Headline: The hidden risks of synthetic CLOs Source: euromoney.com Date: January 22 2001 Author: Jules Evans Synthetic CLOs are the fastest growing form of asset-backed security. Last year brought a record amount of synthetic CLO issuance. Banks have started to use them to sell off the risk attached to their asset portfolio, so that they can free up capital used to cover that risk. But this young and booming market may mask hidden dangers. Jules Evans reports. Without much fuss or publicity at the end of last year, ABN Amro launched the biggest CLO ever - a bond worth $15 billion, secured on loans to US companies. The same month, it launched a bond worth Eu8 billion ($7.5 billion), secured on loans to European companies. Both of these deals were mainly synthetic - the assets on which the bonds were secured stayed on the bank's balance sheet. What was sold was the risk attached. The deal was launched at the end of synthetic CLOs' biggest year so far. And this year looks likely to be bigger: the British Banking Association says that it expects credit derivatives, of which synthetic CLOs make up 20%, to grow from $893 billion in 2000 to $1581 billion in 2001. A new report from Barclays Capital on ABS says that it expects synthetic CLOs to rival even mortgage-backed bonds as the most popular form of ABS. Banks issue CLOs - synthetic or traditional - for one simple reason: regulatory arbitrage. Under the Basle Accord set up by the Bank for International Settlements (BIS), banks have to set aside a certain amount of capital to cover whatever risks they have taken with loans or other transactions. If the loans are taken off balance sheet, as they are with CLOs, this means the banks no longer have to cover the loans' risk with capital. That capital is then freed up to be invested in profitable activities such as making more loans. CLOs are products in which banks securitize loans that they have made. This makes the loans, and the risk attached to them, off balance sheet. In a traditional CLO, the bonds are fully funded, which means the loans that cover the bonds are sold to a special purpose vehicle which then issues the bonds. The relatively new form of CLO, which is quickly replacing the traditional form, is the synthetic CLO. The mezzanine debt in these structures is funded like normal CLOs. But the super senior debt, which make sup the majority of the CLO, is secured on loans or other assets that remain on the bank's balance sheet. But the risk that the loans will default is sold on to the buyers of the CLO. The synthetic CLO combines the securitization techniques of traditional CLOs with credit derivative techniques (specifically credit default swaps). The risk is usually transferred to the bondholder only in the event of certain credit events, such as a default on some of the loans. Hidden risks On the face of it, CLOs - traditional or synthetic - are a good deal all round. Banks get to sell off risk, so that they can reduce their exposure to particular sectors and free up capital to be used elsewhere. The investors - mainly insurance companies, export credit banks and US monoline insurers such as Financial Services Assurance, Ambac and MBIA Inc - receive attractive spreads for doing something similar to what they have always done: insuring companies or bonds. By buying a synthetic CLO, these companies are insuring the seller of the CLO against the risk of default. The market is growing rapidly, as securitization and derivative teams learn from each other and find new ways to do bigger deals. But there are certain dangers to the market. ABS are famous for being low risk. A report published by Moody's last year pointed out that there has never been a default in the European ABS market. But if there is a stain on the ABS market's halo, it is CLOs and synthetic CLOs. The danger is that the loans or the risk sold off will come back to haunt the banks. Says Charles Peabody, a banking analyst at Mitchell Securities in New York: "CLOs are just financial engineering that hide most of the risk that banks are taking. I call them CLOWNs - collateralized loan obligations worth nothing. US and European banks' increased use of CLOs represent a definite danger to banks. They're a risk that the system cannot afford to take now." One of the dangers of CLOs is that banks are not necessarily delinked from their securitized loans because there is still the risk of the issuer ruining its reputation if its CLOs default.. This risk was illustrated in 1995 with credit card securitization trusts in the US. The bonds issued by these trusts are one of the few instances of ABS defaulting, as Peabody explains: "The loss rate on loans in the trust suddenly soared and the trusts started defaulting on their loan obligations. Now supposedly those loans were bankruptcy remote from the banks that had originally securitized them. But in practice the money market funds that had bought the bonds said they wanted 100 cents on the dollar back from the banks for the bonds and that if they didn't get it they would never buy bonds from those banks again. So the banks had to step in and take the credit card loans back onto their assets. If they hadn't their reputations would have been ruined." The danger