Top 12 Pension and Benefits Plan Issues for 2009: Part I
New paradigms for a new year
Market events in the past year have changed the public pensions and benefits landscape for years to come. Plan administrators, benefits professionals and financial officials must re-think almost every basic assumption they have made about how to fund, finance, invest and manage the trillions of dollars of public capital entrusted to their employees' benefits plans.
Here are a dozen timely observations and fresh thoughts to consider and act upon in the coming year. They cover many specific economic topics including several of particular concern to pension managers -- and those who have to fund the plans. Each of them is a mini-column of its own. Therefore, we have divided the full article into two parts, and this Web column contains the index and Part I. Click here for Part II.
A warning: If you are an ideologue or a partisan, I suspect that one or more of these ideas will disturb you, because I prefer to confront realities than to play politics or live on theory. (Perhaps the only people who will love this column will be unpublished academics and doctoral students looking for a real-world topic to research further.) When I claim to introduce new paradigms, I'm quite serious.
This column runs longer than usual, and some readers will only be interested in certain topics, so I have numbered them for the skimmers and speed-readers. Click on the topics below to jump to that section. You can also read an "executive summary" version that capsulizes all of my key points.
2009's Top 12 Issues
1. Financial economics and MVL: rest in peace? In 2008, the hottest topic in the actuarial world was "financial economics," a school of thought advocating that public pension funds should value "at-the-market" both their assets (MVA) and their liabilities (MVL). Further, those liabilities were to be market-valued using a risk-free interest rate because the current practice of assuming long-term diversified investment returns includes a risk premium.
Well, guess what? The financial economists were indirectly right about how grossly public pension plans have underestimated their unfunded liabilities, although it was the funds' asset losses in the market (not the "risky" discount rate) that proved the case of the FE school. The devastating 2008 stock market supported their arguments for sure.
However, their success intellectually in 2008's market meltdown has now dug the FE camp into their graves. Not even the staunchest advocates of FE:MVL dare to press their case in the real world of 2009, because the glaring practical pitfall of their logic has been exposed in spades through this once-in-a-generation flight to quality which has driven the interest rates on risk-free U.S. Treasury bonds to 3 percent. Can anybody who lives in the real world actually expect pension plans to now discount those future liabilities at 3 percent?
At a risk-free 3 percent, the theoretical liabilities of public pension plans are nearly double what they have been estimated to be under traditional conventions. Does it now make sense to double the liabilities when equities have lost 40 percent? That's a double whammy and penalizes the plans on both sides of the actuarial balance sheet. If that weren't bad enough already, the FE purists would further suggest that actuaries should immediately value those assets and liabilities at market, without smoothing, so that taxpayers could "enjoy the benefits" of funding pensions under their theory. Never mind that this implies letting houses burn down and criminals rob communities blind, as layoffs of vital public services become necessary in order to feed the beast of Market Value.
Adding insult to injury, the application of these academic theories today would almost assuredly generate pension surpluses in coming years. And if you look down the chessboard at least five moves, those surpluses will inure entirely to future generations of public employees at the expense of today's taxpayers. In one of my research pieces last year, I noted that there are very few periods in which the rolling 20- and 30-year returns of stocks fail to exceed corporate bonds -- and even fewer when risk-free bonds are the benchmark. Therein lies the fallacy of FE:MVL. The risk-free rates flunk the practicality test and the test of time as a basis for allocating costs between generations of taxpayers. Worse, it invites another round of giveaway benefits increases once the investment portfolios ultimately outperform the risk-free rate, which is highly likely.
2. Pension funding fallout from the market meltdown: the unlisted price of employees' perpetual straddle options. From a peak level in 2007, domestic stock markets have declined by 38 percent through 2008, and the average public pension fund's funding ratio has dropped from roughly 85 percent to less than 65 percent. This doubling of the unfunded pension liabilities will result in a doubling of the annual required amortization of these obligations. For most public plans, this will require employers to raise contributions by 2 to 4 percent in the next year or two, depending on how quickly their smoothing convention requires them to face reality.
This increase in employer contributions is the second round in a single decade. It reflects the true cost of a foolhardy "straddle option" mistakenly granted to many public employees and retirees -- which I will explain below. When markets are good, benefits are increased, and when markets are bad, the taxpayers get the bill. It's a vicious cycle.
