ROIC: a dollar goes in, how much comes back out?

The question it answers

Suppose two companies each earn 100 this year. The first needed 500 of capital to do it; the second needed 2,000. Both report the same profit, and headlines treat them the same way. They are not remotely the same business.

ROIC asks the only question that separates them: for every unit of capital tied up in this business, how much profit comes back each year?

That matters more than the profit figure because it decides what the next dollar is worth. The first company can take 500 of retained earnings and turn it into another 100 of profit. The second needs 2,000 to achieve the same thing. Over ten years of reinvestment those two paths do not converge — they separate violently.

How to compute it

The numerator is operating profit after tax — what the business earned from operating, before the financing decisions. The denominator is the capital actually put to work: equity plus interest-bearing debt, less cash the business does not need to run. Cash is removed because idle cash earns nothing and would otherwise drag the ratio down for a reason that has nothing to do with operations.

You will find three or four slightly different definitions in circulation — some use average capital over the year, some year-end, some add back operating leases. All of them are defensible and none of them will change your conclusion. A business at 22% is not going to be reclassified as a business at 7% by a definitional choice.

Why the bar is 10%

Ten percent is not arbitrary and it is not a target. It is roughly the cost of the capital the business is using. A company earning 10% on capital is running in place: it is paying its providers of capital exactly what they require and creating nothing extra.

• • Below 10%, growth is actively harmful. Every dollar reinvested returns less than it cost, so a bigger company is a poorer one.
• • Around 10%, growth is neutral. Reinvestment neither creates nor destroys value; the business is a savings account with extra steps.
• • Well above 10%, retained earnings are the cheapest capital in the world and management should keep them.

This also reframes the dividend question. A business compounding at 20% that pays out its earnings is handing you capital to redeploy at whatever you can find, probably less than 20%. A business at 6% that retains everything is doing you an active disservice. The payout policy only makes sense relative to ROIC.

Three ways a high ROIC lies

Because ROIC is a ratio, it can be improved by shrinking the denominator rather than growing the numerator. All three traps below are versions of that.

1. Fully depreciated assets. A plant bought twenty years ago sits on the books near zero, so the capital base looks tiny and the ratio looks superb — right up to the year the plant must be replaced at today’s prices. Cross-check: compare capital expenditure against depreciation over ten years. If capex has run persistently below depreciation, the asset base is being consumed, and the flattering ROIC is a loan from the future.
2. Debt and buybacks shrinking equity. Borrow, buy back shares, and the denominator falls while operating profit is unchanged. The ratio rises and the business has not improved by one unit. Cross-check: look at ROIC and the debt load on the same screen, never separately.
3. Timing in the denominator. Capital deployed in December earns for one month but sits in year-end capital in full, so a year of heavy investment prints a depressed ROIC and the following year prints an inflated one. Cross-check: this is exactly why you read ten years. A single year’s figure has a timing artefact baked in that no definition fixes.