Image Source: Pexels

In classes about personal finance, once of the best-known lessons is that investing in US stocks can have considerable short-term risk, but that over an extended period of time, a patient investor has been rewarded for that short-term risk with higher long-run returns. David Chambers, Elroy Dimson, Antti Ilmanen, and Paul Rintamäki discuss this pattern and others in “Long-Run Asset Returns” (Annual Review of Financial Economics, 2024, volume 16). A particular advantage of their study is that, for them, the “long-run” doesn’t mean going back the 1980s or the 1950s, but back to the 19th century and sometimes earlier.

On the issue of stocks having higher returns than bonds in the long-run, they write:

An abundance of evidence has shown that the total returns of equities have exceeded those of (government) bonds over various long-run periods since the beginning of the twentieth century—both in the United States (Ibbotson & Sinquefield 1976; Siegel 1994, 2022) and in the rest of the world (Jorion & Goetzmann 1999; Dimson, Marsh & Staunton 2002). First, we address one of the central questions in empirical asset pricing, namely, whether stocks consistently beat bonds over the long run. As we noted above, this evidence relies primarily on data series starting in 1900 or, in the case of the United States, 1926, when the University of Chicago’s CRSP data starts. In this section, we first discuss the history of stock and bond returns before 1900/1926 … The main message is that the US equity premium over (government) bonds at and above 5% in the 1900s was considerably higher than estimates for the 1800s ranging between +1.6% and –0.6%. … The 224-year estimate of the annualized UK equity-bond premium is in the range of 2–3%, which is again well below the twentieth-century premium of 4–5%.

In short, it’s been a good century for stock returns, especially in the US economy. But some centuries have not been as good.

For bonds, a key question is that they seem to be historically low rates in recent decades.

[B]efore the twentieth century, inflation was very low and there was little difference between nominal and real yields, which both fell to 3–4% around 1900. The twentieth century saw more persistent price rises, particularly the Great Inflation of the 1970s, which pushed nominal rates up to double digits—a level last seen in the 1500s. The post-1982 Great Moderation witnessed global convergence in inflation rates toward 2%—back to the low levels that characterized the pre-1900 economy.

Even after inflation expectations stabilized at low levels in the 2000s, real yields and policy rates kept falling to negative territory. Eventually, even nominal yields in many European countries and Japan turned negative, or reached record low levels in places where they didn’t (the United States and the United Kingdom). The global rise in inflation in 2021–2022 finally turned bond yields higher after a 40-year downtrend. The yield evidence … stretching back over centuries provides confirmation that the recent low (even negative) bond yields were truly exceptional.

What about corporate credit risk, measured by how much the rate of return on corporate bonds exceeds the rate of return on government bonds? They discuss the weakness of the evidence on this point, but argue that it seems as if corporate credit risk doesn’t actually appear in the data until 1973.

What about investments in housing? “When reviewing the evidence on the investment performance of real estate, we discuss the challenges presented by the heterogeneity and immovability of this asset class. The evidence has largely focused on housing and suggests that total returns are below those from equities.”

What about returns to investing in commodities? “We then turn to commodities and critically examine the claims of the recent studies pointing to surprisingly high long-run returns. Here, we conclude that the long-run historical returns on a diversified portfolio of futures have come close to approximating those on equities.”

Generalizing wildly from my own personal experience, I sometimes tend to think about “long-run” returns as the returns happening over the decades in which I am saving for retirement. But of course, the “long-run” does not care about my birthday. A longer view on the long-run opens up new ways of thinking about how possible outcomes, and how and why financial markets distribute their rewards.

More By This Author: