January 3, 2008
Monolines will lose their “AAA” ratings and are egregiously overrated. Both the default risk and the loss given default of their exposure to subprime loans and gravely flawed collateralized debt obligations has been wildly underestimated by the credit rating agencies.
by Janet Tavakoli
On December 14, 2007, Moody’s expressed its concerns about the monoline financial guarantors (“Moody’s Announces Rating Actions on Financial Guarantors”). To achieve the “triple-A” (Aaa) rating, Moody’s looks for an insurer with capital equal to 130% times base case losses and 100% times stress case losses. Using my base case assumptions for subprime losses and Moody’s base case capital criteria, most of the financial guarantors do not merit their current high ratings.
Financial Guaranty Insurance Corporation (FGIC), MBIA Inc. (MBIA) and Security Capital Assurance (XL Capital Assurance Inc. and XL Financial Assurance Ltd. (XL) merit immediate multi-notch—in some cases multi-grade—downgrades (ACA Financial Guaranty Corp. was already downgraded to CCC by S&P). Ambac Financial Group Inc. (Ambac) and CIFG are stronger yet not sufficiently strong to merit the “triple-A” rating, and only Financial Security Assurance Inc. (FSA), and Assured Guaranty Corp. (AGC) would retain a “triple-A” rating (based on subprime exposure). Unfortunately, the latter two are relatively small, so the problems of the larger players will have a huge market impact. Stress test scenarios are even worse. If one considers that an increase in capital cushion may also be required to support other business lines, the situation is desperate.
Since more than $2 trillion in securities are insured by Ambac and MBIA, and another $315 billion insured by FGIC, massive downgrades can be expected on insured bonds. Downgrades will affect CDOs and to a much greater extent, public finance securities. Investment banks will have massive mark-to-market write-downs of collateralized debt obligation (CDO) positions insured by any of the downgraded entities, and other investors will have similar problems.
S&P affirmed the ratings of Financial Security Assurance Inc. (FSA is AAA rated), Assured Guaranty Corp (AGC is AAA rated), and Radian Asset Assurance, Inc. (AA rated). Moody’s also affirmed FSA’s and AGC’s Aaa ratings and Radian’s Aa3 rating with stable outlooks.
Of the “triple-A” rated entities, only AGC and FSA have minimal exposure to subprime backed CDOs, and both have recently provided bond wraps for other “triple-A” entities. Radian (“double-A rated) also has minimal subprime exposure, but it has other mortgage exposure and may or may not be out of the woods.
S&P affirmed the ratings for Ambac Assurance Corp, CFIG’s entities, MBIA Insurance Corp., and Security Capital Assurance (XL Capital Assurance Inc. and XL Financial Assurance Ltd. (XL)) with negative outlook. FGIC is rated AAA but is on negative watch. ACA Financial Guaranty Corp. was recently downgraded from A to CCC with CreditWatch Developing, but the downgrade was long overdue.
Moody’s put the Aaa ratings of FGIC and XL on review for possible downgrade. It affirmed the Aaa ratings of MBIA and CIFG, but with negative outlook. Moody’s had a stable outlook when it affirmed the ratings of Ambac (Aaa).
In late December 2007, Fitch placed the “AAA” ratings of Ambac, MBIA and FGIC on review for possible downgrade giving them more than a month to raise only an additional $1 billion in capital. XL Capital Assurance was put under review, and Fitch said it needs to raise $2 billion (this is higher than my estimate). Fitch indicated Ambac and MBIA could be cut to only to AA+, and in MBIA’s case in particular, it is unclear how it justifies that indication.
In its December 19, 2007 report (“Detailed Results of Subprime Stress Test of Financial Guarantors”), Standard & Poor’s estimated that monoline financial guarantors will have $10.7 billion in aggregate losses under S&P’s subprime loan stress tests for residential mortgage backed securities (RMBS) and collateralized debt obligations (CDOs). S&P acknowledges that from the six-year period beginning 2001 to the end of 2006, the industry had losses of only $2.5 billion representing only 0.14% of a lower average par base outstanding.
S&P’s assumptions may be too optimistic and may have failed to take into account important data. S&P’s stress case for losses is below my base case. For example, for 2006 vintage first lien subprime loans, I am assuming a 30% default rate and a 70% loss rate for net base case losses of 21%. S&P assumes stress case losses of only 15.5%. Moody’s base case assumption for 2006 vintage first lien subprime loans is 11% average losses across securitizations. Moody’s stress case loss of 19% for subprime loans is near my base case of 21% and well below my stress case of 30% for 2006 originated subprime loans.
