Self-Serving Bias: How Attribution Errors Destroy Investment Returns

Key Takeaways

• Self-serving bias attributes positive outcomes to personal skill and negative outcomes to external factors outside your control.
• Barber and Odean found men traded 45% more than women, cutting their net returns by 1.4 percentage points per year on average.
• A common error is reviewing winning trades briefly and losing trades with elaborate macro explanations, asymmetric scrutiny that trains the bias.
• Pre-registering entry and exit criteria with a post-trade attribution audit are the process tools that interrupt self-serving feedback loops.

What It Is

Dale Miller and Michael Ross published the canonical review in their 1975 paper "Self-Serving Biases in the Attribution of Causality: Fact or Fiction?" in Psychological Bulletin. They surveyed a large experimental literature and concluded that people reliably make self-enhancing attributions after success, taking more personal credit than the evidence warrants. Evidence for self-protective attributions after failure, where people push blame outward, was present but less consistent.

Later reviews, including James Shepperd, Wendi Malone, and Kate Sweeny's 2008 overview "Exploring Causes of the Self-Serving Bias," showed the effect is stable across cultures and domains, including academic performance, job reviews, sports, medical judgment, and financial decisions.

The Intuition

Every decision you make produces a result. To learn, you need a clean signal about how the decision contributed to the result. The mind does not like assigning that signal accurately because your self-image is on the line. A clean positive outcome that you did not cause is threatening in a subtle way. A clean negative outcome that you did cause is worse.

Self-serving attribution solves both problems by sliding credit toward you in the first case and away from you in the second. The solution feels emotionally stable. The problem is that you are now tracking a false record of your own skill.

How It Works

Two mechanisms drive the bias. Motivational explanations, emphasised by Miller and Ross, say people distort attributions to protect self-esteem. Feeling competent is pleasant and costly to give up. Cognitive explanations say the pattern falls out of how people form expectations. Most people expect success more than failure, so success lines up with their plan and gets coded as "caused by me," while failure does not match the plan and gets coded as "caused by the situation."

The two mechanisms are not mutually exclusive and modern reviews treat them as complementary.

In financial markets, Brad Barber and Terrance Odean's 2001 paper "Boys Will Be Boys" showed how self-attribution fuels overtrading. Investors who attribute past gains to skill trade more, pay more in commissions and spreads, and underperform buy-and-hold strategies. The 2001 sample found that men traded 45 percent more than women and their net returns were about 1.4 percentage points lower per year. The gap was consistent with self-serving attribution turning lucky wins into confident skill claims.

Worked Example

Imagine you take two positions in the same week. The first is a beaten-down chemical stock on a cyclical thesis. It rallies 30 percent over the next quarter. The second is a small position in an AI hardware name that matches a general thesis you have been running. It falls 25 percent on a supply-chain disclosure.

A self-serving write-up sounds like this. "Chem call validated my cycle model. Entered at the right time, sized correctly, capital allocation is working. AI name got hit by a macro disclosure nobody could have predicted. Thesis still intact."

Now audit both. The chemical rally may have ridden a broad commodity move that lifted most peers by similar amounts. Your contribution might be modest. The AI disclosure may have been hinted at in the last two quarterly calls, filings you did not open. Your contribution to the loss may be larger than "nobody could have predicted" suggests.

Without the audit, your stored record grows a little more flattering with every trade. The next position is sized by a version of you that does not actually exist.

Common Mistakes

1. Reviewing winners and losers with different rigor. A win gets a two-sentence note. A loss gets a page of macro explanations. The asymmetry trains the bias into your process.

2. Using narrative to defend outcomes. "I would have made that trade work if the Fed had not surprised" is a self-protective attribution dressed as analysis. It may be true, but it should be tested against what you actually did, not just what the market did.

3. Confusing beta with alpha after a good year. In a strong equity year, almost every long position makes money. Self-serving attribution reads the whole result as skill. A factor-adjusted benchmark tells a different story.

4. Firing informants who are right when you are wrong. An analyst or signal that flagged the losing position gets demoted in your mind because their accuracy embarrasses your write-up. This is self-serving bias expressed as pushing away uncomfortable evidence.

5. Publishing only the winners. If you post trades to social media, a curated win record trains both you and your audience to believe in a hit rate that never existed.

Frequently Asked Questions

What is self-serving bias in investing? Self-serving bias is the habit of crediting wins to your own skill and blaming losses on external factors. It distorts the feedback loop that would otherwise help you identify genuine edge and correct genuine mistakes, each trade's result gets recorded in a way that flatters the decision-maker.

How does self-serving bias affect investment decisions? It fuels overtrading. Investors who attribute gains to skill trade more frequently, pay more in costs, and underperform passive alternatives. Barber and Odean found that men traded 45 percent more than women, with excess activity producing 1.4 percentage points less in annual net return, a gap consistent with self-attribution converting lucky wins into overconfident skill claims.

What is a real-world example of self-serving bias? A portfolio beats its benchmark in a strong equity year. The investor attributes the outperformance to stock selection. A factor-adjusted analysis shows the excess return is almost fully explained by a growth tilt in a year when growth broadly outperformed. The attribution to skill was self-serving; the return came from factor beta.

How can investors reduce self-serving bias? Run attribution on wins and losses with equal rigor. For each position, decompose the return into market, sector, factor, and idiosyncratic components before taking credit. Keep written pre-trade notes so recall cannot quietly edit the process record. A post-trade journal reviewed quarterly provides the honest signal that in-the-moment attribution cannot.

How is self-serving bias different from healthy self-confidence? Healthy confidence is grounded in a tracked attribution record that consistently shows positive skill contribution after costs and factor adjustments, across a large sample of decisions. Self-serving bias is ungrounded: it assigns credit to skill without checking whether the evidence supports that, and blame to bad luck without checking whether the loss reflected a process failure. The difference lies in whether attribution is tested or assumed.

Sources

1. Miller, D.T. & Ross, M. (1975). "Self-Serving Biases in the Attribution of Causality: Fact or Fiction?" Psychological Bulletin, 82(2), 213-225. https://web.mit.edu/curhan/www/docs/Articles/biases/82_Psychological_Bulletin_213_(Miller).pdf
2. Shepperd, J., Malone, W. & Sweeny, K. (2008). "Exploring Causes of the Self-Serving Bias." Social and Personality Psychology Compass, 2(2), 895-908. https://people.clas.ufl.edu/shepperd/files/SSB2008.pdf
3. Barber, B.M. & Odean, T. (2001). "Boys Will Be Boys: Gender, Overconfidence, and Common Stock Investment." Quarterly Journal of Economics, 116(1), 261-292. https://academic.oup.com/qje/article/116/1/261/1939000
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.