Joel Greenblatt investing strategy

Joel Greenblatt is two schools in one person, and you need both. There's the hands-on Greenblatt of special situations, from You Can Be a Stock Market Genius (1997), and the systematic Greenblatt of the Magic Formula, from The Little Book That Beats the Market (2005). Both rest on one conviction, that the market leaves structural mispricings lying around for the patient to exploit.

They just exploit it in opposite styles. One works by hand in overlooked corners. The other ranks the whole market on two numbers.

Why are spin-offs and other special situations mispriced?

Special situations are where inefficiency gets manufactured. Greenblatt's first book is a field guide to corporate events that create mispricing as a byproduct: spin-offs, restructurings, mergers and merger securities, rights offerings, recapitalizations, bankruptcies, and stub stocks. His favorite hunting ground is the spin-off, where a parent company distributes shares of a subsidiary to its holders. Work out why spin-offs are systematically mispriced and you'll understand how structural inefficiency works in general.

Four mechanisms drive it, and you can verify each one in the filings:

1. Forced, indiscriminate selling. When a large company spins off a smaller one, the parent's shareholders receive shares they never chose and often can't hold. Those holders include index funds and institutions with mandates. They also include investors who owned the parent for reasons that don't apply to the spin-off. An index fund tracking a large-cap index has to sell a small-cap spin-off mechanically, at any price, on day one. That selling ignores price, and it pushes the spin-off down for reasons that have nothing to do with its value. It's the purest form of the forced selling Klarman hunts.
2. No coverage, no attention. The spin-off is small and new, and nobody follows it. No analyst models it on day one, and the parent's investor-relations team ignores it. Greenblatt's edge is information that's public but unprocessed. You do the work nobody else has bothered to do yet.
3. Newly aligned incentives. Spin-off management usually receives fresh equity in the standalone company, often after the initial price has been deliberately kept quiet. Freed from the parent's bureaucracy, they're suddenly running their own company with its own capital allocation. Value that was trapped in an underappreciated division of a conglomerate is now visible, and someone is paid to grow it. It's consolidation and the holdco discount running in reverse.
4. Hidden quality. The spin-off is often the best division of a mediocre parent. The market stays anchored to the parent's mediocre multiple and misprices the freed jewel.

Worked example: a demerger in India

Worked application (India: the demerger). India's version of the spin-off is the demerger, and the structural inefficiency is often sharper than in the US because of index mechanics. Take Jio Financial Services, demerged from Reliance Industries in 2023. Index funds that held Reliance received Jio Financial shares that didn't belong in their large-cap mandates. The stock was given a placeholder price in a special session and then had to be removed from the index. Those inclusion and exclusion mechanics forced price-insensitive selling. That's Greenblatt's mechanism 1, playing out at national scale. India's chronic holding-company and conglomerate discounts are a standing inventory of Greenblatt situations. A demerger is the catalyst that can collapse a holdco discount, and the analyst who has done the sum-of-the-parts before the announcement owns the edge.

One caution is specific to India. Promoter dynamics and related-party history, the ethics thread from the full-filing capstone, mean you have to check why the demerger is happening and who benefits. Sometimes it unlocks value for every holder. Sometimes it's a restructuring that favors the promoter.

The mechanism creates the opportunity. Governance decides whether it's real.

What is the Magic Formula?

Greenblatt's second, more famous idea is the Magic Formula, his attempt to bottle the Buffett principle of a good business at a bargain price into a mechanical rule any investor can run without judgment. It ranks the entire investable universe on two factors and combines them:

• Earnings yield = EBIT / Enterprise Value. This is "cheapness." Greenblatt chose EBIT/EV over the ordinary earnings-to-price ratio on purpose. EV (market cap + net debt + minorities − cash, as you built it in the ratio and relative-valuation work) and EBIT (operating profit before interest and tax) are both neutral to capital structure and tax rate. That lets you compare the operating cheapness of levered and unlevered businesses like for like. A high earnings yield means you get a lot of operating profit per dollar of enterprise value, so the business is cheap.
• Return on capital = EBIT / (Net Working Capital + Net Fixed Assets). This is "quality," the returns-on-capital idea in Greenblatt's own words. He uses tangible capital employed, which is net working capital plus net fixed assets with goodwill and intangibles left out. He wants the return on the capital the business actually needs to run, and the price a prior acquirer paid doesn't count. A high ROC means the business earns a lot on the capital it uses, so it's good.

