Triple-A ratings, the key to the municipal bond insurers' franchise, overstate the strength of the industry, a report released last week says.
"The bottom line is that investors should not assume that bond insurance is fail-safe, [that] this protection reduces but does not eliminate default risk," said Fred S. Cohen, municipal credit research analyst at Sanford Bernstein & Co. and author of the report. "There is a possibility, albeit small, that a severe future recession could create an abnormally high level of defaults. Under those conditions, the insurance safety net would be likely to fail," Cohen said.
Cohen's report says that a double-A rating is a more accurate indicator of the creditworthiness of bond insurers.
Reactions from industry officials were surprisingly mild.
"Obviously, four rating agencies disagree with this gentleman," said Wallace O. Sellers, chairman of the Association of Financial Guaranty Insurers. "He's entitled to his opinion, but the four rating agencies all agree that these companies are triple-A."
Insurance officials said privately that the report makes unsubstantiated arguments and cited both the rating agencies' analyses and their own research as evidence.
"No credit, including the federal and state governments, would ever be rated triple-A if required to meet the catastrophic and wholly unrealistic economic environment envisioned by the report," one industry executive, who asked not to be identified, said last week.
Other industry officials said insured bonds might even be stronger than triple-A credits that earn that distinction on their own merits.
"While the report points out the many strengths underlying insured bonds and their value to investors, we strongly disagree with the conclusion that insured bonds should not be rated triple-a," said David H. Elliott, president and chief executive officer of Municipal Bond Investors Assurance Corp.
"Our own research convincingly demonstrates that insured bonds should perform even better than natural triple-As under severe economic conditions. The industry has vast claims-paying resources that could easily handle any losses that might develop under severe economic conditions," Elliott said.
Renwick A. Paige, vice president of market development at Financial Guaranty Insurance Co. said in a prepared statement: "Mr. Cohen has written a thought-provoking piece. I encourage investors who are familiar with the models and their assumptions to request and read FGIC's paper on capital adequacy. We believe that the myriad of worst-case assumptions that we survive will impress even the harshest critics."
Many bond insurance officials said privately that they did not wish to lend credence to the report by discussing it and that the rating agencies - not the insurers - should defend their criteria.
As a group, rating agency officials stood by their ratings but said they welcome analysis from other parties and generally praised the report.
"Obviously, the rating agencies believe in their own analysis and that the insurers do rate a triple-A, but there's good stuff in the report," said Richard Smith, managing director at Standard & Poor's Corp. "We just don't agree with the conclusion."
"There's room for more than one opinion, although we stand by ours," said Laura Levenstein, vice president and supervisor at Moody's Investors Service.
David Litvack, vice president at Fitch Investors Service, said the report's flaw is its failure to note differences among individual companies. Fitch rates just one bond insurer: FGIC.
"The report is well written and provides an excellent overview of the risks and benefits of bond insurance," Litvack said. "I agree that not all bond insurers are triple-A, but some are. We feel there are big differences in the bond insurers' ability to pay claims."
Standard & Poor's rates all 10 of the major monoline insurers; Moody's rates eight monolines; Duff & Phelps Credit Rating Co. rates four.
The report said the current ratings are undeserved because the rating agencies' stress tests, from which triple-As are derived, do not adequately incorporate recent changes in municipal finance and underestimate the impact that widespread defaults would have on insurers' balance sheets.
Cohen said while the insurers pass the stress tests on paper, the rating agency scenarios do not take into account how the market will react to a dramatic downturn.
"If you had even 5% to 10% defaults, everyone would run for the hills. The situation would be out of control," Cohen said. "When things. start to snowball, people ignore scheduled debt service, and you've got a potential liability. All of a sudden you take a big accrual. That's the mentality that will set in."
The report says repeatedly that it does not expect a deep, protracted recession to hi.t the municipal market. However, it does say that "the probability of one occurring eventually is high enough to warrant a rating below triple-A" for the insurance industry, since a triple-A rating implies a "near-zero chance of bankruptcy."
In contrast to the rating agencies, Sanford Bernstein says the monoline insurers could not withstand a severe depression, implying that the rating agencies analyses are flawed.
The rating agencies' depression scenarios are all based, to varying degrees, on the performance of municipal bonds in the Great Depression of the 1930s, the report says.
One reason the tests are inaccurate is that municipalities "have become much more sensitive to economic cycles" than during the Depression, the report says.
For example, municipalities today rely much more on "economically sensitive" taxes, like those on personal income, retail sales, and corporate profits. Such taxes account for 41% of total revenues, compared to only 9% in 1930. Meanwhile, property taxes, which the report argues are affected more gradually by changing economic conditions, have fallen to 22% of municipalities' total revenues, from 60% in 1930.
Another problem with relying on the Great Depression to determine triple-a criteria, according to Sanford Bernstein, is that general obligation issuance has declined precipitously since the 1930s, and other, untested bond sectors have picked up the slack.
"While many of these [non-go] bonds performed well during the last recession, we do not know how they would respond to depression-like economic conditions," Cohen's report says. "The rating agencies have estimated the potential loss, but their analysis is an educated guess at best. Our suggested double-A rating for the insurers acknowledges the inherent unknowns of these untested sectors." Levenstein of Moody's said the rating agency's stress test does incorporate changes in the market. "The rating test has evolved over time and is still evolving. We always try to be consistent with what's going on today and what will go on tomor- row," Levenstein said. "We have an extremely conservative stress test." The Sanford Bemstein report was written in response to the growing number of inquiries regarding bond insurance that the firm has received from its clients, Cohen said. "It's a buyside piece, going out to our individual clients, mostly high net worth investors and small businesses," Cohen said. "We're trying to give them an overview and then a macro look at strengths and weaknesses of the industry and how we value the insurers in a total return framework."
Throughout the report and in interviews, Cohen stressed the strengths of the industry and the role insured bonds can play in a diversified portfolio.
"Our concegmn about insurance do not imply that insured bonds have no place in an actively managed tax-exempt portfolio," the report says. "It should be noted that the double - a rating we consider appropriate for insured bonds is at the high end of the market's credit assessment" in terms of the trading value of insured bonds.
