How Detroit’s Pensions Will Profit From the City’s Bankruptcy


Much has been written about the unprecedented “Grand Bargain” at the heart of Detroit’s plan of adjustment, and typically with a populist moral chiaroscuro. This arrangement—which assembles $816 million of funding from private foundations, the Detroit Institute of Arts, and Michigan taxpayers over the span of 20 years—seeks to minimize pension cuts and would ostensibly shelter the city’s art collection from “Wall Street huns” (as the city has referred to its capital markets creditors).

A number of prominent charitable foundations and corporations with economic interests in the region have agreed to participate financially. A bill to provide the state’s portion wound its way through the legislature. In brokering the deal, however, the city has exploited stakeholders’ ignorance about how Chapter 9 works and the magnitude of the city’s pension liabilities.

First, a court cannot legally force an insolvent municipality to liquidate assets, much as a court cannot legally force the municipality to levy additional taxes. This is one of the primary differences between municipal and corporate bankruptcies. The suggestion that the art was ever at imminent risk of being liquidated without the city’s consent was merely a political expedient. The city used this narrative to build an artificial sense of urgency around adopting this particular solution to the exclusion of more thoughtful approaches.

Second, if the city wished to consider liquidating assets to offset its liabilities (as Harrisburg, Pennsylvania has done successfully), it has close to a dozen options other than its art collection. These also have the potential to raise more money, especially in the context of the conservative appraisals the city received on its art.

Third, the city’s quantitative treatment of its pension liabilities in the plan of adjustment will ultimately produce above-par recoveries for pensioners, thus rewarding decades of rent-seeking behavior right before the November elections.

Detroit’s Pension Funding Prior to the Issuance of COPs


Municipal bond market participants did not consider Detroit’s pension funds to be poorly funded when the city filed for bankruptcy in 2013. This is primarily because the pension funds received a massive cash infusion through the issuance of certificates of participation (COPs) in 2005 and 2006, when Kwame Kilpatrick was mayor. Ironically, the city is currently trying to wipe out the investors from these transactions.

To understand how Detroit has treated—or rather, used—its pension liabilities in Chapter 9, some historical background is required.

According to the Detroit Free Press, city taxpayers supported an extensive bureaucracy that continued to grow for decades even as the city’s population was in decline. Before Kilpatrick was mayor, the city had one employee for every 51 residents. (By comparison, the city currently has one employee for every 73 residents.) Unions frustrated all efforts elected officials made at downsizing the city’s workforce.

The unions were also the reason Detroit’s two pension funds did not behave much like pension funds. Dating back at least to the 1980s, the pension funds distributed earnings they made above their benchmarks to retirees and active employees in the forms of 13th checks and annuity bonuses. Normally, a pension would reinvest these earnings so that the fund’s assets would grow over time and be sufficient to cover benefits payments as they come due.

Dennis Archer, mayor of the city during the 1990s, attempted to end these “bonus” payments in the interest of keeping the funds solvent. Archer was not able to do this directly, however. Most public pension funds are not controlled by the governments responsible for funding them. They are controlled by independent boards. This was also true for Detroit, although the boards were largely packed with union representatives.

Again, from the Detroit Free Press:

Archer backed an effort to block the payments through a proposed new city charter, which actually passed in August 1996. Enraged, several city unions and a retiree group sued and won. Archer tried again to block payments through a ballot initiative, called Proposal T, but it failed ….
One of the city’s two pension funds, the General Retirement System board, which represents nonuniform employees, doled out $951 million in excess earnings, mostly to retirees and active employees, from 1985 to 2008, according to a report conducted for City Council in 2011 by independent actuary Joseph Esuchanko.
He estimated that the total accumulated cost to the city from the distributions, including lost interest, was $1.9 billion as of June 30, 2008.
It is unclear how much extra money was distributed from the city’s other pension fund, which covers police and fire department employees. Officials have told Free Press, however, it was a much less frequent practice and that it happened in earnest for only a few years. With the stock market booming under Archer’s tenure, the city agreed in negotiations with its police and firefighter unions in October 2000 to hand out $190.4 million in bonus pension payments to current and future employees.

Detroit’s employees had little economic incentive to spare the funds from assuming nothing but downside investment risk here. Unlike most public pension systems, Detroit’s employees were not required to contribute to the funds themselves.

