The worm of doubt: credit-worthiness of monoline insurers


Peter Lee
Published on:

It's a sure sign of a nervous credit market when even a monoline insurer's spreads can blow out by 30 basis points in a matter of days. MBIA says its triple-A rating is unimpeachable. Most people agree. But the worm of doubt has been released. Peter Lee reports.

"TO BE HONEST, it scares the hell out of me." The American bond fund manager had never had to think much about MBIA's credit. The ratings agencies did this for him and for many others. So when the financial guarantor's credit spread almost doubled in April, following an earnings restatement, the shock spread around the global bond markets. The ratings agencies quickly reaffirmed their AAAs on MBIA and its spreads began to narrow back again. Are investors' concerns now allayed? Apparently not.

"It does concern me," says Joe Biernat, director of European Credit Management, a specialist independent credit fund manager set up in 1999. "As a credit investor, we don't usually buy AAA paper. But our ABS fund started holding cash in MBIA-wrapped bonds about six months ago. The whole point of doing so was that we wouldn't have anything to worry about. Now it's on our radar." European Credit has sought a better understanding of MBIA. "We've had meetings with management and they assure us that there's nothing to the rumours. They tell a very credible story. But from the outside, there's no way to analyse these insurance companies. "

Questionable accounting

MBIA's insurance wrap is meant to remove credit risk from bond holdings. And for as long as its own AAA rating looks cast-iron guaranteed, that's just what it does. The ratings agencies have reaffirmed them on the basis of MBIA's strong capital ratios, claims-paying resources and good track record. But that's not what investors are worried about: their concern is with the outcome of regulatory investigations into accounting practices and MBIA's admission that it used dubious means to evade accounting for loss in the past.

The worm of doubt is now burrowing deeper in investors' minds.

In March this year, following an investigation initiated by the audit committee of MBIA's board of directors in October 2004 and conducted by outside counsel, the company restated prior year earnings after conceding that a so-called retroactive reinsurance agreement that had protected it from a big loss in 1998 may not, in reality, have transferred any risk. It was more like a loan.

At a time when the whole US insurance industry is being investigated over the use of finite insurance to smooth earnings and other questionable accounting practices, and with the turmoil at AIG still in the headlines, MBIA then further revealed that it had been subpoenaed by the SEC and the New York state attorney general's office over its relations with an offshore captive reinsurer, Channel Re; its dealings in credit default swaps on its own name; and other accounting issues relating to treatment of advisory fees, loss reserves and case reserves.

We've seen this movie before, investors thought to themselves, and it never has a happy ending. How often does Eliot Spitzer get his teeth into a company and not spill blood?

Their worries about MBIA start with the possibility that there will be a large fine for transgressions already revealed; progress to a fear that it might be required to consolidate Channel Re, through which it now lays off 36% of its total of $85.5 billion in par reinsured risk; and then spiral off into the possibilities of more profound, serial wrongdoing emerging from the investigations.

The company's shares sold off quickly on much higher than normal volume. Shares that had traded at around $60 at the start of March fell to within sight of $50 in early April. Then came the steady drip of class-action suits by law firms inviting investors who had bought shares between summer 2003 and the end of March 2005 to join complaints alleging violation of federal securities laws in relation to fraudulent accounting used to artificially inflate results. At the very least, the company faces mounting legal bills.

Spread blow-out

At the start of April, spreads on MBIA's credit blew out by almost 30 basis points, a breathtaking move on a triple-A name. While five-year CDS spreads of its closest competitor, triple-A rated Ambac, widened from 30bp to 36bp in the volatile markets of late March and early April, MBIA's spreads blew out from 41bp to 70bp, widening even beyond the 58bp on double-A rated insurer Radian.

Investors had reason to worry. A downgrade would have a drastic impact and not just on MBIA. If MBIA were to be downgraded by one notch, logic dictates that the 80% of net par outstanding on which the underlying rating (often called the shadow rating) is below AAA – that is some $468 billion of debt that carries MBIA's AAA wrap – would also be downgraded at a stroke. 