One major servicer asserts that the rating agencies’ subprime recovery rates are much too optimistic, and have noted that modifying a mortgage is highly preferable to recovering zero or negative value after foreclosure fees and depressed asset prices take their toll on recovery of relatively low loan balances. One major servicer told me that my recovery rate assumption of 30% (70% loss rate) is too high, and my recovery rate is already much more pessimistic that the rating agencies’. Furthermore, loan scrubbing firms have reported current bad loans as high as 20% in 2006 vintage subprime portfolios, and increased interest payment resets on adjustable rate mortgages have not yet occurred. When that happens, the default rates will increase.
It is important to note that percentage changes in default and recovery assumptions are now very important, since the moat has been crossed and the fortress walls have been breached. For many deals, support tranches will be consumed by losses, so incremental losses will be a dollar for dollar hit against wrapped tranches.
S&P points out that a reduction in future business volume and lower risk would free up capital. The irony is that lower future business volume—not actually a good thing, unless risk premiums are too low to justify participation—is the only guarantee on which the monolines can rely. The key problem facing financial guarantors is lack of faith in their ratings.
Most with a nominal “triple-A” rating no longer merit those ratings, since they are undercapitalized. Even if these entities raise capital to meet rating agencies’ requirements, they will still be undercapitalized in the eyes of many investors who no longer trust rating agency guidelines. My own estimates for additional capital far exceed those of the rating agencies (with the exception of Fitch’s estimate for XL). Many lower former low-risk customers will not be repeat customers.
While the financial guarantors—other than ACA—do not have to post collateral and seem to have done pay-as-you go credit enhancement structures, they may experience “death by one thousand cuts,” so it is critical to view the overall picture and not near term survival when evaluating the ratings. If unwind triggers are hit and financial guarantors elect to unwind deals as opposed to trying to work out loan recoveries—possibly a futile exercise—both the CDOs and the inevitable losses will be accelerated.
Contrary to S&P’s assertions, underwriting quality for several of the financial guarantors was not high and was actually naïve. Rigorous statistical sampling of the underlying portfolios was not performed, and they overly relied on faulty models. S&P seems to be the most optimistic of the three major rating agencies (including Moody’s and Fitch). [The Wall Street Journal also made S&P famous for refusing to rate securitizations using loans originated in Georgia and New Jersey, after these states enacted restrictive consumer protection laws designed to raise underwriting standards, albeit somewhat ambiguously (“Lender Lobbying Blitz Abetted Mortgage Mess,” December 31, 2007). Yet S&P was content to rate securitizations containing stated income loans.]
S&P notes that the financial guarantors have offered credit enhancement for deals that include inner collateralized debt obligations (CDOs used as collateral within a CDO), and the financial guarantors have no control over whether these inner CDOs can be liquidated if they hit rating unwind triggers. My larger concern is that some of the CDOs for which financial guarantors offered credit enhancement have other features that are included in the worst practices playbook from the point of view of senior noteholders.
I recently reviewed a publicly available prospectus from a deal brought in the last quarter of 2006. It was a subprime mezzanine deal, a “CDO-squared,” among the worst performing of all CDOs. The underlying portfolio includes tranches of CDOs with the lowest available investment grade rating of BBB when the tranches were initially brought to market. The interest cash flows are diverted to the benefit of lower class tranches. If the deal hits an unwind trigger enabling the financial guarantor to liquidate the portfolio, the prospectus language is ambiguous enough that the financial guarantor may have to fight with other investors for cash flows. Furthermore, it is virtually guaranteed that the illiquid overrated tranches used as collateral will fetch poor market prices.
The CDO-squared was rated by both Moody’s and S&P, and it is inexplicable how the rating agencies justified their opinions. The prospectus states: “Reliable sources of statistical information do not exist with respect to the default rates for all of the type of securities represented by the Collateral Debt Securities.” In fact, reliable performance data was sparse for most of the securities in the portfolio. The underwriter should have stated that given the lack of data, the ratings were no more than paid-for labels and should not be relied upon to make investment decisions.
Within a year of coming to market, both rating agencies downgraded the CDO’s tranches several notches with negative outlook, and more downgrades are inevitable. The deal was overrated before it closed.
As noted before, S&P’s stress case is more optimistic than my base case and the financial guarantors look stretched when one considers only the subprime-related credit enhancements. The following comparisons are roughly in order of worst to best case:
ACA: According to S&P’s stress tests, ACA’s stress case after-tax loss would be around $2.1 billion, but ACA has only $650 to $700 million of capital cushion. My estimate is it would take roughly $3.5 billion in capital to reclaim a single-A rating, and the prospects for future business would be poor. ACA has close ties to collateralized debt obligation (CDO) underwriters including Bear Stearns, Canadian Imperial Bank of Commerce (CIBC), and Merrill Lynch. All of ACA’s projected losses are due to credit enhancement of CDOs. The Maryland Insurance Commissioner signed a December 19 consent order saying ACA could be placed under conservation, rehabilitation or liquidation. This seems not merely possible but inevitable.
