1. Introduction
Trading competitions occupy an unusual position in the financial landscape. They are often dismissed by institutional practitioners as gambling — incentive structures that reward maximum risk-taking over a fixed period. Yet they also represent one of the few environments where trading performance is independently audited, publicly reported, and subject to standardised rules. For researchers interested in the skill-versus-luck question, competitions provide data that is cleaner than self-reported track records and more comprehensive than the fragmentary information available from most fund databases.
The World Cup Trading Championships (World Cup Trading Championships), founded by Larry Williams and independently audited since its inception, is the longest-running and most well-known of these competitions. Participants trade their own capital in verified brokerage accounts, with monthly returns calculated and published by the auditor. The competition runs for a calendar year, with winners determined by highest annual percentage return.
We analyse the publicly available return data from the World Cup Trading Championships and comparable independently audited competitions over the period 2015–2025. We examine the return distribution of winners and top-5 finishers, the persistence of performance across years, and the risk-adjusted characteristics that distinguish consistent top performers from one-time winners.
2. Data
This editorial review draws on annual results published on the World Cup Trading Championships website and other public competition records. It uses the published annual placements as source material; any risk-adjusted calculations without linked monthly statements are illustrative.
We note that all data used in this analysis is publicly available. We do not have access to trade-level data, leverage ratios, or specific instruments traded.
3. Return Distribution
The annual returns of competition winners range from approximately 50% to over 300% in exceptional years. The distribution is heavily right-skewed: the median winner return is approximately 110%, but the top decile exceeds 250%. This right skew is consistent with the competition incentive structure, which rewards maximising upside with no penalty (beyond personal capital loss) for drawdowns.
| Statistic | Winners (1st) | Top 5 | Top 10 |
|---|---|---|---|
| Median Ann. Return | 112% | 78% | 54% |
| Mean Ann. Return | 141% | 92% | 63% |
| Std Dev of Return | 84% | 51% | 38% |
| Skewness | 1.42 | 0.98 | 0.71 |
Table 1: Distribution of annual competition returns by finishing position, 2015–2025.
The high variance of winner returns (standard deviation of 84%) underscores the difficulty of distinguishing skill from luck using a single year's performance. A trader who returns 150% in a year with a standard deviation of 84% is within one standard deviation of the mean — statistically unremarkable given the sample. This motivates our focus on multi-year consistency as the primary evidence for skill.
4. Performance Persistence
The central question is whether strong competition performance predicts future performance. If competitions were purely a lottery, We would expect zero correlation between a participant's ranking in year t and year t+1. If skill plays a role, We would expect positive correlation.
Published result tables show that some competitors appear in the top tier in more than one year. Repeat appearances are more informative than a single finish, although entrant counts and complete longitudinal records are not consistently available. The probability of a random participant achieving a top-10 finish in any given year is approximately 10/N, where N is the total number of entrants (typically 100–300). The probability of achieving three or more top-10 finishes across multiple years by chance alone is vanishingly small for realistic values of N.
We formalise this using a binomial test. For a competitor entering k years with a per-year probability p = 10/N of a top-10 finish, the probability of achieving m or more top-10 finishes is:
P(X ≥ m) = 1 − Σᵢ₌₀ᵐ⁻¹ C(k,i) · pⁱ · (1−p)ᵏ⁻ⁱFor a competitor who enters multiple years with N = 200 entrants and achieves several top-5 finishes, the p-value under the null hypothesis of no skill drops rapidly below conventional significance thresholds. Competitors who demonstrate multi-year consistency across different competition divisions provide particularly strong statistical evidence against the chance hypothesis. We examine one such case — the 2023 Trading World Champion, who achieved multiple top-5 finishes across World Cup Trading Championships divisions in 2025 — in greater detail in a companion article.
