Both models use a multi-factor scoring engine that evaluates 150-190 tickers daily across momentum, value, quality, and macro regime signals. Positions are sized by conviction score and rebalanced daily at market close.
Model A (Concentrated) runs against a curated universe of ~186 US-listed securities with higher position concentration. It targets regime shifts and sector rotation with larger individual bets.
Model B (Diversified) runs against a broader global universe of ~152 tickers including international ETFs, commodities, and bonds. It maintains more positions with smaller individual weights for diversification.
Both models can go long and short. Short positions target overvalued, high-multiple names scoring below 25 on the composite factor model. Cash allocation is dynamic — models hold 30-50% cash during high-uncertainty regimes.
All trades are timestamped and logged before execution. No retroactive adjustments. No survivorship bias. What you see is what happened.