Ambac, FGIC May Be Put in Runoff Mode

The Wall Street Journal today says that even if the efforts to raise new funding for the troubled bond insurers are successful, they are unlikely to stave off a ratings downgrade. This story, based in part on reports coming from the rescue discussions led by New York state insurance superintendent Eric Dinallo, indicates that the reality of how bad the bond guarantors’ situation is is finally sinking in.

One proposal that is underway for FGIC and under consideration for Ambac is to put the companies in runoff mode. That means they stop writing new business and continue operations only to see through existing guarantees. That also means they try to find a way to unwind their credit default swaps on collateralized debt obligations.

I must admit I am at a loss to understand how this process would work. The CDS were written to credit enhance collateralized debt obligations. Unless the CDOs are owned in their entirety by the parties involved in the negotiation (almost certainly not), the group will be taking actions which will affect the value of the CDOs. The unwinding of the CDS will also hit some tranches, presumably the lower rated ones, worse than other (those who have worked on CDO structuring and pricing are very much encouraged to speak up).

The article implies that the CDOs in question are on “portfolios of CDOs” raising the possibility that the CDOs in question are only ones that the banks entered into with the monolines, perhaps to hedge trading inventory and unsold underwritings . If so, that wouldn’t involve third parties and wouldn’t be problematic legally. But the article then refers to what appears to be total CDO exposures of FGIC and Ambac, so Journal’s reporting isn’t at all clear.

But independent of possible ramifications to investors, CDOs are heterogeneous. While CDS on corporate reference entities and indices have ready markets to provide pricing, CDS written on CDO tranches are likely to be one-off events. Any credit default swaps written on them would presumably be difficult to evaluate, given both changes in the underlying credit and deterioration in creditworthiness of the guarantors. How different banks with different and legitimately hard to analyze risks reach agreement on a pricing methodology (necessary for determining how to take them out of their CDS exposures) may well be a very complicated task. I hope the banks can reach agreement for this could prove to be a large stumbling block (or perhaps there is a simple way to cut the Gordian knot that, as an outsider, I fail to see).

Note that the Journal story quotes a UBS analyst who has far and away the rosiest view of the industry.

Separately, the Financial Times reports on yet another largely unrecognized hole in the bond insurers’ balance sheets. Wall Street was apparently fond of a so-called negative basis trades. If they bought a bond, hedged it with credit default swaps, then hedged the risk of the guarantor defaulting (generally a monoline) with a different guarantor (generally a different monoline), they could accelerate the expected profits over the life of the deal into the current period. The result is that the bond insurers have an unknown (to the outside world) but potentially significant number of guarantees written on each other. Yikes. The FT report suggests that losses on these CDS are not factored into most estimates.

From the Wall Street Journal:

Rescue plans are starting to take shape for struggling bond insurers, but they aren’t likely to prevent further ratings downgrades for many of the companies…..

Still, some banks and investors working toward salvaging the bond insurers, which guarantee the interest and principal in the event of default, are realizing that even the best plans could require them to settle for less — less risk, less reward and bond insurers with less-than-triple-A ratings in the future, according to several people familiar with the rescue talks.

A group of banks — betting that the insurers still have some value — are working with the management and investors of New York-based Financial Guaranty Insurance Co. on a potential plan for FGIC. They have held a series of meetings and conference calls in recent days. The group, which is led by Calyon, a unit of French bank Credit Agricole SA, includes UBS AG, Société Générale SA, Citigroup Inc., and Barclays PLC. A Calyon spokeswoman declined to comment.

The banks are trying to figure out how to commute, or unwind, their credit-default swaps, which are contracts they entered into with FGIC and other bond insurers to guarantee their portfolios of complex debt securities known as collateralized-debt obligations, or CDOs, according to people familiar with the talks.

A consortium of banks working toward a rescue plan for bond-insurer Ambac Financial Group Inc. also is discussing a similar option, according to people familiar with the matter

The banks, then, would share in the proceeds that the bond insurers would make as they collect premiums and wait for their existing portfolio of policies to expire, or “run off.” In this scenario, the most the banks are hoping for is that the bond insurers’ credit ratings don’t fall below double-A, but they aren’t getting their hopes up for a return to triple-A glory, according to two of the people.

