1. Sales growth: the driver that gets overestimated most

Level 2: The value drivers

Any valuation, however sophisticated it looks, hangs off four numbers: how much sales will grow, what margin that growth will leave, how much capital will have to go in to achieve it, and at what rate the whole thing is discounted. This level walks through them one at a time. And we start with the first because it is, by some distance, the one that has financed the most nonsense in the history of the markets.

The figure that gets written first and thought about least

Sales growth is the first box on any projection and the one most often filled in out of habit. You look at what the company did over the last three years, round it off, extend the line. The problem is not that the method is lazy: it is that it drags along a brutal and undeclared assumption, namely that what has happened is what will carry on happening.

The research on growth persistence points the other way. When you rank companies by their past growth and look at who is still in the leading group five or ten years later, the result looks a lot like what chance would produce. There are famous exceptions, and it is precisely because they are famous that they are remembered. What nobody remembers is the full census of companies that were growing just as fast and fell by the wayside.

Why high growth does not stay put

High rates reverting toward the mean is not bad luck. It is the predictable result of three forces acting at once:

• Competition. A business that grows fast and makes money is a neon sign. It attracts capital, imitators and better-funded competitors, and all of them eat from the same market.
• Saturation. Every market has a size. The first customers are the easy ones: the people who wanted the product most. The last ones are expensive to convince, and at some point there is nobody left to convince.
• The arithmetic of the base. The most ignored and the most implacable, because it does not depend on anybody's strategy.

That third one deserves a section of its own, because it is pure math and admits no argument about the quality of the management team.

The arithmetic of the base, with numbers

Think of a company turning over 100,000 million and assumed to grow at 15% a year for a decade. In the first year it has to add 15,000 million of new sales. Demanding, but plausible.

Now run the clock forward. At 15% a year, sales multiply by a little over four in ten years: the company would end up turning over around 404,000 million. And here is the trap. The increment in the tenth year, the jump between the ninth and the tenth financial year, is some 53,000 million. In other words: to hold the same percentage, the company has to add in a single year more than three and a half times what it was adding at the start, and a good deal more than an entire mid-sized company turns over.

When somebody defends 15% growth over ten years, what they are really claiming is that the company will create, year after year and at an accelerating pace, the equivalent of several large companies from scratch. Put that way, the assumption stops looking conservative.

How many years can genuinely be sustained

Rappaport and Mauboussin call this the competitive advantage period (CAP): the number of years a company can grow while generating returns above its cost of capital, before competition levels the ground. It is not something you observe, it is an assumption you choose, and plenty of valuations bury it inside a ten- or fifteen-year horizon without ever justifying it.

Damodaran insists on an elementary check that almost nobody runs: adding up. Take the ten largest companies in an index, accept the growth rates the market appears to be discounting for each one, and check whether that sum fits inside the expected growth of the economy they operate in. Very often it does not. The individual expectations are defensible one by one and collectively impossible, because one company's sales are, to a large extent, another company's spending.

Two points less: what it does to a valuation

This is where the assumption turns into money. Take a company whose free cash flow next year is 500 million, with a cost of capital of 8% and an assumed perpetual growth rate of 4%. The value is the cash flow divided by the difference between the two: 500 over 0.04, that is, 12,500 million.

Now cut the assumed growth from 4% to 2%. Not one factory has closed, not one customer has left. The value becomes 500 over 0.06, that is 8,333 million. More than 4,000 million has vanished, a third of the value, from moving by two points a number nobody can know.

And there is a detail worth committing to memory: the closer the cost of capital and the assumed growth rate are to each other, the more violent that sensitivity becomes. With a cost of capital of 10% instead of 8%, the same two-point cut would have cost 25% of the value rather than a third. The fragility is not spread evenly.

A real case: growing when you already turn over 650,000 million

Walmart is the textbook example. Through the eighties and nineties it grew at high double-digit rates for a long stretch, and anyone extrapolating that had all the historical evidence on their side. Today it turns over something like 650,000 million dollars a year.

Do the sum: growing 10% would require adding some 65,000 million of sales in twelve months, more than the vast majority of S&P 500 companies turn over in total. Its growth has been in the low single digits for two decades now, and not because its management has got worse: because the base changed scale.

This is education, not a recommendation on any security. The lesson is the arithmetic, not the company.