Why the Investors Most Afraid of a Market Crash Aren't Moving Markets
By Alpesh Patel OBE, hedge fund manager and author
Short answer: For three decades, wealthy and professional investors have put the odds of a catastrophic stock market crash within six months at about 19%. Option prices, where people put real money behind their views, have put the odds of a 20% fall at about 5.6%. New research from Yale shows why the two disagree. The most frightened investors step back from the market, so their views stop showing up in prices. The people who stay take on the risk they've left behind, and they tend to be paid for it. The lesson isn't "buy when others are scared". It's this: build a portfolio and a written process that can survive the crash you fear, so fear never makes the decision for you.
Before you read on, answer one question. Robert Shiller, the Nobel prize-winning economist, has asked investors this question for more than 30 years:
"What do you think is the probability of a catastrophic stock market crash in the U.S., like that of October 28, 1929 or October 19, 1987, in the next six months, including the case that a crash occurred in the other countries and spreads to the U.S.?"
Write your number down. We'll come back to it at the end.
The 19% vs 5.6% gap
Shiller's Investor Confidence Survey is run by the International Center for Finance at Yale School of Management. Each month about 300 questionnaires go out to high-net-worth individual investors and institutional investors. A different group answers each month. The aim is 20 to 50 responses from each type.
William Goetzmann, Dasol Kim and Robert Shiller have now analysed the individual responses from 1996 to 2025 in a new Cowles Foundation discussion paper, "Non-Marginal" Investor Beliefs (September 2026). Here's what they found.
Survey respondents | Option prices (market-implied) | |
|---|---|---|
Chance of a crash in the next six months | 19.0% | 5.6% |
Expected S&P 500 return over six months | +1.4% | +4.2% |
Source: Goetzmann, Kim and Shiller (2026), Cowles Foundation Discussion Paper 2558, sample 1996 to 2025. The option-implied crash figure is the market-implied chance of the S&P 500 falling 20% or more within six months.
Three details matter.
The gap never closed. The survey's crash probability was above the option-implied figure in every single month of the sample.
It's getting wider. The gap widened by about a quarter of a percentage point a year. Almost all of that came from survey respondents getting gloomier, not from option prices changing.
The comparison isn't perfect. Shiller's question asks about a "catastrophic" crash without defining it. Still, a gap that big, lasting 30 years, tells you something.
These aren't nervous beginners. They're wealthy individuals and professional money managers, and as a group they've consistently felt far more danger than the market was pricing in.
Why crash fear runs so high: the headlines effect
An earlier paper by the same three authors, Crash Beliefs From Investor Surveys (NBER Working Paper 22143, 2016, revised 2017), looked at 26 years of the same survey. It started from a simple fact: historically, the base rate for a severe single-day crash is low. Investors put it much higher.
The researchers then asked what drives those estimates up. Their findings:
Recent falls raise crash fears. After the market dropped, respondents thought another crash was more likely.
The financial press raises them too. Bad market events that got more media attention went with higher crash probabilities.
Even earthquakes do it. Unrelated shocks such as earthquakes were linked to higher crash estimates, even though they had nothing to do with share prices.
Fear moves money. Higher crash probabilities went with weaker flows into mutual funds.
The authors say this fits availability bias: if something is vivid and easy to recall, it feels more likely than it really is. A dramatic headline, a scary chart on the news, a friend who sold everything in 2008. None of these change the odds of a crash. All of them change how likely it feels.
You'll recognise the pattern. When markets wobble, the scary headlines multiply. When markets hit a record, people worry about a crash because "it's due". The feeling of danger rarely matches the data.
Who steps back from stocks
The new Yale paper goes a step further. It asks: who exactly is the most frightened, and do their views end up in prices?
Shiller's survey asks respondents whether this statement is true or false:
"Although I expect a substantial rise in stock prices in the U.S. ultimately, I advise being less invested in stocks for the time being because I think that prices are likely to drop for a while."
Read the first half again. "I expect a substantial rise ultimately." That's the classic market-timing impulse: shares for the long run, just not right now.
On average, 31.3% of respondents answered "true". According to Robin Powell's write-up for Index Fund Advisors, that rose to close to 80% at the height of the financial crisis, and it was high again during the 2022 fall.
Now split the two groups and compare what each expected over the next six months, measured on the same days:
Group | Expected six-month return |
|---|---|
Advised being less invested ("step out") | -1.2% |
Stayed invested | +3.3% |
Option prices on the same days | +4.3% |
Source: Goetzmann, Kim and Shiller (2026).
The people staying in the market held views close to the market price. The people stepping out were nowhere near it. As the authors put it: "The participating group holds beliefs that coincide with the price, while the non-participating group is furthest from it."
And crash fear was the key. Respondents who put higher odds on a crash were more likely to advise stepping out, even after allowing for their return forecasts. In plain English, it wasn't mainly that they expected poor returns. It was that they feared a disaster.
The researchers call these investors "non-marginal". In economics, the "marginal" investor is the one whose buying and selling sets the price. Once you step out, your opinion no longer counts in the price, however strongly you hold it.
