Editor: Brian, can you tell us about CDOs and why they became so popular?
Rance: A relatively recent phenomenon, asset-backed collateralized debt obligations ("CDOs") came about in the late 1990s when people first realized that a composite group of debt obligations - each with its own independent probability of default, and each with a recovery rate based upon the type of obligation - behaves much better from a probabilistic standpoint than the rating of the individual bonds or loans.
Many corporate bond and loan obligation CDOs were closed over the course of the 1990s, and eventually people began to ask, if this could be done with respect to corporate loans and corporate bonds, why not use the same technology for asset-backed securities?
Meanwhile, asset-backed securities, which were created in the 1980s, really took off in the 1990s when the SEC adopted rules that fostered the development of the asset-backed securities market. The total asset-backed security market, particularly mortgage-backed securities, swamped the total amount of corporate debt obligations, leading to a great interest in trying to apply this COO technology elsewhereto asset-backed securities.
When CDOs function properly, economically they solve one of the biggest problems that people face in trying to finance financial assets: margin calls. Normally, everything from securities lending to repos to total return swaps - the kind of financings that companies rely on where financial collateral is involved - require that additional collateral be posted or that debt be paid down if the collateral declines in value. Because cash-flow CDOs are the exception to that rule, they became a very attractive financing tool. The ability to use them for asset-backed securities made them an important component of U.S. debt markets.
Editor: Where do mortgages fit into the story?
Rance: The dotcom crash and subsequent low interest rates led to the creation of a large pool of capital around the world (about $70B 70 billiondollars) looking for a home. One of the homes it found was in asset-backed CDO securities, which allowed people to make increased yield. The CDO technology that was employed in turn created a demand for the underlying mortgage-backed securities. The volume of issuance increased dramatically year after year - it mushroomed not only in size, but also in terms of the number of people participating in the market.
Because it was such a successful market, demand for the underlying food that drove it - mortgages - increased. These were not the standard conforming mortgages guaranteed by Fannie and Freddie, which easily find a place in the capital markets, but more "interesting" mortgages with more "interesting" features that paid higher yield - the kind of mortgage instruments that made it possible for many people who had never owned a home to do so. Over time, the volume of ABS issuance increased and the volume of ABS CDOs increased. Most people genuinely believed that these products were in fact safer than corporate bonds, because they thought the underlying mortgage-backed securities that had been rated by the rating agencies had conservatism built into them. If we look at the actual losses on asset-backed securities historically relative to their ratings, asset-backed securities had performed better than comparably rated corporate bonds.
After a few years, Wall Street began to look for ways to make these transactions function more efficiently and to address more customers' needs. The largest segment of the market in asset-backed securities was residential mortgage-backed securities ("RMBS"), and deals over time acquired higher concentrations of RMBS. The structures became more complex in order to reduce the amount of subordinated debt and to make them more attractive from the standpoint of generating arbitrage profits (which is effectively the profit that flows to the equity by being able to finance the pool at an aggregate lower cost than the average spread payable on the mortgage loans). Soon, people began using synthetic technology to create ABS CDOs. Synthetics were used repeatedly, sometimes exclusively, and in some cases multiple deals referenced the same underlying MBS. As the volume of issuance and number of participants involved increased and the product became more popular, competitive forces pushed the spreads down.
Meanwhile, investors continued to believe that the super senior pieces, the highest rated tranches, were very safe. They said to themselves, "This is AAA paper, backed by very safe underlying securities and well-structured transactions. I can earn much more on this than I can earn in alternative investments in the fixed-income space. Even if I can't easily sell it, I'm happy to hold onto that risk." And then something happened: the music stopped, and there weren't enough seats. Everyone on Wall Street found that they had large volumes of these securities on their balance sheets, as well as exposure to the underlying mortgages that were being warehoused in anticipation of closing transactions. This began to suggest that maybe something was amiss.
There were certainly people in the market - hedge fund managers, for instance - who'd taken a very skeptical view of the continued resilience of the U.S. housing market. Prices had gone up too much, too fast.
Editor: What were the first obvious signs of stress?
Rance: The first sign of stress was the ABS index. Then trouble emerged at Bear Stearns, where the principal assets of two very large hedge funds were securities issued by ABS CDOs. In June of 2007 Bear began to halt redemptions in two of those funds, after which they made a number of statements that they were going to support the funds. A highly leveraged fund failed first, and then the other failed. This caused people to wonder what was going on.
Soon a second class of vehicles began to come under attack - the so-called structured investment vehicles, or SIVs, which borrow at the very short end of the yield curve and issue commercial paper and then buy assets at the long end of the market, and which theoretically pay a much higher rate of return and profit based upon the differential between the two ends of the yield curve. The modeling of those structured investment vehicles was supposed to demonstrate that even in a distress scenario where the asset prices are going down, the vehicles would be able to dispose of the assets in an orderly fashion such that maturing commercial paper could be paid off with enough subordination to cover any of the potential losses. Actually, that's the way all large financial institutions make money. The difference here is that world financial institutions have capital and ratings, and structured investment vehicles are grounded upon a ratings agency premise that turned out not to work very well. While there was an initial attempt to rescue these SIVs, the effort broke down, causing people to lose even more confidence in this market.
