A well-documented experiment in behavioural psychology asks subjects to evaluate two research papers on a controversial topic. Half the subjects are given a paper that supports their prior belief; half are given one that contradicts it. Subjects rate the supporting paper as substantially better-argued, more rigorous, and more trustworthy than the contradicting paper — regardless of which paper is objectively higher-quality. This is confirmation bias, and it is the most portfolio-corrupting behavioural pattern most investors carry, because portfolios involve continuous exposure to fresh information about existing beliefs.
The mechanism
The bias operates in three ways simultaneously. First, we seek out information that supports what we already believe — we read the analysts we agree with, follow the accounts we align with, and skim past the counter-arguments. Second, when we do encounter contrary information, we process it more critically than supporting information — we look for flaws, question methodology, and downweight the source's credibility. Third, we remember the confirming information more accurately than the contradicting information, so our recollection of what we've read is systematically biased even when our initial exposure was more balanced.
Each of these is small in isolation. Compounded across hundreds of pieces of investment information over years, they produce a portfolio outlook that has diverged significantly from what any dispassionate observer would produce given the same underlying data.
Why the portfolio makes it worse
An investor who owns a stock has a financial incentive to believe the stock will do well. This is not analysis; it is the alignment of belief and material interest that behavioural economists call "motivated reasoning." When the two align, the mental filter tightens. Analysis that supports the position is welcomed and remembered; analysis that questions it is scrutinised and forgotten.
The classic example is the earnings report. A shareholder reads the same 10-K as a non-shareholder analyst, but the shareholder is systematically more likely to conclude that the report is positive on balance, more likely to weight the quarter's positives heavily and its negatives lightly, and more likely to project the current-quarter trends forward as sustainable.
The concentration trap
Confirmation bias is why concentrated portfolios become more concentrated over time. An investor who owns a large position in a stock consumes more analysis of that stock than of any other. That analysis, filtered through confirmation bias, generates a stronger conviction over time — not because the stock is doing better, but because the mental picture of it has been sharpened and edited into a more coherent narrative. The stronger conviction supports a larger position. The larger position generates more attention. The cycle reinforces itself.
Most concentrated positions that end in large drawdowns end that way because the investor never seriously engaged with the counter-thesis. They saw it, dismissed it, and moved on. When the counter-thesis proved correct, the surprise felt genuine — from the inside of the confirmation bias, the risks really did look manageable up until the moment they didn't.
What actually reduces the bias
Not much reduces confirmation bias in the sense of eliminating it. But three practices measurably soften its portfolio impact.
Actively seek counter-analysis. When you form a view on a stock, spend at least as much time reading the best analysis against it as the best analysis for it. This is uncomfortable, and the discomfort is the point — if it doesn't feel uncomfortable, you are reading counter-analysis you already dismissed at the first paragraph.
Write down the disconfirming conditions in advance. Before taking a position, write down what would prove the thesis wrong. Not "if the price drops 20% I will sell" — that is a stop, not a disconfirmation. Something more like: "if annual revenue growth falls below 10% for two consecutive quarters, my thesis is invalidated regardless of price." Then, when the condition is met, honour it.
Rotate the perspective. Ask yourself, without the position, what you would think of the company. Rotate the pronouns — imagine a stranger told you they were considering the exact position you already have. What questions would you ask them? What warnings would you raise? Almost always, the questions and warnings you would raise for a stranger are things you have not raised with yourself.
Why this bias survives
Confirmation bias is not a bug — it is the mechanism by which humans maintain coherent beliefs at all. Without it, every stray piece of contradicting information would destabilise every prior conclusion, and the mind would never converge on anything. It is a feature that becomes a problem only in specific domains where the stakes of a coherent-but-wrong belief are high.
Investing is one of those domains. The bias survives because it works fine most of the time, and the times it doesn't work — the large drawdowns from unexamined concentrated positions — are dispersed enough in time that no individual investor experiences them as a pattern. You learn once. Some people learn twice. Most learn in a way that costs enough that they never fully recover the invested confidence.
The rule to internalise
You will read what you agree with more carefully than what you disagree with, and you will remember what confirms your existing positions more accurately than what challenges them. This is not going away. What you can do is manufacture forced exposure to the counter-thesis, in writing, and honour it when the specified conditions arrive. Every large investment blowup you will read about could have been softened by that one discipline consistently applied.
Educational content only. Not investment advice.