Pensions Rescued by Borrowing: What Could Go Wrong with That?

A Tale of Two Cities
 
The dominant form of pensions for public sector workers are defined benefit plans. Such plans typically specify a retirement age and a benefit formula that determines the yearly payment to retirees. The benefit formulas are based on the number of years of employment, the average of their last few years’ of earnings, and a conversion factor, with certain adjustments based on family status. The pension plan collects contributions from the public sector employer and employees, invests them in equity and bond markets, and uses the proceeds to pay the retirement benefits.
 
For any defined benefit pension plan, actuaries can provide estimates of the level of total accrued benefits in the pension plan at any given time.  Adequately funded plans have assets in excess or equal to the plans’ accrued benefits.  However, for the reasons outlined below, state and local government pension funds are universally plagued by the problem of unfunded liabilities.  That is, such pensions have a large gap between accrued benefits and the assets set aside to pay those benefits.
 
Consider the examples of Houston and Chicago, the fourth and third largest cities in the U.S.  At the end of 2016, Houston’s three public pension funds collectively had unfunded liabilities of $8.2 billion. The total pension debt for Chicago’s public pension funds was $28 billion based on the city’s 2017 financial report.  Under Mayor Sylvester Turner, Houston completed a pension reform in 2017 that included, along with some natural debt-reducing measures such as benefit cuts and contribution hikes, the issue of $1 billion of pension obligation bonds with varying maturities.  The economic rationale behind the bond issue was that the pension funds can invest the borrowed money, which has an average interest rate of 3.97%, in equities that would deliver an expected rate of return of 7.25%.  Therefore, as the argument goes, the bond issue literally pays for itself without additional taxes, due to the expected net earnings of 3.28% (7.25%-3.97%).  These net earnings will then help reduce the size of the unfunded pension liabilities.
 
Faced with an even larger pension debt, and apparently less willing to make benefit cuts and contribution hikes than their Houston counterparts, Chicago’s Mayor Rahm Emanuel and his team are currently considering issuing $10 billion in pension obligation bonds as the centerpiece of a pension reform that would reduce the city’s public pension unfunded liability.  
 
Can We Borrow Our Way Out of the Public Pension Debt Crisis?  
 
Of course, the expected return on riskier assets, such as equities, is higher than the expected return on less risky assets such as government bonds.  That is why Houston could borrow at a relatively low interest rate and at the same time expect to earn a considerably higher expected return on the borrowed money by investing it in the equity market.  When an individual investor does this, it is called “buying on margin.”
 
However, as the saying goes, there is no free lunch.  The excess of the expected equity rate of return over the interest rate on borrowed funds is the risk premium that the market pays to those who bear the additional risk involved in equities.  Investing $1 billion of borrowed money in equities increases the risk in the investment portfolio of Houston’s pension funds. This buying on margin increases the probability of higher pension fund earnings, but it also increases the probability of lower pension fund earnings.  That is the nature of increased risk. Buying on margin is risky.
 
The taxpayers in Houston are the ultimate source of payment for the city’s unfunded pension liabilities.  Suppose that the taxpayers’ optimal investment portfolio contains both equities and bonds.  The city’s buying on margin reduces the bond component and at the same time increases the equity component in the taxpayers’ portfolio.  To revert to the previous optimal portfolio composition, therefore, the taxpayers would have to increase bond holdings and reduce their personal equity holdings. In principle, this could completely offset whatever effects the city’s buying on margin might have on taxpayers.   
 
Of course, there is nothing intrinsically wrong with borrowing to invest in equities.  However, the belief that public pension financing can be magically improved by taking advantage of the gap between the expected rates of return in equities and bonds is na├»ve at best and dangerous at worst. Pension funds know, and should acknowledge, that buying on margin is a risky investment strategy. The idea of having a government fund use leverage to finance risky positions in order to boost returns is certainly not new.  One spectacular example of a government losing on such a gamble is provided by the “Orange County California bankruptcy” in 1994.  Orange County funds were highly leveraged, with the borrowed funds invested in risky positions that generated high returns for several years.  When the market moved against the funds’ positions, the funds received margin calls that they could not meet, and eventually declared bankruptcy.
 
An economically identical issue came up in the late 1990s in the debate on whether the Social Security trust fund balances should be invested in the then-booming stock market. Proponents argued that this would lead to the trust fund earning above-average stock returns rather than being used to finance federal budget deficits earning about 3%, the average rate of return on long-term treasuries.  Economists studying this issue concluded that such a move would increase risk and would not generate any first-order benefits when the additional risk was properly accounted for.  The federal government is able to borrow at even lower interest rates than state and local governments can, for obvious reasons.  If a scheme of borrowing at relatively low interest rates to invest in high-yield and high-risk equities would not considerably improve the financing of Social Security, it is even less likely to improve the financing of local government pension funds.
 
This is not to say that a pension fund – or Houston – cannot win when it gambles.  Apparently, Houston’s gamble has been a success to date.  But it was and still is a gamble, and winning a gamble one time does not guarantee similar results on other gambles.  
 
 
Why Are Unfunded Pension Liabilities a Universal Problem?
 
Unfunded liabilities are a near-universal problem for government pensions.  In the U.S. alone, the total state and local government pension debts are on the order of trillions of dollars.  Yet these pension debts of subnational governments are dwarfed by the unfunded liabilities in the Social Security (and Medicare) trust funds at the federal level.  Indeed, unfunded public pension liabilities were a major reason for the bankruptcy of several local governments in the U.S.  In 2013, Detroit went bankrupt due to a total of $18 billion of debts and unfunded pension liabilities, of which the unfunded pension liabilities were about 50%.  In 2017, Puerto Rico filed for bankruptcy with a $123 billion of bond debts and unfunded pension liabilities, of which the latter’s share was about 40%. 
 
