Joel Greenblatt's Magic Formula - Does It Still Work?
Joel Greenblatt published "The Little Book That Beats the Market" in 2005 with a deceptively simple premise: a mechanical formula based on just two financial ratios could consistently outperform the stock market. Greenblatt was not a self-help author or a newsletter writer. He was the founder of Gotham Capital, a hedge fund that returned approximately 50% annually from 1985 to 1994 (40% net of fees), one of the best track records in hedge fund history. When someone with that pedigree claims to have a formula that works, the investment community pays attention.
The Magic Formula ranks stocks by combining two metrics: earnings yield (a measure of cheapness) and return on capital (a measure of business quality). Stocks that rank highly on both dimensions, cheap and good, are bought and held for one year, then replaced. The formula is a quantitative distillation of the value investing principle that the best investments are high-quality businesses purchased at bargain prices. Greenblatt's contribution was to show that this principle could be implemented mechanically, without any qualitative judgment, and still produce strong results.
How the Magic Formula Works
The formula uses two specific metrics, defined slightly differently from their standard textbook versions.
Earnings Yield = EBIT / Enterprise Value
EBIT (earnings before interest and taxes) is used instead of net income to make companies comparable regardless of their capital structure and tax situation. Enterprise value (market capitalization plus net debt) is used instead of stock price to account for the value of debt in the capital structure. A company with a high earnings yield is generating a lot of operating income relative to its total cost (equity plus debt).
Return on Capital = EBIT / (Net Fixed Assets + Net Working Capital)
This measures how efficiently the business converts its invested capital into operating earnings. A company with a high return on capital is generating a lot of operating income from a small asset base. This is a proxy for business quality: companies that earn high returns on capital typically have competitive advantages that allow them to charge premium prices, operate more efficiently, or both.
The implementation is straightforward. Take all stocks above a minimum market capitalization threshold (Greenblatt originally suggested $50 million, later raised to $100 million). Rank them from 1 to N on earnings yield (highest yield gets rank 1). Rank them from 1 to N on return on capital (highest return gets rank 1). Add the two ranks for each stock. The stocks with the lowest combined rank are the Magic Formula picks.
An investor buys the top 20-30 ranked stocks, holds them for exactly one year, sells them, and repeats with the new top-ranked stocks. The portfolio is equal-weighted, and no qualitative judgment is applied. The formula does all the work.
The Historical Evidence
Greenblatt back-tested the Magic Formula over the 17-year period from 1988 to 2004. The results were striking. A portfolio of the top 30 Magic Formula stocks, rebalanced annually, returned approximately 30.8% per year before transaction costs, compared to 12.4% for the S&P 500. Even after adjusting for reasonable transaction costs and slippage, the formula significantly outperformed.
Greenblatt also tested the formula over shorter periods and found that it underperformed the market in approximately one out of every four years, and in some years by a significant margin. Over any three-year rolling period, it underperformed approximately 17% of the time. Over five-year rolling periods, it almost never underperformed. The pattern is consistent with other value strategies: short-term underperformance is common, but long-term results are strong.
Several academic studies have examined the formula with out-of-sample data. Research published in the Journal of Portfolio Management and other peer-reviewed publications generally confirmed that the two factors (cheapness and quality) each contributed positively to returns, though the magnitude of the combined outperformance varied by time period, market, and implementation details.
A study by Robert Novy-Marx, published in 2013 in the Journal of Financial Economics, provided indirect support for the Magic Formula by demonstrating that profitability (a close cousin of return on capital) is a strong predictor of stock returns. Novy-Marx found that combining profitability with a value metric produced results superior to either factor alone, which is essentially what the Magic Formula does.
Why the Formula Should Work
The theoretical justification for the Magic Formula combines insights from both the value and quality investment traditions.
The earnings yield component is a standard value factor. It buys stocks that are cheap relative to their operating earnings. The value premium, documented by Fama and French and many subsequent researchers, shows that cheap stocks tend to outperform expensive stocks over long periods. The earnings yield component captures this premium.
The return on capital component is a quality factor. It selects companies that generate high returns on their invested capital, which typically indicates competitive advantages, efficient operations, or both. High-return businesses tend to sustain above-average profitability longer than the market expects, creating a persistent source of mispricing. The market may assign a moderate multiple to a company with a 30% return on capital, not fully pricing in the durability of that return advantage.
The combination is more powerful than either factor alone. A cheap stock (high earnings yield) that is also a good business (high return on capital) is more likely to see its price converge toward intrinsic value than a cheap stock that is a bad business. The quality filter screens out the value traps, the statistically cheap stocks that are cheap because the business is permanently impaired.
Where the Formula Struggles
Despite the historical evidence, the Magic Formula has significant limitations that practitioners should understand.
Sector concentration. Because the formula is purely mechanical, it can concentrate heavily in specific sectors. During periods when financial stocks are cheap, the formula may load up on banks and insurers. During energy downturns, it may load up on oil companies. This sector concentration introduces a risk factor that is not captured by the formula's back-tests, which cover periods with specific sector rotations that may not repeat.
