I was asked recently to speak at the IFEB Conference on the topic “The Collision of Pension Reform and Poor Investment Returns.” I didn’t choose this topic; I inherited it. But I was both intrigued and somewhat disturbed by what it implied.
I was intrigued because of the powerful image it invoked. Here you have pension reform, a monument to financial design, chugging forward just like the Titanic, and it is about to run aground of “poor investment returns” playing the part of the iceberg. All of the noble goals of pension reform—including expanded pension coverage, greater dependability, greater affordability, greater equity between generations—are all put in jeopardy by a financial disaster caused by “poor investment returns.” Presumably, to further flesh out the metaphor, the water around the sinking pension reform is littered with the carcasses of underfunded DB pension plans.
The most disturbing part of all of this is not the image itself, but the scapegoating of “poor investment returns.” This implies an almost serendipitous cause to our pension funding problems, and also possibly a temporary one: ‘We’ve just gone through an unfortunately perverse period in the capital markets. It’s not our fault. Things are bound to improve.’
But have investment returns really been that bad? The Buck Pooled Fund Survey Report at the end of June showed a 10-year return of 6.2% for the median balanced fund. Over this period, inflation has been about 2%, giving us a real return above 4% per annum. Most pension funds have investment policy statements with a real return objective of around 4%, so the disturbing fact is that many funds achieved their real return objective and still ended up in a more precarious funding position.
Over the same 10-year period ending in June, the DEX Universe Bond Index returned around 6.5%, and the long bond index returned about 8.2%. Unfortunately, the premium paid to investors for holding stocks rather than bonds has been negative. This is bad news for pension funds for a couple of reasons.
First of all, the return on long bonds can be viewed as a proxy for the return on the solvency or accounting liabilities. So the problem is not that investment returns have been poor but rather that investment returns have not kept pace with liability returns.
Secondly, the capital gains on bonds are a result of declining interest rates, which has left us in a “low rate environment.” In a world of low interest rates, we should be expecting lower investment returns in future.
We continue to expect higher returns from stocks than from bonds. Pension fund discount rates on a going-concern basis are usually calculated with an expected equity risk premium (typically from 2% to 4%) built in. The higher the equity weighting, the higher our expected return. It is not unreasonable to expect stocks to outperform bonds in future, even if it hasn’t really been true for the past 30 years. But with any higher expected return from stocks comes a wider range of potential results. Just looking at a best-case point estimate, which automatically inflates with a higher equity weighting, rather than considering the full range of what might happen, can be misleading. And if we start to build in expected value-added from active equity management, our return expectations can become hoped-for targets rather than likely outcomes.
It is also worth noting that the estimated return figures tend to be before fees. The money that is left to fund the pensions is after fees. This can be another ongoing source of shortfall to the extent that costs have not been properly provided for in our return expectations. And this shortfall can be increased by allocating more of the fund to investments with higher management fees.
So to return to our original metaphor, pension reform has not so much run aground of poor investment returns as it has been swamped by the waves resulting from a collision between what actually happened and what had been expected. Clearly, if there is a fix to this problem, it lies in improving the way we arrive at our expectations rather than just blaming the actual results.
To be more useful, future funding expectations need to be based upon co-movements of assets and liabilities—in other words, taking into account expected capital market returns relative to expected future changes in interest rates and inflation. And expectations should be stated as ranges rather than just point estimates. Any point estimate of future returns and funding is going to be wrong. The important thing to know is how wrong the estimate might be.
Such improvements in the development of our future funding expectations will act like sonar for the pension fund. They won’t eliminate the icebergs, but they ought to give us more time to react.