Gambler's Fallacy: Why Streaks Don't Predict Market Reversals

Key Takeaways

• The gambler's fallacy assumes past outcomes of an independent process change the probability of the next outcome, they do not.
• Clotfelter and Cook found lottery bets on recently-drawn numbers dropped sharply after a draw and stayed depressed for months afterward.
• Averaging down on a losing streak because it "has to end" converts the gambler's fallacy directly into an active position-sizing rule.
• Genuine mean reversion exists in some series, valuations, spreads, volatility, but it is a statistical property of those series, not a promise for any short run.

What It Is

Amos Tversky and Daniel Kahneman described the underlying error in their 1971 paper "Belief in the Law of Small Numbers." People expect short sequences drawn from a fair process to look balanced. Five reds in a row on a roulette wheel feel like an imbalance that must be corrected, so bettors shift toward black on the next spin. The wheel does not know what the last five spins were. The probability on the next spin has not changed.

Tversky and Kahneman expanded the treatment in their 1974 Science paper "Judgment Under Uncertainty: Heuristics and Biases." They traced the fallacy to the representativeness heuristic, which makes any short sequence feel like it should mirror the long-run distribution of the process that generated it.

The Intuition

The mind treats independent random processes as if they have memory. After three down days in a row, the market "owes" you a green one. After four quarters of a beat, the next one feels more likely to be a miss. For a truly independent process, both beliefs are wrong, and neither belief is harmless. Acting on them changes sizing and timing.

Two conditions must hold for the gambler's fallacy to apply. The events have to be independent, meaning each trial does not depend on the previous ones. And the underlying probability has to be stable, meaning the process is not changing mid-stream. When either condition breaks, the analysis is different, but the reflex to assume balance still fires.

How It Works

Representativeness makes any short run of an outcome feel non-representative. A fair coin should produce roughly 50 percent heads, so a run of seven heads feels unlikely and pulls your estimate of the next flip toward tails. Tversky and Kahneman showed in repeated experiments that even trained statisticians fall into the pattern when the domain is outside their expertise.

Charles Clotfelter and Philip Cook's 1993 study of Maryland lottery data made the effect concrete in finance-adjacent behaviour. After a number was drawn, the amount of money bet on that number fell sharply and stayed depressed for months. Bettors treated a recent draw as evidence the number was less likely to come up next time, even though every draw is independent and the official odds were unchanged.

In financial markets, many return series are close to independent on short timescales. Daily equity returns, in particular, have weak autocorrelation. That makes the gambler's fallacy especially costly there, because it tells a story the data does not support.

Worked Example

Suppose a stock in your portfolio has closed red four days in a row. You start telling yourself the fifth day is "due" for a bounce and size a short-dated call option slightly larger than you normally would.

For an independent series with roughly 50/50 daily odds, the probability of a green day five is still about 50 percent. The four-day streak has no memory. What the streak does tell you is about recent news flow, positioning, or a regime shift, but that is a narrative analysis, not a probability correction. The gambler's fallacy uses the streak itself as the reason.

Apply the same thinking to an earnings track record. A company has beaten estimates six quarters in a row. You conclude the next beat is less likely because "they cannot keep that up forever." Maybe true, maybe not. But the fact of the six-beat streak is not, by itself, evidence for the reversion. You need an actual reason tied to fundamentals. Without it, you are pricing an imaginary self-correcting mechanism that does not exist.

Common Mistakes

1. Treating a drawdown as a timed correction. After three down weeks, assuming the fourth must be up. Drawdowns have no fixed length.

2. Averaging down on streaks alone. Buying more of a losing position because the streak "has to end" converts the gambler's fallacy into a sizing rule. Without an improved thesis, the position is the same bet, just bigger.

3. Rotating out of a winner "before it reverts." The mirror mistake. A stock that has gone up five weeks in a row is not mechanically more likely to drop next week. It may be overstretched on some other metric. The streak itself is not the evidence.

4. Confusing the gambler's fallacy with mean reversion. Genuine mean reversion exists in some series (valuation ratios, volatility, spreads) on long horizons. That is a statistical property, not a promise that any specific short run must turn.

5. Applying the fallacy to weather, sports, or politics and importing the habit into trading. The same reflex that says a hitter is "due" is the reflex that mis-prices a stock on its fifth down day. Notice the pattern in harmless domains and you will catch it faster in the portfolio.

Frequently Asked Questions

What is the gambler's fallacy in simple terms? The gambler's fallacy is the belief that a streak of one outcome makes the opposite outcome more likely next. After five red spins on a roulette wheel, black feels overdue. But the wheel has no memory. For independent events, each outcome has the same probability regardless of what came before.

How does the gambler's fallacy affect investment decisions? It distorts entry and sizing decisions. An investor who believes a four-day losing streak makes a bounce "due" may size a position larger than the actual probability supports. Averaging down on a losing position "because the streak has to end" converts the fallacy directly into a position-sizing rule without an improved thesis.

What is a real-world example of the gambler's fallacy? Clotfelter and Cook's 1993 study of Maryland lottery data found bets on recently drawn numbers fell sharply after a draw and stayed depressed for months, even though each draw is independent and the odds are identical. In markets: rotating out of a winning sector after a run because it "must mean revert" without any evidence that the reversion mechanism has engaged is the same error applied to a portfolio.

How can investors avoid the gambler's fallacy? Separate the streak from the thesis. A losing streak may reflect new information, changing conditions, or a regime shift, those are legitimate reasons to reconsider. The streak itself, for a process with weak autocorrelation like daily equity returns, is not. If the only reason to act is the streak's length rather than an updated thesis, the fallacy is driving the decision.

How is the gambler's fallacy different from mean reversion? Genuine mean reversion is a documented statistical property of specific series, valuation ratios, volatility levels, credit spreads, on long horizons, measured and tested across large samples. The gambler's fallacy is the unsupported assumption that any streak, in any series, on any timeframe, must reverse because it feels like an imbalance. The difference is whether the reversion mechanism is empirically established or simply assumed.

Sources

1. Tversky, A. & Kahneman, D. (1971). "Belief in the Law of Small Numbers." Psychological Bulletin, 76(2), 105-110. https://web.mit.edu/curhan/www/docs/Articles/15341_Readings/Probability_Subjective/Tversky_Kahneman_1971_Belief_in_the_Law_of_Small_Numbers.pdf
2. Tversky, A. & Kahneman, D. (1974). "Judgment Under Uncertainty: Heuristics and Biases." Science, 185(4157), 1124-1131. https://www.science.org/doi/10.1126/science.185.4157.1124
3. Clotfelter, C.T. & Cook, P.J. (1993). "The Gambler's Fallacy in Lottery Play." Management Science, 39(12), 1521-1525. https://pubsonline.informs.org/doi/10.1287/mnsc.39.12.1521
4. CFA Institute. "The Behavioral Biases of Individuals." Refresher Readings. https://www.cfainstitute.org/insights/professional-learning/refresher-readings/2026/the-behavioral-biases-of-individuals

Disclaimer

This article is educational content only and is not financial advice. Nothing here is a recommendation to buy, sell, or hold any security. Consult a licensed advisor before making investment decisions.