What’s the Status of General Obligation Bonds in Municipal Bankruptcy?

Most recent coverage of the Detroit bankruptcy has focused on the issue of pension cuts, and rightly so. Should Emergency Manager Kevyn Orr succeed in his plan to reduce benefits in bankruptcy, other insolvent cities will be encouraged to follow suit. Across the nation, many cities have too much debt and Detroit won’t be the last go bankrupt.

But stakes are also high for Wall Street investors. Orr has classified slightly more than half of Detroit’s general obligation debt as “unsecured,” and thus susceptible to cuts as deep as those proposed for retirement benefits.

Does this classification do violence to the “general obligation” pledge? Or does everything change in bankruptcy? Under ordinary circumstances, GO bonds are considered to be among the safest of fixed income investments. Muni market analysts and participants tend to believe GO creditors’ highly-protected status carries through into bankruptcy. Members of the bankruptcy community, by contrast, have tended to agree with Orr.

So what’s the status of general obligation bonds in municipal bankruptcy?

Kevin Kordana

Kevin Kordana

Kevin Kordana is Professor of Law at the University of Virginia. A graduate of the Yale Law School, he clerked for Chief Judge Richard A. Posner on the 7th Circuit. He has served as a visiting professor at the University of Southern California, Geor...Read More »

Chris Herzeca

Chris Herzeca

Chris Herzeca is an attorney, investor, commercial mediator and ungentlemanly farmer. Mr. Herzeca was a corporate finance partner of Brown & Wood (now Sidley Austin) and counsel at Kramer Levin in New York City. You can follow Mr. Herzeca on tw...Read More »

Day 1 – Opening Remarks


General obligation bonds are, for the purposes of municipal bankruptcy, unsecured.  
Their “full faith and credit” pledge is, as a matter of contract law–and to the (rather considerable) extent that bankruptcy law reflects the “creditors’ bargain,” therefore bankruptcy law as well–merely equivalent to an “and I really mean it” pledge, lacking the discrete and tangible connection to a particular property interest necessary to constitute a secured claim in bankruptcy.  
The pledge is made against the backdrop of, and therefore subject to, federal bankruptcy law.  As a policy matter, this outcome is desirable, as it allows some risk to be shifted to investors, who are better risk-bearers than municipal residents.  As a Depression-era federal court declared, the “purpose” of municipal bankruptcy is to “permit the scaling down of municipal securities, on the theory that half a loaf is better than no bread.”  (Getz v. Edinburgh Consolidated Independent School District, 101 F.2d 734, 736 (5th Cir. 1939)).  The amount of risk borne by general obligation bondholders is constrained by municipal bankruptcy law’s requirement that a reorganization plan be in “good faith” and in the “best interests of creditors.”  These requirements, overseen by a bankruptcy judge, provide protection to general obligation debt-holders.
In a municipal bankruptcy proceeding, the state law contractual rights of bondholders are constrained by the Bankruptcy Code’s “automatic stay.”  Section 506 of the Bankruptcy Code limits secured claims to those “secured by a lien on property.”  Lacking any concrete interest in any particular property of the municipality, general obligation bonds are therefore unsecured.  (In re Sanitary and Improvement Dist. No. 7, 98 B.R. 970, 973-74 (Bankr. D. Neb. 1989)).  This is in contrast to “revenue bonds” which are secured by collateral in a particular income source, such as the tolls from a bridge.  Of course, even secured creditors may not be paid 100 cents on the dollar in bankruptcy.
In the big picture, the fact that municipal bankruptcy law allows some losses to be shifted to general obligation bondholders should be unsurprising.  Just as shareholders in a corporation face limited liability, as do members of a cooperative, municipal bankruptcy law allows residents to shift some losses to investors.  From an ex ante perspective, such risk-shifting can be seen as efficient, as investors may be better-positioned to monitor the municipality’s fiscal affairs as compared to (relatively speaking) rationally passive residents.  This is particularly true given the existence of financial intermediaries such as mutual funds and bond insurance.  Bondholders can also diversify their investments, another risk-bearing advantage as compared to residents.  Lest one be overly concerned about investors in municipal bonds facing risk, it seems sufficient to note that investors in even completely “safe” (from a credit risk perspective) bonds face considerable interest-rate risk.


