Confirmation Bias Investing: Why More Research Makes You More Certain and Less Right

Over the ten years ending December 2024, the average dollar invested in U.S. mutual funds and ETFs earned 7.0% a year while the funds themselves returned 8.2%. That 1.2-percentage-point shortfall — roughly 15% of everything those funds produced — didn’t come from fees or bad fund selection. It came from investors buying and selling at the wrong moments, and confirmation bias investing behavior is one of the biggest reasons those moments feel so justified at the time.

This guide walks through what confirmation bias looks like inside a normal investing routine, what the research says it costs, why doing more homework tends to make it worse, and six specific process changes that make the bias structurally harder to act on. No willpower required — the fixes are all mechanical.

What confirmation bias investing actually is — and what it isn’t

Confirmation bias is the tendency to seek, notice, and weight information that supports what you already believe, while quietly discounting information that contradicts it. Psychologist Raymond Nickerson, in his 1998 review in Review of General Psychology, called it “a ubiquitous phenomenon in many guises” — not a defect of unintelligent people, but a default setting of ordinary human reasoning.

The classic demonstration is Peter Wason’s 1960 “2-4-6” task. Participants were told the sequence 2-4-6 followed a rule, and asked to discover it by proposing their own sequences and receiving yes/no feedback. Most people guessed a rule like “even numbers ascending by two,” then proposed 8-10-12, 20-22-24, 100-102-104 — sequences designed to confirm the hypothesis. The actual rule was simply “any three ascending numbers.” The only way to find it was to propose something you expected to fail. Almost nobody did.

Swap the number sequences for tickers and you have the mechanism exactly. You form a view — this sector is due, this company is undervalued, the market is overheated — and then you go looking for evidence. The evidence arrives, because on the internet it always does. Each supporting data point feels like independent verification. In reality it’s the same hypothesis being echoed back at increasing volume.

Two things confirmation bias is not. It isn’t conviction — having a thesis and holding it through volatility is how long-term investing works. And it isn’t stupidity. The most reliable finding in this literature is that the bias scales with engagement: the more you research, the more raw material you have to be selective with.

What confirmation bias investing costs, according to the data

Behavioral biases are easy to describe and hard to price. Three research strands give reasonable bounds.

The direct study. Park, Konana, Gu, Kumar, and Raghunathan surveyed 502 investors using South Korea’s largest stock message board operator and measured how they processed messages about stocks they held. Investors who preferentially engaged with confirming messages showed measurably higher overconfidence, higher expected returns, more frequent trading — and lower realized returns. The confirming information didn’t improve their decisions; it raised their certainty while degrading their outcomes.

The trading-frequency proxy. Because confirmation bias mostly expresses itself as action, turnover is a usable stand-in. Barber and Odean’s Journal of Finance study of 66,465 discount-brokerage households (1991–1996) found the average household earned 16.4% annually against a 17.9% market return, while the most active-trading quintile earned just 11.4% — a 6.5-point annual gap. Same market, same period, same information environment. The difference was how much they acted on their conclusions.

The aggregate gap. Morningstar’s Mind the Gap 2025 study quantifies the whole population effect, and the category breakdown is the interesting part:

Read the first and last rows together. Allocation-fund investors kept roughly 97% of their funds’ returns — 6.3% of an available 6.5% — largely because those funds live in retirement accounts nobody logs into. Sector-fund investors, who almost by definition bought because they had a specific view about a specific industry, gave up 1.5 points a year. The gap is not about fund quality. It’s about the correlation between having a thesis and trading on it.

For context on how much that compounds: Vanguard’s Putting a Value on Your Value research estimates that behavioral coaching alone — an advisor whose main function is talking clients out of acting on their conclusions — is worth roughly 1.5 percentage points of annualized return, the single largest component of its 3-point “Advisor’s Alpha” estimate. The most valuable thing a professional does, by Vanguard’s own accounting, is prevent you from executing your best ideas.

Four places the bias hides in an ordinary portfolio routine

Confirmation bias rarely announces itself. It shows up as reasonable-looking research behavior.

1. The search box. The phrasing of a query determines the answer set. “Is [ticker] undervalued” and “is [ticker] overvalued” return two entirely different internets. Whichever you type reflects a conclusion you’ve already reached, and the results then function as evidence for it. This is selective exposure operating one keystroke at a time.

2. Source curation. Every follow, subscribe, and mute decision narrows the information environment slightly. Over a couple of years, a feed assembled entirely from sources you found agreeable becomes a machine for producing agreement — and it feels like a diverse set of independent voices because there are dozens of them.

3. Asymmetric scrutiny. This is the subtlest one. You don’t ignore contradicting evidence; you just audit it harder. A bearish report on a stock you own gets checked for author bias, methodology flaws, and short interest. A bullish report gets read once and filed. Both got “evaluated.” Only one got interrogated. Psychologists call this disconfirmation bias, and it’s why smart, skeptical people aren’t protected — their skepticism is simply pointed in one direction.

