A consistent finding from listening to retail investors describe their past mistakes: they almost never describe a pattern. Each failed decision is, in their telling, sui generis. A bad market moment. An unusual circumstance. A piece of information that turned out to be wrong. An exception. A misread. Each story stands alone, narrated as a one-off departure from an otherwise reasonable approach.
Looking at the same investor's history from the outside, the same handful of moves repeat. Sometimes for decades. The averaging-down on a losing position; the panic-selling near the bottom; the over-concentration in a sector the investor knows from their day job; the abandoning of a stated rule during the period when the rule would have mattered most. The mistakes that look like exceptions from the inside are, from any external vantage point, the investor's recurring repertoire.
This is not because investors are foolish. The same investors, in their working lives, are often skilled at recognising patterns in other people's behaviour — they manage teams, they negotiate deals, they read clients. The blind spot is specifically about themselves. The patterns repeating in their portfolios are the same kind of patterns they would notice immediately in a colleague.
This is a piece about the blind spot. What produces it, what would dissolve it, and what changes for the small number of investors who actually do the work.
Why memory does not help
The structural reason patterns stay invisible is that the investor does not have the data. Memory is unreliable in a specific direction — it preserves outcomes and discards process. A decision that worked out is remembered as a good decision. A decision that did not is remembered as the result of unusual circumstances. The third decision — the one made for the same reasons as the first two, with the same emotional state, in the same market conditions — is not remembered as part of a series. The first two are filed in different mental folders.
The folders are usually labelled by outcome rather than by process. "Investments that worked" sits next to "Investments that did not work." Both folders contain decisions made in similar conditions for similar reasons, but the conditions and reasons are not what the folders are organised around. The conditions and reasons would have been the useful thing to file. They are mostly missing.
The investor reviewing their own history later sees a collection of outcomes and an inferred narrative. The narrative is usually flattering. The wins were skill, the losses were bad luck, the patterns are not visible because the data on which patterns would be visible was never recorded.
What the patterns look like, in the rare cases where they're visible
An incomplete sample of recurring patterns observable in retail investors who have actually kept records of their own decisions.
The check-the-price pattern. Some investors check positions far more often than others, and the frequency correlates with poor decisions. The act of checking generates a desire to act, and acting under no new information consistently produces worse outcomes than not acting. The investor who checks four times a day does not see the connection because each check feels independently justified.
The averaging-down pattern. Adding to a losing position is sometimes a disciplined response to a thesis that has not changed. More often, it is a way to avoid acknowledging that the original decision was wrong. The investor experiences the second purchase as conviction. The records, reviewed later, show it was usually denial.
The familiarity-as-conviction pattern. Over-weighting investments in industries the investor works in or knows socially. The familiarity feels like edge. The records usually show the familiar positions performed no better than the unfamiliar ones, and sometimes worse — because the investor was less willing to question information that came from people they knew.
The deal-of-a-lifetime pattern. Acting unusually decisively under fear of missing out. The investor's records typically show that the "rare opportunities" recur every twelve to eighteen months, take similar forms, and produce similar outcomes. What felt unique was actually the third or fourth iteration of the same impulse.
The this-time-is-different pattern. Abandoning a stated rule during the period when the rule would have mattered most. The investor has a rule about diversification, about position sizing, about holding period — and then, during a specific kind of market moment, finds reasons the rule does not apply. The records show that "different" usually meant the opposite of different — it meant the rule was being tested.
These are recurring archetypes, not universal laws. Different investors have different repertoires. The point is not the specific patterns but the general fact: each investor has a handful of recurring moves, and the moves are mostly invisible to the investor making them.
What records make patterns visible
The records that make patterns visible are not the records most investors keep. Brokerage statements show outcomes — what was bought, what was sold, at what price, with what return. They do not show the inputs that produced the decision: what was believed at the time, what was assumed about conditions, what alternatives were considered, what would have changed the decision, how confident the investor was, how they felt.
Pattern recognition requires the inputs, not the outputs. A series of outcomes can look random or skilled depending on the narrator. A series of inputs reveals whether the same logic was being applied each time.
The records that capture the inputs are uncommon because writing them down is uncomfortable. The investor would have to commit, in advance, to a thesis that might later be wrong. They would have to specify what they expected, which means having to confront the gap between expectation and outcome later. They would have to note how confident they felt, which involves admitting to less confidence than the action implied. None of this is information the investor's flattering self-narrative wants to make available.
For the investors who keep this kind of record anyway, what emerges over twelve to twenty-four months is generally a small number of patterns — three to five recurring moves that account for most of the losses and a different three to five that account for most of the gains. The patterns are often surprising, in the sense that they are not the ones the investor had assumed were their problem before doing the work.
What changes for the people who do this
The investors who develop visibility into their own patterns do not, in most cases, eliminate them. The patterns are usually structural — features of personality, response to stress, default behaviours under uncertainty. They do not disappear on being noticed. What changes is the relationship to them.
An investor who knows they are prone to the averaging-down pattern can build a hard rule against it: no doubling a losing position without 30 days of distance from the original purchase. An investor who knows they over-weight familiar industries can require an external second opinion for any position above a certain size in their own sector. An investor who knows their check-the-price pattern is destructive can move positions to a vehicle that produces less frequent updates.
None of these are heroic. They are small structural interventions that protect the investor from their known recurring moves. They work better than willpower, because they do not require the investor to be different than they are — only to design around what they are.
The investors who do not develop this visibility continue to produce the same patterns under different cover stories. The pattern that destroyed money in 2018 destroys money in 2024 under a different name. The investor, asked to describe the 2024 loss, will narrate it as an exception, just as they narrated the 2018 loss as an exception, never noticing that the exceptions form a tight cluster.
What is true regardless of which group an investor is in
Pattern recognition in one's own behaviour is not a universal good. Some investors do better not knowing — the act of self-observation, for some personalities, produces second-guessing that is worse than the original behaviour it was meant to correct. Some investors hold positions through volatility precisely because they do not check, do not measure, do not document. Their lack of records is, accidentally, a form of discipline.
For most investors, though, what changes when the records exist is not just the visibility of patterns but the texture of the relationship between the investor and the portfolio. The portfolio stops being a series of disconnected decisions remembered through their outcomes, and becomes a record of a particular person making particular kinds of moves under particular conditions. The investor reading their own records is, in some way, meeting themselves for the first time.
That meeting is mostly unflattering. It is also, for the investors who can sit with it, the beginning of a different relationship with their own decisions. Whether that relationship produces better outcomes is, in any individual case, unknowable. What it produces reliably is awareness — and for some investors, awareness is what was missing.
The records are unromantic. The patterns they reveal are unflattering. What sits between the two is the investor, and whether they make use of any of it.