It depends, but generally secured.

In a Ch. 9 debt adjustment plan, secured obligations are likely to be “money good,” entitled to repayment in full to the extent of the pledged collateral, while unsecured obligations likely will suffer significant discounts, or impairment, since non-pledged general revenues will likely be insufficient to repay unsecured obligations. Municipal creditors fight to be classified as secured, and to have other creditors be classified as unsecured.

This classification fight regarding whether municipal obligations are secured has been teed up in the Detroit Ch. 9.  The Detroit Emergency Manager (EM) has initially classified certain GOs as unsecured, and the EM has suspended debt service on the GOs and diverted the pledged tax revenues supporting these GOs towards Detroit’s general fund.  Monoline bond insurers, who have guaranteed repayment of certain Detroit GOs, have sued the city, in order to, first, compel segregation of the tax revenues pledged to repay the GOs and, second, require the EM to characterize the GOs as secured in connection with any Ch. 9 debt adjustment plan.

There are three possible characterizations of GOs in Ch. 9:  they may be (i) secured, because they are backed with a pledge of tax or other revenues in a state that automatically places a statutory lien upon those receipts (a Lien State); (ii) secured, because they are backed with a pledge of tax or other revenues in a state that is not a Lien State, but the pledge of revenues nevertheless constitutes “special revenues” in Ch. 9; or (iii) unsecured, notwithstanding that they are backed by a pledge of tax or other revenues, because neither (i) nor (ii) above apply.

California is an example of a Lien State (i).  The California statute states that if a California municipality issues a GO and the authorizing resolution provides for a pledge of revenues, such as tax receipts, to secure repayment of the GO, the GO shall be secured by a first lien on such pledged revenues.  In Re Orange County made clear that Ch. 9 treats GOs with pledged revenues as secured in bankruptcy if issued by a municipality in a Lien State.

Michigan is not a Lien State, but, with respect to (ii) above, the monoline insurers of the Detroit insured GOs have argued that the GOs are secured by a pledge of “special revenues” under Ch. 9 sections 902(2)(E) and 928(a), insofar as the authorizing resolutions for these GOs (a) were approved by Detroit voters, (b) created the levy of taxes and secured these GOs with such levied tax receipts, and (c) financed the construction of specified projects (all prerequisites of section 902(2)(E) of the bankruptcy code).

On the merits, I expect the Detroit insured GOs to be characterized as secured in any debt adjustment plan confirmed by Judge Rhodes in the Detroit Ch. 9.

What about GOs secured by a pledge of tax revenues, but issued by a municipality not within a Lien State, and not satisfying the requirements of Section 902(2)(E)?  Because Section 552(a) of the federal bankruptcy code, which is applicable to Ch. 9, provides that any pre-petition pledge terminates upon bankruptcy as to property acquired after bankruptcy (subject to the Lien State and “special revenues” exceptions discussed above), the municipal debtor may avoid the pledge of tax receipts received post-petition.  This is case (iii) above, rendering these GOs—but only these–unsecured in bankruptcy.

Day 2 – Rebuttals


I found the first round to be quite clarificatory. Indeed, it would appear that there is a considerable (though not complete) convergence between my views and those of Chris Herzeca.  
In round one, I was attempting to answer the “common law” or baseline question, “are general obligation bonds secured” to which I replied no.  That said, I would certainly agree that there could be specific state-level statutes or contractual language that might change this result in particular circumstances.  I take it that Herzeca’s categories (i) and (ii) constitute such examples.  Similarly, I take it that Herzeca’s category (iii) reflects agreement with me as to the baseline or “common law” case.  
With respect to category (ii), I would note that I read Bankruptcy Code sec. 902(2)(e) to reflect what I take to be the common law distinction between revenue bonds, which can provide a secured claim because they pledge revenue from “taxes specifically levied to finance one or more projects or systems” and general obligation bonds, which provide unsecured claims because they are to be repaid from “general property, sales, or income taxes,” the latter not constituting “special revenues” for section 902(2)(e) purposes.  
My reading of section 902(2)(e) provides a good example of how the common law baseline (Herzeca’s category (iii)) might be expected to demonstrate some “gravitational attraction.”  In other words, for doctrinal reasons and perhaps for some of the policy reasons that I touched on in my first post, courts may be relatively reluctant to interpret the provisions in particular general obligation bonds as reflecting the intent to repudiate the common law baseline (i.e., to shift from Herzeca’s category (iii) into category (ii)).


