Wednesday, November 07, 2007

A Practical Look at Bond Insurers

A bond insurer provides credit protection by giving its guarantee to pay any unpaid principal and interest payments in a default. By giving its guarantee, a debt security can then be issued with a higher rating, usually AAA, and this lowers the interest cost of the debt. Now, I'd argue that if you take a close look at the competitive forces these companies face, you'll come to the conclusion that this is a mediocre business at best.

The most obvious competition is from other bond insurers and there really is no moat within this group. The AAA credit-rating is favorable because it lowers the cost of guaranteed securities to almost government treasury rates, but there are a few insurers who have this rating. Otherwise, the business is strictly based on how much capital you have and what price you are willing to pay. MBIA mentions in their annual report that it "also competes with other forms of credit enhancement, including senior-subordinated structures, credit derivatives, over-collateralization, letters of credit and guarantees."

But an even more significant competitor is "Mr. Market" himself. An issuer of debt likes bond insurance because it diminishes the investor's risk, and so it can potentially lower the issuer's interest costs. It all depends on cost of the insurance, and the spread between what the market will charge the issuer and what the market will charge its guaranteed security. If the cost of insurance is less than the spread, then the bond insurer gets new business and the issuer lowers his borrowing cost. But if the cost of insurance is more than the spread, then there is no transaction- the issuer would be better off selling the debt directly to the market. Thus, we see that credit insurers are competing against the market's perception of risk. Effectively, the bond insurer is saying that they have expertise and can profit by being better judges of a security's risk than the overall market.

Now that is plausible. But for one, these insurers are doing this with a large amount of money, making it more difficult to earn out-sized returns. And two, before the recent turmoil in the credit markets, spreads on risky securities were at unprecedentedly low levels.


So taking into account the competition, the large capital base, and the credit spread tightening, you would think that these companies would have been marginally profitable recently. But instead, these companies have been recording record profits with very healthy margins.

MBIA, 2002- 2006 ratios

Loss and LAE ratio

9.7 % 10.0 % 10.0 % 10.0 % 10.5 %

Underwriting expense ratio

26.6 24.7 21.5 22.2 22.9

Combined ratio

36.3 34.7 31.5 32.2 33.4


So that begs the question of how a credit insurer was able to underwrite so profitably(loss reserves at 10% of premiums) in an environment where "Mr. Market" was so willing to buy risky debt. And I guess you can say part of the answer lies in the fact that a bond insurer's loss reserves are very much up to judgment. In traditional property and casualty insurance, you typically have a very good sense of what your total losses will be by the second year. But in bond insurance, the time horizon is much longer. For example, we know that the average of MBIA's insured debt outstanding is over 10 years. That means that at anytime within those 10 years, a credit contraction could reveal that their reserves are drastically inadequate. This makes investing in a bond insurer sort of like a black-box, because you never really know what you are getting without knowing their loss assumptions.

Many bond insurers have lost 50 to 75 percent of their stock value in the last 6 months. (See MBI, ABK, AGO) Perhaps at these prices they are actual values, and maybe I am missing some factor that allows them to underwrite so effectively. All I know is based on these competing forces and the leeway available in their accounting, that it is very likely they have been under-provisioning for future losses, and that I can not tell you by how much. As I was writing this though I did come across this post on Calculated Risk- Egan Jones: Expect "Massive" Losses for Bond Insurers.

6 comments:

Alex said...

Most of the big bond insurers restricted their guaranties to super-senior AAA tranches. Of course, the super-seniors were often built upon other CDOs or, in some cases, CDOs of CDOS, commercial paper, etc. But since the pool of originated debt grows more slowly than the structured market built around it, when you diversify and releverage enough, eventually you start to diversify yourself into the default risk you wanted to avoid! As for the credit ratings, David Einhorn has a good speech that discusses some of the conflicts of interest in that market. So, while subordinate tranches disappear, the big insurer's look profitable. But the losses are mostly delayed, not dodged.

Even if losses turn out to be smaller than expected, I have a tough time viewing the CDS for CDOs industry as anything more than a cigarette butt. There is no indication that the insurers have re-evaluated the soundness of the their assumptions (like an ever rising Case-Shiller), or that they view the current situation as anything more than an unexpected event.

Alex said...

Most of the big bond insurers restrict their guaranties to super-senior AAA tranches. Of course, the super-seniors were often built upon other CDOs or, in some cases, CDOs of CDOS, commercial paper, etc. But since the pool of originated debt grows more slowly than the structured market built around it, when you diversify and releverage enough, eventually you start to diversify yourself into the default risk you wanted to avoid! As for the credit ratings, David Einhorn has a good speech that discusses some of the conflicts of interest in that market. So, while subordinate tranches disappear, the big insurer's look profitable. But the losses are mostly delayed, not dodged.

