MBIA Downgrade Increases Collateral Requirements; Clarification on CDS Acceleration in Insolvency/Custodianship (Corrected and Updated)

Warning: the post below is a bit geeky. Readers might start with our other current MBIA post and then return here for further details.

Please also note that due to a reader catching an error I have looked further into the concept of claims paying resources and made considerable modifications to the paragraphs relating to claims paying resources below, and also added another sentence flagged as “Update”.

Did MBIA’s downgrade impair as much as 25% of its claims-paying resources? Consider these statements, the first from a letter from MBIA yesterday on the Moody’s downgrade of its insurance sub from Aaa to A2 (hat tip to Jason for a useful question):

This ratings action will give certain holders of guaranteed investment contracts the right to terminate the contracts or to require that additional collateral be posted. Again, we have more than sufficient liquid assets to meet any additional requirements arising from any terminations or collateral posting requirements.

The same letter notes that MBIA has $16 billion in claims-paying resources (we’ll return to that concept later). Update: a reader has advised me that teh GICs are not in the same entity as the insurers and hence any changes there would not affect claims paying resources.

Now track back to the most recent 10-Q to see what the effect of a downgrade to A would be. It increases the cost of funding on $400 million of instruments (rounding error). Here is the part related to the GICs:

Investment agreements issued by our asset/liability products segment generally provide for collateral posting or termination in the event of a downgrade of MBIA Corp.’s credit ratings. The maximum collateral posting would occur at a single-A financial strength rating by Moody’s or S&P, and the maximum terminations would occur at a triple-B financial strength rating by Moody’s or S&P. Based on the asset portfolio as of March 31, 2008, the maximum collateral posting requirement at a single-A financial strength rating would be $12.7 billion and free collateral at a triple-B financial strength rating would be $9.2 billion. As of March 31, 2008, the Company’s collateral posting requirement totaled $8.25 billion and free collateral totaled $16.8 billion based on MBIA Corp.’s current financial strength ratings. We believe that the liquidity position of the asset/liability products segment is adequate to meet these requirements related to investment agreements under stress scenarios.

Per the update above, although this nearly $4.5 billion reduction seems significant, it apparently has no impact on claims-paying resources.

I am shortly going to turn the mike over to a former general counsel of a bond insurer, who provides a great deal of informative detail on the issue that the New York Times raised in its recent article on MBIA, namely, what happens to the credit-default swaps contracts in the event of insolvency or regulatory takeover. He very clearly states:

Given the billions of surplus to policyholders reported by MBIA and the mechanics of statutory accounting (which is very different from GAAP), regulatory insolvency seems extremely unlikely in any future I can foresee.

However, the reader input on the GIC question led me to look deeper into the definition of claims paying resources, and I am troubled by what I see. MBIA’s statutory surplus, as shown for its combined insurance companies at the end of 2007, was $3.7 billion (versus $4.1 billion the prior year end). That is a lot less than the so-called claims paying resources. This concept is used by bond insurers (my impression is not by the insurance industry generally; “claims paying ability” which is usually a rating, is a different concept).

CPR is equal to capital (which in insurance lingo is surplus) PLUS the unearned premiums (this definition comes from the financial guarantors’ industry association, the AFGI). . Yet in MBIA’s case (per its statutory filings) the unearned premiums (year end 2007) show $3.5 billion of unearned premiums (see the liablity side of the balance sheet) while the asset side shows $7.3 billion of uncollected premiums. Query how you get to $16 billion if the definition of claims paying resources is surplus + unearned premiums.

Aha, but it’s in the operating supplement, page 7, which as our former general counsel points out, is prepared in accordance neither with GAAP nor SAAP and “should be treated with due caution”. Using the year end figures to make it comparable to the statutory report, they add a $2.7 billion contingency reserve to the surplus and $2.6 billion of present value of installment premiums (I have serious problems with that concept; if you are a bank, you don’t get to discount your future interest spread and count its as some sort of capital). They also show $900 million in “Loss and LAE reserves” and “soft capital” of $850 million.