To see how badly this works, here's some historical data in a chart format, straight from the highly respected 2007 NASRA public funds survey. Analysis of the previous five years' contribution rates show that taxpayers bore the brunt of surging unfunded liabilities -- not employees or retirees. As the chart below shows, employer contributions absorbed the entire increase in costs that resulted from benefits increases awarded in the bubble market years of 1999-2000. From 2002 to 2006, employer contributions increased 40 percent from 6 percent to 8.5 percent of payroll nationally, while employee contributions were unchanged at 5 percent for those in plans that are integrated with Social Security. Results were similar but not as dramatic for plans outside of Social Security because they started at a higher base under their unique "opt-out" structure.
Graphic: NASRA Public Funds Survey 2007
Note: Red commentary above by Girard Miller is his opinion only. It is not representative of NASRA or its staff.
I'll bet even money that the results of the 2008 market rout, which has doubled the unfunded liabilities of public pension plans, will produce equally dismal results in the coming three to five years. Unless the stock market magically recovers its losses in 2009, employer contributions are likely to increase another 2 to 4 percent of payroll. Without a major market rally, there is a good chance that by 2012, many employers' rates will have doubled over 10 years!
Some pension advocates cite actuarial smoothing as the profession's great gift to taxpayers because it enables budgeters to defer the inevitable problem and hopefully wait out the recession. I think this misses the point. This isn't just about cash flow. Pension contribution holidays can accomplish the same result, and we all know where that takes us -- into deeper deficits.
On the positive side, state and local pension plans remain far better funded than the national retirement systems (Social Security and Medicare), and most public pension systems at least have a methodology in place to bring themselves toward actuarial balance. But there now is a cumulative trillion-dollar cost to taxpayers for that protection, and the pension community must face up to its magnitude and address it. Unfortunately, the prevailing attitude to date has been something akin to, "Oopsie, let's change the subject and talk about the things we do well." Nobody in the public pension world seems willing to address an obvious, major structural problem in the prevailing funding model.
The gaping real-time underfunding of public pensions today bespeaks a deep-seated and inefficient market option biased entirely in favor of pension plan beneficiaries at the expense of taxpayers. Here's the payoff matrix of this straddle option, which has been granted to public employees for free (for lay readers, a straddle is a call [up] and a put [down] option combined, which pays the holder whether prices go up or down):
· Heads, Employees Win : Almost every time capital markets outperform and produce extraordinary funding ratios, benefits have been increased -- or even worse, increased retroactively with constitutional protections thereafter. The pension benefits mania of Y2K illustrated this clearly.
· Tails, Taxpayers Lose : When markets underperform, costs are borne by taxpayers, not the employees, through actuarial amortization and increased employer contributions. (See chart above.)
If this is what pension advocates implicitly mean by the term "retirement security," then something is dreadfully wrong with that terminology. On its face, it is terrible public policy. I will quantify this straddle option and its inherent, undeserved cost to taxpayers in Topic #6 below.
3. Sustainability of retirement plan designs: OPEB ostriches. The coming increases in pension contribution requirements will vex public employers whose tax revenues are shrinking with the recession, at the same time they begin to confront their true actuarial costs for retiree medical benefits ("OPEB" plans). For governments that have promised employees and retirees a medical benefit at full cost upon retirement, the typical increase in combined employer contributions can easily approach 10 percent of payroll -- for decades to come!
Most public employers will never actually be able to afford such annual contributions going forward, yet they continue to "pay-as-they go" and thus act like ostriches, hoping that somebody in the future will magically fix the problem if they bury their heads in the sand. Perhaps today's elected leaders are simply cynical and choose to leave these unpaid bills to their successors and future taxpayers.
Some public employers are clinging to the false hope that the Obama administration's national health care plans will magically bail them out of their OPEB deficits. That seems highly unlikely. First, our Medicare system is already a financial train wreck and has no money available to underwrite benefits for state and local workers. The latest Medicare trustees actuarial report says we need to raise Medicare taxes by 3 percent of all U.S. payroll (split between employers and employees), just to make the current system actuarially viable. This pre-existing deficiency virtually rules out a national retiree health care plan of magnitude sufficient to relieve state and local OPEBs.
The smart money in Washington is betting that the Democrats will build upon and work around the existing system. They learned an important lesson about health-care economics during the previously unsuccessful Hillarycare tempest. Why -- and how? -- would the national government take on another $1.5 trillion liability that is constitutionally the responsibility of state and local governments as sovereign employers under a federal system? Especially when the administration's foremost policy focus is to provide access for the millions of uninsured Americans under the age of 65? Nonetheless, there are some who will use the unaffordable dream of "costless" universal health care as an excuse for inaction. They are the true OPEB ostriches.