Analysts describe a downgrade of MBIA as a very low probability, very high impact event for the wider bond markets. The first reaction would likely be a 10bp to 15bp widening on paper with higher underlying ratings and perhaps even more on lower-rated paper, if the fear grew that more downgrades would follow. Because the AAA rating is so important to MBIA's entire business model, one downgrade might well quickly be followed by a second as diminished margins ate into its earnings capacity.

"The demand for Aa credit enhancement is far lower than for Aaa," admits Stanislas Rouyer, senior vice-president in the financial institutions group at Moody's Investors Service. "So any downgrade would have consequences of its own which would have to be reflected in the subsequent rating."

Biernat is more blunt. "There is no AA+ for these guys, it's AAA or nothing," he says.

Even if it was not downgraded, having its debt trade persistently wider than other AAAs would hurt MBIA. Any low-rated issuer thinking about taking out insurance weighs the cost of the insurance premiums to gain the monoline's wrap, and the concession of any protection measures, against the lower coupons and debt service required on AAA paper. A persistently wider trading level for MBIA paper would eat into this margin, so reducing the company's earnings and, over time, its capital. In mid-May spreads on certain MBIA-wrapped paper with non-investment grade shadow ratings, such as Eurotunnel, were 25bp wide of where they had been in March, even though MBIA's spread had narrowed.

For now, Nicholas Ferreri, chief financial officer of MBIA, is not concerned. "Someone trading our spreads in the credit default swaps market – which is where these quotes come from – doesn't impact our business," he says. "Our concern is that an A-rated muni issuer that might pay a 5% coupon on a standalone business can see that it would pay 4.5% with an MBIA wrap, just the same as with an Ambac wrap. In our core business, our spreads are stable."

Right now, investors are checking their limits on MBIA exposure. Some would like to sell and are only prevented because they are already full on the other triple-A names. There are reports in the CDS market of bond investors looking to buy protection on MBIA, though sellers also emerged as spreads went past the AA level.  But what has happened so far would be as nothing compared with an actual downgrade.

A bloody outcome

"There would be a lot of sellers of MBIA-wrapped paper," says Gery Sampere, portfolio manager and head of credit research at credit hedge fund BlueMountain Capital Management, "both in the muni market and in the senior tranches of CDOs in the structured market. What would it mean ultimately? MBIA might have to be replaced or these structures would have to be unwound, with the potential for sales of a lot of assets."

But issuers could not simply invite another AAA-rated guarantor to replace MBIA. The company has locked in future premium earnings from insurance contracts already underwritten. It relies on these stable future earnings to provide it with a cushion in hard times such as these. These locked-in earnings are a great credit positive for the company. They allow it to sit out poor market conditions – when low demand for financial guarantees might tempt newcomers to lower their risk underwriting standards or concede on price.

MBIA make a virtue of falling new business volumes over the past year as a sign that it is sticking to its fabled no-loss underwriting standard in a weak market. "Our level of new underwriting fell 29% from 2003 to 2004," says Ferreri. "But the most critical thing in this business is to maintain your underwriting discipline, and we'll sacrifice growth to do that. Our shareholders tend to be buy-and-hold, looking for steady returns year in and year out. They understand this."

If regulatory investigations did lead to a downgrade, the good news is that 80% of what MBIA insures is already rated A or better. But in the securitization markets it would be pretty bloody. Ganesh Rajendra, managing director in securitization research at Deutsche Bank, says: "Spreads on wrapped bonds would react sharply." Rajendra calculates that just over 5% of AAA public issuance in the European securitization market since 1990 has been wrapped by monolines, with Ambac and MBIA the clear leaders.

The big question about MBIA is  far removed from all the possible smoke-and-mirror accounting issues. If there were something serious at MBIA, who would bear the pain? The growth in the bond markets in recent years has been in structured products. MBIA has facilitated the sale of billions of dollars-worth of these new products. What would happen to these?

Closer look

Right now, there are more questions than answers. So let's take a closer look at MBIA.

