Arranging difficult financings at affordable rates has been a long-revered goal of tax-exempt market participants. It is the foundation upon which the now huge third-party credit enhancement industry has been built, and it has inspired much of the innovation in public finance over the last 15 years.
Several trends are now converging that will require new infusions of creativity--a protracted recession, growing housing and infrastructure needs, deteriorating credit quality among issuers and letter of credit providers, and increasing risk aversion among investors.
The dynamic growth of the asset-backed securities market over the past five years was fueled by many of these same trends, and some of the technology developed there for turning frogs into princes is finding increased application in the tax-exempt market. Three of these techniques will be briefly discussed here.
The asset-backed market was the first to discover that the parts are sometimes greater than the sum of the whole--that borrowing against isolated revenue streams of an issuer is preferable to borrowing on the issuer's general credit where its balance sheet is impaired and the lender can be given an unobstructed claim on the revenue stream. This was no better illustrated than by the late Bank of New England's ability to execute an asset-backed financing at acceptable spreads in the midst of its free fall into the outstretched hands of the Federal Deposit Insurance Corp.
One not so new application of this technique by tax-exempt issuers has been the use of special obligation bonds backed by gasoline taxes, user fees, or toll-road revenues. And it has appeared recently in more innovative structures, where payment streams of rated credits are passed through unrated borrowers. A recent example is the lease purchase certificates of participation financing of a prison in Massachusetts owned and operated by a nonratable entity, which obtained an investment-grade rating solely on the basis of user fees agreed to be paid to it by a rated governmental entity.
The asset-backed market was also quick to understand the applications of the more orthodox converse principle -- that the whole can be greater than the sum of its parts. The larger the pool of financial assets when compared to the average size of each individual asset, the better chance an asset-backed issuer has of obtaining a high rating on the pool.
The rating services have formulated actuarial methods for rating the larger pools, based upon prior performance history, diversity of credits, and small unit sizes relative to the aggregate size of the pool. The tax-exempt market has been slower to embrace this concept, although the idea of pooling municipal leases recently received an endorsement in these pages from representatives of one of the rating agencies.
Rating agencies' recent willingness to use a rating technique other than the weakest-link approach to rate municipal pools, such as the Los Angeles County School District and San Diego Local Government note pool issues, may encourage other pools using similar structures.
Under the Los Angeles and San Diego structures, a collateralization level is negotiated with the rating agencies based upon the percentage of lower-rated credits in the pool. The collateralization is provided through reserve funds or letters of credit.
Efforts to use pooling as a means of upgrading credit quality in the tax-exempt market in other than secondary market transactions will, however, be frustrated by the tax rule governing bank-qualified bonds. These rules disqualify bonds issued as part of a composite issue, even though the bonds of each separate issuer would be bank-qualified. Recent efforts to remove this impediment are well-founded, and if such efforts succeed, pooling of smaller credits could be a growing phenomenon.
The asset-backed market was also first to experiment with alternatives to third-party sources of credit enhancement. Faced with increasing risks of rating downgrades of letter of credit banks, mortgage-and
asset-backed issuers have V3
relied more heavily on "roll your own" forms of credit support, such as senior-subordinated structures and cash collateral, to achieve ratings of investment grade or higher from rating agencies.
With adequately sized subordinated pieces or adequately funded cash collateral accounts, even the lowest-quality assets are capable of achieving high ratings. The Resolution Trust Corp. has had both the necessity and the regulatory freedom to use this technique on a large scale in redeploying the assets of the failed thrift industry.
Most of the recent experimentation in credit enhancement in the tax-exempt market has, understandably, been in multifamily and elderly housing, where third-party credit support is difficult to obtain because traditional credit enhancers are not interested in assuming project risk. One effort to lay off project risks on more traditional real estate lenders resulted in the "gray box" structure.
This structure, much discussed in these pages over the past 18 months, was, at bottom, an effort to entice traditional enhancers into housing financings by providing them with cash, usually from the sale of a mortgage note, sufficient to defease their enhancement exposure.
