Systematic Edge,
Quantified.
Independent quantitative research across US mega-cap equities. Factor models, evidence-based execution costs, and portfolio optimization — all out-of-sample.
Performance Snapshot
Jan 2024 – Dec 2025 · Out-of-sample
Top Signals
Ranked by conviction-weighted predicted return, net of execution costs.
| # | Ticker | Tier | Predicted | Strategy | Win Rate | $10K P&L |
|---|---|---|---|---|---|---|
| 1 | NVDA | HIGH | +48.1% | OTM Call 20 | 56% | $8,407 |
| 2 | AVGO | HIGH | +45.7% | ATM Call | 64% | $7,972 |
| 3 | PLTR | HIGH | +45.4% | OTM Call 20 | 46% | $7,135 |
| 4 | MU | HIGH | +48.3% | ATM Call | 51% | $6,984 |
| 5 | TSLA | HIGH | +41.1% | OTM Call 20 | 46% | $6,729 |
Research Pipeline
Three integrated systems that form a complete quantitative workflow.
Factor Model
9-factor adaptive ensemble with Ridge, ElasticNet, and Gradient Boosting. Regime-aware weights with James-Stein shrinkage.
Execution Costs
Spread model trained on 282 real options chain observations. Half-spread per leg eliminates 14 unprofitable strategies.
Portfolio Construction
Drawdown-adjusted Kelly sizing with conviction-tiered options overlay. Sector constraints and fractional allocation limits.
Conviction Tiers
Strategy selection is driven by prediction confidence and volatility edge.
Pred > 30% · EC > 85%
Calls for upside convexity
Pred 15–30% · EC 65–85%
Vol-edge guided selection
Pred < 15% or EC < 65%
Stock positions only
Transparent. Reproducible. Out-of-sample.
Every prediction is published before the evaluation period. Every execution cost uses real market data.