Why the market often rises when investors expect it to fall
This links to a well-known puzzle in finance. In 2014, Harvard's Robin Greenwood and Andrei Shleifer published Expectations of Returns and Expected Returns in the Review of Financial Studies (vol. 27, no. 3, pages 714 to 746). They studied investor expectations from six different sources between 1963 and 2011.
What they found:
The six measures of expectations moved closely together.
Expectations rose after strong markets and when the market was high.
But they were strongly negatively correlated with model-based measures of expected returns. When investors felt most optimistic, the models said future returns were likely to be lowest.
Their explanation was extrapolation: people project what the market has just done into the future. After a boom, they expect more. After a fall, they expect worse. Greenwood and Shleifer built a model in which "fundamental" traders demand a reward for absorbing the mood swings of extrapolators, and concluded that "markets are not efficient".
The new Yale paper finds the same pattern in Shiller's data. When survey respondents expected higher returns, the market tended to deliver lower returns over the next six months. A one-standard-deviation rise in survey expectations went with about 1.6 percentage points lower returns over the following six months. Expectations taken from option prices pointed the other way and were much better at predicting what actually happened.
But the Yale authors offer a different explanation, and I think it's the most useful idea in the paper.
A survey gives every person one vote. Prices give a vote to the money that's actually invested.
When frightened investors step out, someone still has to own the shares. The risk doesn't vanish. It gets concentrated in fewer hands. The investors who remain are carrying more of it, so they demand a higher expected return for doing so.
The evidence leaned their way. The further survey forecasts fell below the returns implied by option prices, the better the market tended to perform over the next six months, especially when more respondents were advising people to step out. In the authors' words, the measures diverge "because the set of investors bearing risk changes, rather than because the investors setting prices have erred."
Yale's own research brief puts it simply: fearful investors who step out "still influence prices by leaving more risk to those who remain and raising the returns they demand."
The pattern even shows up in individual shares. Stocks that tend to do well when fear is pushing people out of the market act like insurance. Investors accept lower average returns to hold them, just as you accept paying a premium for home insurance.
What this means for you
Put the pieces together and you get an uncomfortable picture for anyone who sells because they're scared:
Your fear is probably overstated. Even sophisticated investors put crash odds roughly three times higher than the market does, and the gap is growing.
Your fear is partly manufactured. Headlines, recent falls and unrelated shocks push crash fears up without changing the actual odds.
Stepping out removes your voice from prices. Your pessimism stops counting in the market the moment you sell.
Stepping out hands the reward to someone else. The investors who stay carry more risk and, on this evidence, have tended to be paid for carrying it.
And you have to be right twice. You have to get out at the right time, then get back in at the right time. Get either wrong and you can end up worse off than if you'd done nothing. I've written about what that costs in the hidden cost of market timing.
What the research can't tell you
To be fair to the evidence, here are the limits.
It's a working paper. The 2026 Yale study hasn't yet been through peer review.
It doesn't follow people over time. Each month's survey reaches a mostly different group.
"Stepping out" means agreeing with a statement. The researchers don't see actual trades or actual returns, so the paper can't prove that those who stepped out lost money.
It's US data. The authors themselves say testing other markets is the next step.
Option benchmarks rely on modelling assumptions.
It is not a trading signal. If gloomy surveys tend to come before better returns, it's tempting to buy whenever everyone else is scared. The relationship is too weak to trade on, and the paper offers no trading rule. Buying because others are frightened is still a forecast. It's market timing in reverse.
Where I agree with IFA, and where I differ
Robin Powell's article for Index Fund Advisors is excellent, and I agree with its central point: the useful question isn't "when is the next crash?" It's "can my portfolio survive the crash I fear?"
IFA's answer is index funds, held for the long term and rebalanced by rule. That's a perfectly reasonable answer for many people. Mine is different in one respect. I think a self-directed investor can do better than owning everything, provided they choose shares with a written, data-driven process rather than a story. In the Great Investments Programme I select shares using value, growth, income, cash return on capital invested (CROCI) and the Sortino ratio, which measures return against downside risk. I explain the framework in how professional investors screen stocks.
Where we agree completely is that a feeling about the future is not a process.
People sometimes ask me about my own Daddy Bear rule, which moves my money to cash when short- and medium-term average prices fall across the S&P 500 and its biggest companies while interest rates are rising, as in 2022. Isn't that market timing? The difference is the one this research is all about. The investors in Shiller's survey were acting on a forecast: "prices are likely to drop for a while." Daddy Bear is a written rule about what has already happened. It lags. It doesn't try to call the top or the bottom. It has triggered roughly once every five to seven years, and I wrote it down long before I needed it. It's what I do with my own money, not a recommendation for you, and most people should not try to time the market at all.
