The mistakes made in your own head

You may have noticed that in six sessions you have never been shown a share price, a ticker or a chart. That is the anchoring defence built into the pages themselves. A price seen first becomes the centre your judgment orbits: a business priced high starts to look like it deserves it, one priced low starts to look like a bargain. Owner session 1 set the habit that answers this. Estimate value first, write it down as a range, and only then look at what anyone is asking.

Written is the key word. A range held in your head will quietly move once you hear a number, and you will not notice it moving. A range written on paper, with the date beside it, cannot move without you crossing something out. Halden's 0.82 to 1.58 million for a 10 percent stake was written in Owner session 2, before the cousin named 1.00 million in Owner session 3. That order was deliberate.

The neighbour's "about a million" is an obvious anchor because it arrives from outside. The harder ones come from your own history. Last year's price makes this year's feel high or low for no reason connected to the business. The price you paid is the most stubborn of all: it feels like a fact about the thing you own, yet the business has no idea what you paid for it, and your purchase price does not enter any future cash flow. A round number such as "a million" pulls estimates simply because it is easy to hold.

The test is the same for all three. Ask whether the number would change any line of the owner's paragraph or any input of the range. If it would not, it is an anchor, not evidence, and it belongs in a different column from your value.

If trained analysts, paid to forecast, drew their spread at a little over half the width of what happened, an owner working at a kitchen table should assume the same of themselves. The practical fix is simple to say and uncomfortable to do: push the bear case down until you would be genuinely surprised to see the business do worse, not just disappointed.

Run Halden's bear case through that test. It says owner earnings fall 3 percent a year for ten years, then hold flat, which gives 8.18 million for the whole at 10 percent (constructed). Is that the worst you can imagine, or just the worst that feels polite to say to your cousin? Two builders already order direct online. The Base Rate Book found 31 percent of large companies had lower net income three years later, and those were the biggest firms in the world, not a three-showroom family business. If a steeper fall would not shock you, the bear case is still a base case wearing a coat.

A useful check on your own calibration: keep the ranges you write, and later count how often the outcome fell outside them. If it never does, the ranges are probably too wide to be useful. If it happens far more often than you expected, they are too narrow, and the Base Rate Book suggests that is where most forecasters sit. A range is a claim about how often you will be surprised. A range that has never once been broken has not been tested.

The same book shows how often the "steady grower" story fails. Over three years, 31 percent of the companies in its sample had lower net income; over ten years, 24 percent did. Earnings growth, it concludes, is not very persistent, and sales growth is more persistent than earnings growth. The full set of base rates, and how to use them as the starting point of a forecast, is Analyst session 7, How often fast growth lasts.

Buffett's first mistake, by his own account, was buying control of Berkshire itself: a textile business bought because it was cheap. He gave the approach a name, cigar-butt investing: buying a poor business at a bargain price in the hope of one last small profit. The trouble is that the bargain is a one-time gain, while the poor economics keep working against you every year you hold it. The longer you own it, the more the business's weakness outweighs the discount you bought it at.

His other example was Hochschild Kohn, a department store bought below its book value that went on to return little. The lesson he drew is the one line quoted on this page: "Time is the friend of the wonderful business". Two more lessons sit beside it. Good managers cannot fix bad economics: put a talented team in a structurally poor business and the business usually wins. And it is easier to avoid problems than to solve them. That is exactly why Owner session 1 put the three business gates before the price gate. A business that fails them is not rescued by being cheap; it is only made cheaper.

The most surprising discovery Buffett reported in 1989 was a force he had not been taught about: the institutional imperative. Organisations resist changing direction, they find uses for whatever money is lying around, they build empires because size feels like success, and they copy whatever their peers are doing, however foolish. None of this shows up in a spreadsheet, and all of it decides how a business spends its owners' money.

For an owner, it changes two questions. About management: when the cousin proposes a fourth showroom, is it because the numbers say so, or because a rival just opened one? About yourself: are you looking at Halden because you understand it, or because people you know are buying stakes in local firms? The 1989 letter adds two short rules worth keeping beside the owner's paragraph. Never a good deal with a bad person: no contract fully protects you from a partner you cannot trust. And the costliest mistakes may be the ones you never see: mistakes of omission, passing on something that sat clearly inside your competence.