Warren Buffett, Bond Insurer

I'm fascinated by the news that Warren Buffett is starting up a new bond insurer. On the one hand, it makes perfect sense: he's an expert in insurance, he already has a triple-A credit rating, and his competitors in the market are all struggling.

But on the other hand, this is a declining market. This time last week I posted a blog entry headlined "Munis Back Away From Ratings-Agency Domination," which celebrated the fact that city and state issuers were increasingly wondering whether bond insurance was worth their while.

After all, in an efficient market, bond insurance shouldn't really exist as a standalone product. That's because it already exists, in the form of tradeable credit risk. Someone buying an uninsured municipal bond is essentially taking exactly the same default risk as a bond insurer who insures that bond. And in a market with thousands of bond investors, it's pretty ridiculous to assume that one specific bond insurer will always have a greater appetite for that default risk than the marginal bond investor would.

Now  historically, there's been another reason why bond insurance exists, and that's credit ratings. There are some investors who will invest only in AAA-rated securities, and who will pay through the nose for the privilege of being able to do so. They don't want to take default risk, and they're happy to pay bond insurers to take that default risk for them.

Well, two things have changed of late. The first is that investors don't care as much as they used to about AAA ratings, ever since a bunch of AAA-rated securities started defaulting not long after they were issued. If the ratings agencies can't be trusted to be right on the subject of how rock-solid a triple-A rating really is, then there's much less justification for paying a premium for AAA-rated paper.

The second development is the explosion of the credit default swap (CDS) market. There's now a very liquid market in default risk, which means that again investors don't need to rely on bond insurers to take their default risk from them. Of course, they still need to worry about counterparty risk, but let's say that they restrict their CDS counterparties to the known bond insurers. In that case, their counterparty risk is no greater than if they'd bought a wrapped bond. As ACA has proven, counterparty risk is hardly unknown in the bond-insurance market.

What's more, Buffett seems to be saying that he's going to charge more for bond insurance not only than the CDS market, but even than his competitors in the bond-insurance industry.

Mr. Buffett said his company will charge more than his competitors because of what he calls the "moral hazard" inherent in bond insurance. That is, governments that have insurance could take advantage of it by borrowing and spending far beyond their means to repay the debt, and simply default, leaving the insurer on the hook.

I think this is a polite way of saying that he will charge more than his competitors because his AAA is real, while their AAAs are looking increasingly fictional. Moral hazard in the muni bond-insurance market is a bit of a non-issue: if a democratically-elected municipality slides into default, it's not going to be because its bonds are insured.

So I'll be surprised, then, if Berkshire Hathaway Assurance Corp becomes a major business. To be sure, it might become a bigger business than Dairy Queen, and Buffett seems to be very happy owning Dairy Queen. But I do have a feeling that the basic bond-insurance business model is probably doomed, even if (or, rather, because) it will continue to be very profitable until its demise.