Fixed-income portfolios have performed as well, if not better, than the U.S. stock market in the past five decades, while exhibiting similar — or higher — levels of volatility. Those two findings of recent research by Wharton finance professor Jules H. van Binsbergen challenge long-held perceptions that equities offer higher returns and are more volatile than fixed-income instruments such as bonds.
The study showed that “investors have received little to no compensation for taking long-duration nominal dividend risk in the past half century,” Binsbergen said in his recent paper, “Duration-Based Stock Valuation.” Further, if anything, stocks seem to have too little volatility (not excess volatility) compared to these fixed income portfolios, the paper noted.
Binsbergen discovered those in attempting to quantify the impact of declining interest rates globally to all-time lows across different maturities, and their effect on the valuation of long-duration fixed-income portfolios. In that exercise, he constructed a series of counterfactual, long-duration fixed-income portfolios, and compared the returns on those with equity returns that include dividends and capital appreciation.
They key factor in the study is in matching the duration of investment across both asset categories, he told Knowledge@Wharton. “[Typically], people always pick a very short-duration fixed income portfolio, like a three-month bond or a one-year bond,” in making comparisons of returns with equities, he said. That explains why the perceptions of superior returns from equities even over long durations have persisted, he explained.
Challenging Conventional Wisdom, Solving Puzzles
The measure of how much stock markets outperform fixed-income securities is captured by the so-called “equity risk premium.” But the short-term lens does distort the long-term picture: Stock market investments have done “much better” than, say, a three-month bond that is reinvested repeatedly over 50 years in a row, Binsbergen said. “But the question is, is that the right counterfactual? Don’t you want to look at a fixed-income portfolio that has the same duration as the stock market? The three-month interest rate is then certainly not the right instrument to pick.”
“We essentially have had disappointing stock market performance for the last five decades, despite the fact that from a first glance, it may not look that way.” –Jules H. van Binsbergen
The right counterfactuals with matching durations make a surprising finding – a first in academic research: “We essentially have had disappointing stock market performance for the last five decades, despite the fact that from a first glance, it may not look that way,” said Binsbergen.
The second surprising finding is that fixed-income portfolios have the same volatility as the stock market, or higher, in some cases. “Maybe you could argue that the stock market is too little volatile, and not excessively volatile,” he said. “Maybe the fixed-income portfolio is excessively volatile.”
With duration-matched fixed-income portfolios, suddenly what seemed puzzling before now doesn’t look that puzzling anymore. Binsbergen’s paper examined four “puzzles” that previous researchers have grappled with. One is the so-called “equity premium puzzle,” where the average return on stocks has been higher than that from a short-duration fixed-income instrument. His study, however, shows that “investors have not received any compensation for taking long-duration dividend risk” from parking their monies in stocks. “One could argue that this simply means that the equity premium puzzle has resolved itself,” the paper concludes.
Another is the so-called “excess volatility puzzle” where the variation in stock prices is larger than the variation in dividends. Earlier research has traced those puzzles to time-varying expected returns specific to equities: time variation in the equity risk premium. Here, Binsbergen’s study brings more clarity by showing that duration-matched fixed income portfolios already exhibit similar or higher volatility as the aggregate stock market. This questions the common belief that time variation in the equity risk premium solves the puzzle.
Will Fund Managers Reset Portfolios?
Will the study’s findings prompt fund managers to recast their asset allocations? “Yes and no,” said Binsbergen. On the one hand, “decreasing interest rates help boost bond prices and give long-duration investors in fixed income securities a very large return and therefore large increases in their wealth,” he noted.
On the other hand, there isn’t much room for interest rates to fall further, so the potential upside is limited, he said. “We are starting to reach something close to a lower bound of how low interest rates can really go. So going forward, we should probably not expect high returns from bonds anymore.”
