“It’s the dumbest idea I ever heard,” Jon Corzine told Bloomberg.com in 2008 when he was still governor of New Jersey. “It’s speculating the way I would have speculated in my bond position at Goldman Sachs.”
Corzine, who followed up his tenure as governor with a $1.6 billion investment debacle as chairman of MF Global, seemed to know a thing or two about risky ventures. In this case, he was speaking of pension obligation bonds. POBs are a financing maneuver that allows state and local governments to “wipe out” unfunded pension liabilities by borrowing against future tax revenue, then investing the proceeds in equities or other high-yield investments. The idea is that the investments will produce a higher return than the interest rate on the bond, earning money for the pension fund. It’s a gamble, but one that a lot of governments are willing to take when pension portfolio returns plummet, causing unfunded liabilities to run dark and deep.
Almost every fund has faced such liabilities from time to time, though current times have been more treacherous than others. As Paul Cleary, executive director of the Oregon Public Employees Retirement System (PERS) points out, since 1970 his state’s pension fund has suffered annual losses only four times. But three of those losses were in the last decade, and one, in 2008, was a catastrophic 27 percent decline.
Faced with such losses -- and with a dearth of state and local revenue to make up for the shortfalls -- POBs have become a favored tool to fix pension woes. Oregon is a big player in the POB market, along with scores of its cities, counties and school districts. Other major POB issuers include California, Connecticut, Illinois and New Jersey.
The bonds took on some notoriety this past summer when two California cities, Stockton and San Bernardino, went bankrupt. Generous pensions awkwardly propped up with ill-timed POBs contributed to both debacles.
Over the years, returns on POBs have often fallen below the interest rate the state or locality paid to borrow the money, digging the liability hole even deeper. Nonetheless, they remain popular with politicians in a revenue pinch. Politically, it is easier to borrow money to pay for pension costs than it is to squeeze an already-stressed budget. While many economists and policy analysts view them as risky gimmicks and question the high market growth assumptions that make them seem viable, POBs have defenders who believe that with careful timing they can pay off.
When Oakland, Calif., launched the first pension obligation bond in 1985, it appeared to be a reasonable strategy. It qualified as a tax-free bond that could be issued at the lower municipal bond rates. A state or city could then pivot and invest the funds in safe securities -- a corporate bond, for instance -- at a slightly higher rate. “That was classic arbitrage,” Cleary says. “You were locking down the difference between nontaxable bonds and taxable bonds.”
The Tax Reform Act of 1986 ended that strategy by prohibiting state and local governments from reinvesting for profit the money from tax-free bonds. When the concept resurfaced, the strategy called for states or localities to issue a taxable bond and leverage the higher interest rate of that bond against higher return but riskier equity market plays. So long as markets boomed, the new tactic seemed savvy. “Some people call this arbitrage, but it’s not,” Cleary says of post-1986 POBs. “It’s really an investment gamble.”
Arbitrage occurs when prices for the same product differ between two markets, allowing a nimble player to exploit the difference. “Real arbitrage is free money,” says Andrew Biggs, a scholar at the American Enterprise Institute. “But it doesn’t hang around very long.”
Safe bonds and risky equities are not the same product, but public pension accounting currently permits state and localities to treat them as if they were. “They are counting the return on the stocks before the return is there,” Biggs says. “If you borrowed money to invest in the real world, you would factor the current value of the debt with the current real value of the stocks.”
Given the inherent risks and possible rewards, how have POBs fared? In 2010, a research team led by Alicia Munnell, director of the Center for Retirement Research at Boston College, ran some numbers to find out. The team took 2,931 POBs issued by 236 governments through 2009. They used each bond’s repayment schedule to calculate interest and principal, and then clustered them into cohorts based on the year issued. They assumed a 65/35 investment split between equities and bonds and tracked the results with standard indexes. They then produced two composite graphs -- one at the height of the market in 2007 and the second in 2009, after a crash and before recovery.
In general, bonds issued in the early stages of a stock boom performed well prior to the crash. Thus, POBs issued in the early 1990s were healthy, ranging from 2 to 5 percent net growth. Borrowings in 2002 or 2003 also looked good.
Those issued in the latter years of the 1990s or 2000, however, were in negative territory even before the 2008 crash, having suffered serious losses to their principal in the 2001-2002 downturn. After 2008, all POBs were under water -- except those issued in the trough of the collapse, which by 2009 were already pushing 25 percent gains.
Oregon’s numbers mirror Munnell’s findings. Local government POBs issued in 2002 at the depth of that market collapse and managed by Oregon PERS gained an annual average of 8.84 percent through 2012, before principal and interest on the bond. Less lucky were bonds issued in 2005. The Springfield School District’s POB earned just 5.53 percent, for example. Since that bond carried 4.65 percent interest, it likely earned roughly one point annually -- not much, but slightly above neutral. Oregon’s 2007 issuers earned just 2 percent on their investments through 2012, and are upside down today after debt service.
The same fate befell Stockton, Calif., which also came to market in 2007. Similarly, New Jersey issued a $2.8 billion POB in 1997 -- on the wrong side of another stock bubble.
“The whole thing is the timing,” Oregon’s Cleary says. “You are trying to issue them when the market has bottomed out and when interest rates are reasonable, because really what you are doing is making an investment bet. If people thought when they did POBs that they were refinancing a debt or doing a locked-in arbitrage, rather than an investment play, I’m sure they have been very surprised by the results.”
And yet that is exactly how they were sold. When Oregon voted on new POBs in 2009, the voter education pamphlet argument in favor of issuance explicitly framed the choice as a “refinance” and cast the projected returns as money “saved.”
