DL 364 · Joel Greenblatt's Magic Formula · Lesson 1 of 5

Two numbers: return on capital and earnings yield

Greenblatt's formula asks two questions about every business. Is it good? Is it cheap? Here's how each one is measured, and why he picked these exact ratios.

Level:Intermediate
◆
Read time:5 min
◆
Course:DL 364
The investor

A hedge fund manager who wrote for his kids

Joel Greenblatt started Gotham Capital in 1985. By most accounts the fund compounded at something like 40% a year for its first two decades, which puts it in a very small club. He has taught value investing at Columbia Business School since 1996.

None of that explains why so many people know his name. That came in 2005, with a short book he says he wrote so his children could understand what he did for a living: The Little Book That Beats the Market. The idea in it fits on an index card.

Buy good businesses when they're cheap, and do it with a fixed set of rules so your feelings don't get a vote.

Anyone who has read Buffett will recognise the spirit. What Greenblatt added was a way to measure "good" and "cheap" with one number each, so that thousands of companies can be ranked in a few minutes.

Is it good?

Return on capital

The first number tells you how much profit a business squeezes out of the money tied up in running it.

Return on capital (Greenblatt)
EBIT ÷ (Net working capital + Net fixed assets)
EBIT is operating profit, before interest and taxes.

Each piece is there for a reason. EBIT lets you compare companies with different debt loads and tax rates, because you're looking at what the operations earn before the lenders and the government take their share.

The denominator is the capital the business actually needs: the money stuck in inventory and receivables (net of what it owes suppliers), plus the factories, stores and equipment, after depreciation. Excess cash is left out. So is goodwill, which is a record of what someone once paid for an acquisition and says very little about the business itself.

A company that earns $30 of operating profit on $100 of working capital and equipment can grow without asking anyone for money. One that earns $8 on the same $100 has to keep feeding the machine.

Why not just use ROIC?
Greenblatt's version is close to ROIC but simpler and stricter. It skips taxes and ignores goodwill, so a company that bought its growth doesn't get to hide the price it paid. Our ROIC lesson covers the standard definition.
Is it cheap?

Earnings yield

The second number turns the usual price multiple upside down and adds debt to it.

Earnings yield (Greenblatt)
EBIT ÷ Enterprise value
Enterprise value = market value of the shares + net interest-bearing debt.

Think of enterprise value as the price for the whole business, debts included. If you bought every share, you'd also inherit what the company owes. A P/E ratio ignores that. Two companies on a P/E of 12 can be wildly different bargains once you notice that one of them carries a pile of debt.

Writing it as a yield (profit divided by price) makes it easy to compare with a bond. An earnings yield of 10% means the operations earn 10 cents a year for every dollar you pay for the whole company.

Worked example

Same profit, very different businesses

Here are two illustrative companies that both report $120 million of EBIT. On a headline basis they look identical. They aren't.

Worked example
Northfield Tools (illustrative)
EBIT$120M
Net working capital$150M
Net fixed assets$250M
Return on capital120 ÷ 400 = 30%
Market value of shares$1,000M
Net debt$200M
Earnings yield120 ÷ 1,200 = 10.0%
Worked example
Sterling Rail (illustrative)
EBIT$120M
Net working capital$100M
Net fixed assets$1,100M
Return on capital120 ÷ 1,200 = 10%
Market value of shares$1,300M
Net debt$500M
Earnings yield120 ÷ 1,800 = 6.7%

Northfield earns three times as much on each dollar it uses, and you pay less for it. The formula would put it well ahead. Sterling might still be a fine company, but you're paying more for a business that needs far more capital to produce the same profit.

Limits

What two numbers can't see

Both ratios use one year of EBIT. That's the formula's biggest blind spot. A steelmaker at the top of the cycle can show a huge return on capital and a low price at the same time, because the market already expects profits to fall.

The numbers also can't tell you whether the moat is getting wider or narrower. Greenblatt knew this. His point was that over a few hundred companies and several years, the averages work in your favour even though some individual picks will look silly.

Banks and utilities are out
Greenblatt excludes financial companies and utilities. Debt is raw material for a bank, and a utility's returns are set by regulators, so neither ratio means what it means for a toolmaker.
Remember

Takeaways

“
Key takeaway
  • Return on capital = EBIT ÷ (net working capital + net fixed assets). It measures how good the business is at turning capital into profit.
  • Earnings yield = EBIT ÷ enterprise value. It measures how cheap the whole company is, debt included.
  • Both use operating profit, so companies with different debt and taxes can be compared fairly.
  • One year of EBIT is the weak spot. Cyclical businesses at peak profits will look better than they are.
Sources

Sources for this lesson

  • Joel Greenblatt, The Little Book That Beats the Market (2005), chapters 6–8 and appendix
Next lesson · DL 364 · 2 of 5 ◆

Why the formula works

Buying good businesses at low prices sounds too simple to beat professionals. The reason it does has more to do with people than with arithmetic.
Continue
Educational disclaimer · This content is for educational and informational purposes only. It does not constitute investment advice, tax advice, legal advice, or a recommendation to buy or sell any security. Always conduct your own research before making investment decisions.