You Keep Finding News That Agrees With the Stock You Already Own. That Is Not a Coincidence
6 min read
Loading...
6 min read
You own a stock. Last night you searched for news about it. Every article you opened was good. The analyst was upbeat. The forum was upbeat. You went to bed feeling smart.
Here is the uncomfortable part. You may not have found the news. You may have picked it.
That habit is called confirmation bias. It is the habit of looking for, and believing, information that agrees with what you already think. This post shows how it works in stock research, what studies found, and four steps to make your research fairer.
Confirmation bias is a mental shortcut. Your brain likes being right. So it pays more attention to facts that support a view you already hold, and less to facts that don't.
A well-known 1998 review by psychologist Raymond Nickerson describes it as seeking or reading evidence in a way that favours your existing beliefs or the idea you are testing. (Nickerson, Review of General Psychology, 1998; I read the abstract details through search summaries, not the full paper.)
It is not lying. It is not even on purpose. That is why it is so hard to catch.
In 1960, psychologist Peter Wason gave people three numbers: 2, 4, 6. He said they follow a rule. Their job was to find the rule by suggesting new sets of three numbers. He would say yes or no to each.
Most people guessed something like "numbers going up by two". Then they tested sets that fit their guess, such as 8, 10, 12. Every time, the answer was yes. They felt sure.
The real rule was simply "any numbers going up". Sets like 3, 5, 7 or 1, 2, 50 would also get a yes. Only people who tested a set that could prove them wrong, such as 1, 2, 50, would find out. Most didn't, and many ended the test confident in the wrong rule. (Wason, 1960, via secondary summaries of the Quarterly Journal of Experimental Psychology paper; I did not open the original.)
Now swap numbers for stocks. "This company is great" is your guess. Searching for good news is testing a set that fits. It always says yes.
Researchers have looked at this on investing message boards. In a 2013 paper in Information Systems Research, Park, Konana, Gu, Kumar and Raghunathan ran a field experiment with 502 investors on a major South Korean stock message board. They found that investors preferred messages that matched their existing view. The effect was stronger for investors who rated their own market knowledge highly. (Abstract at INFORMS)
The same study found that investors who showed more of this bias were also more overconfident. They traded more often and expected higher market returns than the evidence supported. A summary of the earlier working paper also links stronger bias to weaker actual returns, though the final 2013 abstract is more careful and says only that message board use may not lead to better returns. (Summary at CXO Advisory)
An earlier study of about 72,000 Yahoo! Finance message board interactions found a similar pattern. Investors tended to join conversations with people who already agreed with them. It got stronger when there was a lot of information or high uncertainty in the market. (UT Austin news release, 2008)
Note: all figures in this section come from search summaries of these studies. The papers themselves were not opened.
You don't need a message board for this. Your own habits do the job.
You search by your position. Typing "BBCA outlook" gives you articles about outlook. Typing "BBCA risks" gives you a different set. Which one do you type when you own the stock?
You follow people who agree. After a few months, your feed is full of voices that like what you hold. Disagreement quietly drops out.
You read good news and bad news differently. Good news feels like proof. Bad news feels like a "one-off" or "already priced in".
Your watchlist fills up with your own holdings. If every stock on your list is one you own or love, your list only shows you what you already decided.
This is NetWort arithmetic with invented numbers, not real data.
Say you own a stock. Over a month, 20 pieces of news appear. 12 are positive and 8 are negative. A fair reader sees 60% good and 40% bad.
Now say you read 9 of the 12 good ones (75%) and only 2 of the 8 bad ones (25%). You saw 11 articles. Of those, 9 were good.
| What you read | Count | Share of what you read |
|---|---|---|
| Good news | 9 | 82% |
| Bad news | 2 | 18% |
The real mix was 60% good. Your mix feels like 82% good. Nothing was hidden from you. You just clicked in a pattern. That gap is how a mixed picture starts to look like an easy one.
Nothing here tells you what to buy or sell. The goal is only to make sure the evidence you use is fair. Our guide on survivorship bias covers a related trap in historical data, and anchoring bias covers fixation on your purchase price.
Open Explore and browse a few stocks, ETFs or sectors you have never held. Compare how they look next to the ones you own. Ask one question: if I had started here instead, what would I have found? It takes five minutes and it widens your list.