Non-Recourse Debt? Welcome to Public Finance
The recent Bolingbrook, IL sales tax default has re-ignited a debate about what a default of non-recourse debt implies for the credit of the issuing municipal entity.1 “Non-recourse” in US public finance broadly indicates that the repayment pledge is limited, and not subject to an issuer’s “full faith and credit”—a term which itself may not have the legal power many assume. That aside, “non-recourse” is a nearly meaningless datum for gauging issuer credit in public finance. Quite simply, it doesn’t tell us something we don’t already know: many classes of debt in public finance—general obligation, lease appropriation, water or sewer revenue—are legally “non-recourse” to each other.
Short of outright bankruptcy, the implications of a “non-recourse” default upon the issuer‘s overall credit are necessarily fluid and contextual, which can be readily demonstrated from numerous compass points. The key here is that anything the issuer has undertaken by way of liabilities or operations need to be understood and evaluated by the analyst, or the investor. A direct corollary here is the concept of municipal bonds as either comprising “general obligations” or “revenues” is long past its sell-date.
In the case of Bolingbrook, the Village seems to have been heavily involved in its Forest City mall development project from the get-go, financing a full third of the $200 mm project costs with not one but two bonds issues from late 2005—the $47.7 mm 2005 sales tax bonds now in default, and $19.5 mm “Special Service Area” Series 2005-1 property tax bonds. Neither of the 2005 financings were disclosed in audits, even though they were equivalent to half of reported Village debt at the time. (Forest City, by the way, was not the Village’s only real estate involvement.) Curiously, other bonds that were also “not general, full faith and credit obligations of the Village”, were disclosed, including various tax increment bonds, and even industrial development instruments. Why were these disclosed, but not the Forest City related debt? As curious, the Village undertook to refund the 2005-1 SSA debt in 2019 extending its maturities from 2027 to 2041, and presumably protecting these bonds from default. This certainly begs the question as to why one non-recourse bond is worthy of such rescue, but not another.2 Leaving aside the further questionable credit practices of involving itself in costly real estate projects that are speculative—as the Village’s own 2005 offering documents admit, but which its subsequent audits hide—it does seem that Bolingbrook is indeed not a high quality credit by virtue of its poor disclosure alone. No analyst could have reasonably been aware of the Village’s 2005 sales tax debt until the default notice appeared occurred on EMMA. In fact, the only way to discover the Village’s real estate financings was to examine and cross-check all the Village listings in EMMA itself.
The non-recourse character of much municipal debt is a vestige of traditional public “fund” accounting as well as the various constitutional debt limitations that arose from the 1840s state defaults. A typical city water bond official statement will prominently disclose that the revenue bonds are not tax-supported and therefore not considered “debt” as defined by the state. The utility of such distinctions for credit is at best questionable—property taxes and the water bills are paid from the same economic base and the same resident pocket. Indeed, we have seen large scale issuer defaults and bankruptcies cascade across multiple funds and debt types, beginning with Jefferson County—which was triggered by problems in the sewer fund. The longstanding presumed sanctity and primacy of the GO pledge itself has now been upended, to the extent that selective repudiation even within GO debt classes has occurred.3 Further, we see the bond indentures, covenants, and other de jure protections so cherished by public finance immediately play second fiddle before the de facto political supremacy of pensions in municipal bankruptcy and default. Here, effectively, “recourse,” is what you can negotiate in settlement.
The notion that municipal bonds comprise either “general obligations” or “revenues” is long past its sell-date.
A non-recourse financing is reasonably pass-through in nature when the issuer has lent no more than its name to enable tax exempt issuance for an unrelated entity, as with the classic industrial revenue bond. Some cities allow frequent use of pass-through instruments (typically tax increment debt) to facilitate small real estate development projects. Here, the city may eventually see some economic and tax revenue benefit, but there is no direct municipal involvement in nor operational stake in the project. These typically speculative financings, along with elder care projects, comprise much of the non-rated municipal default universe. Such defaults typically tell us nothing about the issuer’s management of its debt and operations, or its relationship to creditors.
At the other extreme of a “non-recourse” default would be the outright failure of a city’s water department enterprise. Mismanagement of any operation that a municipality has chosen to undertake signals risk, and likely poor overall credit because any subsequent path has downside. On the one hand, the city could withhold non-enterprise resources to help the system, thus “protecting” its balance sheet but allowing the revenue debt to default. More likely, it will try to rescue the water system as a core operation, thus risking the general operational deterioration that Jackson, MS underwent. Either way, the city will have badly mis-managed a core operation and its debt, and opened itself up to further operational and possible legal challenges.