By the time Kilpatrick was elected mayor, the city’s pension liabilities had become a major operating issue for the city. The city had fallen nearly $1.7 billion behind in payments to the pension funds. The pension funds had filed lawsuits to compel the city to make up the payments as mandated by the state constitution, although doing so would devour the city’s general fund and likely force large-scale layoffs.

The city’s pension systems would probably not have been underfunded or as severely underfunded at this point if investment earnings had been reinvested rather than distributed to current and former employees. But they were, and so—in a move that perhaps foreshadows the city’s behavior in Chapter 9—Kilpatrick & Co railroaded a controversial financing arrangement through the political process.

Issuance of COPs


From an economic perspective, Detroit’s pension borrowing was similar to those undertaken by dozens of state and local governments in recent years. Essentially, the government borrows money from investors in the bond market and then transfers investors’ money to its pension funds to invest themselves. This is an arbitrage strategy (and the reason why pension bonds are usually issued on a taxable, not tax-exempt, basis): The government hopes the pension’s earnings on its investments will be higher than the interest the government is obligated to pay to investors for the privilege of using their money.

Now, given what you know about the unique way in which Detroit’s pension funds were administered, do you anticipate a problem with how the city might implement an arbitrage strategy?

From a financial perspective, Detroit’s pension borrowing was a Frankenstein of pre-crisis engineering and innovation. Each element of the borrowing wasn’t special in itself given municipal market structure and the Zeitgeist, but their combination in one transaction certainly was.

Establishing shell entities

The Detroit City Council, through Ordinances 03–05 and 04–05, created an “alternative funding mechanism” expressly for the pension deal. These ordinances form two shell entities corresponding to its two pension funds: the General Retirement System Service Corporation and the Police and Fire Retirement System Service Corporation. These entities then formed a third entity—the Detroit Retirement System Funding Trust—through a trust agreement with US Bank.

The Detroit Retirement System Funding Trust is the entity that sold the certificates of participation to investors for the purpose of capitalizing the pension funds and paying associated costs of issuance, including the bond insurance premium. COPs are a relatively common form of financing where an investor purchases shares of lease revenues from a program. This is different from purchasing bonds that are secured by a revenue stream. COPs are considered unsecured debt.

Although a legally distinct entity, the Detroit Retirement System Funding Trust was controlled by a board made up of the city’s finance director, budget director, legal counsel, and two city council members. (The transaction was rife with conflicts of interest, not the least of which was that the city’s finance director at the time, Sean Werdlow, went to work for Seibert Brandford Shank, one of the firms that engineered the deal, less than six months after the transaction closed and is romantically involved with one of the firm’s founders.)

The City of Detroit then contracted with the service corporations to make “service contract” payments to them—which, although they aren’t called debt service payments, happen to correspond to the payments due to COP investors. At closing, pursuant to the trust agreement, the service corporations agreed to transfer, irrevocably, their rights to collect these service contract payments to the funding trust. This gave the funding trust the ability to make payments on the COPs.

You’ll note that the pension systems themselves are not a party to either the service contracts or the trust agreement. The only thing they do is receive the proceeds from the sale of the COPs. This feature of the alternative funding mechanism will probably be relevant in future litigation.

(Here is a link to the official statement for the 2005 deal.)

Circumventing debt limits

While this structure may seem migraine-inducing to non-market participants, the use of shell entities exclusively for the purpose of financing something (usually a capital project, but in this case pension liabilities) is fairly common in public finance. Actually, it’s fairly common in finance generally.

There are a number of reasons why state and local borrowers use independent corporate entities. Sometimes this is an administrative convenience, to separate debt programs from other governmental affairs or from each other. Sometimes this is a legal convenience, to create a vehicle for financing a specific project in a way that is bankruptcy-remote from the rest of government. (Consider a project that does not represent an essential governmental activity that the government may or may not want to support in financial distress, such as a sports venue, convention center, etc.—basically a form of boondoggle insurance for taxpayers.)

In Detroit’s case, the structure was a political convenience in that it allowed Detroit to circumvent debt limits. Michigan law, with few exceptions, prohibits municipalities from borrowing more than 10 percent of the value of the property within their borders. Detroit did not have anywhere near the debt capacity for an offering of this magnitude.