“There’s run-off value as you get rid of the toxic elements of these companies through commutations,” says David Havens, an analyst at UBS Securities, who isn’t involved in any of the bailout talks. “Only about 5% of the business is problematic, and 95% is fine.”

Ambac, FGIC and rivals MBIA, Security Capital Assurance Ltd. and CIFG Holding Ltd. are exposed to about $100 billion in CDOs that were backed by rapidly deteriorating subprime mortgage collateral, according to Fitch Ratings.

Some analysts estimate that sharp downgrades of these bond insurers, or their insolvency, could cause banks to write down as much as $70 billion. In late December, for example, Calyon said €1.2 billion ($1.75 billion) in its write-downs was tied to bond insurers, mainly ACA Financial Guaranty Corp., which was downgraded to triple-C from single-A.

By unwinding the credit-default swaps, the banks would enable the bond insurers to free up capital, which they could use to help preserve their ratings or help prevent further downgrades.

Of about $315 billion in debt that FGIC had insured through to Sept. 30, 2007, about $31 billion was backed by mortgage collateral and $8 billion was backed by subprime mortgages, according to an FGIC presentation. Ambac has about $67 billion in CDO exposure….

Fitch, for example, says it is possible that the majority of bond insurers won’t continue to have triple-A ratings in the future.

“It’s just an indeterminable amount of losses on these assets and the final number could be far more significant that we had been envisioning,” Thomas Abruzzo, managing director at Fitch, says. Last month, Fitch and Standard & Poor’s downgraded FGIC to double-A from triple-A. Fitch also downgraded Ambac to double-A.

A key hurdle for the banks and the bond insurers is determining how much the banks should get in exchange for tearing up their credit-default swaps, and whether owning stakes in companies that could get further downgraded is fair compensation, says one person familiar with the discussions.

Another option being bandied about by analysts and others is to form a new company, funded by the banks, which could take responsibility for meeting the obligations of some of the insurance policies — mainly the credit-default contracts — weighing on the bond insurers.

This model would have the added benefit of enabling banks to unwind some of their credit-default swaps with the other beleaguered bond insurers, and guarantee their debt with the new, triple-A-rated bond insurer.

But there are potential downsides, too. Banks might be choosy about which risks they would be willing to let the new insurer take on, and that could mean bond insurers might be left with the high-risk business, Steve Stelmach, an analyst at Friedman, Billings, Ramsey & Co. noted in a recent research note.

From the Financial Times, “New danger appears on the monoline horizon“:

As the bond insurers, or monolines, have seen their seemingly rock-solid AAA ratings begin to buckle, worries have grown about what downgrades for these companies might mean for banks.

Now, one particular type of trade done between banks and monolines is being seen as an extra hidden danger.

These so-called negative basis trades were done in large volumes in recent years. They allowed both banks and monolines to book apparently “free money” and saw monolines writing guarantees on each other. If they have to be unwound, it will be a costly business for all involved.

The real problem is that almost no one has any idea how significant the profits taken on these trades might be. These trades were profitable because a bond could pay out more in interest than it cost to buy the insurance available in the derivatives market to protect the holder against default. In the world of structured finance, a bank would buy a bond, get it guaranteed, or wrapped, by a monoline to support the bond’s AAA rating, but then also pay another monoline to write a default swap on the first monoline, to guard against it defaulting on its guarantee.

The difference between what the bank paid for the insurance and what it received in yield from the bond could be pocketed as “risk-free” profit – and in many cases banks took the entire value of that income over the life of the bond upfront.

One senior industry insider admits that billions of dollars worth of these trades were done, but insists they were mostly restricted to the arena of utility and infrastructure debt. These were attractive both because they were of long maturities and because they were often linked to inflation, which would increase the returns.

“On a £100m deal over 25 years a bank could conservatively book £5m up front – even more if it was index linked,” says the senior industry executive.

For the monolines, the trades were also seen as near risk-free profit when taking the position of writing protection on peers.

The same executive insists that monoline activity in CDOs was restricted to the hedging of senior tranches that banks had retained on their books after structuring deals and had nothing to do with negative basis trades.

However, others are less sure. Monoline analysts at some of the banks believe a large amount of negative basis trades in the US were done on super senior CDO tranches, but admit they have no idea what proportion of total CDO business for the monolines that was.

