25/20 BIL · MODEL SPECIFICATION
Strategy rules
The model is mechanical: the same market data must always lead to the same decision.
How the model works
9 fixed rules- 01
During Risk On, the portfolio holds the 10 highest-ranked Nasdaq-100 stocks, equally weighted at entry.
- 02
Momentum is measured weekly: over 13 weeks at the beginning of a Stage 2 trend, then from the trend start date.
- 03
A stock is eligible for purchase inside the top 10 and is not sold until it falls below rank 15.
- 04
Risk On requires the SPX to be above rising 10-week and 30-week moving averages; the risk filter uses the 40-week moving average.
- 05
Exit requires two weekly closes below the 40-week moving average, with an immediate exit if the SPX falls more than 5% below it.
- 06
When a position exceeds 25%, it is trimmed to 20% and the proceeds are allocated across the remaining holdings.
- 07
During Risk Off, all stocks are sold and the portfolio is allocated to BIL.
- 08
Signals are set at the weekly close and simulated trades are executed at the next market open.
- 09
The simulation excludes fees and transaction costs.
Fixed parameters
No discretion| Parameter | Rule |
|---|---|
| Universe | Nasdaq 100 |
| Portfolio | 10 stocks |
| Entry threshold | Ranks 1–10 |
| Exit threshold | Rank > 15 |
| Regime filter | SPX Stage 2 / 40-week SMA |
| Exit confirmation | 2 weekly closes |
| Crash override | 5% below 40-week SMA |
| Position limit | Trim from 25% to 20% |
| Risk Off | 100% BIL |
| Execution | Next market open after weekly signal |
| Costs | 0 bps in the simulation |
Research model—not investment advice.
This website presents a hypothetical, rules-based simulation for educational and research purposes. It does not provide personalized advice, a trading recommendation, or an offer to buy or sell securities.
Past and simulated performance does not guarantee future results. The model excludes fees and transaction costs. Actual results may be affected by taxes, commissions, bid-ask spreads, liquidity, execution, and data revisions. Daily adjusted closing prices are reconciled with validated weekly checkpoints. The model uses a current-universe proxy and is subject to survivorship bias.