It is not uncommon for state and local governments to use independent entities for the purpose of circumventing constitutional or statutory debt limits. Usually the limit that is being circumvented, however, relates to debt involving a specific kind of pledge and not overall borrowing. For example, a government might create an entity to issue revenue bonds — which are backed by a specific revenue stream — as an alternative to issuing general obligation bonds — which are backed by the government’s full faith, credit, and authority to levy taxes, and often require a referendum. (So really what is being circumvented is defending a new bond issue to voters.) This is very different than Detroit’s motives, which were to borrow beyond the city’s debt limit in both a legal and economic sense.

By Detroit officials’ logic, the offering did not count against the city’s debt limit because the city was making “service contract” payments rather than “debt service” payments, even though the only purpose of the “service contract” payments was to service the debt.

Some have suggested that the structure was also used to evade state oversight, given that Michigan law also requires debt offerings with certain characteristics to be approved by the state treasury. On this point, I would say that it is somewhat difficult to imagine that state officials were unaware that the offering was taking place. The deal was infamously awarded the Bond Buyer’s Midwest Deal of the Year, which tends to be something officials boast about. Also, according to the Wall Street Journal, “when first extended in 2005, the debt marked the largest municipal financing deal ever offered in Michigan.” I would submit to you that state officials had the opportunity to stop or scrutinize the deal if that had been desirable.

Synthetic fixed rate debt

The other feature of the deal that receives the most attention is the synthetic fixed rate component. The Detroit Retirement System Funding Trust issued variable rate debt and entered into interest rate swap agreements with UBS (which was incidentally also book-runner on the deal) and Merrill Lynch (incidentally also a co-senior manager) as counterparties to convert the debt to a fixed rate.

(As an aside: It is kind of fun to go back and look at pre-crisis official statements and see the number of parties involved then versus now. Really gives you perspective on how much the finance industry has consolidated.)

Back in 2005, UBS was one of the top underwriters of US municipal debt and frequently pitched variable rate deals with and without swaps. (The firm exited municipal underwriting after the collapse of the auction rate securities market in early 2008.) Central to UBS’s pitch was that the bank had strong connections with other European banks, which were all looking for new sources of income. (See here and here for a list of banks that were eventually stung by this transaction.)

Detroit refinanced a portion of the 2005 deal in 2006 to make use of a longer amortization schedule adopted by the pension funds’ boards (from 13 and 20 years to 30 years). Although it doesn’t really fit into the anti-Wall Street posture the city has since adopted, according to offering documents the city was in-the-money on the swaps at the time of this transaction and received a $36 million termination payment. Per the terms of the service contracts, these funds were required to be passed on from the funding trust to the city. It is difficult not to assume that the city was likewise motivated by cash flow concerns to undertake this transaction.

(Here is a link to the official statement for the 2006 deal.)

Synthetic fixed rate debt was popular in the market prior to the financial crisis because it allowed borrowers to exploit the significant difference in interest cost between variable and traditional fixed rate debt. (This difference existed mostly because variable rate structures were short-term debt masquerading as long-term debt, which is a story for another time.) Synthetic fixed rate structures ran into many different snags during the financial crisis, however. The markets for variable rate debt structures collapsed. Historically correlated interest rates diverged. The entities providing letters of credit or credit enhancement experienced financial difficulties of their own and had to be replaced. I could go on and on.

In Detroit’s case, the city’s ratings were cut in 2009, which was a termination trigger on the swap agreements. This happened at the worst possible time. Because a fixed rate swap is a hedge against rising interest rates, it loses value as rates decrease. At this time, the Fed was in the process of driving interest rates into the ground. Detroit’s swaps were out-of-the-money by hundreds of millions of dollars.

Ex post facto collateral

Detroit negotiated a reprieve from its counterparties by pledging its casino revenues (approximately $180 million annually) as collateral on the swaps, although it is far from clear whether this was permissible under state law. (I would submit to you that it was not.) Legally permissible or not, pledging the casino revenues should have removed any incentive the city might have had not to pay what was due under the contracts because the banks could block one of the more reliable sources of revenue the government had.

