Pension Fund Return Projections are Based on Arithmetic, not Just History

Morning Edition had a segment on a change in public pension fund accounting that will show many funds have a much larger shortfall. The piece included comments from a Stanford business school professor, Joshua Rauh, that complained that the discount rate assumed by pension funds assumed that future pension fund returns will be like past returns.

Rauh's statement to this effect is inaccurate, or at least incomplete. The main question mark in pension fund returns is the return on stock, which typically accounts for 60-70 percent of pension fund assets. While stock returns can fluctuate hugely year to year, over the long-term (like the 30-year time horizon of most pension funds) they are a relatively predictable function of current price to earnings ratios and the rate of growth of the economy.

Given current price to earnings ratios in the market, it would require an unprecedented economic collapse for the market to yield substantially lower returns than what pension funds are now assuming. Ruling out a complete economic collapse might be assuming that the future will be like the past, but this sort of extrapolation is pretty much impossible to avoid.

The piece also wrongly implied that the Governmental Accounting Standards Board (GASB) agreed with Rauh's assessment in its proposed changes to accounting standards. This is not true. A pension fund that is fully funded using the 8.0 percent discount rate that Rauh criticized would not see any change in its funding status under the new GASB rules. Only pensions that are underfunded by the old accounting standard that would see a change in their calculated level of funding.

pension returns
written by john, July 20, 2012 9:23

Dean,

Please show me a pension fund that has a 30 year time horizon. If Calpers does not have increased returns over the next decade it will have a devastating effect on state and local finances. Massively higher contributions will be required by local, county and state government. These contributions will be bankrupting for many of these entities.

And public pensions are political animals. When times are good (and stock valuations are high) politicians give away surplus funding. As was the case in California in the 1990s with both Calpers and the UC pensions. Once funding ratios approach 100% their is an irresistible urge by politicians of both parties to shower public employees with increased benefits. This is when you get such insanities as 3/30 pensions for safety employees resulting in a typical fireman retiring in his early 50s with a 6 figure pension (inflation adjusted for life). So 30 year average returns are irrelevant. "Excess" returns during the boom years are not locked away to make up for lean years. There is little smoothing funding over the market cycle. During the lean years the shortfalls are made on the back of the taxpaying public as is happening right now.

-1

...
written by skeptonomist, July 20, 2012 9:45

The NPR piece gives a quote: "...over the last 30 years, our actual rate of return has been 9.3 percent." 30 years takes us back to 1982, when P/E of the major indices was little more than half what it is now. What would the rate of return be over the next 30 years if P/E goes back to the 1982 value? At first approximation an 8% rate of return implies P/E of 12.5 - what is the ratio now? Keeping in mind that bond returns from now on are not likely to be as good as they have been since 1982, the rate from stocks would have to be even higher than 8% to give a total return of 8%.

I do agree with Dean that these estimates are tending to be too low now - most economic estimates including pension-fund rates tend to be too high in good times and too low in bad times - that's why we have cycles.

Pension Fund Return Projections are Based on Arithmetic, not Just History

About Beat the Press