After that headline, a small confession before we go any further: this isn't a professional investor but Gary Jackson, the head of editorial at FE fundinfo.
I spend my working life researching funds, writing about investing and editing analysis of market behaviour. I earned a badge for being in the 'top 1%' of a trading session in the FT's ongoing stock-picking competition and, after three weeks of the game, I've already made mistakes and picked up a few hard lessons.
I joined the competition a day late, so by the time I made my first trades the leading participants were already up around 4%. That gap became my anchor, rather than my actual starting point of zero or a rational assessment of what returns were achievable over nine weeks.
Anchoring bias had taken hold before I bought a single share and it reshaped every decision that followed. The question stopped being 'what should I own?' and became 'what will close this gap quickly?' Those are very different questions and the second leads somewhere much more reckless.
My portfolio ended up as five semiconductor stocks: Marvell Technology, ASML, Intel, Applied Materials and Micron. The honest explanation of why is not that I had conducted careful analysis but because they fit a pattern.
Semiconductors have been one of the dominant market narratives of 2026 and, to many, these stocks look like continued winners as the AI build-out continues. This is the representativeness heuristic: we judge probability by how well something matches a familiar template rather than by rigorous analysis.
Once the pattern is in place, the narrative fallacy follows. A story forms around why these stocks will work. I knew the AI trade had been volatile but the narrative made ignoring semiconductors feel riskier than owning them, which is a reasonable description of how a lot of money gets put to work in markets.
Alongside all of this came a feeling I can only describe as certainty: I felt like I knew these stocks would do well. That feeling had nothing to do with informational edge, because every other participant had access to exactly the same public information.
What I had mistaken for insight was familiarity with the narrative, which is overconfidence bias in its most common form: the illusion of knowledge. Being well-informed about a situation does not produce better judgement when that information is already priced into every share.
Watching the portfolio move daily created a further layer of pressure. Stocks that had risen started to feel safer but stocks that had fallen started to feel like mistakes worth cutting.
Neither feeling reflected underlying value. This is momentum chasing, rooted in recency bias: the tendency to weight recent price movements too heavily when assessing where things are heading. The rational response to a long-term holding that falls 5% in a week is usually to do nothing but the emotional response is to sell it and buy something that has just risen.
The clearest illustration came with SpaceX. I bought it despite having no genuine conviction in it as an investment. I bought it because others appeared to be piling in and because not owning something that was moving felt worse than owning it without a good reason. The trade was textbook herding.
I sold SpaceX and avoided a close-to 20% fall, which was luck rather than judgement.
The competition format made all of this more likely: a compressed time horizon, visible peer rankings, a deficit to chase from day one. But these conditions are not unique to a newspaper game. Fund managers with publicly available performance league tables, quarterly reporting cycles and clients who monitor short-term numbers face a similar environment.
The gap between what I would pick for the long term and what I actually picked is, I think, the most honest measure of how much external pressure distorts decision-making. Given more time and no peer visibility, I would not have built a portfolio of five semiconductor stocks and I would not have gone near SpaceX. My ISA, SIPP and workplace pension look nothing like this.
At the time of writing, I am currently ranked 904th out of 10,634 portfolios, which I suppose leaves me at risk of outcome bias. Outcome bias is the tendency to judge a decision by its result rather than by the quality of the reasoning behind it. If the semiconductor stocks rally and the final ranking looks respectable, there will be a temptation to conclude the process was sound even though it was not.
Knowing the names of these biases did not protect me from a single one of them. I could identify anchoring as it was happening and still anchored. I knew I was herding into SpaceX and bought it anyway.
Whether the semiconductor portfolio rallies or collapses from here, the experience has already been more instructive than expected and none of the lessons are flattering. The question I keep returning to is: does knowing about these biases actually change behaviour or does it mainly give you a more sophisticated vocabulary for explaining your mistakes afterwards?
That question seems worth putting to people who manage money for a living, where the stakes are real and the pressures are considerably higher. If you’re a professional investor and have found something that works, or are willing to be honest about what does not, I would like to hear from you for a follow-up piece.
You can reach me at gary.jackson@fefundinfo.com
Gary Jackson is head of editorial at FE fundinfo. The views expressed above should not be taken as investment advice.