Research Transparency
Full methodology, historical performance and limitations — connected to our frozen, reproducible backtest records.
This is historical backtested performance, not independently audited and not actual brokerage returns. The test period was previously optimized.
Overview
In-Sample Simulated CAGR
28.26%
Max Drawdown
13.34%
Trades
81
SPY CAGR
22.81%
Testing Period
2023-10-11 → 2026-10-09
Stock Universe
73 stocks
Initial Capital
$50,000
Rules Version
phase3-v2-frozen-events
Historical Simulated Equity Curve
Annual Return Breakdown
Drawdown History
Strategy Comparison
| Strategy | CAGR | Max Drawdown | Trades | Win Rate | Avg R |
|---|---|---|---|---|---|
| Ranking Only | 28.26% | 13.34% | 81 | 37.0% | +0.63R |
| Ranking + Rotation | 25.78% | 23.87% | 131 | 46.6% | +0.37R |
| Rotation Only | 9.36% | 31.57% | 184 | 39.1% | +0.13R |
| Original Recovery | 6.46% | 23.36% | 94 | 28.7% | +0.13R |
Methodology
Entry Method
Signals are confirmed on Daily candles: at least 2 of the previous 3 closing-price changes are negative, today's candle is bullish (close > open), and today's close exceeds yesterday's high. Entry is simulated at the next session's open with slippage applied.
Exit Method
Default exit is 3R+Trail: take profit at 3R (3x the initial risk distance), then activate a trailing stop at 2.5 ATR. Stops move up only, never down. Conservative rule: when a candle's low touches the stop and high touches the target, the stop is assumed hit first.
Portfolio Risk Management
2% risk budget per trade. Maximum 5 simultaneous positions. Maximum 20% of equity per position. Shared cash pool. Initial capital: $50,000.
Position Sizing
Risk-based sizing: shares = (equity x 2%) / risk_per_share. Risk per share = signal close minus stop reference (lowest low of 10 candles minus 0.2 x ATR).
Transaction-Cost Assumptions
$0.01 per share slippage applied to every simulated fill. $1 flat commission per trade. Dividend-and-split-adjusted OHLC prices used throughout.
Trading Statistics
Avg Holding Period
46d
Median Holding Period
24d
Cash Utilization
85.4%
Avg R per Day
-0.096R
Limitations & Research Bias
The 73-stock universe was selected after the historical period began. Point-in-time halal eligibility and delisted constituents are not available; survivorship and selection bias remain.
Daily OHLC cannot reveal intraday path. Stop-before-target is conservative when both levels are touched after the open.
All four rolling walk-forward windows overlap previously explored history. Rank agreement alone is not proof of a durable edge.
No threshold is certified optimal. The 20-day rotation parameter is a predeclared baseline, not an optimization claim.
Prospective Paper Trading
Historical backtesting is separate from prospective paper trading. We run a forward-looking model portfolio with the same rules, recording every signal and trade in real time. Paper-trading results may differ from backtested performance.
Past performance does not guarantee future results. These are simulated, in-sample results on a previously optimized period with survivorship bias. Not investment advice. Not independently audited. Not actual brokerage returns.