“Just like many homeowners are refinancing their home mortgages,” the pamphlet read, “the State should take advantage of these historically low rates, which can save Oregon more than $1 billion over the next 25 years. The money saved will help reduce cuts and protect services that all Oregonians rely on.”
Because POBs demand headroom between the interest an issuer pays to borrow and the high returns promised on resulting investments, their investment strategies tend to chafe against safer portfolios. Without a hefty “discount rate” -- as the projected annual gain assumed by a pension fund is known -- the pension bonds would not be possible.
In a 2012 paper, Andrew Biggs argues that the aggressive 8 percent discount used by many states overstates likely earnings and understates risks. A fund that required $100 million in 20 years and employed an 8 percent discount rate would be “fully funded” with $21 million, Biggs notes. But if that same fund were to gain only 5 percent annually, it would need $38 million today to be fully funded in 20 years.
Many experts argue that because public pension obligations are legally binding, pension funds should be discounted at close to zero risk on the front end -- at or near the rates offered by government bonds. “While economists are famous for disagreeing with each other on virtually every conceivable issue,” wrote then-Federal Reserve Board Vice Chairman Donald Kohn in 2008, “when it comes to this one there is no professional disagreement: The only appropriate way to calculate the present value of a very-low-risk liability is to use a very-low-risk discount rate.”
In point of fact, the 8 percent discount rate may be on its way out. The Governmental Accounting Standards Board (GASB) is launching a complex hybrid discount standard in 2014, which will affect the assumptions states make with their funds. Some fear the GASB rule will only create more confusion. Bond rater Moody’s is taking a simpler tack in weighing government pension plans, having recently proposed to shift its pension discount rate down to the level of AA taxable bonds, which are now at 5.5 percent. “Currently, discount rates used by state and local governments are all over the place,” says Tim Blake, Moody’s managing director of public finance. “Most are in the range of 7.5 to 8 percent. We need a uniform rate.”
Not surprisingly, 5.5 percent is very close to the rate at which many POBs are sold to investors.
With aggressive 8 percent discount rates now under attack by economists, oversight boards and rating agencies, issuers who counted on rosier outcomes have learned some hard lessons. Five years ago, when Connecticut State Treasurer Denise L. Nappier announced a new $2.28 billion pension bond, she noted that the state had “achieved a favorable borrowing cost of 5.88 percent, which is well below the 8.5 percent assumed long-term return on assets of the Teachers’ Retirement Fund. This will provide significant cash flow savings over the long term and a potential savings to taxpayers of billions of dollars.”
When the bond was issued in April, the Dow Jones average stood just shy of 13,000. By November, the market was in free fall. It bottomed out the following March at just over 6,600. Connecticut’s timing could hardly have been worse. As the market plunged, Pensions & Investments lit into POBs, singling out Connecticut. The editors argued that POBs shove obligations “that should have been paid as earned” onto future generations, along with the risk of the debt.
By 2010, with the market still emerging from the trough, Connecticut’s finances were as messy as ever. But now there was little appetite for more bonds. POBs “are certainly a risky proposition,” Michael J. Cicchetti, chairman of Connecticut’s Post Employment Benefits Commission, told the CT Mirror. “Things are different now than they were then.”
Even as Connecticut licked its wounds, Boulder, Colo., was launching its own new POB, designed to clean up some tangled threads from pension programs long since closed to new employees. With the market down, Boulder’s CFO Bob Eichem was ready to take a chance. “POBs are not for the faint of heart,” Eichem says. “You have to understand them.”
Boulder’s bond, for a tame $9 million, was issued into a recovering market, and the city got a low 4.29 percent interest rate. Boulder also ran worst-case scenarios and decided it could absorb any likely loss.
“You’ve got to pay that debt every year,” Eichem says. “So if the market goes against you, you will have to decrease your other expenditures to make up for what the market did to you.”
With the bond money in hand, Boulder staged its investment. “There was great uncertainty in the market, as there still is,” Eichem says. The city put the funds in a holding account and “income averaged” them into the market over a year.
“That would hold you down if the market really rose in that year, but it would also protect you if the market really dropped,” Eichem says of the gradual strategy. After stringing out three quarters of the fund through the year, Eichem put the remainder in when the market bottomed in early 2011.
“Issuing a POB may allow well heeled governments to gamble on the spread between interest rate costs and asset returns or to avoid raising taxes during a recession,” Alicia Munnell and her team wrote in 2010, warning that “most often POB issuers are fiscally stressed and in a poor position to shoulder the investment risk.”
“Good for Boulder,” says Munnell when told of how Eichem had gauged its risk. What would she tell others set on a similar path? “You better have deep pockets. It’s gambling, and you have to be able to absorb any losses.”
Munnell is less approving of Oakland. In 1997, San Francisco’s cross-Bay neighbor followed up its 1985 invention of the POB with a $417 million pension bond, designed to buy the city a 15-year “holiday” from its police and fire pension contributions.
The timing was poor. As noted earlier, bonds issued in 1997 were, on average, underwater in 2007, even before the stock market crash. And so was Oakland’s. In 2010 Oakland’s city auditor did a careful analysis of the 1997 POB and found that “the amount still owed by the City is approximately $250 million higher than the scenario where the POBs were not issued in 1997 and the same payments were made to the pension fund instead.”
With the “pension holiday” over, Oakland’s City Council approved a new $210 million POB to be paid off in 2026 -- once again borrowing money -- cash that it would have had in hand had it not borrowed the last time.
And so, back where it all began, the POB wheel takes another spin. As columnist Len Raphael wrote in Oakland North, a local online newspaper, “That’s like a compulsive gambler telling you that he has to bet it all on red to make up for his past losses.”