T. Rowe Price: 125 years of returns: Timeless lessons in investing

Justin Thomson, Head, Investment Institute and CIO

Short-term anomalies bring risks, but remaining on the sidelines is riskier.

Key Insights

• Viewing an asset class’s returns over a decade, or even several decades, can give a distorted impression of future trajectory; a longer perspective is required.
• For 125 years, equity returns have shown that, over time, investors in the asset class have been compensated for higher volatility and the risk of capital destruction.
• Short‑term anomalies such as high market concentration bring risks, but staying in the market during these periods will likely be more rewarding than trying to time it.

Professor Elroy Dimson of Cambridge University recently presented his findings on the Dimson, Marsh and Staunton (DMS) Investment Returns database as part of our T. Rowe Price Investment Institute Speaker Series. Now 125 years old and counting, this dataset is highly authoritative on equity and bond returns. So, what does it reveal that might help us today?

Asset classes reveal their truth over time—not overnight

The quick and dirty readout from 1900 to 2024 shows that equities did very well (U.S. equities delivered 6.6% annualized real (i.e., net of inflation) returns, and ex‑U.S. stocks returned 4.3%); U.S. bonds performed more modestly, delivering 1.6% real returns; and U.S. bills (cash proxy) were a hedge on inflation plus 50 basis points (1 basis point = 0.01%). The best‑performing equity market was the U.S., and the best‑performing bond market was Switzerland at 2.8% annualized real returns (measured in U.S. dollars).

Equity returns over the period show that investors in the asset class were compensated in the long run for higher volatility and—as is well documented in the study—the risk of permanent destruction of capital. It is arguable whether government bond holders were similarly compensated for the permanent destruction of capital that came with bouts of hyperinflation.

"Let us learn from the past to profit by the present."
William Wordsworth

A powerful telescope is required when searching for the truth about asset class returns, which is why having a 125‑year‑old dataset is so useful. Viewing returns through the lens of a single decade, or even several decades, can provide a distorted impression of future trajectory. For instance, in the two decades up to the year 2000, equities produced an annualized real rate of return of 10.5%. This clearly overstated the long‑term real rate of return on equities, and the payback duly came in the next decade as equities delivered a period of subnormal returns. And pity those investors who were long bonds in 1950 (or were starting their careers in fixed income) as bonds in the U.S., UK, and most of Europe were just embarking on a three‑decade stretch of negative real returns up until 1981.

Inflation is even more pernicious than you thought

The DMS Investment Returns Database also provides context for modern concerns over inflation, U.S. exceptionalism, and market concentration. Let’s start with inflation. The database documents episodes in history in which equity and bondholders have been subject to the permanent destruction and/or confiscation of capital. Examples of this during the last century include revolutionary periods in Russia and China when incoming governments repudiated all issuance by the previous rulers, but it was the severe hyperinflation that gripped Germany and Austria following World War I that most clearly demonstrated the pernicious impact of high inflation. It is worth noting, by the way, that the DMS study does not remove these episodes from the database to avoid survivorship bias.

Periods of hyperinflation are rare, but the analysis shows that even relatively small differences in inflation can have major effects. In the U.S., annualized inflation from 1990 to December 2024 was 2.9% per year versus 3.5% in the UK. Thanks to the power of compounding, however, this apparently minor difference meant that, while U.S. consumer prices rose by a factor of 37, UK consumer prices rose 78‑fold. If the question is whether equities provide a hedge against inflation, then the answer is: up to a point. The tipping point appears to be an inflation rate of 4%. At 4%, equities averaged a real return of 6.3% annualized and real bond returns tended to zero. If inflation reached 7%, equity returns tended to zero and real bond returns averaged -5% annualized.

It’s too early to call the death of U.S. exceptionalism

While the case for international diversification is a strong one, in my view, reports of the death of U.S. exceptionalism are premature. The DMS database shows that outperformance of U.S. equities has been a persistent historical trend. From 1900 to 1999, U.S. equities returned an astonishing 7.0% annualized while the rest of the world delivered 4.9% (in U.S. dollar terms) and the second‑best market, the UK, returned 6.2%.

In the first quarter‑century of this millennium, U.S. equities returned 4.9% annualized compared with a miserly 2.0% for the world ex‑U.S.—and while the December 2024 cutoff excluded this year’s volatility in U.S. markets, surely the gap between the U.S. and everybody else cannot be wholly explained by endpoint bias?

Market concentration is not a reason for inaction

At 64.6% of the entire world equity markets and 72.6% of developed markets, the weighting of the U.S. in global markets¹ is now at its highest since the early 1970s, but this is not without precedent. However, with the top 10 stocks now (June 30) accounting for 35% of the S&P 500, the U.S. market is at its most concentrated in 92 years—and growth indices are likely at their most concentrated since the advent of style benchmarks.

"...performance concentration is more of a risk than index construction."

A deeper dive into the figures shows that performance concentration is more of a risk than index construction. As I mentioned at the beginning, the U.S. stock market delivered 6.6% annualized real returns between 1900 and December 2024. However, the years 2022 and 2024 were exceptional for the U.S. market (and by extension global equities) as the top 10 stocks accounted for 63% of overall S&P 500 returns in both years. It is highly unusual for the largest stocks to also be the best performers.

If you were out of the market for the best 20 months of the 125‑year period, your annualized return would drop to 3.6%.² Clearly, over the long term, time in the market is far more important than timing the market.

I’m reminded of a quote from former Berkshire Hathaway vice chairman Charlie Munger: “The big money is not in the buying and selling, but in the waiting.”