Chasing returns: seven decades of evidence on why investors get it wrong
- Robin Powell
- 10 hours ago
- 7 min read
Two investors can look at the same rising market and reach opposite conclusions about what comes next. One reads a strong run as a reason for confidence; the other reads it as a reason for caution. A new study spanning nearly 70 years of data finds that only one of them has ever been reliably right — and it is not the one chasing returns.
When a market has climbed for a while, the natural feeling is reassurance. The portfolio is up, the trend looks friendly, and the pull is to add to what has been working rather than trim it. Chasing returns, in this sense, feels less like a gamble than like common sense. According to a long run of evidence, it is also the thing most likely to cost an investor money.
That evidence has just been pushed further back than ever before. David Thesmar and Emil Verner, both economists at MIT Sloan, have assembled a record of investor expectations running from 1956 to 2024 — 68 years — in a working paper circulated by the National Bureau of Economic Research and not yet peer-reviewed. Their central finding is that the more an investor's expectations are shaped by what the market has just done, the less those expectations say about what it will do next.
What two groups of investors see in the same market
To test that, Thesmar and Verner built a measure of what sophisticated investors expect from the stock market, drawn from the Value Line Investment Survey, and set it against eight other measures of expectation: surveys of individual investors, of company finance chiefs, of institutions and of professional forecasters. Nine readings, in all, of what investors thought was coming.
The two camps read the same market in opposite directions. Sophisticated investors are contrarian: when prices have risen, they expect lower returns from there. Individual investors do the reverse. They extrapolate, expecting a strong run to continue — the behaviour that defines chasing returns. In the authors' account, naive investors read the recent past as a forecast while the sophisticated stay 'close to rational'.
The decisive part is what happened next. Of the nine measures of expectation, only the sophisticated one reliably predicted future market returns. The bullishness of individual investors predicted nothing useful, and over longer horizons it pointed the wrong way: the more confident the crowd, the weaker the returns that followed. One forecast tracked reality. The other tracked the recent past and called it the future.
How the researchers built a 68-year record
The reach of that record is the reason the finding carries weight. Most studies of investor expectations begin in the 2000s, when survey data became plentiful. Thesmar and Verner went back to the source. They digitised the physical volumes of the Value Line Investment Survey by hand, back to 1956, and joined them to modern digital data running to 2024 — close to 92,000 firm-year observations, covering between 500 and 1,600 companies a year and more than 80 per cent of the US market in most of them.
Value Line is an independent equity-research firm. From its published earnings forecasts and price targets, the authors worked out the returns its analysts were implicitly expecting, then aggregated those into a single annual figure for the market. Over the full period that figure averaged about 13 per cent a year, against an average realised return of roughly 11 per cent — a little optimistic, but moving in the right direction at the right times, which is what matters for prediction.
A near-70-year window is not just a longer version of a short one. Predicting returns is statistically demanding, and samples that begin in the 2000s rarely contain enough independent market cycles to say much with confidence. The extra decades are what let the authors test whether expectations actually forecast returns, rather than merely describe a mood.
The same past, read two ways
Put numbers to the divergence and it becomes concrete. The sophisticated forecast moves against the prior year's return: when the market has gone up, the expected return drops. The relationship is statistically clear and consistent across the record. Individual investors move the other way, marking their expectations up after a good year — the familiar habit of chasing yesterday's winners. Same information, opposite response.
This is not a new observation. In 2014, Robin Greenwood and Andrei Shleifer documented the same extrapolative pattern among retail investors, drawing on six separate surveys between 1963 and 2011, in the Review of Financial Studies. When ordinary investors expected high returns, they found, the evidence-based forecast was pointing the other way: high investor optimism lined up with what they called 'low' model-based expected returns.
What Thesmar and Verner add is the part the earlier work could not settle: which reading comes true. Across the nine series they compared, the sophisticated forecast was the only one that positively and significantly predicted future returns, and the relationship grew stronger as the horizon lengthened, clearing the usual tests of statistical significance by a wider margin the further ahead the authors looked. The extrapolative crowd, reading the same past, got the direction wrong.