As suggested in a prior column, prudent administrators and policymakers would be well advised to undertake a sustainability audit of their overall retirement benefits plan design to determine whether they are sitting on a time bomb. When the harsh new facts of retirement plan finance at the individual employer level are put on the table for all to see, the opportunity and the resolve to fix the problem may eventually follow. Each employer will need to decide whether the path to progress involves more funding at a higher contribution cost to employees and taxpayers, or reduced benefits and a shifting of the future point-of-service cost burden to employees.
4. Crocodile "tiers" for new employees. Based on my consulting experiences in 2008, most employers confronting the sustainability issue will first try to take the easiest path to reducing long-term costs. They will "tier" their benefits plan -- especially OPEB and perhaps also the pension plan -- to provide a lower benefits structure or a cost-shift to new employees. Realizing that labor unions are prone to "sell their unborn" in order to preserve benefits for incumbent employees, such tiering arrangements have been most common in the past.
That's a commendable first step, but in many cases, it's insufficient. The cost savings that result will do very little if anything to control or reduce costs in the immediate future. Ironically, the tier strategy cuts benefits for new employees in an effort to save costs that were incurred by incumbent employees. It's another example of public benefits systems dumping their costs on future generations.
The long-term problem with tiered reductions in defined benefits plans is that they often are reversed later. Politicians and chief administrators come and go, but unions hang around forever. So they just wait until amnesia sets in sometime down the road, and regain the original benefits in future bargaining. It's what the great economist John Maynard Keynes and his intellectual successor John Kenneth Galbraith called the "ratchet effect" of wage and benefits bargains which are "sticky downward." They go up, but seldom ever down for long -- regardless of market forces in the real world.
The only tiering arrangement that actually rules out a future reversal is a defined contribution arrangement. Then it's much more difficult for the unions to cut a deal with future politicians for retroactive benefits increases. This is where I part company with the pensionites who think their favorite plan design is heaven's only gift to mankind. They fail to acknowledge the costly chinks in the armor of conventional DB plans in a world of public-sector unionism and short-term politicians.
Don't get me wrong: I would never throw out the pension baby with the bath water, but a half-and-half DB-DC plan for new employees, along the lines of the Washington State retirement plan model, makes much more sense to me as a long-term, sustainable plan design for the many public employers who will increasingly find themselves unable to afford their present plan design. That half-and-half design shares the market risks and rewards equally between employees and employers -- which seems imminently fair to me. Those who are considering a new benefits tier should at least consider this design feature.
Further, the only tiering arrangement that materially reduces current services costs immediately is one that affects incumbent employees. This is difficult to achieve with pension funds, even though most constitutional protections for earned benefits do not actually apply to future benefits -- it's mostly a labor-relations challenge. But for OPEB benefits which lack constitutional protections and are much "softer" obligations, it makes great sense for public officials to bargain much more intensely to reduce the prospective benefits formulas for incumbent employees as a way to reduce current operating costs. One solution is to require a new, increased or matching employee contribution. (That remains a viable option for pension funds, incidentally.) Another is to put a CPI cap on future OPEB benefits increases. A third is to reduce dependent coverage to an actuarial equivalent, similar to well-designed pension plans. And if benefits are untouchable, then a permanent contribution increase for incumbent employees as mentioned above really must be put on the table.
5. POBs and OPEB-OBs: the 2009 "benefits bonds" window. If there is a single silver lining in the demise of the financial markets during this recession, it is the upcoming opportunity for certain high-credit-quality states and municipalities to issue taxable bonds to refinance their pension and OPEB plans. Right now the taxable municipal bond market is frozen up, so the potential for this strategy to be deployed successfully must await an improvement in the credit markets as well as a bottoming of the recession. But later this year, some public finance experts expect a "window" to open for pension obligation bonds (POBs) and OPEB obligation bonds (OPEB-OBs). Here's the picture:
Graphic: PFM Group
Historical studies cited in my previous columns show that the 20- and 30-year prospects of success for this strategy of selling bonds to invest long-term in stocks have become much more promising as the stock market meltdown has taken on historic magnitude. Perhaps the most intriguing question is not whether this window will open, but rather how long it will last: If the stock market rallies strongly in anticipation of an economic recovery, the opportunity to invest in stocks at discount prices (after selling bonds to raise necessary capital) could evaporate. Therefore, the smarter finance officials and plan administrators are starting to develop their strategies and authorization requests in the first half of 2009, and some are working on this strategy already.