Bob McKee, an analyst at Independent Strategy, a London research house, believes that up to $150bn worth of CDO business done by the monolines could be negative basis trades.

Standard & Poor’s, in a note on the potential impact of monolines on banks this week, said it believed some of the CDOs hedged by bond insurers were part of a strategy of “negative basis trades”.

The problem is that if monolines are downgraded and their protection becomes ineffective, profits booked up-front need to be reversed. Restating earnings is a very tricky area for investment banks – not least because the traders involved will have long ago pocketed their bonuses.

Print Friendly, PDF & Email


  1. John

    “(or perhaps there is a simple way to cut the Gordian knot that, as an outsider, I fail to see).”

    Yves, no doubt you are aware that one of the collective Masters of the Universe named its off balance sheet entity “Gordian Knot”.

    These things were designed never to be unpicked.

  2. RK

    Those negative basis trades bring to mind the herarchy of methods used to evaluated securities proposed by James Grant:
    1. Mark to market
    2. Mark to model
    3. Mark to myth
    4. Mark to year end bonus

  3. Francois

    “A key hurdle for the banks and the bond insurers is determining how much the banks should get in exchange for tearing up their credit-default swaps, and whether owning stakes in companies that could get further downgraded is fair compensation.”

    The banks should get anything from that mess? Really!

    Caveat Emptor apply only to the individual, but not to the banks?

    My breakfast is trying to force its way out of my digestive system.

  4. realty-based lawyer

    Have several years’ experience in the bond insurer industry, including working on the CDS under discussion. Don’t think it’s all that complicated.
    – The monolines typically wrapped only the AAA or super-AAA tranches of CDOs. (There may be some exceptions, but not many.) They didn’t want exposure to the lower tranches, and the rating agencies didn’t want them to have such exposure either.
    – The negative basis trades were done with single counterparties, so each swap is with one bank. That’s true for the CDS business generally (if you think about the documentation and the fact that CDS aren’t treated as securities, you’ll understand why). Therefore, this unwind is feasible.
    – “Standard ISDA” documentation for the monolines’ CDS is that the protection buyers can “walk away” from the swap at any point by simply ceasing to pay premiums. (Some monolines, but not I believe the majority, required a termination payment from banks wanting out; but I don’t think that was the “standard” position.) In those swaps, the bank’s only loss is the value of the protection being provided by the CDS.
    – It may make sense for both parties. The banks are currently paying premium (since neither the underlying nor the monoline has defaulted). That premium was based on AAA protection and is now overpriced, with the result that the banks have already suffered a loss on the trade. On the other hand, from the monoline’s perspective the premium turned out to be underpriced for the risk involved. There’s some negotiation room here.
    – As for the swaps on monoline swaps/policies, remember there’s no acceleration. They’re pay-as-you-go. Also, as far as I know the monolines didn’t write protection on downgrades of other monolines – only their default. So such swaps would require only making the payments due on the underlying after the first monoline failed to pay. Still very low-risk due to the dual-default requirement.

    Having said all that, I agree that the negotiations are tricky. The difference in positions here is more than over the amounts involved: it involves the business model and viability of one counterparty. Easy to imagine scenarios in which they won’t reach agreement.

  5. bob

    I don’t think that the losing the CDS on CDOs will affect them that much, because CDOs are basically all crap anyhow. As near as I can tell, the only reason to create a CDO in the first place is to artificially inflate the ratings on some of the lower tranches of MBS – that horse is out of the barn, everybody knows that the CDO king is butt-naked.

    This just confirms what is now pretty widely known.

  6. Yves Smith

    reality-based lawyer,

    Thanks for the help. The unclear drafting of the WSJ piece suggested that the rescue group was contemplating unwinding CDS relative to CDOs (as in ones sponsored by the IBs) and I couldn’t get my mind around that, which got in the way of thinking through the simpler case of hedges on super-senior tranches.


    The issue here is that there are still some holder who haven’t had to write the CDOs down because under their regulations, they aren’t required to (ie, they aren’t marked to market). But they would if they were downgraded, which is what a formal downgrade or removal of the monoline CDS. Pension funds fall in this camp.