My disagreement with Professor Kordana is that while he asserts the pre-petition security pledge of taxes that are levied post-petition does not secure GOs in Ch. 9, I assert that it does with respect to GOs backed by “special revenues”, as defined by Ch. 9, and GOs backed by a pledge of taxes that are issued in a Lien State. A municipal GO bondholder reading this debate might be left uncertain. What should an investor do?

First, remember that (1) states can’t file for Ch. 9, so the secured status of GOs in bankruptcy applies only to municipalities’ GOs, (2) only about half the states authorize their municipalities to file for Ch. 9, (3) most of those states that authorize municipal bankruptcies do so on a conditional basis (usually after the municipality has made a good faith attempt to negotiate a consensual agreement), and (4) a municipality that is authorized to file must be insolvent on a cash flow basis (so mere underfunding of pension plans is not sufficient).  Moreover, GOs have a reported 0.06% default rate.  So don’t run out and sell your GOs!

It has been reported that there are 30 Lien States.  Their GOs, when backed by a pledge of taxes, are secured in Ch. 9.  Moreover, many municipalities in non-Lien States that wish to finance new projects face statutory limitation upon debt issuance and tax levy. These municipalities will issue secured GOs (such as Detroit’s insured GOs) if their bond resolutions authorize the levy of additional taxes securing the authorized GOs, and the GOs are issued to finance specified projects.  

So, this is not a small universe of secured GOs.  But here’s the rub.  How do investors know whether their GOs are these GOs?

Investors can search online by CUSIP for their GO’s Official Statement disclosure, but they are not likely to discover a definitive conclusion regarding the security status of their bonds in Ch. 9.  This investor uncertainty is a good argument for buying insured GOs: let the municipal finance guaranty firms do their specialized analysis and price the risk accordingly.  The investor benefits not only from the financial guaranty, but can piggyback on the guarantor’s superior risk analysis of the GO’s security.

While I agree with Professor Kordana that investors in a municipal GO are more able to diversify risk than the municipality’s residents (although the massive decline of Detroit’s population indicates that residents do diversify their risk, by moving van if not by portfolio diversification), residents should prefer that their municipality GOs are secured rather than unsecured.  GOs finance long-lived municipal capital improvements, which contribute to the welfare of the municipality just like municipal health, safety and recreational services.  The municipal tax base must support debt service as well as payment of current services and amenities.  If it is clear that the GOs are secured, the debt service burden should be less than otherwise, thereby providing more financial flexibility for the payment of municipal services.

Why do we have a situation where the secured status of GOs in Ch. 9 is difficult to assess with certainty?  Is this a failure of our legal system?  Ch. 9 involves a unique interplay between aspects of federal bankruptcy law (not all of Ch. 11 is incorporated into Ch. 9, and certain provisions of Ch. 9 differ significantly from Ch. 11) and state law, which can vary significantly from state to state.  Municipal bankruptcy is a minefield even for the cognoscenti.  My own view is that uniform treatment of municipal debt in bankruptcy is not only within federal congressional power, but also desirable. 

Day 3 – Questions from the Moderator


Kevin, A question about the “common law” or traditional understanding of GOs. Much of the debate thus far has focused on GO bondholders’ security relative to other bondholders. But the most controversial question in the public’s mind is the “Main Street v. Wall Street” question of how much protection retirees should expect from municipal bankruptcy compared with credit market creditors.

What was so startling about Kevyn Orr’s proposal to creditors was that he placed some Detroit GO bonds in the same class as pension and retiree healthcare obligations, which are not backed by any specific revenue commitment. Should GO debt be treated the same as retirement benefit debt in municipal bankruptcy?


