DTRM Fund
+41.3%
Cumulative return
- Max drawdown
- -7.51%
- Calmar ratio
- 3.96
- Annual volatility
- 14.2%
- Sharpe · Treasury 3M
- 1.64
Field 01 / Markets
Beat The Machine
A live research question: can machine search become more valuable when a human still defines the game, allocates finite capital and remains accountable for risk?
The question
The machine ranks opportunities across a defined equity universe. The human does not replace that search; the human decides how ranked opportunities interact inside a finite portfolio and accepts responsibility for the resulting exposure.
The experiment compares that human-governed process with a recurring machine advisory, an AI-managed market instrument and the S&P 500. The aim is not to celebrate a winning prediction. It is to observe whether judgment adds value once return is considered together with realized risk.
Current governed state
- As of
- 2026-10-15
- Ranking
- Top 30
- Universe
- 174
- Risk artifact
- verified
Public snapshot generated 2026-10-15T23:00:00Z. Risk ratios are rendered only from the separately validated backend risk contract.
Current evidence
Return is only half of the score.
Cumulative return is the visible payoff. Maximum drawdown, Calmar, annualized volatility and Sharpe add complementary information about the path and risk taken to reach that payoff. The site never recalculates these risk diagnostics in the browser or presentation layer.
ChatGPT Fund
+34.2%
Cumulative return
- Max drawdown
- -16.54%
- Calmar ratio
- 1.50
- Annual volatility
- 14.2%
- Sharpe · Treasury 3M
- 1.36
AI Managed ETF — AIEQ
+22.6%
Cumulative return
- Max drawdown
- -9.10%
- Calmar ratio
- 1.82
- Annual volatility
- 12.3%
- Sharpe · Treasury 3M
- 1.00
S&P 500
+29.4%
Cumulative return
- Max drawdown
- -9.10%
- Calmar ratio
- 2.36
- Annual volatility
- 12.1%
- Sharpe · Treasury 3M
- 1.35
How to read the experiment
The Adjusted Score is an ex-ante ranking device: it exists before portfolio selection. Sharpe, Calmar, volatility and drawdown are ex-post diagnostics: they describe what happened after decisions were made. They are deliberately kept outside the historical score so future outcomes cannot leak into the search signal.
In game-theoretic terms, this is a repeated decision process in an adaptive environment rather than a single prediction contest. The ranking proposes a search space, portfolio construction selects an action under constraints, and the realized risk-return path is evidence about the resulting strategy. That framing does not make the market a literal zero-sum game and does not establish causality.
Read formulas and conventions →Inspect the evidence
Humans, machines and researchers can inspect the same read-only evidence. Public visibility ends at validated artifacts: it exposes no API key, raw Treasury collection, execution endpoint or capability that can operate the research system.
Research boundary. This page is evidence, not an investment recommendation and not an execution interface. Public visibility ends at validated, read-only artifacts.