Money & Mind
Most people rate their own financial judgement above average, which is arithmetically impossible
Confidence in money decisions consistently exceeds the accuracy of those decisions, and the gap does specific damage in the places where confidence leads to action.
By Aditya Ramaswamy3 min read
Confidence and accuracy come apart
Ask people how good they are with money, at driving, or at judging character, and the distribution of answers cannot possibly match the distribution of ability. This tendency to place oneself above the middle is one of the more robust findings in the study of judgement, and it appears across domains rather than being a peculiarity of finance.
A second and related form is more specific: when people state a range they are confident contains the right answer, the true value falls outside that range far more often than their stated confidence implies. The estimate is not merely wrong, it is wrong while feeling certain.
Neither is a defect of intelligence. Well-informed people are, if anything, more susceptible in their own field, because knowledge supplies the raw material for a convincing account of why this case is different.
Where it becomes expensive
Overconfidence costs money mainly through action. It encourages trading on views that are not better than average, which incurs costs and taxes for no reliable advantage. It encourages concentration in a small number of holdings, on the basis that the holder has identified something others have not. And it encourages underestimating how long a project will take and how much it will cost.
It shows up in borrowing too, through confidence that income will rise, that a side venture will produce, or that a temporary arrangement will be temporary. Optimism about one’s own future circumstances is what makes an aggressive commitment feel manageable at the moment it is entered.
And it shows up as a reluctance to seek advice, since the value of another perspective is discounted in advance by someone who is confident their own is sound. That is the version with the largest consequences, because it removes the correction rather than merely biasing the estimate. Inaction, by contrast, costs relatively little when the underlying position is sensible.
Feedback in money is too slow and noisy to correct it
Skills improve where feedback is fast, frequent and unambiguous. Financial decisions have almost none of those properties. The result of a decision may take years to appear, is affected by everything else that happened in the meantime, and rarely comes with any indication of what the alternative would have produced.
That absence of clean feedback allows confidence to grow without accuracy growing alongside it. It also permits a particular kind of self-assessment error, where good outcomes are attributed to judgement and poor ones to circumstance — a pattern that keeps confidence intact regardless of results.
In domains with tight feedback, experience produces expertise. In this one, experience frequently produces conviction instead, which is not the same thing and is harder to argue with.
A written record is the practical countermeasure
The intervention that seems to help most is unglamorous: writing down, at the time, what you expect to happen and why. Not for accountability, but because memory reconstructs past expectations to match what actually occurred, which quietly removes the evidence that would have been informative.
A short note attached to a decision — what is expected, over what period, what would indicate this was wrong — converts a vague recollection into something checkable. After a few years of such notes, a person has an actual record of their own calibration rather than an impression of it.
The second measure is stating ranges rather than figures, and then deliberately widening them, since the demonstrated tendency is for stated ranges to be too narrow. It feels like weakness and is closer to accuracy.
Where confidence is doing something useful
It would be a mistake to read all this as an argument for constant self-doubt. Confidence is what allows decisions to be made at all, and a household paralysed by awareness of its own fallibility does not thereby make better choices — it makes fewer, later, and often defaults into the worst option by not choosing.
The distinction worth holding is between confidence in a process and confidence in a prediction. Being confident that a sensible, low-cost, diversified approach will do its job over time is a different claim from being confident about what happens next, and the first is far better supported.
None of this is guidance about any particular decision. Circumstances vary widely, self-assessment is unreliable by construction, and one of the more useful functions of a regulated adviser is being someone whose view of your plan was not formed by you.
Common questions
Does knowing about overconfidence reduce it?
Only slightly, and awareness alone tends not to change judgements much. What appears to help more is a procedural change — recording expectations in advance, stating ranges rather than point estimates, and deliberately widening those ranges — because these work whether or not the underlying feeling of certainty shifts.
Is confidence in financial decisions always a problem?
No, and its absence has costs of its own, including decisions deferred until the default outcome takes over. The useful distinction is between confidence in a sensible process and confidence in a specific prediction, since the evidence supports the first far better than the second.
Why does experience not fix this?
Because financial feedback is slow, noisy and rarely counterfactual, so it is unusually difficult to learn from. Outcomes take years, other factors intervene, and the road not taken is never observed. Experience under those conditions tends to build conviction more reliably than it builds accuracy.
Aditya covers spending, saving, debt and the questions readers actually send in and thinks most subjects are more interesting once you know how they work.
