07-08-2025
Want better returns? Forget risk. Focus on fear
An investor will take on more risk only if they expect higher returns in compensation. The idea is a cornerstone of financial theory. Yet look around today and you have to wonder. Risks to growth—whether from fraught geopolitics or vast government borrowing—are becoming ever-more fearsome. Meanwhile, stockmarkets across much of the world are at or within touching distance of record highs. In America and Europe, the extra yield from buying high-risk corporate bonds instead of government debt is close to its narrowest in over a decade. Speculative manias rage around everything from cryptocurrencies and meme stocks to Pokémon cards.
A common explanation for effervescent markets is that investors have become reckless or outright irrational. Or perhaps the relationship between risk and return simply is not there, posits a working paper by Rob Arnott of Research Affiliates, an investment firm, and Edward McQuarrie of Santa Clara University. They argue that over the past two-and-a-bit centuries, risk (as conventionally defined) has done a lousy job of explaining the relative returns of stocks and bonds. In its place, they propose fear—a more complex thing—as the real driving force of markets.
Standard portfolio theory says a stock's uncertain future returns are distributed along a bell curve. The expected return lies under the peak, and risk is equivalent to the curve's variance, or spread. These assumptions make the maths elegant and, more important, tractable. But they are also flawed. Stock returns do not in fact follow a bell curve: they take extreme values too often and are asymmetric. Investors, meanwhile, do not regard the curve's full spread as risky, but just the side of it corresponding to losses. Who, however risk-averse, would be upset by an outsize return?
What is more, risk theory gives an inadequate account of historical returns. A core prediction is the 'equity risk premium', meaning the tendency of stocks, being riskier, to deliver better long-term returns than government bonds. To test this, Mr McQuarrie compiled American stock and bond prices going back to 1793, using data from newspaper archives. Previous studies had seemed to establish the equity risk premium as a persistent, relatively stable property of markets; his new database calls that into question.
An investor who bought American stocks in 1804 would have had to wait 97 years before their return beat that of bonds. By 1933 they would have fallen behind again. A statistical test of the relationship between variance and return, over the database's full timespan, failed even to find a 'modest or inconstant' risk premium. The cumulative equity risk premium (up to 2023) has nevertheless been large. But 70% of it came from an exceptional period between 1950 and 1999; the rest of the time, stocks' relative performance was middling or poor. And these, after all, were results for one of the world's best-performing stockmarkets. Other researchers have shown that, since 1900, those of other countries have on average returned far less.
Realised variance and returns contain both expected and unexpected elements, so no theory is likely to match the data perfectly. Even so, the scale of these departures from what risk theory would predict, over such a long timespan, warrants a search for a new framework. Messrs Arnott and McQuarrie propose that instead of pricing assets by their variance, investors price them according to two fears: fear of loss (FOL) and fear of missing out (FOMO). Whereas risk is measured by variance, FOL refers only to its downside (or 'semivariance'). An asset inspires FOMO if it has the chance of wild, unexpected gains that those shunning it might miss. This is measured by the 'skewness', or asymmetry, of its return distribution.
Rather than working through fear theory's maths, which they admit is formidable, the authors hope to tempt others to investigate it with them. They might just succeed. As well as being a widespread, often rational impulse, FOMO helps explain why people would buy overpriced stocks, or even speculative assets with no fundamental source of returns. Its absence from conventional theory seems like an error. And FOL describes how people actually think of risk far better than variance does. Just like investors' mood and market dynamics, the balance between the two can vary dramatically with time and circumstance. The historical record suggests that portfolio theory needs some new ideas. Fear might be just the thing.