FGIC: S&P’s stress tests show estimated after-tax losses due to subprime RMBS and CDOs combined of $2.2 billion, of which around 40% is due to CDOs. FGIC’s capital cushion is only around $300 to $350 million. I estimate FGIC would need $4 billion in capital to maintain a triple-A rating. Some of its private equity owners including PMI Group Inc., Blackstone Group LP and Cypress Group LLC are trying to raise capital, but rating agencies express skepticism about whether their efforts will succeed in time. One of the other private equity owners is CIVC was part owner of Ownit, the mortgage lender that went bankrupt in December 2006. CIVC’s Dan Helle was on Ownit’s board, so subprime distress should not be a surprise to these owners. It recently engaged Blackrock to model its positions, and it is not a good sign that it is depending on outside resources and doing this late in the game. FGIC insures $315 billion in bonds; and its outlook is problematic. FSA has recently re-wrapped bonds already wrapped by FGIC. FGIC should currently be downgraded not merely several notches, but several grades.
MBIA: S&P estimates after-tax subprime-related stress case losses at $3.2 billion, of which 47% is CDO-related. The capital cushion was only $1.75 to $1.8 billion prior to a $1 billion additional capital infusion. I estimate MBIA would need an additional $3.5 billion in capital—even after its recent additional cash infusion—to maintain its triple-A rating.
XL: S&P estimates after-tax subprime-related stress case losses at $884 billion and the capital cushion is only $600 to $650 million. I estimate XL would need an additional $1.2 billion capital infusion to maintain its triple-A rating.
Ambac: S&P estimates after-tax subprime-related stress case losses at $1.85 billion, of which approximately 50% will be due to CDOs, and the capital cushion is only $1.55 to $1.6 billion. I estimate Ambac would require an additional $2.1 billion in capital. Although Ambac has around $32.2 billion of CDO exposure, Ambac may have avoided most of the recent severe problems of CDO-squared deals and recent (2006 and 2007) subprime vintages. It also reinsured $29 billion in other debt with Assured Guaranty, but that only freed up $255 million of capital, a small fraction of what it needs according to my estimates. FSA has also “rewrapped” bonds already wrapped by Ambac.
CIFG: Natixis transferred CIFG to its two major shareholders: Caisse Nationale des Caisses d’Epargne and Banque Federale des Banques Populaires. They will provide $1.5 billion in additional capital for maintenance of the triple-A credit rating. There is some indication that the shareholders are trying to shore up expertise and may provide additional support viewing this as a business opportunity. CIFG’s shareholders may be switching to a Berkshire Hathaway Assurance-type strategy (see below) in anticipation of the effective demise of other “triple-A” financial guarantors, but CIFG’s intentions are not yet clear.
Furthermore, Alt/A and prime loans originated in late 2005 to early 2007 include negative amortization loans and other risky products that will become delinquent as the soft housing market persists and interest payments reset upwards.
Berkshire Hathaway Assurance’s announcement that it will enter the municipal bond business may become a litmus test for other insurers. Berkshire Hathaway is clear that it is happy to do zero business when the risk premiums make no sense. Its triple-A rating is viewed as a genuine rating. Its stated intention of doing premium business at premium prices may leave the largest of the legacy financial guarantors scrambling for scraps.
In the late 1990’s, when CapMAC Holdings Inc., was in trouble (CapMAC merged with MBIA in 1998), it was given a minimal grace period; there was no kidding around. The rating agencies recent actions reek of desperation, since several financial guarantors have been given more than a month to raise capital.
Why are the rating agencies being so indulgent? There may be several reasons. In its previously mentioned report, Moody’s revealed 90,549 securities—of which, 89,709 are politically charged public finance deals—are on negative watch due to the potential downgrades of FGIC and XL alone. Including all of the financial guarantors, more than $2.4 trillion in insured securities will be affected, most of which are public finance deals.
Investment banks that bought protection from financial guarantor counterparties will have large mark-to-market losses. Canadian Imperial Bank of Commerce (CIBC) already announced $2 billion in write-downs related to ACA alone. Other investors will have to acknowledge mark-to-market losses. Finally, the rating agencies’ lack of vigilance in awarding initial ratings on structured finance deals has drawn criticism from many quarters, and I continue to assert that the Nationally Recognized Statistical Rating Organization (NRSRO) designation of Moody’s, S&P and Fitch should be withdrawn with respect to structured finance ratings.
The situation of many of the financial guarantors is much more desperate than the ratings reflect, and once again, I am at odds with the slow response time and artificially high ratings of the rating agencies.
Read finance articles by Janet Tavakoli