Magic Formula Investing: Joel Greenblatt's Systematic Approach to Beating the Market

Magic Formula Investing: Joel Greenblatt's Systematic Approach to Beating the Market

In 2005, Columbia Business School professor and hedge fund manager Joel Greenblatt published "The Little Book That Beats the Market" — a slim volume written to explain stock picking to his children that ended up becoming one of the most influential investing books of the last two decades.

The central idea, which Greenblatt called the Magic Formula, is disarmingly simple: rank every stock in the investable universe by two measures — how cheap it is (earnings yield) and how good the underlying business is (return on capital) — then buy the stocks that rank highest on the combined measure.

Behind the simplicity is a rigorous framework that avoids several of the most common pitfalls in traditional value investing screens. The backtested results were extraordinary. And the reason most investors cannot replicate those results, even when they know the method, tells you something important about how markets actually work.

This article is for educational purposes only and does not constitute financial advice.

The Two Factors

Factor 1: Earnings Yield (EBIT/EV)

The Magic Formula measures cheapness using EBIT divided by Enterprise Value, not the more common P/E ratio. This is intentional, and the difference matters.

EBIT/EV advantages over P/E:

P/E is Price divided by Net Income. Net income is the bottom line after interest expense and taxes, which means it is heavily influenced by how a company is financed. A business that carries significant debt will show high interest expenses, reducing net income and inflating the apparent P/E. Two identical operating businesses — same revenues, same operating profits, same assets — will show very different P/E ratios if one is funded mostly with equity and the other with debt.

EBIT (Earnings Before Interest and Taxes) is the operating profit of the business, before financing decisions affect the numbers. Enterprise Value (market cap plus net debt) represents what you would actually pay to acquire the entire business — equity and debt together. Dividing one by the other gives you the return you would earn on the total business, regardless of how it is financed.

This makes the comparison between companies far more apples-to-apples. A company with $100M in EBIT and a $1B enterprise value has a 10% earnings yield whether it has $0 in debt or $500M in debt.

Factor 2: Return on Capital (EBIT / Net Working Capital + Net Fixed Assets)

The Magic Formula measures business quality using EBIT divided by tangible capital employed, where tangible capital is defined as net working capital plus net fixed assets.

This is also a deliberate departure from the more common Return on Equity (ROE). ROE is earnings divided by book equity, which is significantly affected by financial leverage. A company can achieve a high ROE simply by using a lot of debt, even if its underlying business earns mediocre returns. ROE tells you how efficiently a company uses equity capital; it does not tell you whether the business itself is a good one.

Return on tangible capital measures how much operating profit the business earns relative to the actual operating capital it requires — the working capital tied up in receivables, inventory, and payables, plus the fixed assets like equipment and facilities. This strips out the effect of financing and focuses on the economics of the core business.

A business that earns 50% ROIC on tangible capital requires very little capital to grow, generates substantial free cash flow, and is likely protected by a competitive moat. A business that earns 6% ROIC on tangible capital is barely covering its cost of capital and creates little value for shareholders.

Why the Combination Matters

Neither factor alone is sufficient.

High earnings yield without high return on capital gives you cheap stocks that are cheap for a reason — commodity businesses, cyclicals at peak earnings, or genuinely impaired companies. This is the classic value trap.

High return on capital without high earnings yield gives you great businesses at expensive prices. You might own a wonderful company and still earn mediocre returns because you paid too much for it.

The Magic Formula looks for the overlap: businesses that are both good (high ROIC) and cheap (high earnings yield) simultaneously. This combination is rare, which is why the ranking methodology is so important — it finds the stocks where the two factors converge.

The Original Backtested Results

Greenblatt's backtest covered U.S. stocks with market capitalizations above $50 million from January 1988 to December 2004 (17 years). The results:

• Magic Formula portfolio: 30.8% annual return
• S&P 500: 12.4% annual return
• Outperformance: approximately 18.4% per year

Even excluding the best-performing 5-10% of stocks (to test whether the results depended on a few outliers), the strategy significantly outperformed.

Subsequent academic studies and practitioner implementations have generally confirmed that the underlying logic holds, though the magnitude of outperformance has moderated as the approach became more widely known. Returns in the 12-15% range have been more typical in more recent periods — still meaningful outperformance, but not the extraordinary gap of the original study.

How to Run the Screen: Step by Step

The implementation is straightforward in principle:

Step 1: Define your universe. Greenblatt recommends stocks above $50M-$100M in market cap. Exclude financial companies (banks, insurance) — their balance sheets work differently, making the capital ratios meaningless. Exclude utilities — they operate in regulated markets where returns are capped by regulation. Exclude foreign stocks if you want a domestic-only screen.