What raises my eyebrows (aside from the notion of the present value of installment premiums) is that the operating supplement shows that their claims paying resources increased by $1.5 billion by the end of the first quarter versus year end The change is largely explained by the change in “Loss and LAE reserves” which increased by $1.2 billion since year end (note that “Loss and LAE reserve” corresponds to the total shown for “Loss” on the year end statutory balance sheet).

Any readers who know this area are encouraged to comment.

Although there have been a number of red flags waved about MBIA, the biggest is its reinsurance. This discussion came from a letter Ackman sent to MBIA in January:

As of September 30, 2007, MBIA has re-insured approximately $80 billion of par value of its exposures. More than $42 billion of this reinsurance was purchased from Channel Re, a Bermuda- based reinsurer whose only customer is MBIA. The two most senior officers of Channel Re are former executives of MBIA. MBIA owns 17% of the company and has two representatives on Channel Re’s board of directors.

On recent conference calls, Moody’s and S&P have stated that they have not yet updated their ratings of the monoline reinsurers including Channel Re. Earlier this week, on January 16th, Partner Re and Renaissance Re, the majority equity owners of Channel Re, wrote off the entire value of their investments in Channel Re due to losses it has recently incurred that substantially exceed Channel Re’s capital, an impairment that Channel Re’s two majority owners have concluded is “other than temporary.”

Despite the fact that Channel Re has negative book equity and $42 billion of MBIA’s credit exposure – $21.5 billion of which is CDOs of ABS or CLO/CBOs – Moody’s and S&P continue to rate the company Triple A with a stable outlook….

Captive reinsurers whose ratings are not regularly updated offer the potential for abuse. We believe that MBIA reinsured on a quota share basis 25% of its 2007 CDO transactions with Channel Re. As a result of Moody’s and S&P not updating its ratings
of Channel Re, these exposures do not appear on MBIA’s list of exposures and have not been included in your calculation of MBIA’s capital adequacy.

MBIA’s second largest reinsurer is Ram Re which has reinsured $11 billion of par as of September 30, 2007. While the rating agencies have not updated their credit ratings of Ram Re, the market appears to have already done so. The publicly traded stock of Ram Holdings Ltd., the parent company of Ram Re, has declined 92% in the last year. The company currently trades as a penny stock with a market value of $32 million.

We believe that Ram Re is substantially undercapitalized and therefore, like Channel Re, is unlikely to be able to meet its obligations to MBIA.

Moody’s downgraded Channel Re to Aa3 in February. The point is that serious impairment of either reinsurer has the potential to substantially lower MBIA’s surplus (the cost of replacing the reinsurance would be substantial, and likely uneconomic).

Let me emphasize that none of the above points to solvency issues in the near term. But they do raise worries about long-term viability, particularly as the credit crunch rolls on and municipalities start showing signs of distress.

Now to the information on how CDS would play out in the arguably not-so-likely event that MBIA were to get in serious trouble. I was wrong in some of the things I said earlier. While the CDS holders do have the right to accelerate (demand immediate payment), it is far more useful to them as a bargaining chip, since forcing an insurer into insolvency is a “nuclear option”. However, the reasons it plays out that way are complex.

From the former bond insurer general counsel:

I agree entirely with your position on MBIA’s obligation to downstream $900 mm to the insurer.

As to the CDS, it rapidly becomes very complicated. I don’t have copies of any of MBIA’s CDS, but accept the accuracy of their statement that “Typically in MBIA’s policies insuring CDS contracts, there are no provisions that allow a counterparty to terminate the insured CDS contract and make a claim under the policy, absent MBIA’s bankruptcy or insolvency or an MBIA payment default on the policy insuring the CDS contract. Each insured CDS contract that MBIA Corp. enters into is governed by an International Swaps and Derivatives Association, Inc. (ISDA) Master Agreement and Confirmation. MBIA’s CDS contracts typically conform to the Monoline Supplement Agreement, where a bankruptcy of the credit support provider, including the initiation of a receivership proceeding by the NYSID, would give the counterparty the option to terminate the swap and receive a termination payment; it would not be an automatic termination event.” The CDS I negotiated for my former employer adhered to that standard (except for the Monoline Supplement Agreement, which according to my recollection covered CDS written on monolines rather than CDS written by monolines, but it’s been a while and I could be wrong or the form could have changed).