What chasing returns costs ordinary investors
This is not an argument that lives only in the data. The behaviour the study describes — chasing returns by buying after good runs and selling when the news turns bad — has a measurable price, and Morningstar puts a number on it each year in its Mind the Gap study.
Over the decade to the end of 2024, the average dollar invested in US funds earned 7 per cent a year, against the 8.2 per cent those same funds returned — a shortfall of 1.2 percentage points a year, given up purely through the timing of investors' own buying and selling. That gap is worth about 15 per cent of the total return the funds produced, forgone not to fees or bad funds but to the decision of when to be in them.
The more investors traded, the wider the gap grew. Funds whose investors moved money in and out most gave up roughly twice as much as those whose investors sat still. Morningstar is careful not to read the whole gap as panic and greed: some of it comes from ordinary habits, such as paying in month by month rather than all at once. But the direction is consistent across every ten-year period the firm has measured.
Why the instinct is so hard to shake
Extrapolation is not a failure of intelligence. It is close to the default way people form expectations about an uncertain future, and the behavioural-finance literature has modelled it for years; Thesmar and Verner build their own account on an extrapolative model of asset prices. Reaching for the recent past when you try to picture what comes next is normal cognition, not a character flaw.
Recent work suggests why it is so durable. A separate group of economists, including Shleifer, has found that beliefs about the economy are shaped by personal experience and emotional context, not only by the facts in front of us — people reach for whatever is most readily brought to mind. That study examined expectations about inflation and house prices rather than stock returns, so it points to a plausible mechanism rather than proving one for equities. The stock-market case rests on Thesmar and Verner, and on Greenwood and Shleifer, not on it.
Where the evidence stops
Three caveats are worth stating plainly. The first is that this is a working paper, not yet peer-reviewed, covering US stocks and using a single research firm as the stand-in for sophistication. The second is that the sophisticated forecast is slightly too optimistic on average, and its predictive power is strongest over several years, not over the next quarter.
The third is the most important. When the authors control for a standard measure of how cheap or expensive the market is, the forecast's edge largely disappears — which suggests it works because it tracks valuation, not because anyone holds a private crystal ball. That matters for how the finding should be read. It is not evidence that the market can be timed, and no one in this data has a reliable way of doing so; if anything, it is a reminder of how losses tend to harden the very habits that cause them. The sturdier point is that of the nine signals studied, the least reliable was the sentiment of ordinary investors — the reader's own included.
The signal worth ignoring
Across nearly seven decades, then, the forecast that came true was the contrarian one, and the forecast that misfired belonged to the crowd — the same crowd most investors join the moment they start chasing returns. The feeling that tempts someone to pile in after a strong run, or to bail out after a weak one, is precisely the signal the evidence says to distrust.
The difficulty is that the feeling is so convincing from the inside. Confidence built on a rising market feels like judgement. Seven decades of data suggest it is closer to an echo of the recent past, and the most useful thing it has to tell you is usually to do nothing at all.
Resources
Bordalo, P., Gennaioli, N., Lopez-de-Silanes, F., Schröder, S. G., Shleifer, A., & van Rooij, M. (2026). The psychology of macroeconomic expectations. NBER Working Paper No. 35214.
Greenwood, R., & Shleifer, A. (2014). Expectations of returns and expected returns. Review of Financial Studies, 27(3), 714–746.
Ptak, J., & Arnott, A. (2025). Mind the gap 2025. Morningstar.
Thesmar, D., & Verner, E. (2026). Beliefs and stock market fluctuations: New evidence from the past seven decades. NBER Working Paper No. 35286.
How to stop chasing returns
The hardest part of evidence-based investing is rarely choosing the funds. It is sitting still once you own them, particularly when a rising market is quietly suggesting you do more. How to Fund the Life You Want, by TEBI editor Robin Powell and Jonathan Hollow, is written for UK investors who want a plan they can hold to rather than one they keep adjusting on instinct. Bloomsbury published a second, updated edition in May 2026, and it covers how to save, invest and draw a retirement income without letting sentiment make the decisions. Buy it on Amazon.
For readers who would rather have a professional alongside them, TEBI's Find an Adviser directory lists advisers who have publicly committed to evidence-based investing.
