Four Percent Of Stocks Made All The Money. Here Is What That Means For Yours.

Start with the part the how-the-market-works primers get right, because it’s genuinely worth an evening: a share of stock is a fractional deed to a real business — its factories, brands, cash flows and debts. The exchange is a continuous auction where anyone can sell that deed to anyone else, and the quoted price is simply the last trade’s argument between one seller and one buyer — not a verdict, not an appraisal, just the most recent handshake, updated thousands of times a minute.

Two things follow from that mechanical fact, and they run the rest of this letter. First, in the short run the price measures the mood of the argument, not the health of the business; a company can grow earnings for a decade while its stock does nothing, because the argument started from too-high a number. Second, over the long run the argument loses and the business wins: a century of stock returns is, almost entirely, a century of dividends and earnings growth, with the mood contributing noise on the way through. Owning stocks is owning businesses. Everything else is commentary with a data feed.

Which raises the question the primers stop short of: if the return comes from the businesses, which businesses supplied it? The intuitive answer — “most of them, more or less” — turns out to be spectacularly wrong, and the true answer, published by finance professor Hendrik Bessembinder in one of the most quietly devastating papers in the field, is the subject of section two.

In 2018, Bessembinder did something nobody had bothered to do: he measured the lifetime return of every U.S. stock that ever traded — roughly 26,000 companies since 1926 — against the humblest alternative on earth, the one-month Treasury bill. The headline finding rearranges how you should think about everything on the stock-picking shelf: just 4% of companies account for the entire net wealth the stock market ever created above T-bills. The other 96% of companies, taken together across nine decades, collectively did no better than parking the money with the government.

Now the uncomfortable syllogism. If nearly all the wealth comes from a tiny, rotating cast of extreme winners, then picking stocks isn’t a game of being right on average — it’s a game of holding the specific handful of peaks, for decades, through drawdowns that repeatedly looked terminal. Miss them, or sell them at double, and it doesn’t matter how sensible the rest of the portfolio was. This is why the professional record looks the way it does: across fifteen-year windows, roughly nine in ten U.S. large-cap funds trail the plain index — not because the managers are foolish, but because the winners are too few, too extreme, and too identifiable only in hindsight.

And it’s why “101 ways to pick winners” is a shelf, not a science. Each heuristic — low P/E, high growth, insider buying, chart patterns — is a trail that once led somewhere, photographed after the fact. The record doesn’t say the trails never work. It says nobody has reliably known which trail, in advance, for ninety years running — while one boring vehicle held every summit the whole time by refusing to choose.

So what does a household actually do with the concentration finding? Three things, in order of importance.

First: make broad ownership the default, and make it cheap. The core of a long-term portfolio should own the market — whole, boring, and at index-fund cost — because that’s the only structure that can’t miss the next generation of peaks. Every dollar of unnecessary fee is a dollar of guaranteed loss deducted from an uncertain return; the hedge fund letter ran that arithmetic to $2.1 million over a career, and it applies at every scale.

Second: stop grading yourself stock by stock. Under concentration math, a diversified portfolio will always contain positions that look foolish in isolation — that’s what owning the 96% alongside the 4% feels like from the inside. The unit of success is the whole portfolio against the whole plan, over years. Any process that reviews individual holdings weekly will eventually sell a future peak at double to fund something that felt smarter, which is the single most expensive transaction in retail investing and completely invisible on any statement.

Third: give conviction a container instead of a veto. The urge to pick is human, occasionally right, and — unmanaged — ruinous, because it expresses itself as concentration precisely when confidence is highest. Our answer is structural, like everything else in this letter: the core owns the coast; a separate, size-capped tactical book holds the named bets, every position with a written reason, every quarter graded in public on our Marked to Market page. When the bets win, the household wins a little more. When they lose, the plan doesn’t notice.