The erosion of pension surpluses is calling into question the guiding principle by which pension funds have been run. The "best practices" have called for the use of what the industry refers to as the Sharpe model. Its main conclusion can be concisely stated: stocks for the long run, bonds as a buffer.
But a new paradigm is on the ascent, the Leibowitz model. Over time, it will become the dominant model for pension investing.
The Sharpe principles are based on a long track record of stocks consistently beating intermediate bonds. Yet, returns on stocks can vary widely from year to year. Consecutive years of double-digit gains can quickly erode with not uncommon market retrenchments of 30% to 40%. To dampen the annual volatility in stocks, part of the fund allocation is dedicated to bonds. Bonds serve the role of a diversifier or an anchor, while the real return is generated from stocks.
This was the conventional attitude toward the nature of pension plans and the role of bonds that prevailed throughout the 1990s bull market. Now, the dive in surplus value from 2000 through 2002 is calling this guiding assumption into question.
The Sharpe model is seeking high and stable asset values, while the Leibowitz model is seeking high and stable surplus values. They are mean-variance optimizers, but with different benchmarks. The risk-free asset for Sharpe is cash; for Leibowitz, it is a bond portfolio that resembles a plan's liability schedule. Their attitudes toward correlation differs: Sharpe values negatively correlated assets; Leibowitz values assets that are positively correlated with the liability benchmark.
There has been no panel of judges assigned to evaluate the merits of each argument and declare a winner. The winner, up to now, has been determined by the actions of pension investors.
A survey by Thomas J. Healy, finance professor at Harvard University, noted a gradual climb in equity allocations over the past decade, to the point where the 60% equity/40% intermediate bond mix is near universal and is "the combination of academic work and consultants' work on asset allocation. It's remarkable that everyone is doing it the same way." He credits Harry Markowitz, the precursor of William Sharpe: "He deserves the Nobel for something everyone in a $3.6 trillion industry uses today." (The two shared the 1990 Nobel Memorial Prize in Economic Sciences.)
The stocks for the long run and bonds as a buffer theory could not have held sway without being supported by an often unnoticed but crucial premise. To believe in the validity of the Sharpe model, one must believe in the perpetual nature of the sponsoring business. You can only hold out for stocks in the long run if you first assume the firm itself will be around for the long run.
A corollary to the "on-going entity" argument is that anyone who forsakes equities for bonds is sacrificing return potential. They are giving up too much return for the misguided notion that they need to come up with the liability cost today. Equity-based assets will provide enough for retirement incomes, so long as sponsors are allowed enough time for their equity performance to shine.
On the other hand, the camp inspired by the work of Martin Leibowitz is calling for a "termination value" approach to managing pensions.
By now this much should be evident: To defend the status quo, its practitioners are required to de-legitimize the reality of the market value of the plan's liabilities. The Leibowitz camp can cite FAS 87, which requires sponsors to measure and record their pension obligations as real debt. But the Sharpe camp counters that you can't run an investment strategy according to the dictates of accountants; you have to conform to the realities of the market.
For the Leibowitz camp to effectively refute Sharpe they must prove there are market-based reasons for recognizing pension liabilities. To the extent that a plan's obligation is marked-to-market, either through withdrawals, conversions, mergers or terminations, the market value of the liabilities is realized. If it can be realized, then it is relevant. Plan liabilities need to be retrieved from obscurity and repositioned with the prominence they deserve - front-and-center to a plan's investment strategy.
The Leibowitz model provides an elegant solution and an objective standard for managing pension plans. Generating surplus value and guarding against volatility in the surplus value are the proper focus of effective pension management.
Pension investing in an efficient frontier universe is not a reliable model because it ignores pension liabilities. Without a liability benchmark, how do plans decide on appropriate asset mixes? Under the Sharpe model, plans are left in the absurd position of using a rate of return target as the standard for deciding which risky assets to choose.
For instance, if the CFO says we need an 9% rate of return on assets and the CIO says we can't expect more than 5% from bonds this year, the inevitable conclusion is that our equity and bond portfolios must be supplemented with alternative assets. This explains plans' heightened interest in hedge funds, private equity and international stocks.
Within the Sharpe framework, first we subjectively choose our return target, and then we find the assets to meet our needs. Whereas within the Leibowitz model, we first objectively evaluate the requirements of the plan, and then we construct an investment portfolio that best matches the needs of the funding schedule.
Adopting the Leibowitz paradigm will lead to reform, not a revolution. Make no mistake; plan sponsors can no longer tolerate risky pension allocations. Pension plans are risky because they don't have enough interest-sensitive assets.
The role of accountants as facilitators will soon end. When the users of financial statements begin reworking the results, it's only a matter of time before the Financial Accounting Standards Board responds. Add to this the pressure on the profession to shape up, and we give FAS 87 a life expectancy of less than three years. You can be sure that whatever replaces it will be crafted with the goal of increasing the transparency of pension results.
This doesn't mean pensions need to go 100% immunized. Plans may purposefully reduce asset duration for the benefit of increasing the surplus value when they expect rates to rise.
Bonds need to be a significant pension asset because the primary driver of their returns is the same driver of liability returns - changes in interest rates. Long-duration bonds as an asset class have a prominent role to play: both reducing interest rate risk and replacing a portion of the surplus value formerly generated by stocks.
As managing the surplus value becomes more common, the usefulness of the Sharpe model will fade. Increasing the pension surplus through responsible surplus management is the new target; using a longer duration within the bond allocation is the best means to that end.