No time to worry about CalPERS

By Felix Salmon
April 12, 2010
responding forcefully to a rather silly Stanford policy brief which gets very alarmist about California's pension liabilities. There are so many enormous and immediate fiscal problems facing California right now that it seems utterly pointless to put out a paper saying that the state should inject $200 billion into its pension funds -- especially when the logic of the paper is as confused as this:

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Well done to CalPERS for responding forcefully to a rather silly Stanford policy brief which gets very alarmist about California’s pension liabilities. There are so many enormous and immediate fiscal problems facing California right now that it seems utterly pointless to put out a paper saying that the state should inject $200 billion into its pension funds — especially when the logic of the paper is as confused as this:

The CalPERS portfolio has had returns averaging 7.91 percent over the last 25 years, with a standard deviation of 11.91 percent. As expected, the high standard deviation means that 68 percent of the time, returns range from –4.0 percent to 19.82 per­ cent. Historically, if CalPERS had simply invested in investment­ grade corporate bonds, the fund could have earned 7.25 percent, only .66 percent less than it has earned with its highly volatile portfolio. This small reduction in earnings would have allowed CalPERS to reduce volatility by a full 7.68 percentage points.

Therefore, in order to avoid future severe underfunded scenarios, we recommend that CalPERS, CalSTRS, and UCRS allocate more of their investment portfolios to fixed income asset classes, thereby reducing risk with a minimal loss of long term investment performance.

I’m not entirely sure where to start on this, but are the Stanford wonks really unaware that the rate of return on fixed-income investments over the past 25 years is largely a function of the fact that interest rates have been declining steadily over that time? And that now they’ve reached zero, they can’t really continue to do so for the next 25 years?

On top of that, the Stanford types seem to think that it makes sense to use a risk-free discount rate to calculate pension-plan liabilities, while even they admit that the assets shouldn’t be invested in a risk-free manner.

I do like the way that the Stanford paper tries to look at the probability of a shortfall of various magnitudes, rather than simply boiling things down to one number. But I’m not convinced by the methodology it uses to arrive at those probabilities, or by the way that the paper ignores all the political realities surrounding pension contributions and payouts.

The fact is that a defined-benefit pension scheme is always going to run the risk that it won’t be able to meet its liabilities as they come due. The California pension plans constitute an attempt to save hundreds of billions of dollars to pay for the pensions of the state’s workers; the attempt might succeed, or it might not.

But right now there are clearly more important and urgent things to do with California’s tax revenues than throw them into a pension pot to support the retirees of the 2040s and beyond. CalPERS might not be perfect, but it’s a lot less dysfunctional than most of the rest of the state government. Let’s get our priorities straight here.

5 comments

Comments are closed.