—George Santayana
Today's public pension trustees and OPEB plan overseers are re-thinking their assumptions about long-term investment returns, as well as their basic investment strategies. With long-term government bond yields touching record lows and stock markets jittery at best, prudent fiduciaries must ask hard questions of their investment consultants and advisors. In the wake of lousy, roller-coaster investment returns since 2000 (see Chart 1 below) , and the haunting parallel of Japan's dreadful experience over the past 20 years (see Chart 2 below), many are wondering where to invest — and what rate of investment returns should they project for actuarial purposes.
If plans continue to anticipate a blended return of 7 or 8 percent but receive far less in the coming decade, their funding ratios will deteriorate even further, and employer costs will continue to skyrocket.
Nobody can predict future investment returns. The best we can do is to understand investment fundamentals and the lessons of history. Investment fundamentals tell us that most prudent investors avoid risk and will give up higher potential returns to obtain safe, secure, guaranteed returns. Conversely, risky assets must be priced low enough to produce superior returns to those who buy them. This should be true especially when the economic environment feels "risky." We expect the prices of risky investments to decline during "risky" periods, so that firms that deliver business results as expected will produce even better returns to the more daring investor. Risk is thereby rewarded. What history tells us is that over longer periods, the (theoretical) fundamentals have been borne out by actual investment experience.
The charts above show what can happen at the end of an era when investors see only roses during the euphoric times and bid up the prices of stocks to levels that cannot be supported by economic fundamentals. Risk is temporarily forgotten as collective amnesia overtakes the market psychology. In Japan's case, the late 1980s were a period when the country enjoyed an export surge globally and a debt-financed real estate bubble; their stock prices quintupled. Then the bubbles burst and 20 painful years of malaise ensued. Here in the U.S., the Internet bubble of the late 1990s was followed by a deep correction. In the last decade, a secondary real estate boom led to the Great Recession of 2008 and a stock market that has returned virtually nothing over the past decade. In both cases, the ebullient investors paid dearly for their rose-colored views of investment risk.
Now, we find ourselves in the opposite environment. Market psychology and global economic fundamentals are so shaky that risk is viewed negatively around the world, making low-risk investments like Treasury bonds a magnet for safety-conscious investors. In fact, T-bonds are now less of an investment to obtain a safe return (in the face of possible inflation risks and rising interest rates eventually) and more of a put option to protect against stock market risk. We hear daily about the markets trading in "Risk-On" and "Risk-Off" cycles in highly volatile, news-driven flurries that look ahead only minutes — not even hours or days, let alone years. Traders buy T-bonds when the herd shifts to Risk Off.
In this context, it may be helpful to put today's global economic malaise into the perspective of the Great Depression, which some folks compare with our current Great Recession. So let's look at the long-term investment results from putting money to work during the Great Depression. I picked three points-of-departure scenarios that seemingly had the odds stacked against them — to study 10-, 20- and 30-year investment returns and compare them with inflation rates during the same periods. The three starting points for this analysis were January 1 of:
- 1933, following a respectable 50 percent recovery from lows of the stock market crash of 1929-32, and thus fairly analogous to today's market following the trough of 2009
- 1937, which reflected a period of economic recovery just before a secondary recession (today we call it a double-dip) which carried into 1938 and was an even deeper contraction. Those now worried about a new double-dip will find this data informative.
- 1941, which preceded the infamy of Pearl Harbor and America's entry into WWII
From prior research, I already knew that the 30-year returns from investments in stocks during this period were higher than the long-term averages that professional investors often cite from the Ibbotson data series that begins in 1926. Those 85-year average annual returns are now a smidge below 10 percent at 9.88 percent through June 2011, whereas the 30-year compound average annual returns from stock purchases emanating from any of the three selected entry points were in the 10 to 13 percent range.
I was a little surprised to see the 20-year returns also well above the longer-term averages — and in most cases even higher. What surprised me most was that in two of the remaining three scenarios spanning the Depression and the war, the shorter 10-year returns were also respectable and in line with longer term averages. Only the 1937 entry point received inferior returns over ten years, reflecting the immediate losses that stocks experienced in the 1937-38 rout (a 49 percent one-year loss) coupled with the economic drain of WWII. That one scenario, however, should be enough to warn investors that a naïve strategy of investing in riskier assets during a risky period is not a sure thing over a single decade.
GASB pension accounting implications. Hence, it is reasonable for pension and OPEB trustees to trim their expected equity returns when they begin to disclose their assumptions and to align them with a shortened investment horizon period that would be more consistent with proposed accounting standards. In the past, retirement plans have used long-term (30 years or longer) investment horizons for estimating their expected rates of return on investments. But if a substantial portion of the plan's funding must be contributed in the next decade, as will be the case under new accounting standards with shorter amortization periods, then the equity and bond investment horizons should be shortened.
In the case of bonds, the results are obvious. With a shorter horizon, yields are lower in a normal yield-curve environment. For example, the 10-year Treasury bond now yields around 2 percent whereas the 30 year T-bond now yields about 3½ percent. If unfunded liabilities must be financed over the average remaining service lives of employees (between 12 and 15 years for most plans) and retirees' average remaining life spans are 10 to 12 years, then the shorter bond maturity with lower yields is the right discount rate component to align with those liabilities. Likewise, the shorter investment horizon for stock investments must take into account the 1937 scenario and shave the probabilistic expectation of equity returns for shorter periods. And even for the longer horizons, a 3½ percent long T-bond won't produce 5 percent returns as many pension funds still expect from that sector. Only the riskier corporate bonds will get yields that high, after deducting their probable default rates.
