Monday, August 06, 2007

I assure you, we are quite safe from your CDOs here

In 2001, when Enron and Worldcom's accounting scandals were rocking the corporate bond market, the question on every one's lips was "who's next?" Who else might be hiding devastating losses in off-balance sheet vehicles? Who else might have booked phantom revenue? Who else was going to go from an A-rated to bankrupt in a matter of months? Companies with even mildly complex financial structures were pounded by aggressive short-selling. I remember intermediate Time Warner bonds trading with a $70 dollar price.

Today is playing out very similarly, although so far on a smaller scale. There are traders combing through the market place looking for anyone with sub-prime exposure. Be it as originator or investor, the punishment bond issues are taking for having any association with the RMBS market is indiscriminate.

My favorite victim here is MBIA. CDS for MBIA have gotten killed lately, rising from a low of 19bps on 2/7 to 232bps today. It traded below 50bps as recently as June 14.

MBIA's basic business is to insure bonds. Just like GEICO pays its customers cash if their car gets totaled, MBIA pays its customers cash if their bond defaults.

MBIA business was originally insuring municipal bond issues (the name used to be Municipal Bond Insurance Association). Today this remains their primary business. Now, let me let you in on a little secret. Municipal bonds never default. Ever. Well not never, but damn close. According to Moody's, about 2.23% of all investment grade corporate bond issues default within 10 years. A mere 0.06% of investment grade municipals default. And this isn't a coincidence of municipals having higher average credit quality. Baa corporate bonds default at a 4.89% rate over 10-years, while Baa municipals default at 0.13% rate.

So of all par value insured by MBIA, 65% are municipal bonds. The other 35% are structured finance, including CDOs, RMBS, and CMBS. 7.5% of their total portfolio is in RMBS. 10% of this is in sub-prime, or about 0.75% of their total portfolio.

Within their CDO portfolio, which is 17.5% of the total, 21% (4% of total portfolio) is in "Multi-Sector CDOs." This is where the company categorizes ABS-related deals that could have sub-prime exposure.

Let's step back for a moment and talk about why MBIA is involved in the structured finance market at all. A large percentage of AAA-rated CDOs are put into what's called a negative basis trade. This is normally works as follows.

  • Investor borrows at LIBOR+5.
  • Buys a AAA-rated CDO at LIBOR +30.
  • Buys an insurance policy from MBIA which costs 10bps (structured like a CDS).

So they make 30bps over LIBOR on the asset, and they have 15bps in costs, for a 15bps profit. So its no surprise that 95% of MBIA's CDO portfolio is AAA-rated, and another 4.5% is rated AA.

Within their sub-prime RMBS portfolio, the results aren't quite as stark: 81% AAA-rated, 5% rated AA, 3% A and 4% BBB. About 7% is below investment grade.

So let's assume things are really extremely bad. Let's say that everything in their sub-prime and ABS CDO portfolios not rated AAA is worthless, and that within the next year, they will have to make good on the insurance policies written. How much in losses would they be taking?

The answer is $1.2 billion. At the end of June, MBIA had over $33 billion in marketable securities.

Would an impairment of $1.2 billion of assets result in a downgrade of their insurance subsidiary below AAA? That's the real question. If MBIA lost their AAA rating, they would effectively lose all competitiveness. FGIC, AMBAC, FSA and others basically provide the same service. Very few would be interested in buying an insurance policy on a bond to get less than a AAA rating.

I'm highly skeptical MBIA will suffer a downgrade. First of all, they have $500 million in loss reserves. Plus, bear in mind that if their investment portfolio earns a conservative rate of return of 5%, they'd earn $1.6 billion in interest income per year. Plus once they pay out on the sub-prime insurance policies, they quality of their portfolio would theoretically be better, not worse. So taking $1.2 billion in sub-prime losses would sure hurt earnings, but it would hardly cause a liquidity problem with MBIA.

But what about the argument that many of these AAA-rated sub-prime deals should never have been AAA in the first place. OK, but MBIA doesn't have to pay out on an insurance policy because it gets downgraded. Only if it defaults. And with the most senior portion of a CDO, there may actually be recovery in the event of default. Once you get to the point where the most senior tranche is in default, all cash flows belong to that tranche. So the odds of a AAA bond actually being worthless are very low indeed.

What about those Bear Stearns hedge funds that went bust? Weren't they mostly in high-grade ABS and CDO's? Yes, but they had mark-to-market problems, causing margin calls. They also suffered due to poor hedging on the fund's part. MBIA would never suffer a margin call, because they have only issued an insurance policy. They don't actually own securities. So even if the value of the insurance policy (in effect a long credit position) declines, there is no risk of the mark-to-market decline turning into a liquidity crisis.

Perhaps my overconfidence is my weakness.

Fair disclosure: I don't own any MBIA corporate securities, although my firm owns many MBIA insured municipal bonds.


Ritholtz said...

Nice explanation of a complex situation.

This blog is rapidly becoming a regular stop for me!

David Merkel said...

Great blog. On the negative basis trade, is it truly riskless? Supposedly, MBIA mainly insures super-senior AAA CDO risk, where there are some AAAs in front of them.

Is it possible that not all of the AAA bonds are getting wrapped here? Also, who gets to borrow at L+5? That seems pretty aggressive. Do they have to offer their AAA bonds as collateral?

Thanks. I regularly read your work.

Accrued Interest said...


Thanks. While it is complex, I wish more MSM reporters took the time to understand the situation before reporting on it. As for the blog, its a labor of love.


MBIA said in their conference call on Thursday that they are mostly in the senior-most tranche, although I don't remember them actually giving a statistic there.

Anyway, the negative basis trade isn't riskless at all. But it becomes more like counterparty risk and less like CDO risk.

As for borrowing at L+5, commercial paper is usually rich to LIBOR. So we're talking about very highly rated banks and insurance companies who are usually involved in negative basis trades.

Anonymous said...

Re: Borrowing at LIBOR+5

Many large financial institutions (buy and sell side) can borrow at even below LIBOR, near the risk-free treasury rate. This is done by offering treasuries they own as collateral, which is essentially risk-free.

Ramesh said...

There's an alternative view about MBIA here -

Ramesh said...

Sorry, here is the link

Anonymous said...

You are wrong interest,very wrong. if the bonds they insure get downgraded or the ABX keeps going down THEY WILL HAVE TO MARK TO MARKET, just not in the same scale as citigroup. that will be a charge against their earnings(indeed q3 was negative for mbia) and that will lower their capital necessary for AAA rating, once they lose the rating MBIA is going to 0 100%¨guaranteed, no one is going to bailout someone who holds billions of CDO exposure.

Anonymous said...

In Dec 2007, Calculated Risk reported on an issue that went from AAA to worthless in less than a year. That's right, they defaulted:

Accrued Interest said...


You win. I've already admitted I was too sanguine on the insurers many times.