The traditional model for POBs was driven by naïve debt managers who relied on their advisers. Apparently none of the paid experts ever bothered to talk with the investment people or to think about asset-liability management and the consequences of their strategies. Sadly, many of these folks never saw a POB they didn't like. Such POB myopia has resulted in untimely and bloated transactions that have called unfavorable attention to the entire strategy. Fortunately, there is a better way. Public employers now have available to them a completely new paradigm for "Benefits Bonds," which include both POBs and OPEB-OBs.
First, benefits bonds should only be issued during recessions or during the early stages of economic recovery, when stock prices are depressed. (See graphic above.) The entire strategy is dependent on stock market returns exceeding taxable bond borrowing costs, and those who sell benefits bonds after the economy has recovered are likely to experience net investment losses in the next recession -- especially when borrowing costs are considered. One need only study the sad experience of previous untimely issues to see this simple fact. Financial advisers who hang their hat on naïve Monte Carlo models are using static analysis that ignores the power of the business cycle and associated bear markets. A blind monkey can push the button to run a Monte Carlo analysis to show, statistically, that POBs should work in the long run. But that monkey can't help a debt manager who loses face or gets fired in the next recession for being dumb enough to rely on that sole metric.
Second, it is foolhardy to fund more than 80 to 85 percent of the total plan liabilities through a debt issue. In the past, POB issuers often were enticed to bond for "full funding" by issuing debt equal to the entire unfunded liability. This naïveté simply guaranteed that once the pension portfolio achieved the stock market returns it sought, the plan's funding ratio exceeded 100 percent -- and the unions and retiree groups were back at the table demanding benefits improvements which then put the plan right back into an unfunded position in the next recession. Bonding more than absolutely necessary invites a vicious cycle.
Third, it makes no sense to sell taxable bonds to buy bonds. A benefits bond issue generally should only fund investments in equities, which historically earn more than the costs of paying off the taxable bonds. To sell bonds and fund a pension or OPEB portfolio that invests a portion in bonds is simply selling bonds to buy bonds, with little or no profit whatsoever especially after considering the costs of issuance, advisor's fees, bond counsel fees and investment managers' fees. Any profit that does result is probably a result of credit risks taken in the bond portfolio. Certainly it's pure lunacy to sell bonds to buy riskier fixed income securities (as in the Wisconsin schools featured in my prior column) or to buy the bonds of neighboring jurisdictions without appropriate diversification (as some financial and investment advisors in the Midwest have recently suggested).
Fourth, the bond proceeds should be governed by a new trust document in order to assure that bondholders and taxpayers are protected. In fact, a POB trust can correct one of the defects of the pension funds' inherent straddle option cited in Topic #2 above. My companion column on this topic explains this concept further. Finally, the possibility of a double-dip recession like 1980-82, or a replay of 1932-33's whipsaw markets, suggests that most issuers should break their bond issues into two parts and enter the market twice rather than betting the ranch on a single market point of entry.
6. DB vs DC: a false dichotomy and a true cost. My column last month on The Wrong Way and the Right Way to Cut Benefits apparently confused a few readers who entirely missed my point regarding the cost features of a lower-contribution defined contribution (DC) plan for new employees. What escaped them was the rationale for employers making a lower contribution for new hires into a DC plan than they presently pay into the pension plan: because that's all they can afford. Several asserted that defined benefit (DB) plans are always better than DC plans because they have lower investment costs and earn better returns than individual investors.
That got me to thinking more deeply about that claim, and here's what I came to realize: The arguments advanced by pension fund DB advocates in the past several years regarding superior investment returns over DC plans are founded on a false dichotomy. There is nothing whatsoever inherent about a defined benefit plan that produces higher returns than a defined contribution plan. The superior investment performance reported in the studies results instead from collective vs. individual investments, which is not a function of defining benefits vs. defining contributions. That's a distinction that makes a difference.
To make this point clearly, let's take a good hard look at what is known as a Collective Defined Contribution plan or trust. Under such an arrangement, an employer (and sometimes employees) makes ongoing defined contributions to a benefits trust fund governed by trustees who engage an actuary to calculate the future benefits that can sustainably be paid from the trust assets assuming that the contributions remain the same. If the fund has insufficient resources to pay all the promised bills, the plan's trustees adjust the benefits. The trustees invest the money just like a pension fund - collectively. Individuals make no investment decisions. Guess what? The investment returns are identical to a similarly sized pension plan (a DB plan). Thus, there is no inherent investment efficiency in DB plans over DC plans.