Having said that, our next stop is indeed parsing the language. I’m sure you’ve already noted the weasel-word “typically”; no need to discuss the implications and resulting questions.

Let’s think about bankruptcy and insolvency. First, MBIA the insurance company is not subject to the Bankruptcy Code and cannot voluntarily file for bankruptcy, but they could be involuntarily filed into bankruptcy by their creditors. Such a bankruptcy filing would almost certainly be dismissed by the bankruptcy court. But it would take time. The standard ISDA bankruptcy event is triggered if the filing isn’t dismissed within 15 days, which just isn’t possible (absent massive intervention with the bankruptcy court); it would probably require 45-90 days, by which time the termination event would have occurred. Second, insolvency, which again is a matter of definition; probably MBIA, like other monolines, limited the definition to entering rehabilitation, conservatorship or other similar regulatory proceedings and excluded a balance-sheet test or other, broader, tests of insolvency in the normal ISDA definition. Given the billions of surplus to policyholders reported by MBIA and the mechanics of statutory accounting (which is very different from GAAP), regulatory insolvency seems extremely unlikely in any future I can foresee.

The other termination event is a default by MBIA on the actual policy insuring the CDS, which requires both an initial default on the underlying security and by MBIA on that policy – they probably didn’t permit cross-default clauses in their CDS. So that again is extremely unlikely, and in any event would be limited to that single CDS.

But let’s assume a bankruptcy termination event occurs. Then the question is whether the counterparties will terminate the swaps (and how long it has to make its decision, which depends in part on whether the termination event merely has to occur or, more likely, occur and be continuing). Termination of all the CDS could require MBIA to pay its counterparties the cost of replacing the CDS with an equivalent protection seller (another interesting question) – if, as is probable, MBIA agreed to make 6(e) mark-to-market payments upon termination, which again depends on the CDS documentation. It’s also probable that being required to pay such an amount would render MBIA insolvent under statutory accounting, which presumably would result in their entering rehabilitation proceedings. At that point, Mr. Dinallo could refuse to pay the mark-to-market payments, which arguably are not the sort of losses intended to be covered by financial guaranty insurance – and MBIA would become solvent again, able to pay claims on defaulting securities as they became due.

My own view is that termination is a “nuclear option” that operates to the detriment of both parties, and that (like any other lender) the CDS counterparties don’t want to force MBIA into insolvency. They want negotiating leverage, and presumably would get it in much the same way ACA did: a forbearance agreement would be negotiated, under which the CDS counterparties would agree not to terminate the swaps – but the counterparties would probably extract something in exchange for forbearing. Of course, all the counterparties would have to agree, because termination by any one counterparty would force the others to terminate as well, so the negotiations would be long and wearing, and probably conducted under the auspices of the NYSID. I don’t think anyone really wants to go there, and in that sense agree with Felix Salmon.

The operating supplement is a filing that’s conventionally made by insurers, but required neither by the SEC nor the insurance regulators and not prepared under either GAAP or stat. It’s available on MBIA’s website and should be read with appropriate caution, although the op supp contains much useful information not available in any other format. MBIA’s statutory financials are also available on its website.

Update 6:20 AM: FT Alphaville raises an issue which complicates the analysis above:

In the case of MBIA and Ambac, we’re not talking about bankruptcy or margin calls. Indeed in many cases, the value of the insurance written is already worthless – in the case of MBIA, many wrapped bonds have an underlying rating now higher than the wrapper. Policyholders, in other words, have everything to gain from terminating the insurance contracts.

So the fact that the policyholders have no downside (except legal costs) may embolden them. If nothing else, since the insurance is worth less than the underlying bonds, it would be worth pursuing courses of action to terminate the CDS so as to end payments on useless guarantees.

It’s going to be an interesting next few months….
Second Update 6/21, 12:05 AM: This paragraph comes from today’s Wall Street Journal story on the downgrades of bond insurers FGIc and Security Capital Assurance:

For both companies, Moody’s noted that a breach of the minimum regulatory capital requirement heightens the risk of regulatory intervention, which could trigger a market value termination of their credit default swap contracts.

