Public Pensions Roll the Dice

By Andrew G. Biggs

How have public-sector pensions responded to the financial crisis? By doubling down, thus jeopardizing taxpayers.

Pensions for state and local government employees lost 27 percent of the value of their investments from 2007 to 2008 and today assets are still below 2007 levels, while liabilities have continued to rise. In part, these investment losses stemmed from a decades-long trend toward riskier investments, which began with a shift toward equities in the 1980s, and today toward so-called alternative investments such as private equity and hedge funds. As a result of investment losses, taxpayers face larger pension contributions, other government programs are starved of resources, and some public employees may face benefit cuts.

How have these losses affected pensions’ investment strategies going forward? Did the financial meltdown result in public-sector pensions backing off these high-risk investment strategies, or have they doubled down in an attempt to make up their losses?

As part of a recent Wharton School/Pension Research Council conference, I analyzed the target investment portfolios of 30 large public pension plans holding over half of total pension assets. Unlike day-to-day portfolios, which can appear low-risk simply because all the high-risk assets lost value, target portfolios indicate the broad direction plan managers wish to take. Differences in target asset allocations from 2007 through 2010 provide insights into changing pension investment strategies.

By taking more risk, pensions impose a contingent liability on taxpayers to bail out pension funds if investments fall short.

In addition to investment losses, a second factor could affect pensions’ attitudes toward risky investments: lower projected investment returns for all asset classes. Since 2007, Wilshire Consulting has lowered projected returns across the board, implying a more than 1 percentage point reduction in expected pension returns. By itself this would increase unfunded pension liabilities by more than a half trillion dollars. Since target investment returns, usually around 8 percent, are often imposed by state legislatures—who are reluctant to pay the increased contributions that a lower return would require—plan managers often have no choice but to construct a portfolio that will generate the desired return, without regard to risk. If projected asset returns fall, then pensions have to take more risk to maintain a targeted 8 percent return.

From 2007 to 2010, the shift in public pensions’ target asset allocations was subtle but important. While the bond and real estate shares held steady, the typical plan shifted around 7 percent of assets out of equities into higher-returning—but riskier—alternative investments. Some plans have gone much further—Maryland, Pennsylvania, and South Carolina increased the alternative investment shares of their portfolios by up to 25 percentage points, and New Jersey’s pension board recently voted to allow its plans to hold up to 38 percent of assets in alternatives.

Using Wilshire’s data on how returns correlate between broad asset classes, it is possible to estimate the risk and return of each plan’s investments. Since alternative investments have higher expected returns than equities, these portfolio shifts raised returns, though not enough to compensate for lower returns projected across the board. However, risk also increases—the standard deviation of annual returns for the average plan increased from 12.1 to 12.6 percent. Put another way, the median plan in 2010 took more risk than two-thirds of plans in 2007. The five plans making the largest changes increased their risk by over 2 percentage points.

The typical plan shifted around 7 percent of assets out of equities into higher-returning—but riskier—alternative investments.

The problem, of course, is that this risk is all off-balance sheet: nowhere in plan reports is it disclosed and nowhere is it priced. By taking more risk, pensions impose a contingent liability on taxpayers to bail out pension funds if investments fall short, and make state and local budgets more susceptible to the shifting fortunes of financial markets.

While experts debate how pensions should invest their assets, economists agree that public pension accounting is wrong to credit the risk premium on pension investments before those risks have actually paid off. Private pensions, for instance, are free to invest in stocks or other risky investments, and many do. And these risks, when they pay off, improve a plan’s solvency by increasing the assets held by the plan. However, private pensions cannot assume that they will receive high guaranteed returns on risky assets in all future years, which is precisely what public pension accounting allows.

Better accounting, which values guaranteed public employee pension liabilities independently of the assets used to finance them, would eliminate incentives to increase investment risk. At the very least, public pensions should clearly disclose how much risk they are imposing on taxpayers in their efforts to avoid other difficult policy choices.

Andrew G. Biggs is a resident scholar at the American Enterprise Institute.