ROIC Against the Cost of Capital: The Value Test

Scope. This article explains the comparison at the centre of corporate finance — what a company earns on its capital against what that capital costs — and what the comparison does and does not establish. A positive spread is a statement about a company's economics. It is not a statement about whether a security is attractively priced, and this article says so more than once because the distinction is the one most often lost. No threshold is given, no company is characterised as creating or destroying value as a verdict, and nothing here is valued.

That test is the reason accounting profit is insufficient on its own. Net income is struck after the cost of debt — interest is an expense — but never after the cost of equity, because equity has no contractual charge. A company can therefore report positive net income every year while earning less than the capital costs, and the income statement will never say so. The value test exists to put the missing charge back in.

The two halves

Return on invested capital is after-tax operating profit over the capital invested to produce it. Using Wexford Instruments and the definitions fixed there: operating profit of 100.0 less tax at the 24.0% effective rate gives NOPAT of 76.0; invested capital is total debt plus total equity, averaged across the year — 850.0 opening and 862.3 closing, so 856.2. The return is 8.9%.

The cost of capital is the blended rate built in the previous article: 7.88%.

The comparison must be like for like, and getting this wrong is the commonest error in the whole subject. Return on invested capital is a return on debt and equity together, so it has to be compared to the weighted average cost of both. Comparing it to the cost of equity alone — 9.0% here — sets a hurdle that is too high by more than a point, because it charges the whole capital base at the price of its most expensive component. On that mistaken comparison this company appears to fall just short; on the correct one it clears by a point. The two answers are on opposite sides of zero, produced by the same figures. The mirror error is comparing return on equity to the cost of capital, which flatters in the other direction. The rule: match the return's capital base to the cost's capital base, every time.

What the spread is worth in currency, and how little it takes to erase it

Economic profit is the spread multiplied by the capital it was earned on — here 1.00pp of 856.2, or 8.6. Set against net income of 62.3, the company's economic profit is 13.8% of its accounting profit. Most of what it earned went to paying for the capital it used.

Now the part that matters more than the result. The cost of capital carries a range, and the spread inherits it.

Worked example — the answer is thinner than the uncertainty around it. Economic profit ranges from 6.2 to 16.1 across defensible constructions of the cost of capital — a factor of more than two and a half — and turns negative on a beta assumption that is one of the portal's own three reference points. Nothing about the business changed in any of those rows. A one-percentage-point error in the cost of capital erases the entire result, because the entire result is one percentage point. That is the honest description of this company: it earns a little more than its capital costs, by a margin narrower than the measurement error in the estimate of what the capital costs. The response is not to abandon the test. It is to treat the sign as provisional, the magnitude as approximate, and persistence across many years as the only version of the question that a single year cannot answer — which is exactly the position the article on economic moats arrives at from the other direction.

What the spread does not tell you

This is the compliance-critical section of the pillar and it is stated bluntly.

A positive spread describes what a company earns on its capital. It says nothing about whether its shares are attractively priced. The two are different questions with different inputs, and running them together is how an educational measure becomes an investment conclusion without anyone noticing the step.

The reason they come apart is that the market already knows. A company visibly earning above its cost of capital is priced by people who can also compute this, so the expectation of continued excess returns is already in the share price. A wide spread on a company priced for a wider one and a narrow spread on a company priced for none produce opposite outcomes for a holder, and the spread by itself distinguishes neither case.

Two further limits. It is historic. The measure describes capital already deployed and a return already earned; what matters for the future is the return on the next capital deployed, which is not disclosed. And it is accounting-dependent. Invested capital moves with how acquisitions were structured, whether development spending was capitalised, and how long assets have been depreciating — an older asset base flatters the return by shrinking the denominator, and nothing about that company got better.