Overconfidence bias

Overconfidence bias is the tendency to have excessive faith in the accuracy of your knowledge and predictions. You overestimate how much you know, how reliable your information is, and how likely you are to outperform. Studies show that roughly 90% of investors believe they are above average, an impossibility that reveals the depth of overconfidence.

The core phenomenon

Overconfidence bias has three components:

Overplacement. You rate yourself above average, even though by definition most cannot be above average. Studies show 90% of drivers believe they are above average, 93% of college professors believe they are above average teachers, and similarly, most investors believe they will beat the market.

Overprecision. You are overconfident in the accuracy of your knowledge. When asked to provide a 90% confidence interval for a quantity (e.g., “S&P 500 return next year”), people typically give ranges that are too narrow. Actual values fall outside these ranges far more often than the 10% you predicted.

Overplacement combined with overestimation of others’ mistakes. You not only overrate yourself; you underrate others. “Most investors are dumb; I am smart.” This combination is especially toxic.

Why overconfidence exists

Several mechanisms drive overconfidence:

Selective memory. You remember your successful calls and forget your misses. Your wins feel more memorable and real. Over time, this biased memory makes you feel more competent than you are.

Confirmation bias. You seek and find evidence that supports your views, which reinforces your confidence in them.

Limited feedback. In complex domains like investing, feedback is delayed and noisy. You might make a bad decision that happens to work out because of luck, reinforcing your confidence in the decision process. The feedback loop is broken.

Illusion of understanding. The more you know about a domain, the more you feel you understand it. But knowledge and prediction ability are not the same. You can know a lot about a company and still have no idea whether its stock will outperform.

Overconfidence and active investing

Overconfidence is the fuel of active investing. Studies show that individual stock-picking does not beat the market. Professional fund managers also mostly underperform. Yet overconfident investors persist in the belief that they can beat the market.

Overconfidence leads to:

Excessive trading. Confident that you can time the market or pick winners, you trade frequently. Trading costs (commissions, taxes, bid-ask spreads) eat the returns, and the overconfident investor underperforms the passive holder.

Concentrated bets. Overconfident that you have found a winner, you overweight a single stock or sector. This increases risk without increasing expected return (overconfidence is not the same as skill).

Market timing. Overconfident that you can forecast market moves, you shift allocations based on your predictions. You tend to move away from the market near lows (when you are most pessimistic) and into the market near highs (when you are most optimistic). The opposite of buy low, sell high.

Overconfidence and confirmation bias

Overconfidence and confirmation bias amplify each other. Overconfident that you are right, you seek confirming evidence, which you find. This reinforces your confidence, which drives more confirmation-seeking. The cycle becomes self-reinforcing and hard to break.

An investor overconfident in a stock will read bullish research and ignore bearish research. The bullish research reinforces her confidence, which drives more bull-seeking, and so on.

Overconfidence and professional investors

Even professional investors are overconfident. Studies of mutual fund managers show they believe in their skill despite evidence that most do not beat the market. The more experience a manager has, the more confident she tends to be — even when the data shows experience does not improve returns.

This might seem paradoxical: how can professionals be overconfident if markets are competitive? The answer is that overconfidence is partly about the complexity of the environment. In complex, high-noise domains, it is easy to mistake luck for skill. A manager who beats the market for three years might attribute it to skill, even though random variation alone could easily produce such runs.

Degrees of overconfidence

Overconfidence is not uniform. It is strongest in:

• Domains with delayed feedback (investing is especially prone)
• Domains with high uncertainty (market timing is especially prone)
• Individuals who are intelligent or well-informed (knowledge increases confidence, but not accuracy)
• Individuals with recent success (recent wins boost confidence)

Defenses against overconfidence

• Track your predictions. Keep a record of your forecasts and actual outcomes. Over time, you will see that your confidence intervals are too narrow and your predictions are less accurate than you believe.
• Use a decision framework, not judgment. Rather than trusting your judgment to pick stocks or time the market, use a mechanical framework based on valuation, diversification, and asset allocation. The framework does not rely on your predictions.
• Use index funds. The simplest defense against overconfidence is to hold a diversified index fund and stop trying to beat the market. Accept that the market return is good enough.
• Get feedback quickly. In domains where you can get feedback, seek it out. Backtesting your strategies on historical data, or tracking a “paper portfolio,” gives you feedback on your decision-making without real money at stake.
• Recognize that luck and skill are hard to distinguish. Even if you beat the market for five years, is it skill or luck? Statistically, luck alone would produce some five-year winners. Be humble.
• Diversify within your investment approach. Rather than concentrating in your best ideas, hold a diversified portfolio. This protects you against overconfident mistakes.

See also