Although it is a small company, with an equity market capitalization of just $7.5 billion – compared with AIG's $138 billion – MBIA is a key player in the US domestic and international bond markets, where it acts as financial guarantor on some $585 billion in net par outstanding. In its home market it is renowned as the leading insurer of municipal bonds, backing $367 billion in outstandings. It also insures large volumes of domestic and international structured finance bonds, including mortgage backed, asset backed and CDOs that together account for 27% of its net par outstanding.

At the end of 2004, in the US structured finance market it had $113.4 billion of net par outstanding, of which CDOs were the largest category, accounting for 38%. In international finance it had $105 billion of net par outstanding, of which CDOs accounted for 39%.

Monoline insurers have been with us for a long time. Their AAA ratings survived the US real estate and leveraged loan problems of the early 1990s, the emerging-market and LTCM collapses of 1998, the revelation of fraudulent accounting at US corporates in 2001 and 2002, and plenty of other credit setbacks along the way. Their track record has been so strong that they have almost been taken for granted. Now, all of a sudden, investors aren't so sure.

MBIA's restatement of earnings going back several years, announced this March, relates to an extraordinary event in 1998. Then it had been facing an unusually large hit on $265 million of bonds issued by troubled Allegheny Health, Education and Research (Aherf), which MBIA had insured. Staring at a $170 million loss, which would have been its biggest single hit in almost 25 years of operations, MBIA suddenly pulled a sizeable rabbit out of the hat. It announced that it had  three excess-of-loss reinsurance facilities with AXA Re Finance (ARF), Converium Reinsurance (North America) and Munich Re.

Too neat a trick

Even though Aherf's bankruptcy was widely known, under these agreements, Converium reimbursed MBIA $70 million and Munich Re and ARF reimbursed $50 million each. Hey presto! There was no need to account a loss, as the company had the present value of future payouts relating to Aherf's bankruptcy covered. In return for this cover, MBIA agreed to cede to the reinsurers, on a quota share basis, a cross-section of new business written by MBIA. This was to have an aggregate par sufficient to generate over the next six years $102 million of cash premiums to Converium, $98 million of adjusted gross premiums to ARF and $98 million of adjusted gross premiums to Munich Re.

It was a very neat trick. Although it might look as if the reinsurers were simply lending MBIA the money to cover its loss and that the ceded premiums amounted to debt service, for as long as there was genuine risk transfer, MBIA could happily present these excess-of-loss facilities as entirely legitimate.

After all, primary insurers regularly seek to expand their capital available by entering treaties to cede new risk and premiums to reinsurers. MBIA's very conservative underwriting and forceful remediation – its efforts to protect against realized losses on  credits that hit trouble after it has insured them by, for example, by seizing collateral, changing servicers or deleveraging CDO tranches – has resulted in a very low loss rate. Indeed its proud boast is that in its 30-year history, in which it has insured 99,000 issues worth in total $1.82 trillion, it has suffered just 3bp of losses on 78 issues amounting to just $586 million (now including $236 million for Aherf). So the reinsurers knew they would be getting good-quality business.

It looked odd. But the ratings agencies didn't seem too worried. Perhaps that's because while a full recognition of the Aherf loss would have severely hit MBIA's quarterly earnings, it wouldn't have erased MBIA's claims-paying resources. It would have been a much bigger deal to equity investors than to creditors.

Analysing insurance companies is very different to analysing industrial companies. To most investors they present an incomprehensible pattern of obscure risks refracted through some pretty dense accounting. "Our typical approach to this kind of business, which is levered, complex and difficult to understand, has been to run a modest short," says Sampere at BlueMountain. "We don't think there's a major problem there, but you're relying on faith and track record. Well, Enron had a very good track record. And we can take no great comfort from the credit rating agencies, because the day the rating goes it's in free fall."

