Privatization proposals shun danger — study

BOSTON — Proposals to privatize the Social Security system ignore the risk inherent in investing in equities, a new study says.

President Bush has said the privatization of Social Security is one of his key legislative proposals for 2005. The issue is being hotly debated in Congress and elsewhere.

The study, written by Alicia H. Munnell, Steven A. Sass and Mauricio Soto, was published in January by the Center for Retirement Research at Boston College. Ms. Munnell is the director and Mr. Sass the associate director for research at the center; Mr. Soto is a graduate student in economics at Boston College.

"Yikes! How to Think About Risk?" first notes the confusion in how pension funds value their liabilities. When calculating a plan's funding status, accounting rules require using a low-risk rate tied to bonds. But when estimating the current year's pension expense, sponsors can use the expected return on pension assets, the paper says. Thus, higher equity allocations result in lower pension expense and higher profits.

That's fine when companies expect to continue their pension plans, but it's a bad idea when the plan sponsor is ill-equipped to hike contributions quickly or can offload the risk to workers, taxpayers or the Pension Benefit Guaranty Corp., the study says.

In contrast, the Office of Management and Budget and the Congressional Budget Office project returns using the long-term Treasury rate for the federally sponsored Railroad Retirement System.

Why? Because this way the government avoids implying that it could "raise money simply by issuing debt and buying stock with the proceeds," the paper says. What's more, using higher expected equity returns would ignore the risk of investing in stocks, the authors add.

Ignoring the risk of investing in stocks is just what the federal government should not do in its discussions about privatizing Social Security, the authors warn.

Last summer, the CBO found that individuals would end up significantly worse off under a partial privatization proposal when returns are risk-adjusted to the expected bond return, the study says.

Based on a portfolio of 20% Treasury bonds, 30% corporate bonds and 50% stocks, the CBO found that an individual reaching age 65 in 2035 would have a risk-adjusted return of -23.7%, based on payouts from a private account. An individual turning 65 in 2055 would have a risk-adjusted return of -37.7%, while one attaining that age in 2065 would have a risk-adjusted return of -45.2%.

"If private accounts earn only the bond rate, Model 2 (of the President's Commission to Strengthen Social Security) produces a significant reduction in benefits as compared to current law," the paper adds.

Alternatively, the CBO simulated a large number of possible returns. Looking at the middle 80% of payouts from both traditional Social Security benefits and individual accounts, the agency found that returns ranged from one-third above to one-third below the expected outcome.

However, Social Security actuaries failed to explicitly address risk in individual account proposals in their own study of the reform proposals. Thus, portfolios including stocks "show very little reduction from benefits scheduled under current law, at least for the next half-century, the new paper says.

To compare apples to apples, it's appropriate to adjust for risk by using a bond-only return in comparing benefits from the current system to the proposed changes, it continues.

The other problem is that workers have little control over their retirement dates, which can have a huge effect on their benefits. "For example, a worker invested in equities would have received 36 percent more by delaying retirement from September 2002, when the market bottomed out, until September 2004.

"The problem is that many people do not have such flexibility," the paper adds. Social Security provides more than 50% of income for two-thirds of those age 65 and above, and "few can control the timing of their retirement," the paper adds.

"The implication is that projecting the expected return on equity investments without any adjustment for risk overstates the contribution of private accounts to retirement security," the paper concludes.