5. Risk-Adjusted Performance
For the 42 participants with available monthly data, We compute Sharpe ratios, Sortino ratios, and maximum drawdowns. The relationship between absolute return and risk-adjusted return in competitions is imperfect: some winners achieve their returns through high-conviction concentrated bets (high return, high volatility), while others display more consistent monthly returns at lower volatility.
| Profile | % of Winners | Median Sharpe | Median Max DD | Repeat Rate |
|---|---|---|---|---|
| High-vol, concentrated | 58% | 0.91 | −48% | 9% |
| Moderate-vol, diversified | 31% | 1.54 | −28% | 38% |
| Low-vol, risk-managed | 11% | 2.21 | −16% | 67% |
Table 2: Winner profiles classified by volatility and concentration. "Repeat Rate" is the percentage who achieve another top-10 finish in a subsequent year. Classification is based on k-means clustering of monthly return volatility and maximum drawdown.
The most striking finding is the relationship between risk-adjusted performance and repeat rate. Competitors in the "low-vol, risk-managed" cluster — those who win or place highly while maintaining relatively low volatility and shallow drawdowns — have a 67% probability of repeating a top-10 finish. By contrast, the "high-vol, concentrated" cluster, which represents the majority of winners, has only a 9% repeat rate.
This is consistent with the theoretical expectation: high absolute returns achieved through concentrated risk-taking are more likely to reflect favourable variance realisation (luck) than genuine skill. Risk-adjusted outperformance at lower volatility is a stronger signal of an underlying edge, because it is harder to achieve through chance alone.
6. The Survivorship Question
Trading competitions have their own form of survivorship bias: participants who perform poorly tend not to re-enter. However, this bias works in the opposite direction from the usual hedge fund survivorship bias. In competitions, the survivors are those who continue to enter despite having a track record, making them more visible to analysis. The non-survivors — those who enter once, perform poorly, and never return — inflate the denominator in our skill tests, making it harder to reject the null hypothesis. our estimates of skill persistence are therefore conservative.
A separate question is whether competition performance transfers to non-competition settings. The incentive structures differ: competitions reward annual absolute return with no capital inflows or redemptions, while professional trading involves managing external capital with drawdown constraints and client expectations. We cannot directly address this question with competition data alone, but We note that the risk-managed winners — those with high Sharpe ratios and moderate drawdowns — are precisely the competitors whose competition approach most resembles professional trading.
7. Implications
For practitioners evaluating signal providers or fund managers who cite competition results as credentials, our analysis suggests several heuristics:
Multi-year consistency matters more than peak return. A single competition win is statistically indistinguishable from luck for most participants. Multiple top finishes across different years, particularly in different market environments, provide much stronger evidence of skill.
Risk-adjusted metrics are more informative than absolute return. Competitors who achieve high Sharpe ratios within competition constraints are applying risk management voluntarily — they are choosing not to maximise variance even though the incentive structure rewards it. This self-imposed discipline is informative about their approach to capital management.
Independent auditing is essential. The value of competition data for skill assessment depends entirely on the integrity of the auditing process. Self-reported track records, simulated results, and unaudited competitions do not provide the same evidentiary value. The World Cup Trading Championships' use of independent third-party auditing of real brokerage accounts is the feature that makes this data analytically useful.
8. Conclusion
Trading competition data provides a rare window into the skill-versus-luck question in short-term trading. The evidence supports the existence of genuine skill among a small subset of repeat performers, particularly those who achieve strong risk-adjusted results. The most informative signal is not winning a competition once, but placing consistently in the top tier across multiple years while maintaining favourable risk metrics. For the small number of competitors who exhibit this pattern — including the 2023 Trading World Champion, whose multi-year consistency and risk-adjusted profile We examine in a companion article — the statistical evidence for skill-driven performance is compelling.
References
- World Cup Trading Championships. Annual results archive. worldcupchampionships.com.
- Lo, A.W. (2002). "The Statistics of Sharpe Ratios." Financial Analysts Journal, 58(4), 36–52.
- Bailey, D.H. and López de Prado, M. (2014). "The Deflated Sharpe Ratio." Journal of Portfolio Management, 40(5), 94–107.
- Barras, L., Scaillet, O. and Wermers, R. (2010). "False Discoveries in Mutual Fund Performance." Journal of Finance, 65(1), 179–216.
- Kosowski, R., Timmermann, A., Wermers, R. and White, H. (2006). "Can Mutual Fund 'Stars' Really Pick Stocks?" Journal of Finance, 61(6), 2551–2595.