Print Friendly, PDF & Email


  1. pd130

    Thanks both to you and your commentator, Yves, for the analyses.

    The addendum from FTAlphaville perhaps answers the remaining question I have for now, i.e. then why did MBIA bother with the trap —and what would it be but a trap in the present situation?

    FTA implies that there might still be a reason for MBIA to fear a black swan in the shape of a seemingly trembling hand, or whatever.

  2. tz

    Am I missing something, or isn’t collateral a funding source to pay claims (the only difference is it would be highly liquid from a market perspective and sticky from a disposition perspective). No claims, no collateral.

    Readers are invited to correct me, but as I read this, the collateral posting requirement increased based on the last 10-Q data from $8.25 billion to $12.7 billion, an increase of $4.45 billion. Now these are maximums; the 10-Q is silent on the real values. However, if we work with the figures given, that would reduce MBIA’s claims-paying resources from $16 billion to roughly $11.5 billion, a decrease of over 25%.

    If the ability to pay claims wend down, the equation expressing it is not linear.

  3. jason

    Thankyou Yves. I wonder if its going tobe a snowball effect, where one policy holder goes for termination,and others follow. Also next month if mbia would pay, their surplus noteholders? What would be the fallout if they do or dont? Will just the questioning, force other interested parties to rush in to grab their share before the rush.

  4. Ken Houghton

    tz – Nor should we expect it to be. A minor decline in credit rating doesn’t trigger much of a change in COF. Dropping five levels at once–and especially dropping below AA–is another story.

    (If your credit rating goes from 755 to 748, you may be in another bracket, but you’re probably not paying enough more for that house that you notice the difference.

    Drop it to 680 and you notice it.

    Drop from there to 605 and you decide to rent.)

    It’s a good thing insurance is a racket where you can assume everyone except your company will remain solvent.

  5. Yves Smith

    Anon of 3:55 PM,

    I quite clearly posed questions and surmises and was quite transparent about my reasoning process. There was no assertion of expertise here. Having worked in financial services for many years, I have seen that the people who do insurance (for instance, as management consultants, lawyers, or accountants) very rarely work in other areas, and vice versa, particularly because the accounting (SAAP) is quite different than GAAP. That is why I was explicit about the reasoning.

    I appreciate your comment and have modified the post accordingly.

  6. Dr. Willy Von Brandtumbaugher

    Bond Insurers, Banks: Partners In Misery

    http://www.nbc4.com/money/152420…004/ detail.html

    Bond insurers wrap a variety of non-subprime related ABS, such as CMBS, corporate CDOs, and credit-card securities. The quality of the collateral on these exposures, while subject to deterioration during the next year, is not likely to lead to the issues currently being experienced with subprime collateral, in our view. Therefore, based on current information, we don’t expect the near-term losses for banks in this area to be substantial.

    The larger concern is that the bond insurance downgrades may affect not just the key markets in which the bond insurers are directly active, but general markets spread beyond the limited sectors affected to date. This could lead to a further prolonged period of generalized market disruption and a loss of confidence that would not be favorable for any financial institution. At present, it would be difficult to offer an opinion on the likely effects on the financial sector of a generalized market disruption. At a minimum, we believe trading results could be affected, as could fixed-income business volumes, not only in the municipals business, but more broadly.

  7. Yves Smith

    Anon of 4:17,

    I am in the process of revising the post. However, I started the entire paragraph with “readers are invited to correct me”. That says I am working through something but not certain as to the validity of my reasoning process. If you have read other posts, my tone is usually very declarative. I do not use conditionals like “seem” or “appear” unless I have doubt, while others use those words routinely.

  8. Yves Smith

    First you have not seen the revised post, so you have no idea how I am handling this. Second, it was flagged AS A QUESTION in the very first sentence. That said I was not sure. I do not know a clearer way of indicating doubt.

    This is what they call in consulting a forcing device. You throw out a theory, indicate you aren’t certain, and wait for feedback. And I got it. The process worked, and I am changing the post and CLEARLY flagging the post has been changed, even in the headline, which was my plan independent of this byplay.

