Two Numbers to Rule Them All
Joel Greenblatt's Magic Formula has beaten the stock market over multiple decades using exactly two financial metrics. Not fifty indicators, not a machine learning model, not a team of analysts. Just two numbers.
Understanding these numbers deeply — not just what they are, but why they work — is the foundation of everything the Magic Formula does.
Metric One: Return on Capital (ROC)
The Concept
Return on Capital answers a fundamental question: how good is this business at turning invested money into profit?
Every business requires capital to operate. A retailer needs inventory and store space. A manufacturer needs machinery and factories. A software company needs servers and salaries. Return on Capital measures how much operating profit the business generates for every euro of capital it employs.
A company that generates €25 million in operating profit using €100 million in capital has a Return on Capital of 25%. Another company generating the same €25 million profit but requiring €500 million in capital has an ROC of 5%. The first company is a dramatically better business. It does more with less.
The Formula
Greenblatt calculates Return on Capital as:
ROC = EBIT ÷ (Net Working Capital + Net Fixed Assets)
- EBIT = Earnings Before Interest and Taxes — operating profit before financing costs or tax effects
- Net Working Capital = Current assets minus current liabilities (excluding cash and interest-bearing debt)
- Net Fixed Assets = Property, plant and equipment after depreciation
The denominator represents the tangible capital actually required to run the business. Greenblatt excludes excess cash and interest-bearing debt deliberately — he wants to measure the returns generated by the operating assets, independent of how the company is financed.
Why Greenblatt Uses EBIT, Not Net Income
Many investors use net income in their calculations. Greenblatt prefers EBIT for two reasons:
1. It removes capital structure effects. A company financed primarily by debt will have high interest expenses, making net income look small. EBIT strips both effects away, allowing meaningful comparison between companies with different financial structures.
2. It removes tax differences. Companies in different countries or industries can show very different net income figures due to tax rates alone. EBIT neutralises this.
What High ROC Tells You
A consistently high Return on Capital is one of the most reliable signals of a durable competitive advantage. Companies with high ROC typically have one or more of the following:
- Brand power that allows premium pricing (luxury goods, consumer staples)
- Switching costs that lock in customers (enterprise software, payroll services)
- Network effects that make the product more valuable as more people use it
- Cost advantages through scale, proprietary processes, or geographic positioning
- Intangible assets such as patents, licences, or regulatory approvals
Benchmark: What Is a "Good" ROC?
| ROC Level | Interpretation |
|---|---|
| Below 10% | Below-average business quality |
| 10–20% | Average to above-average |
| 20–40% | Strong competitive position |
| Above 40% | Exceptional business with significant moat |
Metric Two: Earnings Yield
The Concept
Earnings Yield answers a different question: how much am I paying for the earnings this business generates?
If you buy a company for €10 million and it earns €2 million per year, your earnings yield is 20%. You're getting €20 back annually for every €100 you invest. That's a good deal. If you pay €200 million for the same €2 million in earnings, your yield drops to 1%. That's an expensive deal.
Earnings Yield is the inverse of the Price-to-Earnings ratio (P/E), but calculated more precisely — using Enterprise Value rather than market capitalisation, and EBIT rather than net income.
The Formula
Earnings Yield = EBIT ÷ Enterprise Value
Where Enterprise Value (EV) = Market capitalisation + total debt − cash and cash equivalents.
Why Enterprise Value Instead of Market Cap
When you buy a company, you also effectively take on its debt (and receive its cash). Enterprise Value accounts for the full picture. Consider two companies, both with a market cap of €100 million and EBIT of €20 million:
Company A has €50 million in debt and no cash. Its EV is €150 million. Earnings yield: 13.3%.
Company B has no debt and €30 million in cash. Its EV is €70 million. Earnings yield: 28.6%.
Company B is dramatically cheaper on a true basis — the market cap alone obscures this. Enterprise Value reveals it.
Benchmark: What Is a "Good" Earnings Yield?
| Earnings Yield | Context |
|---|---|
| ~4–5% | Rough level of a fairly-valued S&P 500 |
| 6–9% | Modestly undervalued |
| 10–15% | Significantly undervalued |
| Above 15% | Deep value territory (but verify quality) |
Why These Two Metrics Together Are More Powerful Than Either Alone
The Problem With Value Alone
Screening only for high earnings yield will fill your portfolio with value traps — companies that appear cheap because they're fundamentally broken. A struggling retailer losing customers every year. A cyclical business at the peak of the cycle. The price is low because the future is worse than the past.
The Problem With Quality Alone
Screening only for high return on capital will fill your portfolio with overpriced compounders — exceptional businesses that everyone already knows are exceptional. The market has bid up their prices to reflect their quality, and future returns are therefore modest. You're buying a wonderful business at a terrible price.
The Power of the Combination
When you require both metrics simultaneously — high ROC and high earnings yield — you dramatically narrow your universe to companies that are genuinely good businesses priced as if they were mediocre or troubled. This gap between quality and price is where investment returns are generated.
Common Mistakes When Calculating These Metrics
Mistake 1: Using market cap instead of enterprise value for earnings yield. This significantly distorts comparisons between debt-heavy and debt-light companies.
Mistake 2: Using net income instead of EBIT. Net income is affected by capital structure and tax rates in ways that make cross-company comparison unreliable.
Mistake 3: Including financial companies and utilities. Banks and insurance companies use capital very differently — their ROC and earnings yield figures are not comparable to operating businesses.
Mistake 4: Using a single year of data. One-year EBIT figures can be distorted by unusual items. Where possible, using normalised or multi-year average EBIT provides a more reliable picture.
The Bottom Line
Return on Capital and Earnings Yield are not arbitrary choices. They represent the two most fundamental questions in investing: is this a good business, and am I paying a fair price for it?
By quantifying both and combining them into a single ranking, the Magic Formula turns these timeless questions into an actionable, systematic process that any investor can follow — without requiring individual stock analysis, market predictions, or financial expertise. The formula is simple. The underlying insight is profound.
If you're new to the Magic Formula, our 5-minute beginner's guide explains the full strategy step by step. Once you're ready to find stocks, check out our comparison of the best Magic Formula screeners available today.
