About Verdict Trading

The approach

Verdict Trading is a dual-regime system. In calm markets, an ML model ranks 1,000+ US stocks on momentum and trend features and holds the top 20. You rebalance monthly. In volatile markets, a statistical model detects the shift and the system moves to cash, closing all positions until calm conditions return.

Both regimes use volatility-based position sizing, sector caps, and crash stops. Over a 6-year test period, MTUM (the leading momentum ETF) returned +126%. The system returned +485%, with a 35% CAGR.

Everything is validated with walk-forward testing across 12 non-overlapping windows spanning 6 years of real market data. That includes COVID, rate hikes, and bear markets. Every window is shown, not just the good ones.

Who this is for

Investors who want systematic, data-driven exposure to US equity momentum without managing it themselves. The strategy is available on Collective2, where trades are executed automatically in your linked brokerage account. The system handles entries, exits, and regime switches on its own. You don't place orders or monitor positions.

Most of the time the portfolio rebalances once a month. When markets turn volatile, the system moves to cash automatically and re-enters when conditions improve.

Why I built this

Momentum investing has decades of academic evidence behind it, but running it well requires daily data processing, disciplined rebalancing, and a risk management layer that actually responds to market conditions. I built Verdict Trading to handle all of that systematically. Every decision is model-driven, and every part of the system (data pipeline, model, backtesting framework, and execution) was built from scratch and validated on six years of real market data.

About me

I'm Devin, a software engineer based in Ontario, Canada. My day job is building reliable, scalable systems. Verdict Trading is where I apply that same discipline to investing.