Factor Investing: Building a Systematic Edge
Value, Momentum, Quality, Size, Low-Vol — the academic factors that work, and how to combine them.
Factor investing is the empirical foundation of quantitative stock selection. Decades of research have identified persistent characteristics that predict higher returns.
The Big Five
- Value — Cheap stocks (low P/B, P/E, EV/EBITDA) outperform over time. Fama-French (1992). Premium is ~3–5% annually but comes with long droughts.
- Momentum — Stocks going up tend to keep going up (12-1 month return). Jegadeesh & Titman (1993). Strongest short-term predictor but crash-prone in regime shifts.
- Quality — Profitable companies with strong balance sheets outperform. Novy-Marx (2013) showed gross profitability predicts returns as well as value.
- Size — Small-caps outperform large-caps over very long horizons. The premium has weakened since publication — possibly arbitraged away.
- Low Volatility — Low-vol stocks deliver better risk-adjusted returns, defying CAPM. Persists because institutions chase high-beta for benchmark outperformance.
Why factors work
Risk-based: Premiums compensate for real risks (distress, slow information diffusion).
Behavioral: Investors systematically overpay for glamour, under-react to news, neglect boring compounders.
Our 5-pillar implementation
- Quality pillar = Quality factor (ROIC, margins, balance sheet)
- Growth pillar = Growth factor (revenue/EPS acceleration)
- Valuation pillar = Value factor (P/E, EV/EBITDA, sector-relative)
- Momentum pillar = Momentum factor (12-1 return, risk-adjusted)
- Revisions pillar = Earnings revision factor (analyst estimate changes)
Weights are dynamically adjusted for sector context and macro regime — our edge over static factor ETFs.
The interaction effects
Cheap + improving momentum = deep value turning around (most profitable). Expensive + decelerating momentum = crowded growth unwinding (most dangerous). Our scorer captures these through pairwise pillar interaction terms.