Model Accuracy & Empirical PerformanceOUT-OF-SAMPLE VERIFIED
Realized trade outcomes vs. predictions, calibration diagnostics, and benchmark alpha comparison.
Aspirational Target vs Scientific Measurement:95% is an aspirational theoretical research boundary, never a guaranteed outcome or commercial marketing claim. All metrics are computed using realistic walk-forward validation and simulated paper trading with slippage and execution costs.
Measured Win RateREALIZED
65.2%
Aspirational Goal: 95%
Directional Hit RateOUT-OF-SAMPLE
69.0%
Out-of-sample directional accuracy
Sharpe RatioRISK-ADJUSTED
1.92
Risk-adjusted return measure
Max DrawdownDOWNSIDE
-7.4%
Historical peak-to-trough
Precision (UP Forecasts)
71.4%
True Pos / (True Pos + False Pos)Recall (Opportunity Capture)
68.5%
True Pos / (True Pos + False Neg)F1-Score Confluence
0.699
Harmonic precision/recall meanProbability Calibration
0.165
Brier Score (< 0.20 = well-calibrated)Multi-Environment Segregation
Strict separation of Backtested, Paper-Simulated, and Live Brokerage results. Live execution is locked.
Cumulative Alpha vs Buy-and-Hold SPY
Net portfolio trajectory compared directly against S&P 500 ETF (SPY) buy-and-hold.
Smart Money AI (+21.6%)
SPY ETF (+7.1%)
Aug 2026 (Inception)Oct 2026Nov 2026 (Latest Verified)
Performance by Forecast Horizon
| Horizon | Directional Accuracy | Trade Win Rate | Profit Factor | Brier Score (Calibration) | Sample Trades |
|---|
Time-Stamped Predictions vs. Subsequent Realized Outcomes
Empirical audit comparing immutable timestamped model predictions against subsequent market settlement.
| Forecast Time | Ticker | Horizon | Forecast | Calibrated Prob | Entry | Realized Time | Realized | Return % | Directional Hit | Environment |
|---|---|---|---|---|---|---|---|---|---|---|
| No realization logs recorded yet. | ||||||||||
Event-Driven Walk-Forward Backtester
Simulate historical execution with realistic slippage (8 bps), trading fees, and stop/target boundaries.