In fact, many defined benefit plans have suffered a fatal inherent investment inefficiency: They have granted a perpetual straddle option to employees that is a wickedly costly investment for taxpayers ("heads, I win; tails, you lose"). This straddle option is priced at zero by DB plans -- yet its actual cost is immense. Certainly these perpetual options cost far more than any fees paid to mutual funds vs. pension money managers. In fact, they carry a cost far greater than all the investment inefficiencies of DC plans cited in the pension advocacy research.
What's the intrinsic value of the employees' perpetual straddle option? Ask yourself this: How much would you pay to an insurance company to guarantee that your IRA would always return 8 percent (the average pension discount rate), yet still give you all the upside if the markets do better than that? Your answer will depend on your risk tolerance, but I am confident that it's far more than the fractional difference in fees between mutual funds and institutional money management!
Just for the record, there is not a long-term straddle available anywhere in the world with at-the-market exercise values that has a price of zero, which is how the pension world has valued these. And the greater the volatility, as in last year's market, the greater the value of the straddle option. So what I'm explaining here is a really, really big deal.
Theoretically, this option's intrinsic price to a risk-conscious investor should be the present value of the difference between the pension fund's current discount rate and the risk free rate of return. Remember those pesky Financial Economists? Here's where their insights have policy value. The risk-free rate today is 3 percent, so the straddle's intrinsic put-option value of the DB guarantee today could be as high as the present value of 5 percent annually on today's $2 trillion held in pension portfolios, compounded annually. Plus, we also need to add the straddle's call-option value of potential market appreciation above 8 percent compounded, which has historically been given away to employees in the form of benefits increases, and not the underwriting taxpayers who bear all risks in this totally asymmetrical arrangement. Then, as a technical formality, we need to discount the option for the holder's tolerance for risk (which is relatively low for public employees).
A second way to evaluate this arrangement is to ask what would be the cost-at-market to buy a swap from a AAA-rated counterparty to trade the returns of a diversified pension portfolio for a fixed interest rate of 8 percent over 30 years. Once you think about it, that swap would require a gigantic premium. Just for the fun of it, ask your chief investment officer what that swap would cost your plan. As a third point of reference, a pension portfolio could be swapped with no premium for a fixed-rate contract -- but at an interest rate that would be well below 8 percent.
Hopefully by now you get my point here. Whichever way you value this inefficient straddle, its real-world cost nationally clearly totals 11 figures (i.e., tens of billions) annually!
The number would be smaller if employee groups were willing to share the financial risk for market underperformance through matching contributions for unfunded liabilities -- but that suggestion would be laughed out of the room in collective bargaining. The taxpayers hold the bag, although they rarely voted this fate upon themselves. It's all been done to them by well-meaning but naïve pension officials, short-sighted politicians and self-interested labor groups.
Corporate America has largely abandoned DB plans precisely because of this dynamic, and their payoff matrix was only a put option. At least in their case, an overfunded plan could be valued as a corporate asset (a call option for the employer) -- unlike the public sector, where market appreciation typically benefits employees, not taxpayers, in far too many cases to be dismissed as exceptions rather than the rule.
A close cousin of these under-priced options are the DROPs (deferred retirement option plans) concocted in some states and localities that pay employees interest at the pension fund's assumed rate of return in their individual DROP accounts, regardless of actual returns. Talk about investment inefficiency. First of all, the average life of a DROP account is far shorter than the infinite investment horizon of a pension fund and therefore should receive a lower rate of return simply as a matter of bond and capital markets math. Second, the employee is again receiving a risk-free investment at above-market rates, and shifting all investment risk to the employer. (Again, the variable-to-fixed swap concept outlined above provides insight into the fair pricing of DROP interest credits.) Most of these arrangements have lost money for the pension funds, especially in light of actual capital market returns since the inception of DROP plans. Reforms are clearly needed in many of these plans as explained in one of my original columns in 2007.
Of course, if the debatable issue is strictly between pension plans and individually managed (that is, participant-directed) 401 plans, then the past comparisons made by the pension plan advocates are fair. But that obviously does not need to be the only choice available to plan sponsors in the future. The collective defined contribution plan model deserves much more attention than it has received to date.
Now, before leaving this topic, let me be clear that I am not suggesting that employers eliminate public pensions when sustainable. That's not my point. Without public pensions, I dread to think what faithful public servants my age or older would do for their retirement after last year's market meltdown. I strongly favor index funds and professionally designed target date funds and other ways to improve the efficiency of participant-directed DC plans. What we need to do is find the best aspects of all available systems and better manage their drawbacks.