Even if losses turn out to be smaller than expected, I have a tough time viewing the CDS for CDOs industry as anything more than a cigarette butt. There is no indication that the insurers have re-evaluated the soundness of the their assumptions (like an ever rising Case-Shiller), or that they view the current situation as anything more than an unexpected event.

Sivaram Velauthapillai said...

As someone who is mildly bullish on the sector and thinking of taking a position in Ambac--the one with the biggest exposure to CDOs--here is how I see things. I'll outline my thinking and hope some readers can poke some holes in my thinking...

Although the questions raised in the original post are worth contemplating, I am not so sure that this business model doesn't work. I'm just a newbie and just started looking at this recently but everything I read seems to imply that debt insurers provide a valuable service.

Let's just ignore the present chaos with housing securities (like CDOs and RMBS) and look at what monoline insurers have historically done. They primarily insured government infrastructure projects or similar corporate projects. If there was no need for this debt insurance, it s unlikely the market would have paid the insurer over the last few decades (first debt insurance by a monoline was back in 1973 or something like that.) If you look at the prior disaster for mononlines, it was probably the eurotunnel fiasco. I would argue that such a project may not even have gotten off the ground without some debt insurance.

Like nearly all insurers, I think these monoline insurers work off probabilities. So in some sense they have no competitive advantage but the competitive barrier is larger than it seems. Very few can enter the market (otherwise a million companies, not to mention the large multiline insurers would have entered the sector).

So I am pretty confident that these companies do offer a value-added service. But the real question is whether they took too much risk that they couldn't handle. That is really the question for potential investors and one just has to make their own call. Some companies may go bankrupt but most should be able to survive IMO. We just don't know which ones won't make it...

A lot of reinsurance companies got whacked after Katrina but that doesn't mean reinsurance doesn't work. I view this situation as being similar.

Anonymous said...

MBIA has a very good track record of underwriting, and then "mitigating" when a default does occur, so that they pay out very little. Look at MBIA's experience with Katrina municipal exposure and their exposure to the Eurotunnel. In both cases there was a default but MBIA was reimbursed their payments.
Their asset backed insurance contracts are very different than the CDS sold by investment banks. The big difference -- they can't be sold or put back to MBI for cash. They only pay if there is a default, and then, they only pay out over the contractual life of the underlying security. Further, they typically only insured very senior tranches of ABS.
The other issue is they are now entering a very “hard market”. According to their 3rd Q report, pricing has improved significantly. The key to MBI is, will they retain their AAA rating, or at least an AA rating. If so, then they are going to make a lot of money going forward in a market that has a new appreciation for credit risk.
-RIV

Nnejad said...

So, to mention a few other things I think are relevant. One is that I went back as far as I could on MBIA's earnings, which was 1992. And since then they have always been underwriting with these low loss estimates. So that is a relatively long period on which they have been doing business very profitably.

Two, I've looked for these barriers to entry but they do not seem to really exist. The only thing seems to be this AAA rating. It is difficult for new companies to get this AAA rating, and there are very few companies that already have it. So perhaps this is a barrier to entry. We know that AIG, another AAA company, jumped into the bond insurance business as well. But, we know that MBIA's 537 million in loss reserves is minuscule compared to its 840 billion in insured debt, and that the credit ratings are beginning to come under doubt. The senior AAA tranches of mortgage debt that MBIA was insuring is currently trading at 70 cents on the dollar.
http://bp1.blogger.com/_pMscxxELHEg/RzOG003pFhI/AAAAAAAABJg/L_ouZnunw-I/s1600-h/ABXHEAAA072Nov8.png

Finally, something which is different about MBIA from most insurers is that they discount their reserves. I'd presume other bond insurers do the same. From the 2006 10-K:

The amount of the case basis reserve with respect to any policy is based on the net present value of the expected ultimate losses and loss adjustment expense payments that the Company expects to pay with respect to such policy, net of expected recoveries under salvage and subrogation rights. The amount of the expected loss, net of expected recoveries, is discounted based on a discount rate equal to the actual yield of the fixed-income portfolio held by the Company’s insurance subsidiaries at the end of the preceding fiscal quarter. The Company believes this yield is an appropriate rate of return for present valuing its reserves as it reflects the rate of return on the assets supporting future claim payments by the Company. The discount rate used at December 31, 2006, 2005 and 2004 was 5.1%, 5.0% and 4.8%, respectively.

Nnejad said...

Link was cut off. Here it is:
http://bp1.blogger.com/
_pMscxxELHEg/RzOG003pFhI/
AAAAAAAABJg/L_ouZnunw-I/s1600-h/
ABXHEAAA072Nov8.png