“One possibility is that the U.S. has gotten stuck in a Japan-like scenario where economic growth rates are just structurally lower than what we had before.” –Jules H. van Binsbergen
The study’s findings have important implications for retiree investors. Defined benefit pension plans around the world have been in an underfunding crisis for several decades now, Binsbergen noted in his paper. “These defined benefit plans that have long-duration, often risk-free, promises (liabilities) to pension holders who have been trying to gamble their way out of their underfunded status by investing in long-duration equities,” he added.
“The results in this paper show that this strategy has not worked for the past 50 years,” Binsbergen continued. “Under the assumption that the duration of the promises is the same as the duration of the equity market, the return differential these pension plans were betting on is exactly the long-duration dividend risk premium I compute in this paper, which has been zero or even negative in the past half century.”
Binsbergen noted that the dividend yield for the S&P 500, for example, is barely 1.8% now. “That is very low,” he said. The upshot for retirement savers is to save more. “If the rates of return on all assets are going to be this low, then, of course, to reach a certain amount of savings for your retirement, you need to put away much more money,” Binsbergen said. Of course, in equilibrium, if everybody thinks this way, this will lower rates of return going forward even further.
The study’s findings also have implications for how corporations could raise money. Traditionally, in corporate finance, the rate of return on equity has been seen as higher than that on debt as a compensation for the higher risk that equity carries. But the latest findings challenge that belief by showing that “at least for the last 50 years, the realized rate of return on debt and equity is pretty much the same, and possibly the rate of return on debt has been higher than that of equity,” said Binsbergen. “Secondly, the duration-matched debt is just as risky as equity in terms of its volatility.”
Drawing from that, equity potentially becomes more attractive than debt for corporations, Binsbergen argued. “If the rate of return on equity is lower than it should be, that means that the value of equity is too high. So, as a corporation, you get more money per stock that you issue.”
“Long-duration investments should be compared to other long-duration investments, and short-duration investments should be compared with short-duration investments.” –Jules H. van Binsbergen
Is the U.S. Headed Japan’s Way?
According to Binsbergen, the experience of Japan may offer pointers to the U.S. on what to expect going forward. Nearly a quarter century ago — in 1996 — interest rates in Japan were as low as they are today in the U.S. In the intervening period, fixed-income portfolios in Japan have outperformed the stock market. The Nikkei 225 index today is at about the same level as it was in 1996, he pointed out. “One reason for that was that, yes, fixed-income portfolios in Japan delivered very low returns, but the stock market delivered even lower returns.”
Binsbergen found the same results for Europe when he analyzed data from January 1996 to April 2020. “The results are even starker [in Europe] than for the U.S. in the sense that even the 10-year constant maturity, zero coupon bonds have outperformed the Euro Stoxx 50 index over this sample period,” he wrote in his paper.
“[In that case], we should just expect all asset classes — including bonds and stocks — to have lower returns going forward,” said Binsbergen. The only way to achieve attractive returns from stocks is to have “substantial growth” in the economy, but that faces headwinds such as the current recession and decreasing population, he added. “If we want to have higher returns, we will have to have growth. The question is: What is the long-term growth outlook for developed economies such as the U.S., Europe and Japan?”
Match Assets by Duration, Not Their Class
Binsbergen made a strong case for discarding the long-held wisdom of fund managers in distributing investments by asset classes, such as stocks, bonds, real estate, bullion or commodities. Instead, they ought to be grouped by their duration to allow for meaningful comparisons of returns, he argued. “Long-duration investments should be compared to other long-duration investments, and short-duration investments should be compared with short-duration investments,” he said. “If you mix and match, you can get very surprising answers.”
In fact, multi-asset class portfolios are starting to gain some traction, said Binsbergen. Balanced funds, for example, offer a mix of stocks and bonds, as do the so-called “life cycle” funds available for retirement savings, he noted. Yet some old habits die hard: “The life cycle funds are very aggressive on the equity share that they put in the portfolio, exactly under the assumption that equity will outperform fixed-income securities,” he said. “[They believe] that investor retirees need put in a lot of equity, because equity has these wonderfully high returns that fixed-income doesn’t have.”