WALK-FORWARD OUT-OF-SAMPLE
The selected record is a historical simulation generated by annual expanding-window evaluation. Aggregate metrics are paired with the yearly record and explicit implementation limits.
ANNUALIZED NET RETURN
48.2%
simulated, after modeled costs
NET SHARPE
1.298
overall walk-forward record
NET SORTINO
2.678
downside-risk adjusted
MAXIMUM DRAWDOWN
-49.6%
simulated peak to trough
AVERAGE EXPOSURE
0.534x
governed exposure
ANNUAL TURNOVER
17.81x
research estimate
AVG TRANSACTION COST
0.071%
per rebalance observation
EVALUATION PERIOD
2021–2026
139 rebalances
SHADOW MANDATE / NON CANONICAL
A high-water-mark capital-cushion controller lowered the simulated maximum drawdown below 25% without changing the ranking model or Top-10 portfolio constructor. It is approved for forward paper tracking, with no live capital, brokerage connection, or order generation.
CANDIDATE DRAWDOWN
-21.6%
-49.2% cash-yield comparator
CANDIDATE RETURN
35.1%
70.0% of comparator
CANDIDATE SHARPE
1.435
+0.099 vs comparator
PAPER MANDATE
$93,943
63% equity · 3 completed intervals
Soft floor: 78% of the running high-water mark · cushion multiplier: 3.0x · block bootstrap: 10,000 paths. Historical results will not be backfilled into the shadow ledger. The temporal confirmation segment is not a pristine holdout, the soft floor cannot prevent gap losses, and canonical release metrics above are unchanged.
ANNUAL RECORD
| Year | Core return | Core Sharpe | Aggressive return | Defensive return | Core drawdown |
|---|---|---|---|---|---|
| 2021 | -18.2% | -0.472 | -4.4% | -15.6% | -33.4% |
| 2022 | -19.6% | -0.614 | -44.7% | -14.3% | -26.9% |
| 2023 | 129.1% | 2.314 | 301.1% | 134.2% | -16.2% |
| 2024 | 77.1% | 1.670 | 97.7% | 71.2% | -11.9% |
| 2025 | 166.9% | 3.320 | 298.9% | 128.8% | -8.2% |
| 2026 | 64.7% | 1.620 | 164.4% | 52.3% | -12.6% |
Simulated historical results. Includes modeled transaction costs where specified. Does not represent live investment performance. The research configuration was selected through iterative experimentation and is therefore subject to model-selection bias.
The homepage compounds the governed detailed return stream into a hypothetical $100,000 account and compares it with a buy-and-hold SPY adjusted-close total-return proxy over identical dates. This page keeps the yearly record tabular so the aggregate curve does not obscure annual variation.