There is, at some level, a frustrating circularity to the question of whether General Obligation bondholders take before pension-holders or share equally with them in municipal bankruptcy.  GO bondholders either take before, because “the law” (which is ultimately “us”) says they are secured, with a property interest in (at least) future tax revenue, or they share equally, because the law says they are unsecured.  But what is really driving this answer?  Rules of property law in general, and legal doctrine regarding security interests in particular, can be structured in numerous ways. Why might we choose one way over another?

An important insight that I will argue is relevant to the GO bond question was provided by Harold Demsetz in his article “Towards a Theory of Property Rights” (57 Am. Econ. Rev. Pap. & Proc. 347 (1967)).  Demsetz recounts how Native Americans who lived both in forests in eastern Canada and on the Great Plains held real estate communally prior to European contact.  Post-European contact, however, the easterners switched their property regime such that individual families now controlled private plots of land, while the plains-dwellers continued to own their land communally.  Demsetz explains the logic of these property regimes as follows: prior to European contact, furs were not particularly valued, and an ample supply was typically obtained as a by-product of hunting for meat.  While there was some danger of over-hunting, there was also a significant cost to defining and enforcing property-lines at the family level, so land was held communally by groups who lived together.  Europeans, with their intense desire for furs and willingness to exchange valuable new products such as guns for them, changed this calculation for the forest-dwellers.  The risk of over-hunting increased.  It now made sense to bear the costs of subdividing the land and enforcing these new boundaries.  Each family would be less likely to over-hunt their particular plot.  (That is to say, a “Tragedy of the Commons” was ameliorated.)  However, while fur-bearing animals in the forest had a limited range and so could on balance be “conserved” on family plots, bison on the Great Plains ranged widely, beyond the ability of any family to “conserve” them.  Thus, the costs of family-level plots were not worth bearing and communal ownership persisted–with the bison, of course, nearly going extinct.

Demsetz’ analysis calls attention to the fact that different property regimes have different costs–especially those of monitoring and enforcing property rights.  In the context of municipal bankruptcy, it seems plausible that allowing GO bondholders to take before pension-holders could be costly.  This is because bondholders seem to be better-positioned to monitor municipal borrowing and to force municipalities to internalize the expected costs of their risk-creation through higher interest rates.  This is not so much because of the difference between the average bondholder and the average municipal employee as it is because of the presence of large and sophisticated financial intermediaries in the municipal bond market.  These intermediaries include mutual funds and bond insurance companies.  These intermediaries would have a greater incentive to monitor municipalities (because they have more money on the line) and lower-cost of doing so (because they have more expertise) than would pension-holders.  Thus, reducing the incentive of bondholders to monitor, by giving them priority over pension-holders in bankruptcy, seems unwarranted.  Thus, I would argue, a Demsetz-style analysis suggests that we should construct our property regime so as not to (at least as a default matter) grant secured status to GO bondholders.


Chris, so, just to be clear, are you saying that GO bonds are secured to the extent that they’re not really GO bonds, at least not according to the traditional understanding? Or are you saying that the traditional understanding of a GO has always been misleading?

Also, do you think that the public benefits from such increasingly complicated debt structures, about which there seems to be so much confusion even amongst the “cognoscenti”? 


The “traditional understanding” of GOs is they are safe investments because of a promise and pledge:  the issuer has promised its “full faith and credit” to levy taxes without limit as to rate and amount to repay the GOs, and has pledged these future tax receipts as security.  The promise and pledge of this “traditional understanding” is enforceable in state court.  In federal bankruptcy court, things get more complicated. 

So, let’s examine this by means of two hypotheticals (and since it is topical, I will also include pension obligations).  Let’s posit that a municipal GO investor, the mayor, and a pensioner walk into a judge’s courtroom (I promise that this is not a variation of the standard barroom joke). 

1.  Let’s first assume that this is state court, not federal bankruptcy court. The mayor says to the judge, my taxpayers are tapped-out your honor, my town can no longer afford to pay debt service on the GOs and make sufficient contributions to the pension plan to keep it fully funded.

The GO investor, while extending a handkerchief to the mayor, demands that the judge enforce the municipality’s promise and pledge, and order the mayor to raise the tax rate, and collect taxes sufficient to pay debt service on the GOs.  The judge grants this order and municipal taxes go up.  The mayor makes a note to ask the town lawyer if there is another way, as an election soon approaches.