Why are defined benefit public pension plans so prone to the problem of underfunding?  Politics has a lot to do with it.  Unfunded public pension liabilities are essentially government debts, and politicians’ incentive to resort to debt financing is easy to understand: pleasing current voters with spending (benefit) increases and tax (contribution) cuts while shifting the financing costs to future generations who do not have adequate representation in a current election.        
 
In addition, defined benefit pension plans, whether they are public pension plans or private ones, seem to have an intrinsic defect that is related to the stochastic nature of investment returns.  After a period of good years with superb fund investment performance, an otherwise adequately funded pension plan would appear overfunded, trigging benefit hikes and/or contribution cuts.  However, the stochastic nature of investment returns suggests that a period of bad years is bound to follow sooner or later, revealing the unfunded liabilities caused by the earlier premature benefit hikes and/or contribution cuts.
 
The origin of Houston’s unfunded pension liabilities helps illustrate this point.  Floated by the bull market of the 1990s, Houston adopted in 2001 massive benefit increases and contribution decreases in its pension plans simply because they were “affordable”.  The financially irresponsible move quickly turned sour amid the dot.com bust which began in 2000 and deepened in 2001.  The situation was made even worse by the decline in asset values during the Financial Crisis and Great Recession.
     
This defect of defined benefit pension plans was very much behind the transition of private pension plans from defined benefit pension plans to defined contribution plans, especially because the bankruptcies of private companies seem to be a regular market phenomenon.  On the other hand, pension plans in the public sector remain predominantly of the defined benefit type, perhaps due to the public’s perception that government agencies last “forever” and find it more difficult to renege on pension promises. 
 
Finally, and quantitatively most important, is the fact that state and local pensions have historically been allowed to discount the future payments of accrued benefits at a rate equal to the return on the plans’ assets. In practice, this meant that the nominal discount rate was in the range of 7% to 8%.  Discounting future obligations at such high rates makes their current or present value smaller than what might be appropriate on economic grounds.  Recently, the use of such high discount rates has been challenged on the grounds that government pension promises are similar to government promises to pay interest and principal on bond debt, and should be discounted at those rates.  The argument is that pension promises are hard for governments to avoid due to the treatment of these promises by the courts and due to the difficulty governments have politically in reducing such promised payments.  Thus, pension promises are regarded as relatively low risk to pension recipients, and low risk promises of future payments are appropriately discounted at lower rates.  With this understanding of the more appropriate lower discount rate, Stanford University Professor Joshua Rauh found that, in 2015, the unfunded state and local pensions’ liabilities based on the lower government borrowing rate was not $1.38 trillion, already huge, but actually almost triple that value, or $3.85 trillion (Hidden Debt, Hidden Deficits: 2017 Edition).
 
After much debate, the Government Accounting Standard Board (GASB) issued new guidelines on the discount rate pensions use.  The new guidelines produce a weighted discount rate that falls between the rate of return on the pension assets and the government borrowing rate. Pensions are allowed to discount the funded portion of their liability at the rate of return on its assets and the unfunded portion at the government borrowing rate. The new blended rates produce a value of liability between that calculated by Professor Rauh and that calculated using the rate of return on pension fund assets. 
 
Solutions to the Public Pension Debt Crisis

Borrowing at relatively low rates and investing the borrowed money in high-yield yet high-risk equities is no magic solution to the wide-spread public pension debt crisis.  It is a gamble that might or might not solve the problem depending on chance and the realization of investment returns.  What are some other possible solutions to the crisis?

First, the mathematics behind unfunded pension liabilities is simple: over-promising and under-funding.  A straightforward solution is to cut benefits to match funding levels (which could include increasing the retirement age) and/or raise contributions from government employers and employees to match pension promises.  Of course, increasing funding from government employers is essentially a call for increased taxes.  In any case, it is better to cut benefits and/or raise contributions sooner rather than later, so that the costs can be more evenly spread over a longer period of time.

Second, this fundamentally political problem calls for a political solution.  Since taxpayers will eventually foot the bill for public pension liabilities, it is only fair, and incentive compatible, to subject any major changes to pensions including benefit increases, contribution decreases, and issues of pension obligation bonds – to voter approval.  Houston has set a good example in this regard by seeking voter approval on its issue of $1 billion pension obligation bonds.  If the city officials in Chicago eventually choose to rescue its under-funded pension funds by borrowing, Chicago’s residents should demand a vote on this matter.

Third, the public sector should seriously consider switching to defined contribution pension plans as a long-term solution to the repeated problem of public pension debts. This would follow the trend in the private sector.  Indeed, a highly touted “corridor mechanism” in Houston’s pension reform is exactly the mechanism to switch from the current defined benefit plan to a hybrid plan in which employees share some downside risk with the taxpayers in the case equity investment fails to produce the expected returns. 

The sheer scale of this country’s public pension debts, and in particular the bankruptcy experience of Detroit and Puerto Rico, have convincingly established an unfortunate fact: government agencies are not necessarily more fiscally responsible than private companies. In fact, quite the contrary may be true.  Further, taxpayers may balk at the demands of ever-higher taxes to pay for underfunded liabilities due to promises made in the past. It is now high time to start the transition to defined contribution pension plans in the public sector, giving employees the ownership on their own retirement accounts and getting politics out of the way once and for all.     

Posted: October 02, 2018 by Dennis W. Jansen, Liqun Liu, Andrew J. Rettenmaier