• Choose a category
• All
• Classic-Fiction

Quite irrationally, in recent years, insurance regulators have also allowed municipal bond insurers to provide coverage against default risks on subprime mortgage securitizations and related collateralized debt obligations (CDOs) and credit default swaps (CDSs). It is unclear why the insurance regulators allowed the insurers to mix the relatively limited credit risks on municipal bonds with the high risks on subprime mortgages and their derivatives, since this clearly violated the monoline principle on which the insurers were chartered.

Worse yet, losses on the subprime mortgage derivatives now threaten the solvency of the municipal bond insurers.9 The failure of these firms would have significant negative externalities in two regards. First, it would become difficult for many state and local government to issue new bonds in the absence of credible insurance. Second, many municipal bonds are held by depository institutions, motivated in part by the relatively low capital requirements allowed on bonds backed by highly rated insurers; if these insurers lost their high ratings, or even failed, the depository institutions could face significantly higher capital requirements. Moreover, if both the bonds and insurers defaulted, the investors would face the losses directly.

In this setting, the government faces the question of how to maintain a functioning market for municipal bonds, dependent as it is on a functioning market for municipal bond insurance. An immediate issue is how to separate the losses that are already present on the existing books of subprime and municipal bond insurance from the task of insuring newly issued municipal bonds.

For this purpose, I propose a good insurer/bad insurer model, and I first consider how to insure newly issued municipal bonds (the good insurer). The basic plan is to create a new government excess of loss reinsurance program that would backstop the risk on newly insured municipal bonds.

The TRIA program already in place for terrorism risks provides a useful template. The basic feature is an excess of loss insurance contract, with a deductible high enough to place all of an insurer’s contributed capital in the first loss position. The monoline bond insurers already face significant capital requirements, and these should be continued, possibly even expanded, for the newly restructured insurers.10 A further refinement, and one that differs from TRIA, would make the reinsurance payouts a minimum of the insurer’s actual losses (above its deductible) and a prorated share of the industry losses based on the principal amount of insured bonds. This would provide insurers an incentive to hold diversified—“market portfolio”—books of business, for otherwise a firm would face a “basis risk” whereby its actual losses might significantly exceed its reinsurance payout.

Dealing with the losses that are embedded in an insurer’s existing book of business represents a more complex problem. This has not been a difficulty for the existing government programs that insure natural disaster and terrorism risks because, by and large, all the extant claims had been settled by the time the government program started. But for the current municipal bond insurers the problem is particularly complicated by the fact that their existing books combine moderate losses on municipal bonds with vast losses on CDO and CDS subprime mortgage derivatives. If the two lines could be separated, then it might be sensible to bail out the existing municipal bond policies while allowing the insurer to default on its CDO and CDS policies. Indeed, had the regulators properly enforced the monoline principle, this would be an available option. In the actual case at hand, the government must either bail out the entire firm or allow it to default on all of its policies. In the latter instance, the government could still bail out the municipal bond investors.11

SUMMARY AND CONCLUSIONS

In this chapter I have proposed that an economist’s view be applied to the United States