The judge then asks the pensioner whether pension benefits are being paid to retirees as promised.  If not, the judge is willing to order the mayor to do so, as the state’s constitution mandates that vested pension rights cannot be reduced.  Pensioner responds that benefits are being paid from the plan, but municipal contributions to the plan have been reduced, and the plan is underfunded.  Judge responds that there is nothing he can do about that, as this is a defined benefit, not contribution plan.

2.  Mayor consults with counsel, and now the mayor, GO investor and pensioner walk into federal bankruptcy court.  The judge points out that the “full faith and credit” promise to raise taxes sufficient to pay the GOs is just a promise which cannot be enforced by GO investors in bankruptcy.  The pledge of taxes as security for the GOs also does not survive the bankruptcy filing.  The judge notes that the municipality is not in a Lien State, and there is no pledge of “special revenues”.  The judge orders that the GOs may be impaired in connection with a debt adjustment plan.

Since the mayor wants to continue to make reduced pension contributions, which won’t sustain the payment of future promised payments, the mayor requests authority to reduce pension plan benefits.  The judge notes that these pension plan benefits are also mere promises not secured by any collateral, and orders that the pension plan may be impaired as well.  When the pensioner replies that his pension plan benefits are constitutionally protected by state constitution, the judge answers that he answers to a higher authority, the Supremacy Clause of the US Constitution, which authorizes federal law to trump state law when in conflict.

Since states can’t file Ch. 9, the state GO investor will face scenario #1, not #2, so that his traditional expectations regarding GOs will be intact (until the state defaults anyway, and commences an out-of-court debt restructuring).  Municipality GO investors who can face scenario #2 will be secured in Lien States and unsecured in non-Lien States, and their outcomes will differ dramatically.

Making investors analyze this enforceability obstacle course makes no particular sense.

Day 4 – Closing Remarks


Chris Herzeca ends his Day 3 commentary by posing the interesting and necessary question of how much sense this all makes.  In general, discrepancy between state law (here, allowing General Obligation bondholders who are victims of a default to pursue a “mandamus” action compelling a municipality to increase taxes) and federal law (here, Chapter 9 of the Bankruptcy Code, which may allow General Obligation bonds to be impaired) stands in need of justification.  Traditional justifications for bankruptcy law include both the procedurally-oriented amelioration of a creditors’ “race to the courthouse” collective action problem (Thomas Jackson, The Logic and Limits of Bankruptcy Law (1986)), and the substantively-oriented correction of a systemic defect in state law (Barry Adler, “The Questionable Axiom of Butner v. United States,” in Bankruptcy Stories (Robert Rasmussen ed., 2007)).

If one considers the history of municipal bankruptcy, the state-law tax increase remedy for General Obligation bondholders suffering default does appear to have considerable drawbacks.  Indeed, it was characterized by the U.S. Supreme Court as “an empty right to litigate.”  (Faitoute Iron & Steel Co. v. City of Asbury Park, 316 U.S. 502 (1942).)  The Court pointed to “the spectacle of taxing officials resigning from office in order to frustrate tax levies through mandamus, and officials running on a platform of willingness to go to jail rather than to enforce a tax levy . . . and evasion of service by tax collectors, thus making impotent a court’s mandate.”  

In another case, citizens, in the role of taxpayers and voters, voted out the justices of the Iowa Supreme Court, replacing them with “judges already committed to [the anti-bondholder] viewpoint.”  (A.M. Hillhouse, Lessons from Previous Eras of Default, in Municipal Debt Defaults: Their Prevention and Adjustment (Carl Chatters ed., 1933).)  The U.S. Supreme Court overruled these new Iowa judges, in spite of the state law nature of their rulings, stating “[w]e shall never immolate truth, justice, and the law, because a State tribunal has erected the altar and declared the sacrifice.”  (Gelpke v. City of Dubuque, 68 U.S. 175 (1864).)  An attorney who had represented both municipalities and bondholders reported on the bleak record of mandamus actions.  (Edward Dimock, Legal Problems of Financially Embarrassed Municipalities, 1935 A.B.A. Sec. Mun. L. Proc. 12.) An SEC report on municipal bonds noted that “high taxes, excessive [property-tax] delinquencies, or both, constitute the normal proximate cause of default, [so] . . . prolific use of [mandamus to raise taxes] must be productive of still only higher tax levies and higher delinquencies.” (Report on the Study and Investigation of the Work, Activities, Personnel and Functions of Protective and Reorganization Committees: Part IV 19 (1936).)  Indeed Hillhouse, cited earlier, reports that the property tax rate reached the seemingly absurd level of 42.5% of assessed value in West Palm Beach.  In light of this historical record, Chapter 9′s attempt to forge a coordinated reorganization plan, in which General Obligation bondholders remain protected by the requirement that a reorganization plan be in “good faith” and in the “best interests of creditors,” may well be justified.

