Topic survey
Quality
Profitable, well-run, conservatively-financed firms beat junk — the best-curated topic in the library, and the value investor's essential complement.
TL;DR
Firms that are more profitable, higher-quality, and more conservative (in investment, accruals, and financing) earn higher risk-adjusted returns than low-quality "junk." Quality is the natural complement to value — it screens cheap-good firms from cheap-bad ones — and it is the most heavily curated topic in our library.
A 30-year arc
Sloan (1996) — the accruals anomaly: earnings backed by cash flow are more persistent than accrual-driven earnings, and the market overprices accruals.
Piotroski (2000) — the F-score: 9 fundamental signals separate winners from losers among value firms.
Mohanram (2005) — the G-score: the growth-stock mirror of F-score.
Fama & French (2006); Novy-Marx (2013) — the profitability premium: gross (Novy-Marx) and operating (FF) profitability predict returns, given value.
Cooper, Gulen & Schill (2008) — asset growth / investment: firms that expand aggressively subsequently underperform.
Fama & French (2015) — RMW (profitability) and CMA (investment) become factors in the five-factor model.
Sub-threads (facets of quality)
Profitability (gross / operating) · earnings quality & accruals · investment / asset growth · F-score & G-score fundamental composites · leverage & financing · earnings stability. "Quality minus junk" bundles several of these into one composite.
Why it works
Fundamental mispricing — investors fixate on earnings levels and underweight their quality (accrual vs cash composition, sustainability), so high-quality firms are underpriced.
Priced risk / valuation identity — Fama-French show that, given price-to-book, higher profitability and lower investment must imply higher expected returns; profitable, conservative firms also resemble a defensive tilt.
The dark side
Definition sensitivity — "profitability" (gross vs operating vs net) materially changes results; the measure matters.
Accrual decay — the accruals anomaly has weakened markedly since the 2000s as it was arbitraged.
Overlap — quality correlates with value, low-volatility, and size; standalone alphas shrink after controls; many fundamental signals are data-mined.
Does it survive out of sample?
Profitability/quality is among the more durable premia — robust across our replications and in post-publication data — while the accruals leg has compressed. Our replications recompute gross profitability, operating profitability (RMW), accruals, and investment, scored on the holdout (profitability reads alive).
Run it yourself
Replications — Novy-Marx gross profitability, Fama-French operating profitability, Sloan accruals, Cooper-Gulen-Schill investment.
Curriculum — the synthetic market ships a planted qual_roe factor; the cost-of-trading mission shows how quality's lower turnover helps net returns.
Playground / Competitions — build a quality screen and score it OOS.
