Four filters. One factor. Asymmetric drawdown protection.

Quality is not "the stock I trust". It is a four screen filter you can write in one line on a spreadsheet. The premium it pays is small in calm years and very large in the years you remember for the wrong reasons.

Filter 1 . Profitability

ROE > 15

Five year average. Generated on equity not on debt.

Filter 2 . Leverage

D / E < 0.5

Net debt to equity. Balance sheet does the heavy lifting.

Filter 3 . Stability

EPS sd < 25

Std dev of EPS over five years. Earnings volatility low.

Filter 4 . Accruals

CFO / NI > 0.8

Cash backs the profit. Sloan 1996 accrual test.

Regime window

Quality cohort

Junk cohort

Gap

10Y CAGR 2016 to 2026 . illustrative backtest

14.6pct

11.9pct

+2.7pp

Max drawdown Mar 2020 . COVID shock

-24.8pct

-38.4pct

+13.6pp shallower

Recovery days From COVID low to prior high

187days

364days

half the dwell

Sharpe 10Y . excess of T bill

0.92

0.51

+0.41

The math of the edge

A 13.6 pp shallower drawdown is the real product. Recovery is asymmetric. -25 pct needs +33 pct to break even. -38 pct needs +62 pct. The quality cohort does not need to outperform on the way up. It is already ahead because it never went as far down.

Cohorts built from NIFTY 200 universe sorted on the four filter screens. Asness Frazzini Pedersen "Quality Minus Junk" Review of Accounting Studies 2019. Sloan 1996 Accounting Review on accruals. Past performance . backtest only . not a guarantee.