Data sources
Market data is cleaned, adjusted for corporate actions, and checked for survivorship bias before a candidate is tested.
AI-automated research
AI proposes candidate signals and strategies. Human validation evaluates candidates out of sample and determines which advance. Approved strategies are executed systematically by live algorithms.
Research, validation, execution.
The roles are deliberately separated. AI proposes candidates. Human validation applies the framework. Live algorithms execute approved strategies.
01 / Discovery
AI research
Candidate signals and strategies are proposed and varied within the existing research framework.
02 / Validation
Human review
Human validation evaluates candidates out of sample, reviews walk-forward behavior and costs, and determines which advance.
03 / Execution
Live algorithmic trading
Approved strategies are deployed to live algorithmic execution. Individual trades are not overridden by discretionary judgment.
Data preparation and integrity.
Every candidate uses cleaned market data, realistic execution assumptions, and a reproducible test configuration.
Cost modeling
Commissions, spread, slippage, market impact, financing, margin, and liquidation are modeled where they apply. Results are evaluated after estimated costs.
Reproducibility
The same data, code, and parameters can be used to reproduce a test. This keeps comparisons consistent across research attempts.
Validation criteria.
Candidates must demonstrate out-of-sample performance, survive walk-forward analysis, remain viable under realistic cost assumptions, and be examined across market conditions before deployment is considered.
01
Fixed research periods
Complete calendar months are reported separately as training (Apr 2003–Dec 2018), validation (Jan 2019–Dec 2023), and holdout (Jan 2024–Jun 2026).
02
Validation-led selection
Candidates are ranked by validation average monthly NAV return after training review and hard accounting/risk checks. Holdout is displayed only after selection and never used to choose a record.
03
Freeze before holdout
A candidate, configuration, and source/cache context are recorded before the holdout is evaluated. The holdout is reported once; a failure is recorded rather than retuned away.
04
Out-of-sample testing
The candidate is tested on data excluded from development. Data access and test timing are reviewed to avoid contamination and look-ahead bias.
05
Cost and slippage stress
The result is recalculated with conservative trading-cost assumptions. Candidates that depend on optimistic execution do not advance.
06
Regime robustness
The candidate is examined across different market conditions. A result confined to one regime is treated as regime-dependent.
Human-built infrastructure
Infrastructure
- Data and market-data processing
- Backtest and risk infrastructure
- Out-of-sample and walk-forward validation
- Live execution stack (Ares)
- Reproducible research records
AI-assisted research
Candidate research
- Proposes candidate signals and strategies
- Explores multi-strategy combinations
- Tests consistency across market conditions
- Investigates absolute alpha hypotheses
- Analyzes failed candidates
- Suggests the next research attempt
Human validation
Validation and deployment
- Maintains the validation framework
- Checks for data contamination, look-ahead bias, and overfitting
- Reviews out-of-sample results
- Decides what advances
Systematic execution.
Strategies that satisfy the validation criteria are deployed to live systematic execution. The algorithm follows the approved rules. Individual trades are not overridden by discretionary judgment.
Current research uses a single AI model. Model attribution is provided for transparency. Comparative model testing will be reported when conducted under the same research framework.
Research record.
The research log documents candidate ideas, tests, failures, and results as the project develops.