Overconfidence Bias

Overconfidence bias is the mental habit of believing your knowledge, forecasting accuracy, and skill are greater than they actually are. It's especially dangerous in investing, because the stronger the confidence, the more you underestimate risk and overestimate expected returns.

A prime example is 'driving skill.' In various studies, about 90% of drivers answered that they were 'above average.' But by definition, only half can be above average. It's arithmetically impossible for more than half to be better than average. This is called the 'better-than-average effect.'

Another is the illusion that 'my predictions are accurate.' When people are asked for a range they're '90% confident' about, the actual rate at which the true answer falls within that range is only about half. In other words, we tend to believe our knowledge is more accurate than it actually is.

The price of this bias appears not as an emotional problem but as 'numbers.' The most famous study is Barber and Odean's 'Trading Is Hazardous to Your Wealth' (2000, Journal of Finance).

Analyzing 66,465 individual households' accounts at a large U.S. discount brokerage from 1991 to 1996, they found the most actively trading group earned about 11.4% per year on average, while the market benchmark over the same period was about 17.9%. That's a lag of nearly 6.5 percentage points.

What's striking is that the 'quality' of the stocks the two groups bought was similar. What made the difference wasn't stock-picking skill but 'how often they bought and sold.' The more they traded, the more costs like fees and taxes piled up, and that ate into returns. The researchers attributed the cause of this excessive trading to 'overconfidence.'

The same research team's follow-up paper, 'Boys Will Be Boys' (2001), analyzed about 35,000 households from 1991 to 1997. Here too, the most frequently trading group produced performance about 7 percentage points below the market per year.

What's interesting is the gender difference. Men traded about 45% more often than women, yet their net returns were actually lower. The stronger the conviction that 'I know better,' the more often they hit the trade button, and the more costs they paid.

The core message is simple. The very conviction that you can frequently beat the market creates costs. The fees, taxes, and bid-ask spread attached to each trade look small, but piled up dozens of times a year, they eat into returns greatly.

You can't eliminate it entirely, but you can build devices to reduce it.

First, record your predictions. When you feel 'this stock seems likely to rise,' write down the reason and the date, and later compare with the outcome. When you confirm with your own eyes that you were wrong more often than you thought, your confidence becomes humbler.

Second, limit the number of trades itself. Trading less reduces both costs and impulsive judgments. That's exactly the lesson of the Barber–Odean research.

Third, know the option of relying on 'the whole market' instead of 'my skill.' Diversification and long-term holding reduce the blow even when individual predictions are wrong. When you check with data how the market fell and recovered in past crises, you can set realistic expectations instead of baseless confidence.