It was another hedge fund manager, Bill Ackman at Gotham Partners, who drew attention to the strange Aherf reinsurance deals in a scathing report on MBIA posted in 2002 that questioned various aspects of the company's financial reporting. As well as questioning the Aherf retroactive reinsurance, the Gotham report also questioned MBIA's use of off-balance-sheet SPVs, alleging that the company was disguising the true extent of its debts and using the SPVs to provide secondary market support and advance credit to companies whose debt MBIA had insured that were now running into trouble. This all smelled rather rotten at a time when off-balance-sheet vehicles came under intense scrutiny.

Credit quality distribution as at Dec 31, 2004
Net par outstanding – $585.6 billion total

Credit quality distribution as at Dec 31, 2004
Business written 2004
Bear raid exposes doubts

MBIA consolidated the assets and liabilities of its conduits in the third quarter of 2003, though it still does not disclose the names of the issuers or identity of the assets purchased in these conduits and funded by issues of asset-backed CP and MTNs. It is clear that the underlying notes in these conduits carry MBIA guarantees.

Ferreri denies that the conduits are somehow propping up troubled MBIA-insured credits. "The conduits have no such role," says Ferreri, "we underwrite inside the conduits to exactly the same standard as outside the counduits. What goes into them is rated at just the same standard as everything else we underwrite. It is standard practice that sometimes there will be issuers looking for another source of finance and who do not want their identities to be known, who will put assets into the conduits."

In 2002, Ackman had shorted the company's stock before publishing the report, allowing Gotham's actions to be characterized as a self-interested bear raid. Amid complaints by mainstream US companies against hedge funds taking short positions against them and then seeking to influence the markets, Eliot Spitzer investigated and Gotham Partners suffered its own redemption problems. But the questions didn't go away.

Central to the investigation of the Aherf excess-of-loss reinsurance trades was the claim that a senior executive or executives of MBIA had offered oral assurances that the company would make whole any losses suffered on business subsequently retroceded by Converium to ARF. If a senior MBIA official had offered such an undertaking, the contracts could not be classified as true risk transfer. The investigation by outside counsel eventually concluded that: "MBIA could no longer represent without qualification that the alleged oral agreement did not exist and that therefore there was sufficient risk transfer under its quota share reinsurance agreement with Converium."

The admission raised an obvious question. If MBIA had cheated once might it not have done so on other occasions? Ferreri says not. "Aherf was a unique transaction for this organization, the only retroactive insurance contract we have ever done. There is nothing else like it and nor have we done any finite reinsurance." MBIA will not be pulling out any more retroactive reinsurance agreements. "We get reinsurance on a transaction when we write it, not after it hits trouble," says Ferreri.

"I believe them," says Gary Ransom, insurance analyst at Fox-Pitt, Kelton. It's important to note that the senior executives now running MBIA are not the same people who were running it back in 1998. "Remember that they only found out about this oral agreement when one of the reinsurers came back to them about it. I think whatever management has learned about this, they have told us. And when there's a question about them, instead of going quiet, they answer it."

Geoffrey Dunn, insurance analyst at Keefe Bruyette and Woods, takes a similar line. "I think they're fine," he says. "We've gone through these issues with them in some depth and I have a very tough time coming to the conclusion that there is anything material out there. Aherf  happened under the old management team and the new team has cleaned it up. As to Channel Re, we can't find anything wrong with it and neither can the rating agencies; treatment of advisory fees, loss reserves and case reserves are questions of accounting interpretation; there's no law against dealing in your own CDS and that was all disclosed anyway two years ago. They are, in my opinion, far away from any risk of being downgraded."

It always comes back to ratings. Whether they like it or not, the credit rating agencies take on a pivotal role in all this. Many bond investors take comfort from the fact that the rating agencies have run the exposures of MBIA and other financial guarantors through their models and found them more than adequately capitalized to withstand extreme stress.

For anyone used to studying the capital structure of banks, monoline insurers appear to inhabit a weird, parallel universe. Supporting its $585 billion in net par outstanding, MBIA had claims-paying resources of $12.9 billion, roughly half of that being statutory capital, with the remainder largely comprising reserves, unearned premiums and back-up lines of credit. It reports a capital ratio of 147:1, a claims-paying ratio of 81:1 and a leverage ratio of 41:1.

