Loose Accounting Rules Keep Defaults Off Books -- for Now
By JONATHAN WEIL and HENNY SENDER Staff Reporters of THE WALL STREET JOURNAL
So much corporate debt. So many defaults. So why haven't the losses shown up yet on the books of companies exposed to the debt?
Chalk it up in part to some elastic accounting rules that companies easily can game to mask their losses. And that could lead to some mean surprises in the months and years ahead, when the companies with the most exposure no longer can bury their looming losses and finally have to come clean.
Recently, many large banks have been crowing that they successfully shifted much of the corporate-debt-related risks off their books in time to protect themselves, through complicated packages of sliced-and-diced corporate borrowings and related side deals. But given the paltry size of the losses reported by the companies that took that risk off the banks' hands, it would seem as though the risk has disappeared into a giant black hole --which, of course, it hasn't.
"We hear the banks praising their gains, but nobody is standing up on the other side with big losses," says Robert Grossman, chief credit officer of Fitch Ratings. "It is a real disconnect."
At least $400 billion of so-called collateralized-debt obligations has been created since about 1996. These CDOs basically are pools of loans or bonds, sold off in chunks as securities to investors. Much of the debt was issued by nascent telecom companies, which promised high interest rates to compensate for their risky -- and often, in retrospect, fatally flawed --business models. As bond defaults soared through 2002, many of the companies that issued the debt have stopped making payments on their obligations.
For investors who embraced the risk of that debt and were counting on regular interest and principal payments, that means their holdings right now are just a fraction of their original value. Getting rid of them on the secondary market may not be easy, because trading in this arena often is thin. About 40% of all collateralized-bond obligations, a type of CDO, issued in the late '90s that originally were rated above triple-B now stand at triple-C. That suggests mounting defaults are a matter of time. Yet reported losses, or even disclosures warning that big losses might be on the way, so far have been virtually nil.
How come? With some exceptions, the conventional thinking goes, the risks associated with the bad debt were spread so widely among so many different large institutions -- especially major financial-service and insurance companies -- that the impact of the recent wave of defaults on any single holder is immaterial. Of course, that will be small comfort to any investor who owns stock in a company that winds up being an exception to the rule and surprises the market with a significant hit to earnings. Fears that there are at least a few ticking time bombs out there may prove to be warranted.
Indeed, what helped make these debt issues attractive to institutional investors is the fact that the accounting rules governing how they are treated in financial reports allow for so much discretion when it comes to valuing them. "The subjectivity of the accounting is one of the appeals of CDOs," says Ray Kennedy, a managing director at Pacific Investment Management Co. in Newport Beach, Calif. "The accounting is all over the map."
For example, some CDOs, which commonly have a 12-year maturity period, are classified as long-term investments held to maturity, depending on a company's declared intentions of how long it plans to own the CDO. In that case, prevailing U.S. accounting rules may let the CDO's value sit untouched on a company's balance sheet for years without being written down. That is because the CDO doesn't have to be written down unless it has permanently declined in value.
Tom Linsmeier, chairman of the accounting and information-systems department at Michigan State University in East Lansing, notes that the accounting involves multiple judgment calls, including whether the value has declined significantly, and, if so, whether that is a temporary condition or not. "The discretion and subjectivity in making an impairment-test determination is huge," Mr. Linsmeier says.
Reasonable people may disagree on these sorts of judgments. But the people who get to make them are the company's management, who may decide to err on the side of optimism.
Other times, if a CDO is held in short-term trading accounts, its value must be adjusted on the balance sheet up or down every quarter, with the changes reflected as losses or gains on a company's income statement. This process, known as "mark-to-market" or "fair value" accounting, also gives companies lots of discretion.
If trading in the secondary market for a particular CDO is thin, and quoted market prices from different dealers vary widely or don't exist, the CDO's real value is anybody's guess. In that event, companies get to use mathematical models to form valuations based on lots of estimates of future market conditions. These also lend themselves to the temptation for companies to value their holdings while wearing rose-colored eyeshades.
CDOs created from bonds and loans provide just one of the accounting playgrounds. Other highly engineered financial products that are linked to corporate borrowers' credit performance include derivatives that let banks and insurers, for instance, offload specific credit risk for a fee. Unlike the most basic CDOs that are held to maturity, credit derivatives must be marked to market. Absent liquid markets with readily available price quotes from dealers, their value also is often in the eye of the beholder.
How much room for disagreement abounds? Consider MBIA Inc.'s book of business guaranteeing CDO tranches totaling a net of about $66 billion. The financial-services concern last year became the target of New York hedge fund Gotham Partners Management Co., which had bearish bets on MBIA's stock and credit. To date, MBIA has reported $36 million in losses related to its CDO guarantees. (Gotham now is winding down its principal funds.) But based on information provided by MBIA about the guaranteed CDO slices as of Aug. 30, Gotham sought estimates from two dealers on the mark-to-market losses the MBIA-guaranteed CDOs faced. One dealer estimated the losses at $5.3 billion, another said $7.7 billion, according to a report published by Gotham.
Of course, Gotham had much less information to work with than MBIA, which vigorously denies the report's accuracy. Still, two dealers with identical information came up with estimates varying by nearly 50%. Even some Wall Street analysts agree that Gotham might have had a point. "Though market dealer quotes have their shortcomings, it is possible that the loss could be greater than the $36 million recorded by the company," analysts at Morgan Stanley concluded in a Dec. 16 report assessing the Gotham report.
Robin Selvey, a senior portfolio manager for the nonprofit American Automobile Association's Auto Club Group insurance unit, offers another cautionary example. In 1998, her fund spent about $5 million on a collateralized-bond obligation. The $5 million slice carried a triple-B credit rating, a low investment-grade rung but not the CDO's riskiest sliver. She thought the pool's diversity meant safety. Little could she imagine that 1998 would produce some of the worst-performing junk bonds of all time.
Today, Ms. Selvey has marked down the investment to $87,500, though some dealers told her it was valued at more than 10 times that. "I am widely regarded as the Rodney Dangerfield of the CBO world," she told a conference on the subject last month in Southern California.
Write to Jonathan Weil at jonathan.weil at wsj.com and Henny Sender at henny.sender at wsj.com