Now back to bonds in the Depression era. Returning to the 10/20/30-year table of returns above, which shows the returns on bonds purchased on those same Depression-era dates, we see that as the malaise continued and yields at purchase declined, the longer-term holding period returns were inferior, especially when inflation was considered. Except for bonds purchased in 1933 and held for a decade, the long-term returns in all other periods studied fell below the rate of inflation, which resulted in negative real rates of return. You can't run a pension fund on negative real rates of return, if salaries track with inflation.
This could well be the scenario we face once the U.S. eventually emerges from the current malaise, and should inform trustees that bonds belong in the portfolio nowadays chiefly for their diversification value. For both pension and OPEB funds, it will no longer be reasonable to project 5 percent bond market returns in the next two decades, given the low intermediate yields in today's markets coupled with the shortening of the investment horizon to match liabilities and funding patterns. Only the well-funded pension funds can actually say their overall investment horizon is 30 years — and thus justify an expected bond return that exceeds today's yields. (For long bonds, interest-on-interest becomes more important and rising yields can actually boost total returns through coupon reinvestment; this is not true for shorter maturities.)
Conclusions. First, past performance of any investment is no assurance of future returns — and despite Santayana's admonition, history never repeats itself exactly in the financial markets. In fact, a good body of research tells us that market historians help today's investors avoid some of the errors of the past and make modern markets more efficient because historical scenarios are back-tested by many big-league strategists. However, there is no doubt that there are some important similarities between the Great Recession era's debt hangover and that of the Great Depression. The lessons from the bond market should not be forgotten. Institutional real estate would also be favored if inflation resumes or exceeds historical levels.
When drawing comparisons, it's important to remember that the U.S. emerged victorious from WWII and enjoyed a postwar boom that may be incomparable to whatever may ensue stateside in the next decade or two. Some analysts would say that just as Britain ceded global leadership to the American superpower in the 20th century, we debt-burdened Yanks must now pass the baton to the Chinese whose century appears to be here — unless their local government debt turns out to be as rotten as our banks' mortgage portfolios. In that centurion context, international equity diversification with a strong Asian regional bias and selected multinationals controlling production inputs (ranging from technology to materials) would be another global long-horizon strategy to consider.
The data displayed here cannot provide a simple answer for pension and OPEB trustees to adopt, but rather provide some factual perspectives that can help focus a more meaningful dialogue about expected returns vs risks and the investment horizons of the fund. That said, I cringe now when I hear novice investors, biased pension critics and journalists challenge the long-term investment return assumptions of retirement plan fiduciaries simply because of the Lost Decade in stock market returns that we recently experienced. Obviously I can't rule out an extended malaise similar to Japan's, but there will most likely be a positive reversion to the mean for longer-horizon assets within our lifetimes. The risk now is that we could over-react by looking too hard into the near-sighted side-view mirrors that only show the last car we passed.
A more realistic and systematic approach would be to segment the plan's liabilities by their demographic duration (retired lives, active employees' remaining service periods, new-hire employment career expectancy, and unfunded liabilities amortization periods) and align asset allocations, investment horizons and return projections with those components. For most plans with funding ratios now below 80 percent, the blended average projected rate of return using this segmented expected-value methodology will likely fall somewhere between 6 and 7 percent. That's well below the 7.5 to 7.75 percent range in vogue for today's pension funds. Needless to say, the funding implications of this realization will be profound, and problematic. Taking into account the impairment of capital market returns caused by the global debt overhang, employer costs for underfunded retirement plans will likely increase yet further in the coming decade. Annual pension costs could increase another 20 percent on average in addition to the impact of proposed accounting changes and catching up for previous market declines. My prior columns about pension attrition and long-term pay freezes will sadly be even more accurate.
With all these buried negatives, I would not be surprised to see mounting public or political pressure to replace defined benefit (DB) plans with defined contribution (DC) or hybrid programs in coming years, as the visible costs of public DB plans escalate. If not done carefully, that would of course shorten the investment horizons of the pension funds even further, and perversely compel higher allocations toward low-yielding bonds. That would only worsen the funding-cost problems for public employers. And then, ironically, if investment history follows the post-Depression scenarios above, the new entry-level employees hired into a DC plan could end up reaping extraordinary returns once the current malaise ends — as did the brave hearts who invested for 30 years beginning in the Depression. To the extent those younger workers will be supporting my generation's unfunded Social Security benefits while also suffering the chronic budget pressure caused by their predecessors' unfunded pension and OPEB liabilities, I suppose it would only be fair that they enjoy a silver lining somewhere.
Disclosure and disclaimer: Girard Miller's comments, suggestions and views herein are his own general opinions and do not constitute specific investment advice nor an offer to buy or sell securities. His independent views do not necessarily reflect those of any organization with which he was or is presently affiliated, including his employer PFM Asset Management LLC. PFM Group statistical information displayed above is copyrighted with data obtained with attribution from sources considered reliable, but cannot be guaranteed for accuracy and does not constitute investment advice or a point-estimate basis for projections.