Egan-Jones Ratings stands out by asserting that "MBIA Inc and MBIA Insurance Corp are not 'AAA' credits because of the slim capital (MBI has only $6.6 billion of equity book value compared with approximately $600 billion of guarantees) and soft demand and margins." For good measure, it expects fines and curtailment of business lines to result from the investigations.

Yet the big three ratings agencies assure us, having stress tested the risk exposures of the monolines and looked at the distribution of losses out to the 99.9 % probability level, that MBIA has sufficient capital to more than cover the amount it might have to pay out even in such dire circumstances. "We calculate that in such circumstances the present value of losses would be no more than 130bp, which is significantly below their financial resources," says Rouyer at Moody's. Such stress tests imply a credit market meltdown. MBIA would no doubt be downgraded, but so would every other issuer in the market. The key is that it could meet its obligations.

The ratings agencies now sense that somehow their own reputations have been put at risk alongside MBIA's. They, after all, have affirmed its high ratings and have a stream of ratings business from the structured debt markets whose growth the monolines have supported. Now they are keen to stress that they are not somehow in cahoots with the monolines.

The received wisdom among outside investors is that the ratings agencies somehow vet and have a say over every new risk the monolines put on. Not so, say the agencies. "We're not part of the approval process," says Jack Dorer, managing director in the financial institutions group at Moody's. "They make their own underwriting decisions. We factor their approach to underwriting and the resulting portfolio quality into our overall rating assessment of the company." He explains: "What we're on the alert for is any deterioration in the credit rating of what they underwrite." He adds: We're also looking for any weakening in structures, or for any subtle changes in triggers, other protections,or legal remedies."And  while there's no easy way of tracking that, nor is there any sign of it happening at MBIA.

MBIA is overcapitalized

Net par outstanding as at Dec 31, 2004
By bond type (%)
It might surprise some credit investors to hear that Fitch does not, in fact, adjust its capital model for MBIA with each new deal the guarantor underwrites. "We assess capital adequacy quarterly, not on a transaction-by-transaction basis," says Nathan Flanders, analyst at Fitch. "Maybe we would do if MBIA were running close to the margin of AAA, but it is not. They're running with enough capital to pay all their claims under our five-year stress scenarios with a cushion as well." Fitch is preparing to roll out a new Monte Carlo based capital model that will permit analysts and the company to judge changing capital requirements arising from each new commitment.

Meanwhile, Ferreri believes that, given the company's very conservative risk underwriting and stable earnings generation, it is in fact overcapitalized. "We want to hold capital to cover more than what we might have to pay out in the worst stress scenario to the 99.9% probability level. How much should that cushion be? Maybe 1.2 times or 1.3 times. So when we get to 1.5% times or higher, it is my view that we're in an overcapitalized position."

MBIA has been buying back shares, having sought board approval for a 15 million share repurchase programme. "We bought back 3.4 million shares in the first quarter and 1 million in April alone," says Ferreri. The company was allowed by state insurance regulators to upstream earnings for a special $375 million dividend in the fourth quarter of 2004 from the AAA-rated insurance subsidiary to the AA-rated holding company. Application for another such special dividend is pending.

If the company thought itself to be even coming close to losing its all-important AAA, it would stop writing new business and start raising capital. It would not be returning it, as MBIA is now. "I'd say the odds on a downgrade are close to zero," says Fox-Pitt, Kelton's Ransom.

Control of the captive

As to the far trickier question of the reliability of MBIA's financial statements, ringing endorsements are harder to provide. "While we cannot be certain of any action to be taken on the part of the regulators, we have received no information that would cause us to change the rating outlook of MBIA inc and MBIA insurance," David Veno, analyst at Standard & Poor's told investors in a note in early April. "As the current investigations continue, Standard & Poor's will closely monitor the progress."

All ratings agencies offer similar caveats. "It is important to note that investigators may ultimately gain access to information not currently available to Fitch and thus not currently factored into our ratings," the agency told investors in May.

