Public methodology ยท snapshot 2026-08-25
How the investing system works.
This page exposes the system's current methodology and evidence boundaries in formats that people, search engines, and answer agents can inspect directly. It describes implementation and reproducibility. It does not claim proven investment outperformance.
3faadabdf71da180(spy_price / spy_ma200) - 1What the system is trying to do
The system supports disciplined recurring portfolio allocation and rebalancing across diversified ETF risk buckets. It uses a deliberately low-dimensional market regime model, bounded portfolio tilts, within-bucket scoring, explicit exposure limits, turnover controls, data-integrity gates, and separate execution permissions.
It is not designed to day trade, predict individual-stock winners, guarantee returns, eliminate drawdowns, or turn free-form AI conversation into trading authority.
Inspect the source material
- Methodology overview (Markdown)
- System card (Markdown)
- Executable implementation audit (Markdown)
- Public evidence standard (Markdown)
- Historical evidence publication gate (Markdown)
- Methodology manifest (JSON)
- Current public-safe policy snapshot (JSON)
- Public core metric calculations (Python)
- Standalone synthetic reproducibility example (Python)
- Official Treasury bill input snapshot (JSON)
- Standalone Treasury cash benchmark example (Python)
- Publication/export boundary (Markdown)
Current evidence state
| long term independently verified live track record | not_established |
|---|---|
| proven superiority to simple passive portfolio | not_established |
| independent third party methodology review | not_completed |
| synthetic reproducibility example | standalone_and_private_parity_tested |
| real market cash benchmark example | standalone_official_treasury_input |
| public policy snapshot | available_and_drift_tested |
| dated public evidence ledger | scaffold_available_no_performance_entries |
A reproducible methodology is not a performance result. A backtest is not a live track record. The public artifacts explicitly distinguish implementation, simulation, paper/shadow evidence, and live evidence, and retain unfavorable evidence under the same publication rules as favorable evidence.