Of course, if it makes sense for Chapter 9 to allow the impairment of municipal General Obligation bonds, it is not immediately obvious why state General Obligation bonds cannot avail themselves of the same treatment (particularly given that some municipalities have a higher population than many states).  Yet states cannot file for Chapter 9.  One suspects some combination of politics and federalism concerns to underly this policy decision. 

Thanks to Chris Herzeca and our moderator, Stephen Eide, for an interesting and productive discussion.


Investors often look for “tells”.  A “tell” is an event occurring in the marketplace that has significance beyond the individual circumstance.  Of course investors may disagree whether an event is a “tell”, and what the significance of any “tell” may be.  This is what makes a market.

For example, in the high yield corporate finance area, one “tell” may be when corporate issuers increasingly issue PIK (payment in kind) securities, where the issuer is not required to pay cash interest.  For some investors, this is a “tell” indicating that it is time to sit out a frothy market; yet, for other investors PIK debt has its attractions (such as higher yield).

Are there “tells” in the municipal finance arena, represented by the Detroit, Stockton, Harrisburg, Vallejo, and Jefferson County bankruptcies? 

Should investors demand higher interest rates for municipal GOs, especially if the issuers are not located in a Lien State and that state also authorizes municipalities to file Ch. 9?  And where exactly can an investor find out the lien status and authorization to file Ch. 9 about that state and municipality, if that investor was so inclined?  Should public unions confront city hall and demand their defined benefit pension plans be secured by collateral (such as city hall, for example)?

My personal view is that there is a silver lining in these municipal bankruptcies, and that silver lining is that these questions are being raised with more urgency than before. 

Municipal default rates have historically been much lower than corporate default rates, and in my view this has spawned a certain laxity among municipal finance participants. Municipal offering statements have historically been less detailed than corporate prospectuses, and municipal credit research has historically been somewhat more forgiving than corporate credit research.  Additional disclosure by municipal issuers and more detailed analysis by municipal research firms can be expected and is welcome.

It is high time for searching (and even “tell”ing) distinctions to be made among the creditworthiness of the various states and municipalities.  Hopefully, this more detailed municipal finance analysis will not be available just as paid research for institutional clients, but become part of the civic as well as investment discussion among everyone who has a stake in state and municipal finance and governance…which means everyone. 

For example, see this interesting study by the Mercatus Center of George Mason University ranking the financial condition of the 50 states.  It might be interesting to layer onto this study’s map a map of the “red” and “blue” states, but this is the subject of an entirely different debate.  In any event, municipal finance needs more studies like this, and walls of municipal finance fame and shame need to start naming names before, not after Ch. 9 filings and municipal defaults occur.  Rating agencies are well intended, but they are not up to the task of filling the municipal finance research void by themselves.  Municipal finance research needs to put its big boy pants on.

Kevin Kordana

Kevin Kordana is Professor of Law at the University of Virginia. A graduate of the Yale Law School, he clerked for Chief Judge Richard A. Posner on the 7th Circuit. He has served as a visiting professor at the University of Southern California, George Washington University, University of Munster, and the Peking University School of Transnational Law.

Chris Herzeca

Chris Herzeca is an attorney, investor, commercial mediator and ungentlemanly farmer. Mr. Herzeca was a corporate finance partner of Brown & Wood (now Sidley Austin) and counsel at Kramer Levin in New York City. You can follow Mr. Herzeca on twitter @cherzeca and read his blog at http://mbibaclitigtion.blogspot.com/.

Add Comment Add yours ↓

Your Comment