How to handle crash fear: six steps
1. Put a number on the fall you can sit through
Forget the probability of a crash. Ask a better question: how big a fall could I watch without selling? In 2022 the S&P 500 fell about 25% from peak to trough and took around two years to get back to its old high. Plan for future falls to be at least as bad as past ones. If a 25% fall would make you sell, your portfolio is too risky for you, and that's worth fixing once, calmly, not in the middle of a panic. Past performance is not a reliable indicator of future results.
2. Hold two years of spending in cash
If you're drawing an income, keep about two years of what your investments need to provide in cash. I keep my own safe money in high-interest bank accounts. When markets fall, you spend the cash instead of selling shares at the bottom. That single step removes most of the pressure that turns fear into a sale.
3. Write your crash rule before the crash
Decide now, while you're calm, exactly what you'll do if the market falls 10%, 20% or 30%. Write it down. For many people the rule will be "nothing, plus top up the cash buffer when markets recover". The point is that the decision has already been made. I go through the different types of fall in the Goldilocks strategy for every type of bear market.
4. Go on a headline diet
The 2016 paper showed that media attention and even earthquakes push crash fears up. You can't control the news, but you can control how much of it you consume. Check your portfolio on a schedule, not every time a headline scares you.
5. Judge your holdings by their downside, not their story
Ask two questions of every holding: what did it do in the last big fall, and how much return has it delivered for each unit of downside risk? That's what the Sortino and Calmar ratios measure. See Sharpe vs Sortino vs Calmar for how to use them.
6. Use a benchmark to manage regret
Fear of a crash and fear of missing out are two sides of the same coin. One approach we teach in the Great Investments Programme is to keep part of your money in an S&P 500 tracker as a familiar core and run your chosen shares around it. When the index rises, you're not left behind. When your shares do better, you see it clearly. Investor psychology matters as much as stock selection.
Watch: what historical data says about stock market crashes
Back to your number
So, what probability did you write down at the start?
If it was well above 5.6%, you're in good company. Shiller's wealthy and professional respondents averaged 19%. But Shiller's question asks you to predict the market, and nobody can do that reliably, including me.
The better question is the one only you can answer: what is the biggest fall you would sit through without selling? If you don't know, find out before the next fall finds out for you. And if you discover your honest answer is smaller than the falls history shows, the fix isn't to sell everything and wait for a better moment. It's to build a cash buffer, a written process and a crash rule, and then let the investors who panic hand the reward to you.
Frequently asked questions
How likely is a stock market crash in the next six months?
Nobody knows. Option prices on the S&P 500 implied about a 5.6% chance of a 20% fall within six months on average between 1996 and 2025. Surveyed investors put the chance of a "catastrophic" crash at about 19%. Neither is a forecast you should trade on.
Should I sell my shares if I think a crash is coming?
The research suggests that the investors most worried about a crash are the furthest from market prices, and that those who stay invested have tended to be compensated for the risk they take on. Selling on a feeling means you have to be right twice, about when to get out and when to get back in. If a fall would force you to sell, it's better to reduce your risk once, deliberately, and hold a cash buffer.
What is a "non-marginal" investor?
It's the Yale researchers' term for an investor who has stepped out of the market. Prices are set by the "marginal" investors who are still buying and selling, so the views of those who've stepped out no longer register directly in prices.
Why does the market often rise when investors expect it to fall?
Greenwood and Shleifer (2014) put it down to extrapolation: investors expect recent trends to continue. Goetzmann, Kim and Shiller (2026) add that when fearful investors step out, the remaining investors carry more risk and demand a higher return, which helps explain why widespread pessimism has often come before better returns.
Is buying when everyone is scared a good strategy?
Not as a rule. The Yale paper says the relationship is too weak to trade on. Buying because others are frightened is still a forecast. A written process that tells you what to do in advance is more reliable than trying to read the crowd.
Not sure your portfolio could survive the fall you fear?
Book a call with my team. We'll look at what you hold, how you choose your investments now, how they'd cope with a 2022-style fall, and whether the Great Investments Programme makes sense for you.
Sources
Goetzmann, W. N., Kim, D. and Shiller, R. J. (2026). *"Non-Marginal" Investor Beliefs*. Cowles Foundation Discussion Paper No. 2558, Yale University.
Cowles Foundation (2026). Why Does the Market Rise When Investors Expect It to Fall? Research brief, 11 September 2026.
Goetzmann, W. N., Kim, D. and Shiller, R. J. (2016, revised 2017). *Crash Beliefs From Investor Surveys*. NBER Working Paper 22143.
Greenwood, R. and Shleifer, A. (2014). *Expectations of Returns and Expected Returns*. Review of Financial Studies, 27(3), 714 to 746.
Powell, R. (2026). Why the Investors Most Afraid of a Market Crash Aren't Moving Markets. Index Fund Advisors, 12 October 2026.
Important: This article is for education only and is not personal financial advice. Alpesh Patel Ventures Ltd provides general investment education and is not authorised by the Financial Conduct Authority to give regulated investment advice. The value of investments can go down as well as up, and you may get back less than you invest. Past performance is not a reliable indicator of future results. The research discussed uses US data and may not apply in other markets.
Comments