Warren Buffett: The Approach Changed
Warren Buffett's approach is to buy whole businesses or shares in them at a price below what they are worth, and hold for a very long time. He began buying statistically cheap companies and shifted to buying high-quality ones at fair prices, which is the most instructive part of the record.
How it works
The central idea is ownership rather than trading. A share is a fraction of a business, and the question is what the business will earn over decades rather than what the price does next.
There were two distinct phases. The early one bought companies trading below the value of their assets; the later one bought companies with durable advantages, at reasonable rather than bargain prices.
The shift is the most instructive thing in the record. Statistically cheap companies are frequently cheap for reasons that persist, and the later approach accepts a higher price for a business that keeps earning. A method that changed once is a method somebody tested, which is more than most published approaches can claim.
The filters
The first filter is durability. Something that stops competitors taking the profit away — a brand, a cost advantage, a switching cost, a network. Without it, a good business attracts competition until it is an ordinary one.
The second is a boundary around what he understands. Declining an opportunity because it sits outside that boundary is a decision rather than a failure, and it is the discipline most easily copied by anybody.
And the holding period is the part that gets left out. Positions held for decades through declines that would have ended most people’s conviction — which is a behavioural achievement rather than an analytical one.
What cannot be copied
Berkshire’s insurance businesses provide capital to invest before claims are paid. That is a structural advantage unavailable to an individual, and it is a genuine part of the record rather than an incidental detail.
His stated advice for people who are not doing this professionally is a low-cost index fund. Which is worth taking seriously precisely because it comes from somebody who did the other thing successfully.
In practice
Nothing in the method uses a chart. Volume, patterns and indicators play no part; the inputs are financial statements and judgements about competition.
The compounding is unremarkable annually and extraordinary cumulatively. No year in the record looks like the thing people imagine when they hear the name.
A gap down is an opportunity in this framework. The business did not change; the price did, and the whole approach depends on treating those as separate facts.
No stop appears anywhere. The exit condition is the business deteriorating or the original reasoning being wrong, and neither is a price level.
Very low turnover removes almost all trading cost. Each avoided round trip saves 2% of a median bar’s range on this site’s shared history, and over decades that accumulates into a real part of the result.
What is actually transferable
Three things transfer to an ordinary investor and one does not. The two filters — durability and understanding — transfer completely and cost nothing. The long holding period transfers if you can manage it. The insurance float does not transfer at all, and any comparison that ignores it is incomplete.
The most honest reading of the record is that it combines a good method with a structural advantage and an unusual temperament. Copying the method without the other two produces something different, which is worth knowing before treating the results as a target rather than as an illustration.
What the approach is not
It is not stock picking by intuition. It is reading accounts.
It is not buying anything cheap. That was the earlier version.
It is not replicable in full. The float is not available to you.
And it is not what he recommends to most people.
When it fails
A flat decade is the case the approach handles best, because the return comes from the businesses rather than from the market rerating them — and it is also the decade in which almost nobody sticks with it.
The second failure is copying the holdings without the reasoning. A position bought because somebody else owns it has no exit condition, because you never had an entry condition.
A third is applying it to companies you cannot assess. The boundary is the method’s central discipline and the easiest one to ignore.
A fourth is expecting the returns without the size. The early record was made on a small amount of capital in a less efficient market.
And a fifth is treating patience as passivity. The holding period is the result of conviction that was built beforehand, not of not looking.
The original data
Of the 31,760 trading and investing videos in this site’s corpus, 77 have “warren buffett” in the title at
a median of 71,546 views across 54 channels, with a maximum of 4,306,456. “Charlie munger” returns 12 at a
median of 307,798, “peter lynch” returns 13 at a median of 10,131, and “howard marks” returns 9 at a median
of 178,341. The counts are in research/corpus-coverage.json, produced by site/measure_corpus.py.
Seventy-seven videos at a median of 71,546 views makes this one of the highest-demand names in the whole corpus, and the answer to that final question is the one his own record actually changed on. The early version said no; the later version said that a fine business at a fair price beats a fair business at a fine price. Reading the shift is worth more than reading either half of it, because it is the only documented case of somebody successful publicly revising their own method.
Related
Value investing is the framework the approach sits inside. Intrinsic value is what the estimate is of. And Charlie Munger is the partner who argued for the change.
What is worth taking from him is not stock picking, it is the two filters. Would I be happy owning this for ten years, and can I explain what it does in a sentence. Almost everything I have got badly wrong failed one of those before I bought it.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.