    There are other bloggers who are prominent who get things badly wrong in areas I know. I even see it in the media. I don’t see that the vitriol is justified. I do get things wrong and I am making a correction. If I weren’t making a correction, then yes, you’d be justified in your ire.

  9. zak822

    The vitriol is unjustified. Yves made it clear that he was asking a question and searching for the answer.

    Not searching for luggies.

  10. Yves Smith

    The proof is that I am not a “coward” have tolerated your comments despite the unnecessary name calling. Other readers have found errors, upon occasion substantial, and not chosen to work in ad hominem attacks.

    But one more comment with gratutioius name calling and I will delete ALL your comments. Your forget I control the delete button.

  11. Yves Smith

    You will note I have removed a series of comments above. I had warned the reader in question, who had usefully pointed out that my original take on GICs was incorrect, that if he persisted in including ad hominem attacks in his comments (and you can see per the number above that he was persistent), I would delete all his comments. He decided to test me nevertheless.

    FYI, I’m more liberal on this matter than Barry Ritholtz: but think his policy should be viewed as the web standard:

    This may be a free country, but The Big Picture is my personal fiefdom. I rule over all as benevolent dictator/philospher king/utility infielder. Fear my wrath, mortals!

    I will ban anyone whom I choose from posting comments — usually, for a damned good reason, but on rare occasions, for the exact same reason God created the platypus: because I feel like it.

    I encourage a broad range of perspectives, philosophies, sexual orientations. Dissent is good. I want to see a debate of views, a battle in the market place of ideas. (Thomas Jefferson wasn’t so dumb after all). You can post on nearly anything, so long as it is at least tangentially related to the topic at hand.

    On occasion, I will “unpublish” a comment if I feel it is too impolite, harsh, ad hominem, inappropriate, or off-topic. Off-topic posts have been rising, and I have taken to unpublishing them en masse. Publish too many comments on a given post (3 or 4 relevant comments out of 30 are fine, 10 out of 30 is excessive). It takes me ~10 seconds to un-publish 10 comments.

    If you find yourself publishing way too many comments, consider this: This humble blog is my forum for expressing my ideas. Get your own damned blog.

    A few things that will get you permanently banned from commenting at The Big Picture. The fastest way to lose posting privileges is to misrepresent your host’s complex and nuanced views in some inane bumper sticker comment. Those who do this, be advised: I’ve read your prose and considered your thought process: Suffice it to say the literary world will suffer no harm for your deletion (Robert Frost’s legacy is safe).

    Other fast tracks to getting banned:

    – Knowingly posting false or malicious material;
    – multiple postings under different names;
    – generally engaging in troll-like behavior;
    – misquoting your host/overlord;
    – being impolite in the extreme;
    – ad hominem attacks;
    – being an asshole.

    Right now, someone is reading this and saying to themselves “What does he mean, being an asshole?” If you wondered that to yourself, well the odds strongly favor that you yourself have sphincter-like qualities. Thus, you should consider it likely that you will be banned as a rectoid from posting comments sometime in the near future.

    Note that I am gratified by the quality of the comments and (aside from removing the occasional spam) have only very rarely deleted comments (this is only the third non-spam incident in the year and a half this blog has been up).

  12. Yves Smith

    Actually, I need to correct my last comment, which I had lifted in large part from an earlier incident. This is the sixth non-spam comment-removal in a year and a half.

  13. Been there


    Suggestion: Don’t feed the trolls with attention. Just ignore and delete. Actually, I’d think that would be quite therapeutic.

  14. etc


    You may want to post on this. Big news re securities disclosure. The Children’s Investment Fund used them in its campaign with CSX to share in the economics of shares without subjecting itself to securities law disclosures. This is a common tactic by investment funds, not just TCI.

    Anyway, a couple days ago a court found that the total return equity swaps tripped the securities disclosure rules. As one would expect, the SEC hadn’t previously frowned on this strategy, merely saying it raised novel issues. Hedge funds also borrow stock to avoid tripping securities law disclosure.

    Coverage by the law firm McDermott Will:

    Coverage by the law firm Akin Gump:


    Ignore the trolls, Yves. You do a great job and are willing to revise your position when you’re wrong. That is ten times better than many public figures.