The procedure is simple. Take every stock above a market-cap floor except financials and utilities, whose balance sheets make EBIT over tangible capital meaningless. Rank them from 1 to N on earnings yield, and separately from 1 to N on return on capital. Sum the two ranks. Buy a basket of the names with the lowest combined rank, typically 20 to 30 of them. Hold for about a year, timing your sales with taxes in mind. Then re-rank and rotate.

You're systematically buying where cheap and good overlap. Paying up for good fails the earnings-yield leg, and buying cheap junk fails the ROC leg. Worked Example 1 runs a full ranking so you can do it by hand.

What does the evidence say about the Magic Formula?

What the evidence actually says. Read this part twice. The Magic Formula is real. The way it's usually sold isn't, and you need to carry the accurate version around with you:

• The headline backtest is a starting point. The Little Book reported roughly a 30% annualized return over its 1988 to 2004 test, against low teens for the market. That result has been legitimately debated. People have questioned the data source and the market-cap threshold, since small illiquid names can inflate paper returns you can't actually capture. They've questioned the rebalancing assumptions too, and there's the ever-present risk of a rule discovered by looking at the same history it's tested on. Treat the 30% as an illustration of the idea. It isn't a number you'll earn.
• It underperforms for years at a time, and that's why it works. Greenblatt is refreshingly frank that the formula can lag the market for two or three consecutive years, and he doesn't treat that as a bug. A mechanical value-and-quality rule that felt comfortable to hold would get held by everyone, and the premium would vanish. The edge is behavioral. It survives because most people abandon the formula during the stretches when it looks broken. So it only "works" for an investor with the temperament to keep running it through the lean years. That's the Marks and Klarman point in a systematic wrapper.
• Much of it is just value × quality: known factors. Modern factor research shows the Magic Formula is, in effect, a clean combination of a value factor (cheapness, the earnings-yield leg) and a quality or profitability factor (the ROC leg). Both are documented sources of long-run excess return. Studies that break the formula down find that those two factors capture a large share of its edge, and you could buy both through low-cost factor funds. That doesn't make the formula wrong. It makes it understood. It's a sound, rules-based way to harvest two real premia, and there's no alchemy in it. Since publication, and with the growth of quant investing, the raw premia have also compressed from crowding, so expect less than the book.
• Use it as a screen and a set of rules, the way Greenblatt himself suggests. That gives you a hopper of cheap-and-good candidates to analyze by hand, with the quality-of-earnings checks from your forensic work, since a high EBIT/EV can be a value trap hiding a shenanigan or a melting cyclical at peak earnings. It's also a mechanical antidote to your own emotions. Treat it as a black box you trust blindly and it'll eventually feed you a cluster of value traps and test your faith at the worst moment. Use it as a sourcing tool with judgment on top and it's one of the most useful screens ever published.

When does Greenblatt's approach work, and when does it fail?

So when does Greenblatt work? Special situations pay off when you're willing to do unglamorous plumbing in corners institutions ignore. They fail when the "hidden value" was hidden for a good reason, like a spin-off loaded with the parent's debt and worst assets. Always read why the parent shed it.

The Magic Formula works over long horizons for the patient. It fails the impatient, and it fails the under-diversified, because any single mechanical pick can be a trap and a 20 to 30 name basket is essential. It also fails anyone who runs it on financials or utilities, where its ratios lie.