Research approach
Evidence, not a promise.
Research is judged by what withstands scrutiny. The process is staged, documented, and open to being wrong.
How research is tested
Research starts with a question and a clear evaluation plan. Out-of-sample results are kept separate from development, while assumptions are set before conclusions are drawn.
Checks performed
- Out-of-sample validation kept separate from development.
- Realistic cost assumptions and sensitivity testing.
- Stress tests and simulations to assess robustness.
- Audits for potential information-use and data-selection bias.
- Results are checked against stated rules, not selected after the fact.
What is not published
This site does not disclose strategy details, signal sources, traded assets, parameters, execution details, or raw data.