Step 2: Calculate earnings yield for every stock. Compute trailing twelve-month EBIT and current enterprise value (market cap + total debt - cash). The ratio is the earnings yield. Rank all stocks from highest earnings yield to lowest. Record each company's earnings yield rank.

Step 3: Calculate return on capital for every stock. Compute trailing twelve-month EBIT, net working capital (current assets - current liabilities, excluding cash and interest-bearing debt), and net fixed assets. The ratio of EBIT to their sum is the return on capital. Rank all stocks from highest return on capital to lowest. Record each company's ROIC rank.

Step 4: Add the two ranks. For each stock, sum its earnings yield rank and its ROIC rank. The combined rank is the Magic Formula score. Lower combined rank is better (it means the stock ranked near the top on both measures).

Step 5: Buy the top 20-30 stocks. Spread purchases over 9-12 months to reduce timing risk. Greenblatt recommends buying 2-3 positions per month rather than all at once.

Step 6: Rebalance annually. After approximately one year (for tax efficiency in taxable accounts, selling losers just before the one-year mark and winners just after), liquidate each position and reinvest in the current top-ranked companies. This systematic rebalancing is a critical component of the strategy.

The ValueMarkers Magic Formula Calculator automates Steps 1-4 across the U.S. equity universe, showing you the current top-ranked stocks and their individual factor scores.

The Implementation Gap: Why Most People Cannot Stick to It

Here is the uncomfortable reality about the Magic Formula: the strategy underperformed the S&P 500 in roughly one out of every four calendar years in Greenblatt's original study. In some of those years, it underperformed by 10-15 percentage points. If you started running the strategy in one of those years, you would have experienced significant pain before the long-run advantage asserted itself.

This is what Greenblatt calls the "implementation gap" — the gap between the theoretical returns of a systematic strategy and what investors actually earn because they abandon the strategy during periods of underperformance.

The psychological mechanism is predictable: the Magic Formula screen tends to buy companies that look ugly. Value stocks that also have high returns on capital are often in beaten-down sectors, recovering from short-term earnings disappointments, or simply boring businesses that have been temporarily ignored. Holding these while the market rallies on tech momentum or other fashionable themes requires genuine conviction.

When the strategy underperforms for 18 months, it starts to feel broken. The natural response is to "override" the screen — adding qualitative filters, avoiding the most uncomfortable names, or gradually shifting to more conventional stocks. Each override might seem justified in isolation, but collectively they transform a systematic strategy into an ad hoc one that no longer has the same properties.

The key insight from Greenblatt is that the strategy works partly because it is uncomfortable. If it were easy to hold, more investors would hold it, and the returns would be competed away. The emotional difficulty of following it mechanically is the moat around the strategy's returns.

Limitations of the Magic Formula

Excludes financials and utilities. The return-on-capital calculation does not work for banks (where leverage is the business model) or utilities (where regulated returns cap ROIC). This eliminates a significant portion of the investable universe. Separate approaches are needed for these sectors.

Backward-looking earnings. EBIT is historical. A company with high trailing earnings yield may have just reported peak-cycle earnings that are about to mean-revert. The screen does not distinguish between durable above-average earnings and cyclically elevated earnings.

Requires annual rebalancing. This is not a buy-and-hold strategy in the traditional sense. It requires systematic turnover of the portfolio each year, which generates transaction costs and, in taxable accounts, capital gains. The strategy's net-of-tax returns will differ from the gross backtest.

Works best with patience over 3+ year horizons. Short holding periods amplify the noise in performance. The strategy's statistical edge becomes meaningful only when evaluated over multi-year periods. This requires tolerance for interim underperformance.

Magic Formula vs. Other Value Screens

The Magic Formula occupies a unique space by explicitly requiring both quality and value simultaneously, rather than screening on either dimension alone.

Putting It Into Practice

For investors interested in running a Magic Formula approach, the practical recommendations from Greenblatt are worth following closely:

• Use the screen as a starting point, not an endpoint. Reading the annual reports of the top-ranked companies takes the screen from mechanical to informed.
• Diversify across 20-30 names minimum. Individual stocks within the top-ranked group will have highly variable outcomes. The edge is statistical, not certain for any single stock.
• Commit to the annual rebalancing. The discipline of systematic selling and rebuying is what keeps the portfolio anchored to the strategy.
• Do not override the screen based on short-term performance. The times it feels most wrong are often when the long-term edge is greatest.

The Magic Formula is not guaranteed to outperform in any given year, sector, or market environment. What it offers is a systematic, repeatable process grounded in two durable economic relationships: businesses that earn high returns on capital tend to maintain those returns, and businesses trading at cheap prices relative to those returns tend to re-rate upward over time. That combination has proven to be a persistent source of long-term advantage.