4. Retroactive scorekeeping. After the fact, memory reorganizes itself so that what happened seems like what you expected. That’s hindsight bias, which we cover in detail in its own case study, and it’s confirmation bias’s cleanup crew: it converts a coin-flip outcome into evidence that your process works, which strengthens the prior you’ll defend next time.

Why more research makes confirmation bias investing worse

The intuitive fix — read more, dig deeper, get more sources — is counterproductive, for three structural reasons.

First, volume creates false corroboration. Forty articles making the same argument are not forty pieces of evidence. Financial media is highly correlated; a single analyst note propagates through aggregators, newsletters, and social posts within hours. The subjective experience is overwhelming consensus. The actual information content is one source.

Second, effort converts into entitlement. Having spent twelve hours on a thesis, abandoning it means writing off twelve hours. That’s the sunk cost fallacy interacting with loss aversion, and it means the deeper your research, the higher the psychological price of updating. Confirmation bias then does the merciful thing and stops disconfirming evidence from arriving in the first place.

Third, research raises confidence faster than accuracy. This is the finding that generalizes across the literature, and it’s the same engine documented in our piece on why overconfidence bias in stock picking gets stronger the more homework you do. Certainty and correctness are separate variables, and only one of them responds reliably to effort.

The market-wide scoreboard is worth keeping in view here. S&P’s SPIVA scorecard found that 79% of active U.S. large-cap funds underperformed the S&P 500 in 2025, and roughly 92% of domestic funds trailed their benchmarks over the trailing 20 years. These are full-time professionals with Bloomberg terminals, analyst teams, and management access. The base rate for beating the market through superior conviction is not encouraging, which is a useful prior to hold before deciding your weekend reading has produced an edge.

Six habits that make confirmation bias structurally harder

Trying to “be less biased” doesn’t work — the bias operates below the level you can introspect on. These change the process instead.

1. Write the thesis down before you buy, with a kill condition. One paragraph: what you believe, why, and the specific observable event that would prove you wrong. Date it. The kill condition is the whole point — it has to be defined while you’re still capable of imagining being wrong.

2. Search the opposite case first. Before “why [X] is a buy,” run “why [X] is a sell” and read it properly. You’re not obligated to agree. You’re obligated to know what the argument is before you dismiss it.

3. Rewrite your questions to invite falsification. The framing does most of the work:

4. Keep a decision journal, not a returns log. Record the decision, the reasoning, and the confidence level at the moment you act. Reviewing this a year later is the only reliable defense against memory quietly rewriting your track record. Most people discover their hit rate is far closer to a coin flip than they remember.

5. Automate the default. This is the highest-leverage item on the list, and the Morningstar category data is the proof: the funds with essentially no investor return gap are the ones held on autopilot. Scheduled contributions into a simple three-fund index portfolio remove the decision surface that confirmation bias needs in order to do damage. You can’t act on a biased conclusion about a trade you never had to consider making.

6. Rebalance on the calendar, not on conviction. A fixed schedule — annually, or on a set drift threshold — forces you to sell what’s run up and buy what hasn’t, which is exactly the trade your current narrative will argue against. Calendar-based portfolio rebalancing is a rule that overrides the story.

What I changed in my own process

I’m a software engineer by trade, which mostly means I’ve spent a long time in a field where being confidently wrong is expensive and gets discovered quickly. Code has a useful property here: it either runs or it doesn’t. Investment theses have no equivalent — you can hold a wrong one for years and get paid anyway, then conclude you were right.

My own portfolio is boring by design — broad index funds inside tax-advantaged accounts, no advisor, contributions automated. But the boring allocation didn’t stop me from spending an embarrassing number of evenings researching individual positions I’d already half-decided on. What changed my mind wasn’t a book about behavioral economics; it was writing down predictions with dates attached, because I’d gotten curious about whether AI forecasting tools were actually better than my own judgment. After about eighteen months of logged calls, my accuracy on anything market-directional was close enough to chance that the comparison stopped being interesting.

The uncomfortable part wasn’t the hit rate. It was rereading the reasoning and finding it genuinely persuasive — each call was well-argued, sourced, and defensible. The research had been real. It had just been recruited in service of a conclusion I’d already reached, which is exactly what confirmation bias is supposed to feel like from the inside: not sloppy, but thorough. I still read as much as before. I just stopped letting the reading authorize trades, and moved the money to a schedule instead.

This article is for general information and is not investment advice. Past performance of any strategy or fund does not predict future results.

Photo by Markus Winkler on
Unsplash

Written by Chris Steve

Chris Steve is a software engineer with a deep interest in personal finance, behavioral economics, and AI. He started Money & Planet to share clear, research-backed money guides — the kind that explain the math instead of pushing products. His writing focuses on long-term wealth building, the psychology behind spending and investing decisions, and the practical tools regular people can use to make smarter financial choices.