Defined benefit plans now find themselves the target of unflattering comparisons with the savings-and-loan industry of the 1980s. The comparison might have some merit. Both attempted to boost income by buying assets with much higher expected returns but with little correlation to their liabilities. Both legally acted as fiduciaries but treated their fiduciary role as a source of profit. Both were backed by a federal insurer responsible for covering any shortfalls. Both responded to an unfavorable performance environment by reaching for riskier assets with higher expected returns, and both used accounting rules to cover their errors.
Pension asset surpluses have been devastated by the combination of a bear stock market of falling equity prices coinciding with a bull bond market of rising bond prices and falling interest rates. Many observers in the industry blame the high hurdle investment rate assumption and call on fiduciaries to lower it. But changing the actuarial assumptions would have done nothing to prevent pension plans from falling into deficit.
The problem with pension funds is not with accounting assumptions but with strategic assumptions. Pension plans are managed using the wrong objective. Pension funds invest using a model that compares one asset to another. The real life problem, though, is to choose appropriate assets to fund their employees' retirement. The essential issue is to recast pension strategy as an asset vs. liability decision not an asset vs. asset decision. Outperforming a generic market benchmark is not the objective. Outperforming liabilities at an acceptable volatility level is the objective.
Typically, pension funds define their problem as maximizing the total return on assets over an extended time horizon. As a result, pension funds skew allocations into assets with the highest expected returns, regardless of the term structure of the pension liabilities. The average pension fund allocation to equities has grown to 61% in 2001 from 43% in 1990. It's probably no coincidence that equity allocations have increased in step with the exceptional equity market returns.
Current return assumptions have become more difficult to reach with today's low earnings and low dividend yields. As equity returns have dropped dramatically in the past two years, pension funds, instead of looking to fixed-income and liability-matching strategies, have searched for more equity-like return in real estate, private equity and hedge funds.
Advocates of high equity exposure emphasize that you need to look at their results over the long term. But the record tells a different story. Over the 13 years ended in 2001, the cumulative total return difference of pension assets minus liabilities is -7.32% - no value added. Even with a bull stock market for most of those years and with an asset allocation skewed heavily toward equities, pension assets underperformed a portfolio of high-quality, low-yielding liabilities.
This experience calls for plan sponsors to seriously re-evaluate standard notions of an appropriate investment strategy for pension assets. It should be apparent that the whole debate concerning an appropriate expected return and the "correct" actuarial assumption is really beside the point. The investment results depicted in the funding ratio completely and accurately describe the plan's position. Actuarial assumptions only determine how fast or slow these investment results are reflected in the company's reported income.
Pension fund accounting rules were designed to allow pension plans to consider the best long-term values without concerning themselves about short-term earnings fluctuations. The unintended consequence is that many pension plans have lost sight of the highly rate-sensitive nature of the liabilities they have been charged with funding. Is there a change in the wind? So far the industry's movement in the direction of asset-liability management has been modest.
The trouble with the traditional view of asset management is that it ignores the pension plan's reason for being - to pay their employees' retirement income. From this perspective, the pension plan's liability schedule becomes the only appropriate benchmark in measuring the performance of the pension assets. An asset is risky to the extent its return behavior diverges from the liability it is funding.
Equities have delivered exceptional performance, yet plans have lagged their benchmark because of their underweighting in bonds. Bonds have an 88% correlation with the liability benchmark. Equities have a 22% correlation. Stocks fluctuate widely vs. the pension benchmark and show no particular relation to changes in interest rates. This is the reason for wide swings in performance during the past 13 years.
Fluctuations in the funding ratio provide an accurate measure for evaluating the risks of a particular asset allocation. Corporate earnings stability is the natural result when the objective is appropriately focused on managing the volatility of the funding ratio.
In an asset-liability context, the typical pension plan's interest rate exposure appears quite large. For the past 13 years, the yearly investment results measured against the annual interest rate changes indicate a mismatch of more than 12 years. The S&L's strategy of funding mortgages with demand deposits resulted in a mismatch of four to six years. A typical plan' s asset-liability mismatch is more than twice the size of the S&L's mismatch.
Pension fund sponsors ignore their true benchmark, pension liabilities, at their peril.
David A. Hershey is a senior portfolio manager at Lotsoff Capital Management, Chicago. The complete report on which his commentary is based, including supporting data, is available at www.lotsoff.com.