In the winter of 2007-8, it became apparent that there were a number of financial guarantee insurance companies, monolines, showing extreme stress. The monolines had written protection on a lot of these products, and it became clear that they really didn't have the financial wherewithal to meet their obligations. Likewise, the financial institutions that had bought protection from these monolines began to wonder if they were really protected. Finally, the credit contagion spread from the MBS market to the corporate loan market. The year 2006-7 saw probably the largest leverage loan market ever in terms of deals done and the amount of private equity chasing corporate deals. This in turn drove corporate asset prices up, and the asset bubble in the corporate market grew. Companies were purchased at record prices, and the leverage was made available on the same basis.
Editor: Adam, what do you see down the road?
Glass: Obviously, we can look forward to some new regulation. Some of the things that we think are likely to be seen include: a central clearinghouse or central counterparty for CDS, with daily or more frequent margining, in which counterparties take clearinghouse credit risk rather than each other's risk; federal regulation of CDS to increase transparency, perhaps by requiring delayed position reporting of volumes and pricing of CDS, as well as the reference entity names the contracts are being written on, by swap dealers or even non-dealer market participants; and federal limits on the extent to which leverage can be created through CDS. In addition, we will likely see the regulatory forced exit of monoline insurers from the market (assuming that their creditworthiness recovers sufficiently for them to have an appetite for this business again), because the terms which the NY superintendent of insurance has imposed on insured CDS, such as not allowing the insurance of ISDA termination payments, are probably terms on which none of the monolines' counterparties will want to trade.
With the standardization that comes with central clearing, the regulatory approach to swap agreements in the US will have to undergo a major rethink, as the current structure and its antecedents have always been based on over-the counter derivatives being nonstandardized bilateral contracts subject to individual negotiation, in which the counterparties take each other's credit risk. Central clearing will impose standardization and eliminate individual negotiation and counterparty credit risk. This will require a reconsideration of the statutory exemptions for swap agreements contained in the Commodity Exchange Act, for example.
It is likely that we will see fewer exotic or bespoke structures because of a distrust of the financial models used to value those structures, which turned out to be subject to massive swings in value. One of the things we expect to see is government intervention to set limits on permissible leverage under swap agreements, as the "market discipline" of private counterparties doing individual credit analysis on their trading partners has failed so spectacularly to control excess leverage and speculation. How this will be implemented, and which regulator will take the lead - the Fed, the SEC, or someone else - is not really clear,
Editor: Please tell us about the New York Insurance Department and Circular 19.
Glass: Recently, the New York Insurance Department sent Circular 19 as a letter to all of the financial guaranty insurance companies that it regulates - which is basically all such companies, since most of them are domiciled in New York, and if domiciled elsewhere (like Ambac), they are still licensed in New York and must comply with Article 69 (the portion of the New York Insurance Law that governs financial guaranty insurance companies, or "monolines"). Circular 19 recommends "best practices," which are effectively new rules that monolines must follow, including various provisions intended to be protective of monolines - increased capital requirements and limitations on insurance activities. For instance, it won't be possible for monolines to be protection sellers under a credit default swap that has a restructuring credit event. Only losses resulting from a failure to pay or a bankruptcy will be allowed.
Circular 19 is aimed at shoring up financial guaranty companies by preventing them from engaging in activities that have produces losses in the past, and most of its 14 or 15 pages solely focus on improving monoline regulation, without reference to the broad derivatives market. However, one short paragraph states that the New York State Insurance Department will regulate credit derivatives. This would be effective January 1, 2009, but there will be no retroactivity.
When credit derivatives became popular, monolines wanted to participate because it was clearly related to their core business. Potential participants asked the Insurance Department for clarity on whether an insurance license would be required to offer these contracts as protection to the seller. In 2000 the Office of the General Counsel published an opinion that adhered to the generally accepted view that the key distinction between insurance and a credit derivative is that insurance requires that the insured to suffer a loss in order to obtain payment, while a credit derivative does not. In other words, an insurance contract is a contract of indemnity. The insured must suffer an actual loss of pecuniary value in order to be paid. If the contract doesn't make that loss a condition to payment, even though the buyer of credit protection may be hedging an obligation it holds and may actually in fact suffer a loss if a credit event occurs, that contract will not be treated as insurance and it won't raise "doing insurance business" and licensing issues.
In Circular 19, the superintendent stated that the 2000 letter left an unanswered question: what if, even though the requirement to hold the reference obligation is not a term of the credit derivative, the protection buyer actually holds or reasonably expects to hold the reference obligation at the time of entering into the contract. The implication is that the new position will be that, if that's the case, then the contract in fact is insurance. Many consequences will flow from that, the most important of which is that the protection seller - be it a bank or a hedge fund or a corporate - will now also have to be a licensed insurer if it is doing this business in New York State.