In what is now proving to be an unrelenting series of initiatives on the part of both Treasury and Congress to channel more funds into housing, no matter what the costs or risks to taxpayers, Treasury has put out a “Best Practices” guide with the aim of stimulating the market for covered bonds.1 A previous Cumberland commentary about covered bonds (www.cumber.com) generated some comments from Europe. In particular, they suggested that I had been less than sympathetic to the attractiveness and utility of covered bonds. Covered bonds have been a successful vehicle in Europe, and this is due in part to the fact that when bank-issuers fail, they typically are nationalized, protecting all creditors.
The situation is different in the United States. A few points of clarification are in order, together with some additional observations.
First, given how the covered bonds would be structured, they clearly would be attractive to investors, perhaps as good as, or maybe even stronger than, insured deposits but not subject to the $100k deposit insurance coverage limit. The bonds would be dynamically collateralized; in fact, they would be over-collateralized by 5% with some of the highest-quality mortgages. Treasury “Best Practices,” for example, call for loan-to-value ratios of not more than 80%. The dynamic collateralization comes from the fact that the collateral pool would be actively managed, and any loan that violated the loan-to-value ceiling or was more than 60 days non-performing or delinquent would have to be replaced. Thus, the bond holders would essentially have both a put and a call: a put on the bad assets in the collateral pool back to the issuing bank and a call on the bank’s good assets for replacements that might be other mortgages, cash, Treasuries, or agency securities. While the puts and calls are likely to be out-of-the-money initially, they both become valuable during times of general financial distress – like now- or when an issuer experiences financial difficulty. In that respect, they transfer value (and reduce risk) to covered bond holders from other uninsured creditors, exacerbating any tendency they might have to flee.
Second, the covered bond would also have a priority claim on the bank’s entire revenue stream, and not just the cash flows from the collateral, when it comes to meeting the required payments on the bonds. Any remaining revenue after paying covered bond holders would then be available to pay interest on deposits or other debt or to make payments to preferred stock and equity holders. This added priority serves to reduce the risks to covered bond holders relative to their promised returns but increases risks to other claimants and uninsured depositors. Overall, the scheme should serve to lower the relative costs of covered bonds, but this should be offset by the added costs of the additional risks that would now be visited on other types of claims and uninsured deposits. These risks become especially evident in the event of insolvency. Covered bond holders would not only have their collateral, but also would have a residual claim, just like any uninsured depositor or other creditors, on the assets of the bank. This is a form of double dipping, at the expense of uninsured depositors.
As an aside, the FDIC statement on how covered bonds would be treated in the event of insolvency provides added assurance of the priority protections covered bonds’ claims would be treated relative to other creditors.
Third, promotion of covered bonds also will impose greater risks on the FDIC and, ceteris paribus, increase deposit insurance costs to banks and ultimately taxpayers. Just like collateral arrangements associated with asset pledging to secure municipal deposits, Home Loan Bank advances, or discount window borrowings, high-valued assets would be unavailable to the FDIC as receiver or conservator to liquidate to cover either its own costs or the claims of other uninsured creditors or depositors. Moreover, since covered bonds will initially be put in place after other creditors have already purchased debt or deposit claims, the interest rates they receive will not fully reflect the risks to which the non-covered bond creditors and uninsured depositors are subsequently exposed by the addition of covered bonds to the bank’s funding pool.
Regulators, of course, have their CAMEL risk rankings and are in position to potentially understand and monitor those risks. But these risk assessments are confidential, and they are not available to depositors who are less informed and must instead rely upon private risk-rating services to assess their risk exposure.
The higher insurance costs and increased risks are just another of the unintended costs of a piecemeal approach to the current problem and of policies indirectly to subsidize housing. Overall, the constraints imposed by the covered bond contract should, in the aggregate, increase rather than lower bank funding costs, because some of the benefits of diversification would be taxed away by the covered bond holders.
The bottom line is that from the investor’s perspective, covered bonds might be as good as sliced bread, but that is only part of the story. The payments and constraints they impose also come at a cost, and those costs will be born by other creditors, the FDIC, and taxpayers. Such instruments put an even higher premium on well-functioning, prompt corrective action and early intervention by banking regulators. Unfortunately, the poor performance of OTS in the case of IndyMac and other regulators in the case of the failure of First National Bank Corporation don’t give one confidence that the system will work as intended.
1 U.S. Treasury, “Best Practices for Residential Covered Bonds,” July 28, 2008.