    And the other worry is that bond insurer downgrades would force sales of CDOs since certain types of holders can only hold investment grade paper (again pension funds and for different reasons, insurers; while both can hold non-investment grade assets subject to certain limits, they’ve generally used them up on alternative investments).

  7. Karl

    To Reality based laywer/Yves —

    My understanding is that there has been quite an active market for CDS, and while the premium payer remains static, the ownership of the CDS can go through multiple iterations. If the insured walks away from the contract, how does that ultimately affect the CDS owner? Wouldn’t that make the contract itself worthless?

  8. Yves Smith


    The CDS that trade are typically written either on what in the trade is referred to as “single names,” meaning individual corporate issuers (say General Motors or Proctor & Gamble) or indices. The CDS here don’t fall into either category.

    CDS are also written to credit enhance collateralized debt obligations. I’ve never heard of them being detached and traded separately (not that conceptually that couldn’t happen, but the trust agreement with the legal entity that owns the CDO assets would presumably prohibit that unless a substitution was made).

    The CDS that the investment banks are discussing are ones they had written for them to hedge super senior CDOs they are holding. Again, since these would be one-off deals (in each case, the bank that was stuck with these tranches sponsored the issuing entity), so I doubt these would be traded, although the hedges might be adjusted as the markets continued to fall (hedges are almost never perfect; the investment bank might have reduced the size of its hedge by writing a CDS to the monoline, or conversely, has the monoline write more CDS if it though it needed to increase the hedge).

  9. Gawain

    The WSJ article is typical of a lot of coverage of the monolines — i get the feeling that it is poorly written because its writer did not quite understand what he was writing about.

    The MBIA conference call (and the attendant presentation) outlines the extent of mutual exposure among the monolines; it isn’t much (relative to the entire portfolio).

    Also, the numbers sound more scary than they are: a 30 bn insured portfolio sounds HUGE, but default payments on the portfolio come only 12-15 months after a credit event, after deduction of the deductible (sometimes as much as 20%), are limited by reinsurance and recoveries on the collateral, and in any case are stretched out over a 30 year period. Thus, if an entire 30 bn portfolio were to fail today, a typical monoline might have to make its first payment in 12 months, and that payment may perhaps be 400 mln. And it would then acquire all the underlying assets and set about recovery, so the cost of these annual payments would eventually go down.

    None of which is to suggest that the monolines have no problems; or to claim that their survival is assured. But I at least hold both monoline stock and debt.

  10. Karl

    Yves —

    Google “Trading of Credit Default Swaps.” I think maybe my confusion is that derivatives of CDS’s trade, and not necessarily the CDS’s themselves.

    To echo Gawain..

    Everybody, it’s important to understand the length of the time over which actual economic monoline losses will happen. Think cash here — put yearly claims payments on one side and premium income/portfolio income less expenses on the other side and compare the outcome. Really, keep this in mind: All the mark to market adjustments killing the banks are irrelevant to a bond insurer. The risk the monolines face is whether investors lose principal and interest on the underlying investments and trigger their guarantees. Even if the guarantee is called, the monolines pay out claims on principal only at the original redemption date so the NPV exposure is a fraction of the notional sums insured. And then there’s the possibility of asset recovery, too.

    That means you need to look at these monolines on both an ongoing concern basis and in a runoff scenario. See what happens — even if it means, believe it or not, that these entities can make it as ongoing, though crippled, concerns.

    What is the probability of a truly armageddon type of scenario in the housing market and the overall economy? Warburg did exactly this and figured that the conservative value during a runoff for MBIA would be about 30 dollars.

    Another way to think about it is this: monline exposure can be compared to a written out-of-the-money put option to protect investors against extreme market events.

    Now, the monoline business model is a complete wreck. And from a cash point of view, it could unravel for them, and fast.

    But, to reiterate, so far we have had mark to market losses in a portfolio that is supposed to be kept until maturity. To date all the brouhaha has been about covenant triggers, valuations, and accounting entries — green eyeshade stuff. Of course, the level of capital required to maintain an AAA is far larger than that required to merely assure that claims are paid for the next year or two. But again, that’s a labeling issue, not an REAL economic issue.

    The real economic losses will occur when claims are paid.

Comments are closed.