Whether these transactions served a valid public policy by making financable projects which otherwise were not, and whether nuances of difference among variations on the gray box structure made some transactions tax-abusive and others not, are questions which have yet to be explored. All transactions with similar characteristics were thrown into the gray box, where they have remained to this date, pending clarification by the Internal Revenue Service.
More recently, a new controversy has arisen over what has been dubbed the "shadow box" structure. The only transaction given this name as of this writing is the Marengo County, Ala., Port Authority transaction in which an unenhanced high-yield zero coupon tax-exempt bond was issued on Day 1. On Day 2, the bond was resold as a triple-A rated security, having received the benefit of a guarantee from an off-shore company funded with U.S. Treasuries sufficient to defease its guaranty obligations.
If the Marengo County transaction has been accurately reported, the tax abuse there is readily apparent. This is because the proceeds of the remarketing, on Day 2, which were invested in the U.S. Treasuries, are analyzable as proceeds of the original bond issue, invested in impermissibly higher-yielding securities.
Because the remarketing of the enhanced bond on Day 2 was clearly part of the initial plan of financing, a reasonable tax analysis would be to treat the yield at which the newly enhanced bonds are remarketed on Day 2 as the original bond yield. Since the yield on the Treasuries exceed this yield, the bonds could be treated as arbitrage bonds.
Many variations on the Marengo County structure are possible, however, which raise more subtle questions of tax analysis and policy.
Suppose, for example, a bondholder who has held unrated high-coupon bonds for a respectable interval after the original issue date desires to enhance them for sale, in a transaction independent of the original issuance of the bonds, by pledging a pool of U.S. Treasuries to secure the repayment. If the bondholder is able to sell the bonds at a premium over his original investment, because of the credit enhancement, and uses some of the excess sale proceeds to reimburse himself for the price of the U.S. Treasuries, is that abusive?
If treated as an invested sinking fund, is it satisfactory if they are restricted to the original yield at which the bonds were issued, or must they be restricted to the yield at which the bonds were subsequently remarketed?
If, in order to have the bonds rated, the seller arranges to have the Treasuries owned by an independent and bankruptcy-remote tax-exempt entity, should the transaction be analyzed differently? Is it important whether the guarantee by this entity is a "qualified guarantee" as long as the Treasuries are restricted to the original yield on the bonds, and no attempt is made to increase that yield by amounts paid to such entity?
Does the treatment of payments pursuant to such entities guarantees as bond insurance (and therefore as a tax-exempt substitute for the issuer's payment obligations) depend on whether such entity takes an insurer's risk, or does it instead depend on the reasonableness of the assumption that the issuer of the bonds will be able to repay its bond obligation from proceeds other than the guaranty?
Even if the new investor has purchased the newly enhanced bonds at a significant premium, the amount of interest that will be excludable from gross income will be unchanged. If the interest excludable from gross income after the transaction is the same as before, then does the remarketing of the bonds at premium constitute an "overissuance"?
Is there any analytical difference between enhancing the value of a bond through the pledge of collateral and a financial institution's enhancing a pool of tax-exempt bonds in its own portfolio with its own guarantee or repo? In either case, the bonds will be resold at a premium and the sales proceeds will be invested at unrestricted yields, but the amount of interest which is excludable from gross income over the term of the bonds does not change.
These questions, and others not noted here, need to be examined calmly and rationally before the latest box eruption casts an overly long shadow over the market's ability to differentiate between tax-abusive deals and legitimate exercises of creativity in providing liquidity to the tax-exempt market. That the line between creativity and tax abuse may be difficult to draw does not mean that no effort should be made to draw it.
IRS clarification of these issues could have the beneficial result of ending the abusive transactions while permitting the acceptable ones to go forward. Without such clarification, the next step in the evolution from the gray box to the shadow box could be a totally empty box or no box at all.
Mr. Borod heads the Public and Structured Finance Group at Brown, Rudnick, Freed & Gesmer, in Boston.
