Zero earnings, growing business: how to see what a company earns before it reinvests

Some of the best businesses report almost no profit for years. They aren't failing. They're spending nearly every dollar they make on growth: new products, new customers, new capacity. The earnings line shows $0, and every tool built on earnings (P/E, PEG, EPS growth) stops working.

So what do you measure instead? You need a base unit: one number that keeps counting the business's real earning power, whatever the company chooses to spend on growth this year.

The idea: split spending into "keep" and "grow"

Every dollar a company spends does one of two jobs:

• Keep: what it costs to serve today's customers with today's products. Run the servers, pay support, replace worn-out equipment, keep existing customers from leaving.
• Grow: what it spends to be bigger next year. Build new products, win new customers, add capacity.

Under US accounting rules, most "grow" spending counts as a cost of this year. Research and marketing are expensed right away, even when they pay off for a decade. So a company that grows hard shows low profit by choice.

The base unit is what is left after the "keep" spending only:

Earnings power per share = (gross profit − keep spending), after tax, ÷ diluted shares.

It answers one question: if this company stopped trying to grow tomorrow and just served the customers it has, what would each share earn?

Two rules keep it honest:

1. Stock pay is a real cost. Accounting already counts it as an expense. Never add it back.
2. Per share, always. A company that grows earnings power 20% but issues 20% more shares has grown nothing for you.

One worked example: two companies that both report $0

Two made-up companies start in the same place. Year 1: revenue $100, gross profit $70, keep spending $20, grow spending $50. Reported operating profit: $0. Both have 10 million shares, a 25% tax rate, and investors want a 10% return.

On the reported numbers, these are the same company three times in a row: $0, $0, $0. On earnings power per share, A is compounding at about 25% a year and B has gone slightly backwards. Same story ("we're reinvesting everything"), opposite results.

Three numbers built on the base unit

1. What the business is worth if it never grows again. Divide earnings power by the return investors want: $37.5 ÷ 10% = $375, or $37.50 a share in year 1. This is an old idea, Bruce Greenwald's earnings power value. If the share price is $60, then $37.50 pays for the business as it is today and $22.50, or 37.5% of the price, is a bet on growth. That split tells you how much of the price your growth story has to earn.

2. What each reinvested dollar turned into. Over years 1 and 2 each company spent $116 on growth ($50, then $66), or $87 after the tax saving. By year 3:

• Company A's no-growth value rose from $375 to $615. Each $1 it reinvested became about $2.76 of business value.
• Company B's rose from $375 to $390. Each $1 became about 17 cents.

Above $1, reinvesting creates value, and you want the company to spend more on growth. Below $1, every growth dollar destroys some of your money, however good the revenue growth looks.

3. Where the growth came from. Break earnings power per share growth into three parts: revenue per share, the keep-cost margin, and share count. A company growing only by issuing shares, or only by cutting the keep spending it will need later, shows it here.

Why not just use free cash flow growth?

Free cash flow is harder to dress up than earnings, because it counts cash rather than accounting choices. But it subtracts the same growth spending. In the example, with capital spending equal to depreciation, A and B both show free cash flow of about $0 in years 1 and 3. Free cash flow growth can't tell them apart either, and a growth rate that starts from $0 or from a loss tells you nothing.

Free cash flow has blind spots of its own:

• Stock pay is left out, because it's paid in shares rather than cash. It is still pay.
• Equipment paid for through leases can sit outside capital spending.
• Cash timing can be moved, for example by paying suppliers later.

So start from cash, then make the same split. Subtract stock pay and lease payments, and add back the growth spending after the tax it would owe. That gives the cash version of earnings power per share. In the example it lands on the same numbers: $3.75 → $5.86 for A, and $3.75 → $3.71 for B.

Standard tools that point the same way

None of these ideas are new on their own. Investors have used them for decades:

Earnings power per share puts these together in one per-share series you can chart over ten years.

Where it misleads

• The keep/grow split is an estimate. Companies don't report it. Research is the easiest to split (capitalize it and amortize it over its useful life). Marketing is the hardest: you need to know how much of it only replaces customers who left. When you're unsure, count the spending as "keep". That keeps the number conservative.
• "Grow" spending only counts if it works. Company B shows the trap: calling a cost an investment doesn't make it one. The value-per-reinvested-dollar check is the test.
• Gross margin sets the ceiling. If gross margin is thin, there is little left over even when growth spending stops.
• It's a measure, not a price target. Use it to see whether the business underneath is growing, then use a valuation model to decide what that is worth.

The takeaway

When earnings say $0, don't stop at the earnings line. Separate what the company spends to stay where it is from what it spends to get bigger. Then count what's left, per share, every year. If that number compounds, and each reinvested dollar turns into more than a dollar, the business is growing even when the income statement says it isn't.

This article is education, not investment advice. Companies A and B are fictional and their numbers are made up to show how the method works.