The market is not efficient — it is systematically irrational. Investors consistently overpay for exciting stories, hot sectors, and recent price momentum while ignoring cheap, boring businesses with strong balance sheets. Benjamin Graham documented this in 1934. Warren Buffett exploited it for six decades. Wesley Gray and Tobias Carlisle quantified it rigorously in 2012, showing that a four-factor composite beats human stock-picking across every market studied — not because the formula is clever, but because it is consistent and unemotional.

The core insight is simple: cheapness alone is not enough. A stock can be cheap because it is genuinely undervalued — or because it is a value trap heading for bankruptcy. The composite filters both. It ranks every stock on an earnings yield (TEV/EBIT), a capital efficiency metric (ROIC), a financial health score (Piotroski F-Score), and an earnings quality signal (Accruals). The equal-weighted average of four percentile ranks is the QV Score. The machine doesn't have opinions about management or industry trends. It reads the numbers.

TEV/EBIT is Gray's primary valuation metric. Unlike P/E, it includes the entire capital structure — equity plus debt minus cash — so it can't be gamed by leverage. Unlike EV/EBITDA, it doesn't add back depreciation, which is a real economic cost. A multiple below 10× means you are paying less than ten dollars for each dollar of operating earnings. The lower the multiple, the higher the implied earnings yield. Below 10× is typically attractive. Above 20× requires extraordinary growth to justify.

ROIC measures how efficiently a business deploys the capital entrusted to it. A company earning 25% ROIC on a 10% cost of capital is creating value with every dollar reinvested. A company earning 4% is destroying it. Combined with cheap valuation, high ROIC separates quality value from value traps. Gray calls this the “magic formula in reverse” — you want the cheapest stock that is also the best business. Above 15% is the threshold where compounding becomes meaningful.

Joseph Piotroski's 2000 paper showed that nine simple yes/no tests of financial statements reliably separate improving companies from deteriorating ones. Score 7–9 signals expanding margins, falling debt, and improving efficiency. Score 0–3 is a distress flag. Gray uses F-Score to eliminate value traps — stocks that look cheap because they're heading for bankruptcy or dilutive capital raises. It's the filter that stops you buying the wrong kind of cheap.

Richard Sloan's 1996 paper documented the accruals anomaly: when accounting earnings exceed operating cash flows, the gap systematically mean-reverts. Companies reporting high accruals often have aggressive revenue recognition, channel stuffing, or one-time gains that cannot repeat. Negative accruals — where cash earnings exceed accounting profit — are a quality signal. The model penalises high accruals directly, reducing the weight of expensive-looking profitable companies that are not generating real cash.

No single factor is sufficient. Cheapness without quality is a value trap. Quality without cheapness is an overpriced growth stock. The QV composite captures both simultaneously. Each stock is ranked on all four factors within the universe, assigned a percentile (0–100), and the four percentiles are averaged with equal weight. The result is the QV Score — a single number that captures relative value and quality simultaneously.

Quantitative value investing has worked historically. It will not work perfectly, always, in every market. Honest disclosure of its limitations is a prerequisite for using it responsibly.