Stocks will underperform bonds over coming years, reversing long-term historical trends and throwing a monkey wrench into funding levels and asset allocation policies, according to two prominent investment professionals.
While many investors are predicting an era of single-digit returns for U.S. stocks, Robert D. Arnott, managing partner of First Quadrant LP, Pasadena, Calif., and Ronald J. Ryan, president of Ryan Labs Inc., New York, are projecting the death of the risk premium given equities over bonds.
They assert that the true real return for stocks going forward is between 3% and 4% a year - far less than the 5% to 7% real return assumption used by most actuaries, and below the real return assumption of 4% generally used for bonds.
In a paper that is slated for publication in the winter issue of the Journal of Portfolio Management, Messrs. Arnott and Ryan argue that pension executives should re-examine how they manage pension assets in relation to plan liabilities. They also should lower today's 70% to 80% equity levels, adopt policies that match assets to existing liabilities, and manage surplus assets separately, the paper contends.
A negative equity risk premium also would support increased diversification in such asset classes as real estate, international stocks and bonds, global asset allocation, absolute-return strategies and inflation-indexed bonds, Mr. Arnott said in an interview.
Both authors have axes to grind: First Quadrant offers absolute-return strategies; Ryan Labs sells a customized liability-driven index recommended in the paper.
Many academics and investors concur that the equity risk premium is shrinking; the question is by how much. Michael Peskin, managing director, Morgan Stanley Dean Witter & Co.'s asset-liability advisory group, New York, said he has lowered the equity risk premium in their studies to 200 to 300 basis points from 400 basis points three years ago.
But few, if any, experts appear to buy the assertion that stocks will underperform bonds for a lengthy period of time. Said Robin Pellish, senior managing director of BARRA RogersCasey Inc., Darien, Conn.: "I'm not sure at this point we are willing to accept a negative risk premium, but it certainly is a well-written article that challenges conventional thinking.'
"It just doesn't make sense for stocks to have lower returns than bonds. Stocks are more risky," said Peng Chen, vice president in the research department at Ibbotson Associates Inc., Chicago.
"Whether it is negative is a very provocative thesis that would require a lot of justification, and I don't know if I see enough justification," said Richard Michaud, president of New Frontier Advisors LLC, a Boston-based money manager.
Using Ibbotson data, Messrs. Arnott and Ryan deconstructed the real stock-market return of 8.4% a year over the past 74 years, and challenged whether the factors that led to that return would continue. During that period, stocks outperformed government bonds by 5.1 percentage points a year.
Two percentage points of the 8.4% real return have come from dividend yields, which have fallen to a record low of 1.2%. In other words, investors now pay 80 years' worth of dividends to buy stocks, vs. 18 years' in 1925, the paper said.
The drop in dividend yields cannot be extrapolated into the future, they wrote. Without this revaluation, real returns would have been 6.4%.
Another 4.2 percentage points of the real return came from the reduction in dividend yields over the 74-year period, partially offset by economic growth. Eliminating that source of gains would reduce the real return to a paltry 2.3% a year, they wrote.
Projecting forward 74 years, the article says that even if stock buybacks, earnings reinvestment and increased productivity boost real dividend growth from its historic level of 1% a year, it's hard to justify more than 2% to 3% real dividend growth. Prospective real return from stocks would be 3.2% a year, based on 2% real dividend growth and assuming valuation levels remain constant, they wrote.
Mr. Arnott concedes, however, "I think the Achilles' heel in our argument is quite simply that if you buy the new economy argument, then economic growth will take place at an unprecedented pace" of around 6% vs. 2% to 3% historically.
But if the new economy argument is flawed and real stock returns average between 3% and 4%, that is much lower than the 5% to 7% real return assumed by actuaries.
If real return assumptions are cut by two or three percentage points, they said, "then most funds would find their ABO (accumulated benefit obligation) funding ratio drops by 25% to 40%," the authors wrote.
Using projected benefit obligations, which include future benefit accruals, the result would be worse, Mr. Arnott said.
Pension executives should manage assets in relation to liabilities, not against a market index, they wrote. The pension fund should be split into two parts: one employing a cash flow matching strategy to meet known liabilities, using a "custom liability index" pioneered by Mr. Ryan's firm; and a surplus management strategy that can take more aggressive and longer-term bets with the remaining assets, they added.
But observers think some of the authors' assumptions are faulty. For one thing, today's dividend yield of 1.2% may not be the right figure for going forward, Ibbotson's Mr. Chen said. Rather, the 25-year average of 3.7% might be more appropriate.
Second, assuming growth in real dividends of 2% may be too low, he added, suggesting a figure of 3% to 4% would be more reasonable.
Third, assuming valuation levels remain constant may be erroneous, since the proportion of stock-market capitalization to the total economy has been increasing steadily over time, Mr. Chen said. If the stock market's share of the economy doubles during the next 50 years, a one-percentage-point premium could be added to the economy, he said.
Based on those differing assumptions, Ibbotson calculates an equity risk premium of 4.6%.
New Frontier's Mr. Michaud argued that the discounted cash flow framework the authors used has predicted previously that the stock market was going to contract, and has not proved to be a reliable predictor.
One factor: The role of dividends has been shrinking, and companies have been reinvesting assets that generate higher returns. Another issue is that investors have become more sophisticated in terms of equity risk. While the risk premium may shrink to closer to the historical average, Mr. Michaud doesn't believe it will disappear.
Carl Hess, global director, asset allocation for Watson Wyatt Worldwide, New York, didn't think the paper's recommendations were practical: "It's quite an interesting view, but I'm not sure I've found a plan sponsor who is willing to put his money where Rob's mouth is."
The risk for plan sponsors is that the tax system allows contributions to go into the plan, and not back out. "If they're wrong, you've eliminated a good source of corporate cash - permanently," Mr. Hess said.