Most lists of behavioral finance biases give you a definition and move on. That is close to useless. A definition does not tell you what the bias looks like when it is costing you money in real time. So each bias below is paired with the clearest documented case available: a study of real brokerage accounts, a market episode, or a well-known experiment.
Overconfidence: the traders who proved it with their own money
Brad Barber and Terrance Odean examined the accounts of 66,465 households at a large US discount broker between 1991 and 1996. The households that traded the most earned 11.4 percent a year after costs. The market returned 17.9 percent. The average household, trading far less, earned 16.4 percent.
The heavy traders were not less intelligent. They were more confident that their next trade would beat the last one, and every trade carried a cost that the presumed skill never recovered. The title of their 2000 paper in the Journal of Finance, "Trading Is Hazardous to Your Wealth", says it plainly.
Representativeness: when adding ".com" was worth 74 percent
Between mid-1998 and mid-1999, dozens of listed companies changed their names to something internet-related, usually by adding ".com". A 2001 study published in the Journal of Finance by Michael Cooper, Orlin Dimitrov and Raghavendra Rau found these announcements produced cumulative abnormal returns of roughly 74 percent over the ten days surrounding the name change, with no sign of the gain reversing afterward. The effect was similar regardless of how involved the company was with the internet. Investors were not evaluating the business. They were matching a label to a category they associated with huge returns.
Herding: 2021 needs no introduction
GameStop ended 2020 trading a little under 20 dollars a share. On January 28, 2021, it hit an intraday high of 483 dollars, according to an SEC staff report on the episode (both prices are from before the company's 2022 stock split). The move was driven overwhelmingly by coordinated retail buying on social media, not by any comparable change in the company's business or earnings prospects.
Whatever you think of the episode's other dimensions, a short squeeze among them, it is a textbook case of herding: individually reasonable people making a collectively extreme decision because everyone around them, online, was making the same one. Three weeks after the peak, the stock was trading near 40 dollars.
Framing: the same outcome, described two ways
In a 1981 experiment published in Science, Amos Tversky and Daniel Kahneman asked people to choose between two programs to fight a disease outbreak expected to kill 600 people. When the options were described in terms of lives saved, most people picked the certain option. When the identical options were described in terms of lives lost, most picked the gamble. Only the wording changed.
Fund marketing, and the case made for active funds, runs on the same effect. "Profitable in nine of the last ten years" and "lost money one year in ten" describe the same track record, and they do not land the same way. Expect to see whichever version sounds better.
Hindsight bias: "I knew it all along"
Psychologist Baruch Fischhoff documented the effect in the 1970s: once people learned how an event turned out, they judged that outcome as having been far more predictable than they had rated it beforehand. After the 2008 financial crisis, a striking number of people described the crash as something they had seen coming, even though very few positioned their money for it at the time. The danger is not the false memory itself. It is that the memory quietly erodes the humility that would otherwise make someone question their next prediction.
The narrative fallacy: when a tidy story explains too much
A stock drops 3 percent and financial media confidently attributes it to a specific piece of news that happened that day. Often that explanation is plausible. Just as often, a comparable stock with no connection to that news also dropped 3 percent the same day. Nassim Nicholas Taleb popularized the term narrative fallacy for this tendency to build a satisfying causal story after the fact, whether or not the story explains what happened.
Loss aversion: the puzzle it helped solve
For decades, economists struggled to explain why stocks have historically outperformed bonds by more than standard risk-aversion models could justify, a mismatch known as the equity premium puzzle. Shlomo Benartzi and Richard Thaler proposed an answer rooted in loss aversion: investors who evaluate their portfolios frequently experience the pain of short-term losses so often that they demand a large premium to hold stocks at all, a pattern they called myopic loss aversion.
The practical takeaway survives even if you find the academic debate dry. Checking a volatile portfolio daily manufactures emotional pain that checking it once a year does not, over exactly the same underlying returns.
Anchoring: why the 52-week high matters more than it should
A stock's 52-week high is displayed on nearly every trading platform, and investors treat it as a reference point for whether a price is "high" or "low". Thomas George and Chuan-Yang Hwang showed in 2004 that how close a stock trades to its 52-week high predicts its future returns better than its past returns do, and that those gains do not reverse later. Their explanation is anchoring: near that familiar number, traders hesitate to bid the price higher even when good news justifies it, so the price only catches up with a delay.
Self-attribution bias: skill when you win, luck when you lose
Ask an investor to explain a winning trade and you will usually hear about research and judgment. Ask about a losing one and the explanation shifts toward bad luck, an unpredictable news event, or market conditions nobody could have foreseen. This asymmetry prevents the honest post-mortem that would improve decision-making over time.
Recency bias: chasing last year's winner
Erik Sirri and Peter Tufano's study of mutual fund flows, published in 1998, found that investors pile disproportionately into the previous period's top performers while pulling money out of poor performers far less aggressively. Research on performance persistence, meanwhile, generally finds limited evidence that recent outperformance predicts future outperformance.
The result is a familiar pattern. Money arrives after the good years, just in time to miss them.
The ten biases at a glance
| Bias | What it means | Documented case |
|---|---|---|
| Overconfidence | Overestimating your own skill and information | Most active traders earned 11.4% a year against 17.9% for the market |
| Representativeness | Judging something by how much it resembles a familiar category | Dotcom name changes, about 74% abnormal return in ten days |
| Herding | Following the crowd instead of your own analysis | GameStop, from under 20 dollars to a 483 dollar high in January 2021 |
| Framing | Reacting to how an outcome is described rather than the outcome | Tversky and Kahneman's 1981 outbreak experiment |
| Hindsight bias | Believing past events were more predictable than they were | Fischhoff's experiments in the 1970s |
| Narrative fallacy | Accepting a tidy causal story built after the fact | Daily explanations of ordinary price moves |
| Loss aversion | Feeling losses roughly twice as strongly as equal gains | Myopic loss aversion and the equity premium puzzle |
| Anchoring | Judging a price against an arbitrary reference point | The 52-week high and momentum |
| Self-attribution | Crediting wins to skill and losses to luck | Post-mortems that blame the market for every loss |
| Recency bias | Overweighting the most recent results | Fund flows chasing last period's top performers |
For where these ideas came from, from prospect theory to the Nobel prizes that followed, see the introduction to behavioural finance.
Frequently asked questions
Which behavioral bias costs investors the most money?
There is no single answer that applies to everyone, but overconfidence-driven overtrading is one of the most measurable: academic research tracking real brokerage accounts has found that the most active individual traders significantly underperform less active ones, largely because trading costs compound while the presumed skill advantage rarely materializes.
Can you train yourself out of these biases?
Awareness reduces but rarely eliminates them, since most of these biases operate below conscious deliberation. Practical countermeasures tend to work better than willpower alone: pre-committing to a rule (such as rebalancing on a fixed schedule regardless of feelings), and having someone else, or a written checklist, review a decision before it is made.