  15. Anonymous

    Yves,Thank you for the update,and the consistent quality of your Blog.You write with clarity about difficult subjects.I visit daily and have since Calculated Risk linked to you.

  16. Anonymous

    I like that word, rectoid. Fits. The guy gets offensive, gets a warning, keeps going, gets his comments deleted, and tries six more times? Obviously has a screw loose.

  17. Anonymous

    “(I have serious problems with that concept; if you are a bank, you don’t get to discount your future interest spread and count its as some sort of capital).”

    This is exactly what happens in the case of bank origination of securitizations.

  18. mmm

    10:29, I disagree. When banks originate securitizations, they on-sell assets on their balance sheet (credit card receivables, mortgages, car loans) to a securitization vehicle. The origination income effectively does front load some of the total spread income, but it comes about in the course of a sale to a third party (and that hocus pocus of the originators and packagers taking more and leaving the investor with less of the vig doesn’t look so hot these days). There is real cash changing hands.

    Plus the bank only receives revenue; it still must recognize various expenses against that (including taxes) before getting to net income, which then increases equity.

    Now I’m not sure about revenue recognition. It’s an issue with physical goods, when to book sales (when the order was placed? the good shipped? payment received?) Probably similar questions here (yeah, I can imagine banks booking securitizatiion revenue before a deal closed). But the assets tagged to go into the securitizaton remain as earnings assets on the bank’s books until the transfer is recognized, or so I’d think.

    Yves’ comparison to NPV of interest spread (interest income less the cost of funds) sounds reasonable. First, the policies in question remain on the insurer’s books; there’s no cash coming in the door and assets exiting. Second, no allowance is made for operating expenses (possibly including taxes). Third, only a portion of bank assets are subject to securitization; MBIA looks to be roping a lot, maybe all, of its future premium income in this treatment

  19. doc holiday

    Praise be to Yves for hanging in here so long and being on the right side of this war!

    Love Yah baby!

  20. Anonymous

    Oh my God your blog has become successful, repleat with Doc Holiday. Good luck with that my friend.

  21. Tom Lindmark

    Let this be a lesson to you. Don’t get so nerdy again. LOL.

    Actually and interesting post. I suspect that in the end it will come down to a dueling lawyers battle. Good or bad outcomes in those situations usually depend on who loses their cool and makes an emotional decision. Clearly it seems to be a situation heading for a precipice and one bad decision could cause a cascade of unanticipated results.

    Remember the Faldo to Norman advice. DLTBGYD.

  22. Anonymous

    In reference to claims paying resources of MBIA. Yves stated the following:

    “surplus and $2.6 billion of present value of installment premiums (I have serious problems with that concept; if you are a bank, you don’t get to discount your future interest spread and count its as some sort of capital).”

    The statement above blurs the definition of claims paying resources by including the phrase “some sort of capital.” I do not believe MBIA is stating that the PV of future installments is capital. (The Capital Base calculation is defined as capital and surplus + contigency reserve.) Future installments does affect their ability to make due on future payments.

    It is not outrageous for an insurance company to consider the present value of future contractually obligated payments when considering their ability to pay future claims. Especially when you consider the fact that many of the future claims are in 20-30 years from now. You may argue that there may be some negotiated settlements that could reduce the present value of future installment premiums (as MMMM… did), but these settlements may or may not be advantageous to MBIA. Remember most of the installment premiums are from the structured finance book. MMMM – under what circumstances would these policies leave MBIA’s book?

    I do not believe the comparison to a bank including future interest spread is appropriate. The bank’s source of funding (deposits) can run out the door or the cost of these deposits can increase dramatically. A bank can not go into run off – there is no guarantee that they will have future funding. MBIA can go into runoff – its premiums are contractually obligated.

    On another note, I have spent a lot of time reading on the web (including Ackman’s stuff) about MBIA and the other guarantors. However, I have read predictions on the size of claims and risk of insolvency. However, I have not seen a cash flow analysis – i.e. the amount and timing of future outflows compared to the amount and timing of future inflows.

Comments are closed.