CEPR - Center for Economic and Policy Research


En Español

Em Português

Other Languages

Home Publications Blogs CEPR Blog The Financial Health of Public Pensions

The Financial Health of Public Pensions

Written by Dean Baker   
Friday, 03 May 2013 10:45

In a post for the Roosevelt Institute's Econobytes, Dean Baker, co-director of the Center for Economic and Policy Research, on the state of public pensions:

Last month, Moody’s, one of the big three bond-rating agencies announced that it was changing the way that it will treat public pension liabilities. While there are several notable features to announced changes, two are especially important.

  • First, instead of accepting projections of pension fund returns based on the assets they hold, Moody’s will evaluate their liabilities using a risk-free discount rate.
  • The second major change is that it will evaluate pension assets at their current market value rather than using a formula to smooth volatile asset prices over a period of 3-5 years as is the current practice.
  • Together Moody’s changes in accounting will have the effect of making pension funds seem in considerably worse financial condition.
  • They could also lead pension fund managers to follow investment strategies that will provide far lower rates of return on assets. This would mean higher costs for taxpayers with little obvious benefit in reduced risk. 

 The practice of using a risk-free discount rate to assess liabilities is a departure from the current practice of effectively discounting liabilities using projected rates of returns on assets.

  • Most pension funds project rates of return on assets of between 7.0-8.0 percent. By contrast, the rate that Moody’s would apply at present would be around 4.5 percent.
  • This makes a huge difference in assessing liabilities. If a pension fund was obligated to make $1 billion in payments each year for the next thirty years its liabilities would be calculated at $16.5 billion using a 4.5 percent discount rate.
  • By comparison, they would be just $11.9 billion using a 7.5 percent discount rate. This is a difference of almost 40 percent.
  • Put another way, a pension that is fully funded using the 7.5 percent return assumption would be almost 30 percent underfunded using Moody’s new approach.

While the implication for an assessment of funding levels is clear, there are grounds for debating which method is appropriate.

  • Supporters of the Moody’s approach argue that it is appropriate to discount for the risk associated with pension returns, and especially stock returns.
  • However this argument is problematic since it is not clear that there is much risk for pension funds on projected returns when they are properly calculated.
  • The reason is that a pension fund, unlike individuals, does not need to be concerned about the stock market’s short-term fluctuations.
  • State and local governments do not have retirement dates where they have to start drawing on stock holdings. They need only concern themselves with long period averages, without worrying about short-term fluctuations.

From this vantage point, there is relatively little risk when pension funds calculate returns correctly.

  • Many pension funds did get themselves in serious trouble because in the 1990s and 2000s they assumed (with Moody’s acquiescence) that stocks would provide historic rates of return even though price to earnings ratios were between 50 percent and 100 percent higher than normal.
  • This was clearly impossible as some us tried to point out at the time (here and here).

However with the ratio of stock prices to trend earnings now close to its historic level, it is quite reasonable to expect that stock returns over the next three decades will be close to their historic average.

  • In fact, it is almost impossible to envision scenarios in which stock prices would fall much below their historic average over the next three decades.
  • Had pension funds in the past adjusted their projected returns for price to earnings ratios they would have faced almost no unhappy funding surprises even given the sharp swings in the stock market over the last century.
  • By contrast, if pension funds were to maintain the current investment strategies (they typically hold 70 percent in equities in some form) and follow Moody’s discounting method, then they would follow very erratic funding patterns.
  • This approach would lead them to effectively accumulate large surpluses in pension funds as they reached full funding levels by Moody’s methodology, then they would be able to contribute little or nothing to the fund, as the return on assets would be sufficient to maintain full funding.

While the idea of not having to contribute to a pension fund might sound desirable, this outcome means that we would have effectively over-taxed or underprovided services to reach this point.

  • It would be nice to have our schools or police service fully paid for in two decades also, but no one would seriously suggest building up a fund now so that the interest would be able to cover these costs after 2033.
  • This is essentially the implication of a pension following the Moody’s methodology in assessing its funding needs.

Of course there is another even worse possibility that follows directly from Moody’s decision to eliminate asset smoothing.

  • The Moody’s methodology gives pension funds no credit for the decision to invest in a more risky asset like stock.
  • It scores pension fund liabilities in the same way regardless of whether they have their fund largely invested in stock or money market funds that pay near zero interest.
  • Since stock returns actually are volatile from year to year, pension fund managers would be taking a serious risk that investments in stock will fall in price in any given year.
  • This would make the fund appear to be in much worse shape, using Moody’s methodology. 

Needless to say, pension fund managers would not like to be embarrassed by seeing their fund’s finances deteriorate sharply.

  • Since they get no benefit under the Moody’s methodology from investing in stock, and they do take real risk in terms of year to year fluctuations, many pension fund managers are likely to make the decision not to invest in stock at all.

This would lead to the incredibly perverse outcome that long-lived funds that are largely indifferent to market timing would not be invested in the stock market at all. By contrast, individuals with 401(k) type accounts, who are enormously vulnerable to market timing, would be invested in the stock market. This is horrible public policy, but may be the result if the Moody’s methodology comes to be the new standard for pension fund accounting.

Tags: public pensions | retirement

Comments (6)Add Comment
written by ltr, May 03, 2013 1:26
Since you valuation of stocks differs markedly from that of Robert Shiller and I respect both of you but have found Shiller highly reliable from an historical perspective, I would like to have you justify your valuation method.

You have never explained your stock market valuation method clearly, and I write this having a doctorate from a tip top school.
written by ltr, May 03, 2013 3:18
You should be your:

Since your valuation....
Valuation method
written by Dean, May 07, 2013 9:56

Shiller, at least in his standard method, is focused on the S&P 500. I look at the whole corporate sector, which I would argue is more appropriate since pension funds invest in a broader range of assets. For valuations, I take the number from the Fed's Flow of Funds account (it would take me a moment to find the table, but it is cited in my papers). For profits I take the average share of after-tax profits in GDP over the last 40 years and multiply it by potential GDP (about 6 percent higher than current GDP). This gives me a trend PE. This has been rising a lot over the last year, which means that return assumptions would have to be revised a bit downward.
written by ltr, May 07, 2013 7:01
I really appreciate the response and am thinking it over carefully though I am still inclined to think Shiller's valuations would be better for fund managers to use if they actually pay proper attention to valuations.
written by ltr, May 07, 2013 7:21
I understand your valuation method now, and as always you you completely clear in explanation, however turning to the just compiled data from Shiller I think you are significantly underestimating the expense of American stocks now and so overestimating future returns.

Also, we should turn to John Bogle and there we would find an agreement with Shiller. I have no idea what will happen to stock prices, but this market is historically quite, quite expensive.
Stock valuations
written by Dean, May 08, 2013 6:17
I think the market is getting a bit high with this recent run-up. We're still way better off than the late 1990s, but it might be worth adjusting expectations downward a bit.

But isn't Shiller's data all for the S&P (haven't looked for a while). That is a more narrow group of stocks and since pensions do generally have substantial PE investments, it is probably not the best reference group.

Write comment

(Only one link allowed per comment)

This content has been locked. You can no longer post any comments.


Support this blog, donate
Combined Federal Campaign #79613
budget economy education employment Haiti health care housing inequality jobs labor labor market minimum wage paid family leave poverty recession retirement Social Security taxes unemployment unions wages Wall Street women workers working class

+ All tags