Less than two months of 2024 have passed, but the year has already been a pleasing one for stock market investors. The S&P 500 index of big American companies is up by 6% and has passed 5,000 for the first time ever, driven by a surge in enthusiasm for tech giants, such as Meta and Nvidia. Japan’s Nikkei 225 is tantalisingly close to passing its own record, set in 1989. The roaring start to the year has revived an old debate: should investors go all in on equities?

Less than two months of 2024 have passed, but the year has already been a pleasing one for stock market investors. The S&P 500 index of big American companies is up by 6% and has passed 5,000 for the first time ever, driven by a surge in enthusiasm for tech giants, such as Meta and Nvidia. Japan’s Nikkei 225 is tantalisingly close to passing its own record, set in 1989. The roaring start to the year has revived an old debate: should investors go all in on equities?

A few bits of research are being discussed in financial circles. One was published in October by Aizhan Anarkulova, Scott Cederburg and Michael O’Doherty, a trio of academics. They make the case for a portfolio of 100% equities, an approach that flies in the face of longstanding mainstream advice, which suggests a mixture of stocks and bonds is best for most investors. A portfolio solely made up of stocks (albeit half American and half global) is likely to beat a diversified approach, the authors argue—a finding based on data going back to 1890.

Why stop there? Although the idea might sound absurd, the notion of ordinary investors levering up to buy assets is considered normal in the housing market. Some advocate a similar approach in the stockmarket. Ian Ayres and Barry Nalebuff, both at Yale University, have previously noted that young people stand to gain the most from the long-run compounding effect of capital growth, but have the least to invest. Thus, the duo has argued, that youngsters should borrow in order to buy stocks, before deleveraging and diversifying later on in life.

Leading the other side of the argument is Cliff Asness, founder of AQR Capital Management, a quantitative hedge fund. He agrees that a portfolio of stocks has a higher expected return than one of stocks and bonds. However, he argues that it might not have a higher return based on risk taken. For investors able to use leverage, Mr Asness argues it is better to choose a portfolio with the best balance of risk and reward, and then to borrow to invest in more of it. He has previously argued that this strategy can achieve a higher return than a portfolio entirely made up entirely of equities, with the same volatility. Even for those who cannot easily borrow, a 100% equity allocation might not offer the best return based on how much risk investors want to take.

The problem when deciding between a 60%, 100% or even 200% equity allocation is that the history of financial markets is too short. Arguments on both sides rely—either explicitly or otherwise—on a judgment about how stocks and other assets perform over the very long run. And most of the research which finds that stocks outperform other options refers to their track record since the late 19th century (as is the case in the work by Ms Anarkulova and Messrs Cederburg and O’Doherty) or even the early 20th century.

Although that may sound like a long time, it is an unsatisfyingly thin amount of data for a young investor thinking about how to invest for the rest of their working life, a period of perhaps half a century. To address this problem, most investigations use rolling periods that overlap with one another in order to create hundreds or thousands of data points. But because they overlap, the data are not statistically independent, reducing their value if employed for forecasts.

Moreover, when researchers take an even longer-term view, the picture can look different. Analysis published in November by Edward McQuarrie of Santa Clara University looks at data on stocks and bonds dating back to the late 18th century. It finds that stocks did not consistently outperform bonds between 1792 and 1941. Indeed, there were decades where bonds outperformed stocks.

The notion of using data from such a distant era to inform investment decisions today might seem slightly ridiculous. After all, finance has changed immeasurably since 1941, not to mention since 1792. Yet by 2074 finance will almost certainly look wildly different to the recent era of rampant stockmarket outperformance. As well as measurable risk, investors must contend with unknowable uncertainty.

Advocates for diversification find life difficult when stocks are in the middle of a rally, since a cautious approach can appear timid. However financial history—both the lack of recent evidence on relative returns and glimpses at what went on in earlier periods—provides plenty of reason for them to stand firm. At the very least, advocates for a 100% equity allocation cannot rely on appeals to what happens in the long run: it simply is not long enough.

© 2023, The Economist Newspaper Limited. All rights reserved. From The Economist, published under licence. The original content can be found on www.economist.com