This article appears in the March 2008 issue of Condé Nast Portfolio. Please visit the site on Tuesday, February 19, for the full contents of the issue.

In a crisis, the guy with the cash gets the call. Late last year, Eric Dinallo, New York State's top insurance regulator, picked up the phone and persuaded Warren Buffett's company to rescue the municipal-bond world.

The normally staid business of municipal bonds was on the verge of panic. Dinallo hoped that Buffett and his gold-plated name could calm things down. The credit crunch that has consumed the real estate and banking worlds is now threatening $2.6 trillion worth of municipal bonds, which help pay for everything from bridges and tunnels to hospitals and schools. (See what some recent bonds have purchased.) Two-thirds of those bonds are owned by retail investors, many of whom want and need the sort of safety that's supposed to go along with investing in government debt. Regulators like Dinallo fear that a panic would crimp the ability of state and local governments to raise money, leading to service cutbacks and canceled community-improvement projects.

The problem isn't with the muni bonds themselves but with the insurance companies that guarantee them. These bond insurers—such companies as MBIA and Ambac—are supposed to ensure that the bonds are safe. Trouble is, the insurers themselves are in crisis. This makes it clear that the entire business of muni-bond insurance is a giant taxpayer rip-off.

In theory, bond insurance lowers the cost of borrowing money. It's not unlike having your mother-in-law co-sign your mortgage: With her good credit alongside yours, the interest rate you pay will most likely drop. Let's say that Walla Walla, Washington, wants to build a dam. The rating agencies—Moody's, Standard & Poor's, and Fitch—assign Walla Walla a credit rating that determines how risky an investment the city is and, therefore, how high its interest rate should be. Assume Walla Walla's rating corresponds to a 5 percent interest rate in the market. If the city buys bond insurance, however, its bond becomes triple-A, reducing its interest rate to 4.75 percent—good for Walla Walla, which can keep its water at bay, and good for local taxpayers, who won't have to pay as much to protect themselves from it.

The business has also been very good for insurers. Until last year, bond insurers ranked among the most profitable companies in the world, with Ambac having the highest operating profit margins of any firm in the S&P 500. But then competition rushed in. In an effort to make more money, bond insurers made a terrible mistake. They branched out beyond their core business to invest in all sorts of exotic mortgage securities and other structured finance products. This has turned out to be the equivalent of insuring your brother-in-law Robbie for his sure thing involving that email from Nigerian royalty. As the bond insurers' deals have inevitably soured, the companies and their credit ratings have come under pressure and their stock prices have been tanking. Indeed, in January, Fitch stripped Ambac of its triple-A rating, which is tantamount to crushing its business model. If the bond insurers go out of business, it could cause widespread devastation. On the other side of these firms' trades are banks, which could suffer losses if the insurers don't make good.

Hence the call to Buffett's Berkshire Hathaway, with its Himalayan mountains of cash and unassailable triple-A rating. Who wouldn't want Buffett as their insurer? It's a great break for a brilliant investor who managed to be in a strong position when others were weak and now may reap the rewards. Other savvy investors, such as Wilbur Ross, have contemplated joining Buffett in jumping in.

The reason the business can be so rewarding is that government bonds are intrinsically safe: Municipalities almost never default. From 1970 through 2006, only slightly more than 0.1 percent of all muni bonds went bad, according to a large Moody's study. By comparison, corporate bonds in the same period had a 9.7 percent default rate. In other words, corporate bonds are 94 times as likely to fail as muni bonds are.

I asked David Schultz, a professor at the Graduate School of Management at Hamline University, in St. Paul, Minnesota, what he made of Buffett's entry into the muni-bond-insurance business. "First off, I would say it's still a useless product," Schultz said. "Buffett's decision demonstrates that it's probably very profitable. What a great market to get into. Since I think you don't need the insurance, it becomes almost pure profit."

How is it that municipalities almost never default on their debt? When communities face trouble and consider filing a claim on their bond insurance, they think twice. They don't want to risk being downgraded or shut out of the capital markets for future bond sales. It's the same rationale that might keep you from filing an insurance claim if you have a fender bender. You could figure that it's better to eat the cost of the repair than risk having your premium go up.

