordination in the future.



Topics:  Financial markets

Tags:  credibility, Subprime, subprime crisis, liquidity, financial stability, subprime mortgage lending, credit crises

Chairman of CEPR, Member of the Group of Thirty, and author of "Europe at the Crossroads".