3. The expectations gap: where returns actually come from

Level 1: The price is a forecast

Here is the idea that reorganizes everything else, and that is hard to accept because it runs against instinct: your return is not produced by the company doing well. It is produced by the difference between what happens and what was already discounted in the price you paid. A magnificent company can ruin you and a mediocre one can make you rich, without either of them doing anything out of the ordinary.

Return comes from a difference, not from a level

If you buy at a price that discounts 15% annual growth and the company grows 15% a year, your return will be exactly the return you demanded when you bought, not one point more. You picked the right company and earned nothing extra, because there was nothing extra to earn: the market charged you up front for every bit of that growth.

So there are only three ways to earn a return above the one you demand:

• The company beats the assumptions built into the price.
• The market revises upwards what it expects beyond the horizon it was discounting.
• The required return falls, usually because interest rates or perceived risk come down.

And there is only one way to lose money on a company that is growing: for it to grow less than the price took for granted. None of this has anything to do with the quality of the business in the abstract. It has to do with the relationship between that quality and its price.

Two companies, two gaps, one counter-intuitive result

The numbers make it obvious. Take two companies, both earning €1 per share.

Company A is excellent. It trades at 50 times earnings, that is, €50, and that price discounts 25% annual growth over five years. The company almost delivers: it grows 20% a year. After five years its earnings are €2.49. Since the market has seen that its pace was slower than forecast, it now values it at 25 times: €62. Whoever bought at 50 earns around 4.5% a year.

Company B is dull. It trades at 10 times earnings, €10, and that price discounts no growth at all: the market takes it for granted that the business is stagnating. The company grows a modest 4% a year. After five years its earnings are €1.22, and the market, surprised that it did not collapse, values it at 13 times: €15.8. Whoever bought at 10 earns around 9.6% a year, more than double.

Company A grew five times faster than B and returned half as much. There is no trick in the example: there is a negative expectations gap in one case and a positive one in the other.

The expectations treadmill

There is a cruel corollary. Success raises the bar. When a company beats expectations, the market builds that improvement into the price, and from then on the new price demands better performance still. To go on delivering extraordinary returns it is not enough to be excellent: you have to be systematically more excellent than what is already discounted.

This explains a pattern that baffles amateurs: why the most admired companies of any given moment tend to return less than their quality suggests, and why the most despised sometimes return more. It is not a moral reward for pessimism. It is the arithmetic of the gap: where expectations are rock-bottom, the bar is easy to clear.

A good company and a good investment are not the same thing

It is worth saying plainly, because it is the most widespread confusion in the industry. The quality of a business is a property of the company. The return on an investment is a property of the relationship between that company and its price.

An investor who only learns to tell good businesses from bad ones has learned half the trade. He is missing the half that decides the outcome: whether that goodness was already paid for. Bruce Greenwald insisted that the margin of safety does not live in the model, but in the distance between what the price demands and what the business can sustain.

Real case: IBM versus Standard Oil, 1950-2003

Jeremy Siegel documented a comparison that sums up this entire module. Over 1950 to 2003 he set IBM, the emblematic technology company of the second half of the twentieth century, against Standard Oil of New Jersey, an oil company in a mature and unglamorous sector.

IBM won on every operating variable. Its annual growth in revenue per share was around 12% against roughly 8% for the oil company. Its growth in earnings per share was around 11% against something over 7%. Its growth in dividends per share was higher too. Had you been handed those figures in advance in 1950, you would have bought IBM without hesitating.

Total annualized return came to around 14.4% for Standard Oil and 13.8% for IBM. Over fifty-three years, that apparently trivial difference leaves the oil company's investor with roughly a third more money at the end.

The cause lies entirely in the starting price and in what that price was discounting. Over the period IBM traded at an average multiple in the region of 26 times earnings, more than double Standard Oil's roughly 13 times. And the oil company's dividend yield, around 5% against something over 2% for IBM, was reinvested year after year at low prices, buying ever more shares in a business the market persistently undervalued.

IBM met its expectations. Standard Oil beat its own, which were exceedingly modest. That is the whole secret, and nothing about it changes because we are talking about half a century rather than a quarter.