Accounting quality. The formula uses reported financial data without adjusting for accounting quality. Companies that manipulate earnings through aggressive revenue recognition, capitalization of expenses, or off-balance-sheet liabilities will appear to have higher earnings yields and higher returns on capital than they actually deserve. The formula has no mechanism for distinguishing genuine profitability from accounting artifice.
The post-2005 reality. The formula's back-test covered 1988 to 2004. In the years since the book's publication, several factors have changed. The strategy has become widely known, potentially reducing its effectiveness as more investors attempt to implement it. The market has shifted toward asset-light technology companies whose returns on capital are difficult to measure using the formula's definitions. And the prolonged period of low interest rates from 2009 to 2022 may have altered the relationship between earnings yield and stock returns.
Real-world implementations of the Magic Formula have produced mixed results since 2005. Greenblatt's own fund, Gotham Asset Management, which manages several formula-based strategies, has had periods of both outperformance and underperformance relative to the S&P 500. The formula-based strategies available on Greenblatt's website (magicformulainvesting.com) have similarly shown periods of strong results and periods of disappointing results.
Turnover and taxes. The annual rebalancing requirement generates significant turnover, which creates tax liabilities for taxable accounts. In a tax-deferred account (IRA, 401k), this is not an issue. In a taxable account, the tax drag can reduce after-tax returns by 1-3 percentage points annually, depending on the investor's tax bracket. This is a meaningful friction that the back-tested results do not reflect.
Greenblatt's Own Evolution
Greenblatt himself has evolved beyond the simple Magic Formula. At Gotham Asset Management, he runs long-short strategies that go beyond the two-factor ranking system. His more recent book, "The Big Secret for the Small Investor" (2011), focuses on equal-weight indexing as an alternative for investors who do not want to pick individual stocks. And his investment lectures at Columbia Business School cover a much broader range of valuation techniques than the Magic Formula alone.
Greenblatt has acknowledged that the formula is a simplified version of how he actually invests. It captures the core principle (buy cheap, quality businesses) in a form that anyone can implement, but it omits the qualitative judgment, the industry knowledge, and the special-situations analysis that characterize his more sophisticated approach at Gotham Capital.
This is an important nuance. The Magic Formula is not Greenblatt's best idea. It is his simplest idea, designed to be accessible to individual investors who do not have the time, skill, or inclination for deep fundamental analysis. For investors who do have those capabilities, the formula is a starting point, not a destination.
Implementing the Formula Today
For investors who want to apply the Magic Formula, several practical considerations deserve attention.
Screen size matters. The formula works best on stocks above approximately $200-500 million in market capitalization, where the data is relatively reliable and the stocks are liquid enough to trade without excessive market impact. Applying the formula to micro-cap stocks introduces additional risks (accounting quality, liquidity, survivorship bias) that may offset the selection advantage.
Diversify across sectors. Consider imposing sector limits on the portfolio (no more than 25-30% in any single sector) to avoid the concentration risk that a purely mechanical approach can create. This is a deviation from the pure formula, but it addresses a real risk that the back-tests may not capture.
Be patient. The formula underperforms in roughly one out of four years. An investor who abandons the strategy after a year or two of underperformance will miss the subsequent recovery. Greenblatt himself emphasizes that the formula requires a minimum of three to five years to demonstrate its advantage, and even then, there is no guarantee for any specific five-year period.
Consider tax implications. In taxable accounts, explore tax-loss harvesting strategies and consider extending the holding period slightly beyond one year to qualify for long-term capital gains rates. The annual rebalancing is not sacred; adjusting the timing slightly can significantly reduce the tax burden.
Use it as a screen, not a system. The most productive use of the Magic Formula for a sophisticated investor is as a screening tool that identifies candidates for further research, not as a complete investment system. Running the formula produces a list of cheap, high-quality companies. The investor then applies qualitative judgment, industry knowledge, and balance sheet analysis to decide which of those candidates to actually buy.
The Bigger Lesson
The Magic Formula's enduring contribution to investment thinking is not the specific two-factor ranking. It is the empirical demonstration that the core principle of value investing, buying good businesses at cheap prices, can be quantified and tested. The formula provides evidence that value investing works not because its practitioners are geniuses, but because the strategy exploits systematic behavioral biases (the tendency to overreact to bad news and underappreciate durable business quality) that produce persistent mispricings in the stock market.
Whether the exact formula will produce 30% returns in future decades is unknowable. Whether the underlying principle, that cheapness combined with quality produces superior long-term returns, will continue to work is a much safer bet. The behavioral foundations of the strategy (overreaction, neglect, short-termism) show no signs of diminishing, which suggests that some version of the Magic Formula's logic will remain profitable for investors with the discipline to apply it consistently.
Put these principles into practice. Track fundamentals, build portfolios, and analyze stocks with AI-powered insights.
Start Free on GridOasis →