Two interesting examples from a decade ago underscore these difficult choices. The city of Monticello, MN, tried to build a public internet system at the behest of citizens unhappy with the private service; the revenue bond issuance was voter-approved, and the offering documents took pains to underscore the risky, non-recourse nature of the project. The city ultimately spent general reserves to help the system succeed, but the citizens who voted for the project stayed with the private provider, who apparently improved their service while delaying the city system with litigation. Monticello ultimately saw a credit downgrade because of its subsequently weakened balance sheet. The citizens of Griggs County, ND, thrice voted down a building project, for which county commissioners then issued appropriation debt, which required no voter approval, and whose issuance documents failed to disclose any project controversy. Voters appropriately recalled the commissioners; the new county board threatened to non-appropriate when the new debt came due, but ultimately relented, avoiding default and a potential credit downgrade.4 One can perhaps sympathize with the officials of Monticello abandoned by their fickle citizens, as well as the voters of Griggs County who were abused by their elected officials, but in either case the result was the same: general credit was placed at risk because new debt was issued for projects that were ultimately not tenable, whether economically or politically.
The “non-recourse” character of much public finance debt has widened with the widespread adoption of appropriation financing since the 1980s, which was engineered specifically to skirt debt restrictions. Just as a city’s water revenue bonds may legally not comprise “debt” under a state definition based upon tax-support, the appropriation bond is also not “debt”, thus avoiding the voter approval. Here the definitional sleight of hand is the appropriation itself: failure to pay once an appropriation has been made is a legal default, but non-appropriation itself severs any recourse and lets the issuer off the legal hook. If New Jersey, to take one example, suddenly decided to non-appropriate for what is the bulk of its debt, would the market not consider this a major default and credit event?
Appropriation debt where a leased asset is involved (more typical at the local government than state level) can even more problematic, because the structures often replicate equipment financing instruments where the recourse is limited to the leased asset itself, which may well be illiquid and impractical for the general investor. If a city decides it no longer needs a leased bulldozer, and surrenders it to the leasing company with liquidated damages, no one would properly consider this a default as long as contract terms were met. Yet each year, it seems, investors in terminated equipment leases point to the same city’s lease of its city hall and wonder why the bond rating isn’t affected. It is, ultimately, not a bad question, perhaps acknowledging an uncomfortable “wink and a nod” element to municipal appropriation instruments. When is a lease just an annual contract lease, as opposed to a capital market instrument, when the terms are so alike?
The ease with which an appropriation financing can suddenly become non-recourse can lend itself to abuse. A number of nominally robust, high-quality issuers—i.e., not in under any general stress at all—have walked away from lease financings, to the detriment of their general credit. The city of Vadnais Heights, MN and Platte County, MO both issued appropriation debt and where municipal support was, arguably, strongly implied based upon issuance documentation and audit treatment even though the projects were not critical to entity operations;5 in Missouri, furthermore, lease appropriation debt is a public finance mainstay given state debt restrictions. These entities could well afford to pay, but chose not to, and arguably mislead investors.
Between the extremes of the pure pass-through obligation and the outright failure of a core operation, then, the effect of a “non-recourse” default upon general credit is a matter of contextual gradation, defined by the nature and extent of the issuer’s involvement in the financing, its importance to the issuer, the issuer’s intentions, and its signals to investors. Above all, it is a matter of what the municipal entity has chosen to undertake in terms of operations and liabilities Investors need to be diligent about all the debt an issuer has incurred.
Al Medioli is a seasoned veteran of the capital markets, and investor, and Senior Fellow at The Transparency Foundation. This piece originally appeared in Smith Research & Gradings - a newsletter for investment professionals.
1 First published in Smith’s Research and Gradings, March 23 2023 as “Non-Recourse Debt: Welcome to Public Finance”
2 This refinancing was noted in the 2019 CAFR, the first time the 2005-1 debt was even acknowledged; the 2005 sales tax debt was still not disclosed.
3 E.g., Puerto Rico GO: nine classes of recovery
4 by a quirk of state law, non-appropriation was not an option because the project had a public safety element, enabling the skirting of voter approval.
5 In the case of Platte County, the project was a parking garage in support for a private mall development. Descriptions these examples, as well as those of Monticello and Griggs referenced earlier, can be found in Moody’s US Municipal Bond Defaults and Recoveries 1970-2020 (April 2021)