The Magic Formula
Warren Buffett's whole philosophy — a wonderful company at a fair price — boiled down to two numbers, applied to every stock at once, and ranked. No forecasts, no narratives, no judgment calls. Just: how cheap is it, how good is the business, and how do the two combined stack up against everything else? A hedge-fund manager turned that into a strategy that beat the market for two decades.
01Where it comes from
Joel Greenblatt ran Gotham Capital to a reported ~40% a year from 1985 to 2005 — one of the great track records in investing. Then he did something unusual: he tried to distill it into a rule a teenager could follow, and published it in The Little Book That Beats the Market(2006). He called it the Magic Formula, half tongue-in-cheek, because it's really just discipline wearing a catchy name.
The insight is Buffett's, made mechanical. Buffett buys wonderful businesses at fair prices. Greenblatt asked: what if you scored every stock on exactly those two dimensions — quality and cheapness — ranked them, and simply bought the top of the list? No opinions required.
02The two ingredients
It doesn't look for the cheapest stock or the best business. It looks for the best combination — and that corner of the market is where the bargains with a future tend to hide.
03The two axes, on real stocks
Plot our two recurring names on the quadrant the formula is really searching — cheapness up the side, business quality across the bottom:
weak business
(value trap)
the sweet spot
& weak
(avoid)
expensive
Linamar is the Magic Formula's ideal: genuinely cheap on operating earnings and a decent return on capital — high on both axes, so it ranks well. Constellation is a superb business (returns most companies can only dream of) but a low earnings yield — it wins the quality contest and loses the price one, so its combinedrank suffers. The formula isn't wrong about Constellation's quality; it's just unwilling to pay up for it. That's the whole discipline in one picture.
04When the Magic Formula lies
It's a ranking engine, not an oracle, and treating it like a stock picker is where people get hurt.
Four things the ranking won't tell you
- It's a basket, not a pick. Greenblatt's method is to buy the top 20–30 and hold a year, then rotate — the edge is statistical across the group. Any single name on the list can still be a disaster.
- It's blind to financials and utilities.Banks, insurers, and regulated utilities have balance sheets that distort EBIT/EV and return on capital, so the app (like Greenblatt) excludes them — the ranking simply doesn't apply there.
- Trailing EBIT can be a peak.A cyclical at the top of its cycle looks cheap and high-return right before earnings roll over. The formula can't see the cycle; you have to.
- It only works if you can endure it.Since the 2006 book, the strategy has had long stretches of underperformance. It “works” partly becauseit's psychologically hard to hold through those stretches — most people quit at the worst time.
Used as Greenblatt intended — a disciplined shortlist of cheap, high-quality businesses to research further — it's one of the most elegant ideas in investing. The app runs the ranking across the whole universe so you don't have to, then hands each name to the rest of the toolkit — the Graham checks, the F-Score, the Altman Z — because a high rank is where the work starts, not where it ends.
Rank the market by the Magic Formula
The screener scores the full North American universe on earnings yield and return on capital and sorts by the combined rank — the Magic Formula shortlist, updated from current data, with each name one click from its full valuation.