Nathan Flanders, analyst at Fitch, adds one simple reason to accept the company's assurances that it is not concealing more such oral agreements to make whole risk supposedly transferred away. It hasn't faced anything like the same temptation again. "The magnitude of the Aherf loss was quite unique. It was a wake-up call to MBIA and to the industry and MBIA worked down its healthcare exposure as a result. It's not as if there have been a number of more instances such as Aherf."

The absence of temptation might be no great foundation on which to affirm a belief that the company has not strayed again, but the company's management in 2005 has to submit to Sarbanes-Oxley and all manner of enhanced internal controls that management did not face in 1998. In the end, much comes down to faith. "We've asked if there have been any similar oral agreements and they say no. We have to go on what the company tells us," says Flanders.

As the overriding concern for bondholders shifted away from MBIA to the downgrade of GM and Ford and credit deterioration and correlation blow-ups in CDOs, the brief panic about the monoline insurer quickly receded. Its share price recovered to as high as $58 in early May and then sank back to $55. Its five-year credit default swap spreads narrowed back to 53bp at the start of May and continued to narrow. The spread to other AAAs narrowed to 5bp early in May and the gap had almost vanished by the middle of the month.

Investors took comfort from the agencies' affirmation of the ratings and the judgement of analysts that the company could withstand a fine relating to Aherf as big as the original $170 million transaction itself. In fact, it could absorb one twice that size and still be triple A.

Tricky questions remain about whether it exercises control over Channel Re, its captive reinsurer. MBIA gained capital credit from the agencies for transferring reinsured risk from other reinsurance firms suffering downgrades in 2003 to a new firm, Channel Re, set up in Bermuda in 2004 and rated AAA. MBIA owns 17.4% and a further 45.5% of capital is contributed by Renaissance Re, 27.7% by Partner Re and 9.5% from Koch Financial Re. This is real risk capital put up by independent third parties. However, MBIA is Channel Re's only customer and this, when taken together with its 17.4% shareholding and its ability to appoint two of 12 directors, raises the question of how much control it exercises over the company.

Again, analysts rationalize that in the worst case, if MBIA had to consolidate Channel Re, its credit ratings would not be hurt, though it might have to scramble to secure new reinsurance capacity with similarly high-rated entities. "It would even be accretive to earnings," says Dunn at KBW, thinking from an equity investor's point of view. "It would be a positive use of capital."

Ferreri says: "Given the size of Channel Re, if we had to unwind it or put it on the balance sheet, we wouldn't even need to raise new capital to do that." Meanwhile he affirms that there is true transfer of ceded risk. "There are no side agreements with Channel Re." If MBIA chose to sell its stake, it's reasonable to assume it would find takers.

Ransom thinks it won't have to consolidate it. "However you look at it, somebody else's capital is at risk here. Based on everything I know about it, treating it as reinsurance is the most informative and transparent way of presenting it."

Fear of the unknown

For now bond investors have more pressing things to worry about than MBIA. But they might not be sleeping quite as comfortably as they once did. The New York attorney general's office declined a request to brief Euromoney on the investigation. It does not want to tip its hand. Spitzer's office is not in the habit of summoning the press and announcing that a company that has been the subject of an investigation is in the clear.

Insurance companies remain a focus of attention. MBIA is just one of many that has received subpoenas from various investigators. David Martin, senior credit researcher at BlueMountain Capital Management, ties the questions that have swirled around MBIA to the more pungent stories at AIG and Marsh & McLennan. "We have had a series of insurance practices exposed that were previously not widely known among the investment community. Somehow the insurance regulators, the auditors, the ratings agencies either missed them or condoned them, but they are now re-examining them and the whole area of finite reinsurance is unravelling. Is there an unknown area of the bond insurance business that might become apparent and which will not be condoned? The more the regulators dig, the more they tend to find."

The risk manager in the ABS group at a large commercial bank puts it another way. "Scrutiny of one bond insurer I can cope with, though we're getting a few questions from senior management. What would really worry me, though, would be some controversy or investigation that drew them all in."