4. The mistakes that invalidate a reverse-engineered valuation

Level 3: Reverse valuation

A reverse-engineered valuation done badly is worse than none at all, because it produces a specific figure that inspires confidence without earning it. These five mistakes come up again and again, and all five have the same consequence: they make the price look as though it demands less than it does.

Mistake 1: confusing profit with money

Net income is an accounting opinion about a period; free cash flow is what is genuinely left after paying for the investment the business needs in order to keep running. You work it out starting from profit, adding back depreciation, subtracting investment in assets and subtracting whatever working capital takes.

The difference is not a nuance. A company whose annual investment comfortably exceeds its depreciation is burning cash just to stay upright, and its reported profit overstates what it hands out. Capitalizing profit instead of cash flow inflates the no-growth value and therefore artificially lowers what the price appears to demand.

Mistake 2: paying twice for share-based pay

When a company pays its team in shares, the cost does not show up in the cash: it shows up in your ownership percentage. Plenty of adjusted earnings figures exclude that expense, on the grounds that no money leaves the company. It is true that no money leaves, and it is irrelevant: ownership leaves, and the ownership is yours.

The mistake gets worse when the company buys back shares to offset the dilution and those buybacks are presented as returning money to shareholders. They are not: they are the deferred, cash payment of an expense that had already been excluded from profit. Counting them twice, first as clean profit and then as cash returned, is the commonest shortcut for making something look cheap when it is not.

Mistake 3: mixing market capitalization with enterprise value

Here the failure is one of plumbing, and it is fatal. The operating cash flow of the business belongs to everyone who funds it: shareholders and creditors. So if you are going to capitalize operating profit, the price you compare it against has to be the enterprise value: market capitalization plus financial debt minus cash.

• Using market capitalization with operating cash flows understates the demand in indebted companies, because you leave out a part of the price that somebody is paying.
• Subtracting cash without adding debt, or the other way round, produces figures that mean nothing at all.
• And a subtle case: in a company with a lot of cash, that cash generates no operating cash flow and should not have to be justified by growth. You separate it out, and solve the business on its own.

Mistake 4: extrapolating a peak margin

We saw this with the cyclical, but the mistake is not exclusive to cyclicals. Any company coming out of an exceptional period — a supply shortage, a competitor gone under, a favorable exchange rate, a fashion — shows a margin that is not its normal margin. Capitalizing it in perpetuity sends the no-growth value through the roof and makes the price look as though it demands nothing.

The correction is uncomfortable because it forces you to choose by hand what a normal year is, and there is no objective answer there. But choosing badly and saying so is infinitely better than not choosing: whoever uses last year's figure is choosing too, they just have not noticed.

Mistake 5: believing that precise means accurate

The last mistake is one of attitude. When the calculation finishes, out comes a number with decimals, and decimals are seductive. But that number inherits all the uncertainty of the assumptions: move the cost of capital by half a point, or the horizon by two years, and the result shifts noticeably.

An honest result is expressed as a range and as a threshold, not as a figure: 'the price requires growth of between 9% and 14% a year for about a decade, with the current margin intact'. That sentence can be tested. 'It requires 11.4% for 9.3 years' cannot be tested against anything, and it conveys a false authority that is exactly the opposite of what the method was after.

A worked case: the five mistakes in one company

A teaching example with rounded figures. A company reports 600 million of net income, has 500 million shares at 20 euros (a market capitalization of 10,000 million), 3,000 million of cash and 1,000 of debt. Its depreciation for the year is 250 million, its investment in assets 400 and working capital has consumed 50. Cost of capital, 8%.

The naive version: 600 million capitalized at 8% gives a no-growth value of 7,500 million. Against the 10,000 of market capitalization, only 2,500 million — 25% — is expectation. Sounds manageable.

The corrected version, mistake by mistake. Free cash flow is 600 plus 250 minus 400 minus 50, that is 400 million: the no-growth value falls to 5,000. The right price is not the market capitalization but the enterprise value: 10,000 plus 1,000 minus 3,000, that is 8,000 million. Now the expectation to be justified is 3,000 million out of 8,000: 37.5%, exactly the initial 25% multiplied by one and a half.

Dilution is still missing. If the company issues shares each year equivalent to 2% of the capital and buys back 1%, your holding gets multiplied by 0.99 ten times over: after a decade you own 90.4% of what you thought you owned. The demand per share rises by roughly 10% on top of everything else.

And the final layer: if the cost of capital were 8.5% instead of 8%, the no-growth value would drop from 5,000 to 4,700 million and the expectation to be justified would go from 37.5% to a little over 41%, almost four points more. That is why the result is stated as a threshold, not as a decimal.