The misidentification of luck as skill is one of the most consequential errors available to the investor, because it produces a positive feedback loop: lucky outcomes generate confident self-assessment, confident self-assessment generates increased risk-taking, increased risk-taking amplifies the eventual losses when luck reverses. The investor who has been lucky and believes himself skilled is not merely overconfident in a static sense; he is systematically increasing his exposure to the variance that will eventually produce a large negative outcome.
The fundamental difficulty is that luck and skill are genuinely difficult to distinguish in real time, over the sample sizes that individual investment experience provides. A skilled investor and a lucky investor may produce indistinguishable returns over five or even ten years, because the variance in investment outcomes is high enough that lucky investors can plausibly produce five or ten years of strong performance. The sample size required to distinguish skill from luck at conventional levels of statistical confidence is substantially larger than most investment careers, which means that most investors will never have enough data to know, with confidence, which of the two explains their track record.
The practical implication is that all investment performance, however impressive, should be held with appropriate epistemic humility. The investor who has performed well for five years has evidence that is consistent with genuine skill; it is also consistent with luck operating in a favourable direction. He cannot know which explanation is correct, and the honest acknowledgment of this uncertainty should constrain the risk-taking that confident self-assessment would otherwise encourage. Position sizes should reflect genuine uncertainty about one's own skill; portfolio concentration should be calibrated not to how good one feels about one's recent performance but to how good the evidence for that performance actually is, adjusted for the market environment in which it was produced.
The specific manifestation of luck-as-skill that produces the most dramatic eventual losses is the trader or investor who has experienced a run of success in a particular market environment and has progressively increased position sizes and leverage in response to the apparent confirmation of his skill. In a trending market, many approaches appear to work consistently, creating the impression of edge where the edge is actually the trend. When the trend reverses, the concentrated, leveraged positions that the apparent edge encouraged produce losses that are disproportionately large relative to the gains that preceded them—because leverage is symmetric, and the losses it amplifies are larger than the gains it amplified, given that the leverage was increased over time as apparent skill was confirmed.
The corrective is the adoption of a prior probability about one's own investment skill that begins at low and updates slowly in response to evidence. The default assumption, for any investor who has not demonstrated sustained outperformance over long periods and through multiple market cycles, should be that her returns are primarily driven by market beta and variance rather than by genuine alpha. Evidence that updates this prior—genuinely sustained, risk-adjusted outperformance through bull and bear markets—should accumulate over years before the prior is meaningfully revised upward. This is not a counsel of permanent diffidence; it is an accurate reflection of the difficulty of the investment skill assessment problem.