Building Quality Minus Junk
QMJ builds a single 'quality' score by averaging together profitability, growth, safety and payout — four different ideas about what makes a company good, combined into one number and traded long-short.
Prerequisites: The Quality Factor
"Quality" is a word every investor uses and few define the same way. Asness, Frazzini and Pedersen's Quality Minus Junk (QMJ) factor turns the fuzzy idea into a mechanical, four-part recipe, then goes long the companies that score highest and short the companies that score lowest — building a tradeable long-short portfolio out of a concept that used to live only in analyst reports.
QMJ combines four sub-scores — profitability, growth, safety, and payout — into one composite z-score per company, then sorts stocks by that composite and trades the top decile against the bottom. Quality here is defined operationally, as "the four things a rational buyer would pay more for," not as a vague label.
The four pillars
Profitability measures how much profit a company squeezes out of its assets or equity — gross profit, margins, and return on equity. Growth measures the trend in that profitability over the past five years — is quality improving or eroding? Safety captures balance-sheet and market risk: low leverage, low earnings volatility, low stock-price beta. Payout measures how much of its profit a company returns to shareholders rather than diluting them through issuance, a proxy for management's confidence and discipline.
Worked example
Take two companies. Company A has high margins, growing earnings, low debt, and pays a steady dividend — its four z-scores might read profitability +1.4, growth +0.8, safety +1.1, payout +0.6, averaging to a composite of about +0.98. Company B has thin, shrinking margins, high leverage, and no payout, with z-scores of profitability −0.9, growth −1.2, safety −1.0, payout −0.7, averaging to about −0.95. Sorted into deciles across the whole market, Company A lands near the top ("quality") and Company B near the bottom ("junk"). QMJ's return is the average of the top decile minus the average of the bottom decile, rebalanced periodically as scores shift.
What this means in practice
Because QMJ is a composite, a company can score high overall while being weak on one pillar — a fast-growing, richly leveraged firm can still land in the top decile if profitability and payout are strong enough to offset a mediocre safety score. This makes QMJ different from a single-ratio factor like gross profitability: it's less exposed to any one accounting line being gamed or noisy, but also harder to explain with a single sentence, since the composite score doesn't map cleanly back to any one financial statement item.
Quality and low-beta are correlated but not identical — safe, low-leverage firms tend to have lower market beta too, so QMJ's returns partially overlap with the betting-against-beta factor. Running both in the same book without checking the overlap can mean paying for the same underlying tilt twice.
Related concepts
Practice in interviews
Further reading
- Asness, Frazzini, Pedersen, 'Quality Minus Junk' (Review of Accounting Studies)