But bond insurers have it even better than your car insurance company. If you totaled your car, you'd have to submit the claim, future premium increase or not, because you couldn't afford to shoulder the cost of a whole new car. But when municipalities are hit with a disaster or a bridge collapse, they can always get the money from taxpayers. They raise taxes or find other, often draconian ways to stave off default through service cutbacks. The high-finance term for this type of insurance is the free lunch.

So why do the governments buy bond insurance in the first place? Mainly to get the higher credit rating, which lowers the interest rate and reassures investors. The rub, though, is that they should have received lower rates anyway. The municipalities' credit ratings are too low. If rating agencies properly assessed them according to investors' true risk of loss, muni bonds would have lower interest rates without the expense of insurance.

We know this because Moody's has been conducting an epic research project over the past decade to figure it out. The result is a secret decoder ring, provided by Moody's, into which an investor can plug a muni-bond rating. Out pops what the corporate rating would be. According to Moody's table, almost every muni bond would get a higher rating. About two-thirds would probably be triple-A if they were rated with the same criteria used to rate corporate bonds.

The obvious conclusion is that Moody's, as the most influential of the credit-rating agencies, should simply start lifting its ratings on municipalities. But Moody's doesn't have any plans to do that. Why it won't is one of the great mysteries of the muni-bond world. Is it merely a historical artifact? Are the rating agencies plagued by memories of government debacles like the Orange County, California, derivatives blowup or the strike-riddled and garbage-strewn streets of New York during the 1970s budget crisis? According to Moody's, investors and issuers like the system the way it is. But that's not credible. Why would issuers want to pay more than they should? More likely, the system remains unchanged because bond insurance is good for everyone in the market—except for municipalities, that is, but they have no choice except to go along with it.

Bond insurance companies, which love the current setup in the relatively risk-free muni business for obvious reasons, happen to be among the rating agencies' best customers. You have to wonder whether the agencies have a conflict of interest that prevents muni-bond-rating reform.

Connecticut Attorney General Richard Blumenthal is starting to investigate this very question, issuing subpoenas over the past few months. "The question is why there are two standards for ratings and why Connecticut towns and cities should be required to obtain bond insurance when the likelihood of default is virtually zero," Blumenthal told me. "Is there conduct that violates the antitrust laws in terms of anticompetitive practices or collusion among rating agencies or the bond insurers?"

Investors in muni bonds like insurance because it lets them buy bonds based on their triple-A rating rather than having to pick and choose from an immense universe of offerings. "Bond insurance provides a way to commoditize the market rather than doing credit analysis and checking each credit," explains UBS muni strategist Kathleen McNamara.

And Wall Street firms like insurance, McNamara adds, for essentially the same reason: They don't have to employ lots of people to check a place's creditworthiness before selling a muni to investors. "It lowers the marketing expenses of the bond when the issuer goes to tap the capital markets," she says. In other words, investors and Wall Street outsource credit analysis to rating agencies—not always the best move, as we've seen elsewhere. Credit-rating agencies have blown the analysis on bond insurers, which is why the market is in crisis in the first place. Well, you win some, you lose some.

How the rating agencies handle municipalities really ought to be a bigger scandal. Think about it: Credit-rating agencies screwed up the mortgage-securities business because they were too lax in their ratings. They're screwing up munis because they are too punitive. In both cases, they manage to benefit from their "mistaken" ratings.

Dinallo should be commended for bringing Buffett in, but it's a Band-Aid. It would be better to eliminate muni-bond insurance for good. That's what makes his January efforts to broker a bailout of the bond insurers ill-advised. The financial guarantors got into this mess by making bad bets on structured finance. They shouldn't be helped simply because they are perceived as vital to an entirely separate market. Regulators should let things play out. If bond insurers go under, so be it. If banks traded with them imprudently, let the banks take their lumps. Instead, regulators should force the rating agencies to bring their muni-bond ratings in line with the rest of their ratings. And they could create a governmental mechanism to back muni bonds in the event of default. Municipalities are effectively the insurers of muni bonds already. Why not make it explicit?

What's nice is that some municipalities are finally figuring this out. In recent months, the states of Wisconsin and California, as well as New York City and 300 other issuers, have sold bonds without buying insurance, according to Bloomberg News, which has been on top of the muni-bond scandal. Cities and states would save billions if they didn't buy insurance but received proper ratings.

So Warren Buffett may not reap his windfall after all. What good news for taxpayers that would be.