A Plague of Subprimes
Part II: Provenance
Five Factors:
4) Hazards of the Game
Mortgage obligations are by their nature very complex and quite opaque. This is problematical always, but the presence of federal oversight (and assumed federal backing) in the CMO market provides impetus to maintain standards in underwriting and in marketing the debt instruments. Absent such oversight, chinks appear at the interfaces of the system. Certain moral hazards arise. Such chinks, and the attendant moral hazards, were inherent to the structure of the “disintermediated” (i.e., unsupervised) CDO market. The system nevertheless had worked well through its early years, but success brought enlargement. Expanding a thing by a factor of five, as happened in the CDO market, increases pressure enormously. When the expansion takes the form of several hundred billion dollars of new cash—as happened in the subprime market from 2001-2006—greatly increases the probability of things going awry.
Collateralized Debt Obligations and other vehicles carrying subprime mortgage components were by all measurable signs as stable as bonds backed by prime mortgages, but provided a higher yield. There was no deception in their value proposition—increased return in exchange for assuming increased risk—but there was an intoxicant: The increased return was quite clear, whereas the increased risk was nowhere apparent. This is an enticement difficult for a bond buyer to resist, particularly during a time when a good yield was hard to come by. It was not to be expected at this point in the process that many purchasers—even professional investors—would pause to consider such abstractions as the degree to which the good performance of subprime mortgage bonds had depended upon a continual rise in housing prices. It was assumed, with good reason, that all pertinent risk factors had been impounded in the pricing of the various tranches of the structured vehicle, which is precisely their purpose.
CDO instruments looked good not only because of their performance record, but the impeccable pedigree most of them carried from Wall Street’s blue ribbon investment banks. The crown atop the package, however, was the most impressive item: These bonds typically carried an official stamp from one and sometimes two major bond-rating organizations—such as Moody’s, Fitch and Standard & Poor’s—whose word is taken by the markets as holy writ.
Rating agents are specially designated by the SEC (“Nationally Recognized Statistical Rating Organizations”) to assess bond risk. For them to misjudge an entire category of debt instruments on a consistent basis is a thing literally unheard of. But the ratings on CDO’s often were wrong. Rating agencies, having no historical data with which to model the behavior of these bonds under adverse economic conditions, wrongly extrapolated from the data available. As a result, grossly generous ratings—80% of subprime mortgages were converted into AAA pools (Greenlaw, et al 2008, using date from Moody’s)—were conferred undeservedly upon subprime vehicles (Wall Street Journal, August 15, 2007). Even CDO’s containing piggybacked mortgages were eligible for a triple-A rating, providing room was made for them in the highest tranche, and the zero-equity mortgages composed no more than 20% of the total mix, the Wall Street Journal reported.
Though absurd in retrospect, the reason for the misplaced confidence in piggybacked loans was simply that their performance record was no worse than that of any other form of subprime mortgage. And that performance record was superlative, measured on a risk-adjusted scale—providing a return substantially greater than that of CMO’s, at (what seemed) only a slightly greater risk. The Wall Street Journal quotes representatives of Standard & Poor’s saying the 20% policy on piggybacked loans “remained appropriate for several years.”
When the “several years” ended, they did so with a bang. The behavior of subprimes changed under stress, and no such factor was to be seen in data gathered prior to 2006. Furthermore, the opaque and complex structure of the instruments themselves made it difficult to establish an appropriate rating, as multiple assessments were required to be made simultaneously on the CDO as a package and on each underlying component. Aside from the inherent difficulty of this task, a “wild card” factor entered the subprime picture, completely unsuspected, and in fact never before seen on Wall Street.
Perhaps the one inviolate reputation among the great army of agents in the financial industry had been that of Wall Street’s bond rating houses. However, among the structural chinks exposed in the highly pressurized subprime market was that between bond issuers and rating agencies.
The root of the problem was the complexity of collateralized debt obligations, which provided an opportunity for shrewd investment bankers to create a link between higher revenues for bond rating agencies and higher ratings for the bank’s mortgage bonds. CDO’s became a very lucrative business for rating agencies, as they required complex, fee-intensive inputs to establish the number and size of tranches. As the subprime frenzy gained momentum, CDO’s became a pre-eminent source of profits for the rating agencies, being the line of business with not only the highest margin, but also the fastest growth.
By unhappy coincidence, Moody’s, perhaps the most sacrosanct name among the rating agencies, was in the process of an institutional change of character. A new generation of management had decided to evolve from its traditional guise of a stodgy and aloof old-school practitioner to a savvy, growth-minded enterprise. The big crush in subprime ratings was therefore an extremely welcome opportunity. The firm acted aggressively, and its CDO business grew from 28% of the firm’s revenue in 1998 to 46% in 2006 (Wall Street Journal, April 11, 2008). Moody’s realized more revenue from structured instruments in 2006 ($881 million) than it realized as a firm five years prior.
The investment banks that were creating the supercharged market in CDO’s were not of a type culturally averse to sharp bargaining. When ratings firms—particularly Moody’s (Wall Street Journal, April 11, 2008)—began to target CDO business, the banks took note. The size of risk tranches in a CDO is a subjective question that has a large and direct influence on the market price of the aggregated debt instrument. Issuing banks were quick to see that there was room, and leverage, for negotiation in rating CDO’s that traditionally had been absent in rating standalone bonds. This had the effect of transferring power to the investment banks, which they used to prize open the rating process to discussion—though no one has admitted on the record that these discussions were in fact negotiations. Such disclaimers notwithstanding, discussions involving matters of judgment on one side and large sums of money to be dispensed (or not dispensed) on the other side could be called negotiations by some observers—SEC investigators, for example.
What is not debatable is that the highly optimistic ratings on CDO’s were an important factor in setting the stage for the secondary market for below-prime mortgages to go into a rage. The proportion of disintermediated mortgages sold in secondary markets rose from about 29% in 2001 (the year Mian and Sufi say CDO’s were “invented,” and certainly the year they became a force in the marketplace) to a peak of 54% in 2005 (Mian and Sufi, 2008). Through those same years, the share among all mortgage originations for subprimes and their near-kin Alt-A loans more than tripled, rising from 9.7% to 32.2%; their dollar volume more than quadrupled, from $215 billion to just more than $1 trillion (Greenlaw, et al, 2008, citing as sources Inside Mortgage Finance, and Morgan Stanley).
Ironically, the rating agencies also played a part in the unwinding of the subprime market spiral. Among the first tolls of Bear Stearns’ funeral bell was an announcement by Moody’s—citing delinquencies and foreclosures in underlying subprime mortgages—that 163 tranches of collateralized securities issued by Bear Stearns Alt-A Trust from 2005-2007 had been downgraded. The rating agency said 78 of those bonds were still on review for further (downward) revision, and that 155 other tranches from securities from the Alt-A Trust were under review for a first downgrade. Bear Stearns stock dropped 8.8% on the news, to $63.89 (Wall Street Journal, March 10, 2008). It was the first step of a short-walk off the plank. As is now notorious, approximately a week later the company was taken over at $10 a share.