By June 2013, the city’s finances were so dire that it defaulted on the COPs. Syncora (formerly XL Capital Assurance), which insured the COPs and swaps, attempted to trap the casino revenues that had been pledged as collateral. This move provoked a race to the courthouse, whereby the city filed for Chapter 9 bankruptcy protection.

Detroit’s Use of Pension Liabilities in Bankruptcy


After Detroit filed bankruptcy, two interesting pension-related things happened.

City Attempts to Repudiate COP Transactions

The first interesting thing that happened was that the city filed a lawsuit against the service corporations and the funding trust in an attempt to repudiate the COP transaction. You will recall that these are simply shell entities that the city established itself and governs itself. So the city is essentially suing itself saying that the city’s own actions violated state law. And because the city’s own actions violated state law, the city’s obligations should be unenforceable. The city seems to believe that the pension funds should be able to keep the money they received from COP investors and have gainfully invested in the meantime though.

(This strikes me as one of the more significant risks to the city in trying to survive emerging from bankruptcy. Suppose the city is successful in this endeavor and COPs investors sue the pensions for ill-gotten gains or something along those lines. That would essentially return the city to Kilpatrick-era dysfunctional operations. Who knows?)

While the city wants to keep the proceeds from the COP transaction and wipe out those investors, it settled with its swap counterparties early on in its bankruptcy. You will recall that the swap counterparties were firms that engineered the COP deal in the first place.

In another strange twist, the terms of the city’s settlement with its limited tax general obligation (LTGO) bondholders (i.e., Ambac and Blackrock) give those bondholders and pensioners higher recoveries if the city prevails in wiping out the COP investors. The settlement assumes Detroit will transfer its rights and interest in the COP lawsuit to a litigation trust. The funds the city would have used to repay the COP investors would be divided 65 percent to retirees, 20 percent to the LTGO bondholders, and 15 percent to various other unsecured claimants.

So retirees are essentially getting compensated twice on the same transaction—they received the proceeds that capitalized the pension funds in the first place and a percentage of the revenue stream that was supposed to be used to repay the debt.

City Uses Negotiated Pension Assumptions

The other interesting thing that happened is that Detroit’s unfunded pension liabilities grew—by multiples—from the city’s last actuarial valuation.

This did not happen because anything changed in reality, but because the emergency manager chose to use different performance and amortization assumptions than the actuary had used. There isn’t much of an explanation as to why these assumptions were changed, except that the changes were “negotiated” with labor representatives.

The city used a 6.75 percent discount rate for both plans versus the 8.0 percent for the Police and Fire Retirement System (PFRS) and 7.9 percent for the General Retirement System (GRS) used in the last (2011) actuarial report. The city also cut the amortization period from 30 years to 15 years for the PFRS and from 30 to 18 years for the GRS.

The result of these changes is that the funded ratio for both funds declined substantially. In 2011, PFRS had a funded ratio of 99.9 percent. Under the new assumptions, that declined to 67 percent. In 2011, GRS had a funded ratio of 82.8 percent. Under the new assumptions, that declined to 50.5 percent. The combined funded ratio went from 91.4 percent to 59.0 percent. (Most market participants consider an 80 percent funded ratio to be adequate.)

The aggregate unfunded liability for the plans went from $643.8 million in 2011 to $3.47 billion in 2013.

Why is all of this significant? Creating phantom liabilities distorts recovery percentages. From an economic perspective, taking a cut on an exaggerated liability means that the pension funds actually come out ahead of where they were going into the bankruptcy. This seems to have fooled most of those reporting on Detroit’s bankruptcy, as most reported in surprise that retirees voted to support cutting their benefits. It also seems to have fooled the charitable foundations and policymakers that have been trying to raise money to offset phantom liabilities.

The pension funds come out in an even more advantageous position when one considers that most of the settlements in the bankruptcy involve additional contingent recoveries for pensions.

In many ways, Detroit’s bankruptcy demonstrates that Chapter 9 needs some sort of standard for dealing with pension liabilities because the open-ended and opaque nature of valuation creates non-negligible opportunities to discriminate unfairly among creditors.

It also shows the extent to which Detroit has squandered the opportunity to reform how the city does business. The city has incurred tens of millions of dollars of legal fees and the stigma associated with bankruptcy (magnified by the scope of its cuts to bondholders) just to reinforce the political failures that put the city in Chapter 9 in the first place.