is that a similar event happens with CLOs, and one of two things happens: Firstly, a bank is forced to take loans or risk back onto its balance sheet when it doesn't have the capital to cover it. Alternatively, investors find themselves with defaulted bonds. Tim Drayson, head ABS trader at MSDW, says of this danger: "With ABS, you really want the assets securitizing the bonds to be ringfenced and separate from the company, so that if anything happened to the company it wouldn't affect your bonds. But with credit-linked bonds, like synthetic CLOs, that isn't the case." The report from Barclays Capital says: "The majority of downgrades in ABS have occurred in connection with credit-linked transactions and particularly in respect of CDOs [collateralized debt obligations]." Another danger, specific to synthetic CLOs, is that investors and issuers disagree as to what type of credit events trigger the transfer of risk to the investors. What constitutes a credit event, or an event that triggers a default swap, is not clearly defined. The danger here is illustrated by the example of Conseco, which restructured its debt in the last quarter of 2000. This restructuring provoked debates between banks that had bundled Conseco debt into a synthetic CLO and the insurance banks that had bought the CLO. The insurance banks said that restructuring did not constitute a credit event as Conseco had not defaulted. The banks that issued the CLO said restructuring was a credit event, and triggered the default swap. The argument is not likely to be the last of its kind, and could lead to long and costly legal cases. A third, wider, danger for investors, and indeed for the banking system at large, is the opacity and complexity of synthetic CLOs. Says Drayson: "Many investors don't understand synthetic CLOs." Nor do many CEOs or CFOs. Basle Accord One of the impacts of the BIS' new regulatory proposals, which will come into effect in 2002, is that banks will assess their own risk. There is some debate as to how this will affect CLOs, but it is likely to lead to more CLO issues based on lower graded loans. The consequences if CLOs defaulted would also be greater. The reason for this is that banks will manage their own risk, so can more easily undertake instruments such as CLOs, which avoids regulation. Banks' risk management, far from becoming more transparent, as the BIS hope, is likely to become more opaque as banks invent more and more complicated models for analyzing their own risk.. Self regulation brings some obvious dangers. For one thing, the risk analysts are paid by the institutions they are analyzing. They will be paid more if they can free up more capital. The system encourages risk. Secondly, banks are not known for their self-discipline in market booms. Peabody says: "Self-policing requires a self-discipline that banks do not have. Economic cycle after economic cycle, banks overreach themselves and over-leverage. They manage only for the up part of the cycle." As their loan portfolio gets bigger, and the risks they are exposed to become more obscure, the greater risk grows that banks will find themselves in a slump with a large amount of under-performing loans that they are not able to cover. Ed Mayo, executive director at the New Economics Foundation in London, says: "The Basle Accord was essentially the result of industry lobbying. It was the big banks pushing for changes. The problem is that the changes favour larger banks. They are the ones better able to perform regulation arbitrage like synthetic CLOs. That means they monopolize business away from smaller banks, and so become over-exposed themselves." The danger is not confined to the sell side of the market. On the buy side, it is notable that CLOs, unlike normal ABS, tend to be sold to a handful of investors. Insurers, particularly monoline insurers, snap up many of CLOs in the market. These insurers are already heavily involved insuring other securitization deals - giving, for example, low grade debt an AAA wrap. The principle of CLOs is that they diversify risk. But that is not the case if the risk is bought by just a few companies. How much risk are they actually taking on? Defenders of credit derivatives point out that the main monoline insurers depend on their AAA rating for their success. If they are rated AAA, they must be financially sound. But this ignores the number of smaller monoline insurers in the US that are increasingly active in securitization. These insurers, rated A or AA, do not have the supposedly cast-iron finances of the main insurers. The danger of banks managing their own risk with derivatives was highlighted just this month when Bank of America was reported to have large losses in FX derivatives. The bank had to close trading in its shares when stockbrokers on the NYSE believed the reports and sold its shares heavily. Slightly further back, for another morality tale on the pitfalls of risk analysis, one has only to think of Myron Scholes, who won the Nobel Prize for economics for inventing the Black Scholes options pricing model with Fischer Black. Scholes went on to join a hedge fund called Long Term Capital Management. Enough said. Go to Risk Management page Go to Asset-backed securities