The Haystack and the Needle

“When stated in terms of lifetime dollar wealth creation, the best-performing four percent of listed companies explain the net gain for the entire U.S. stock market since 1926, as other stocks collectively matched Treasury bills.”

— Hendrik Bessembinder, Do Stocks Outperform Treasury Bills?

There is a comforting story investors tell themselves about exchange-traded funds: buy one, and you have diversified. You are no longer picking stocks; you are holding a basket, and baskets are safe. The story is half true — and the missing half is where wealth quietly goes to die.

The uncomfortable arithmetic

Start with the fact that unsettles everything downstream of it. Since 1926, more than half of all US common stocks have underperformed a one-month Treasury bill over their lifetimes. The market's entire net gain — every dollar of wealth the US stock market has created above cash — traces to the best-performing four percent of companies. The other ninety-six percent, taken together, matched T-bills. The distribution of long-run stock returns is not a bell curve. It is a lottery with a few enormous winners and a long tail of losers, and the average is carried entirely by the tail's opposite end.

This is not a US quirk. The global extension — 64,000 stocks across four decades — found the same shape everywhere: most stocks lagged T-bills, and the top 2.4% of firms accounted for all net global wealth creation. Germany is starker still: seventeen of some seven hundred listed companies carried the entire market. The skew is a law of markets, not an accident of one.

The lesson everyone draws — and the one they miss

The correct lesson is well known: own the haystack. If you cannot know in advance which four percent of stocks will be the needles, hold all of them — a broad, cap-weighted index — and you are guaranteed to own the winners along with everything else. This is the honest case for passive indexing, and Bessembinder's math is its strongest argument. Own the whole distribution and the skew works for you.

The lesson most investors miss is what happens the moment you stop owning the whole haystack. The instant a fund narrows the universe — picks stocks, tilts to a theme, restricts itself to a slice — it re-enters the lottery. And the skew that dooms most single stocks now dooms most funds.

Bessembinder came for the funds

This is not speculation. In Mutual Fund Performance at Long Horizons, Bessembinder ran the same lifetime-wealth lens over professionally managed US equity funds from 1991 to 2020 — managers who can buy any stock in the market, with research budgets and full-time analysts. The findings are a quiet massacre:

• Most underperform the market. Over two-thirds of US equity funds underperformed a basic S&P 500 index ETF over the thirty-year period.
• Time makes it worse, not better. Over a one-month horizon roughly 47% of funds beat the index — a coin flip. Over a thirty-year horizon, only 5.5% did. The longer you hold, the more certainly the skew asserts itself.
• Many lost to cash. Just like individual stocks, more than 20% of funds failed to beat one-month Treasury bills over the sample.
• A trillion dollars, gone. Bessembinder tabulated an aggregate wealth loss of roughly $1.02 trillion to fund investors over the thirty years, versus simply holding a broad S&P 500 index fund — the bill for fees and imperfect stock-picking, paid by the people who thought they were being prudent.

Read that fifth-and-a-half-percent number again. Full-time professionals, free to own the entire market, beat a plain index ETF only about one time in eighteen over a lifetime — precisely because holding the magic four percent of mega-winners, and keeping them through every drawdown, is nearly impossible to do on purpose.

Why this is the nail in the coffin for thematic ETFs

Now put the thematic ETF in that light. If a professional who can pick from the whole market has a 5.5% chance of beating the index over thirty years, what chance does a fund that deliberately restricts itself to a tiny universe have — just robotics, just cannabis, just clean energy, just one country's chip makers?

By construction, a thematic ETF makes a double bet: that the theme is right, and that the winners live inside it. If the best wealth-creating companies of the next decade turn out to sit in healthcare or financials or somewhere the theme never looked, the thematic fund is mathematically guaranteed to miss them. It has narrowed the haystack precisely where the needles may not be. It keeps all of the skew's downside — the long tail of losers — and forfeits the diversification that made the skew survivable. That is not diversification. It is a stock pick wearing a fund's clothing.

The industry's own record agrees. On Morningstar's accounting, only about 16% of global thematic funds both survived and beat global equities over the trailing five and ten years — a rate that fell toward 9% at fifteen years, by which point most had simply closed. Reported in ETF terms: roughly one in five thematic ETFs beat its broad benchmark over five years, and the average one lagged by 8.5 percentage points. Different data, same verdict as Bessembinder's: narrow the universe, inherit the skew, lose to the haystack.

The honest caveat

One qualification, because it separates a useful idea from a slogan: this is not “ETFs underperform.” A broad, cap-weighted passive ETF is built to track the haystack, not beat it — and tracking the haystack is the winning move Bessembinder's arithmetic argues for. The wealth destruction is specific to narrow, thematic and actively managed funds measured against a broad benchmark. The S&P 500 ETF in his study is not the villain; it is the yardstick that the trillion dollars was lost against. Own the index and you are on the right side of the math. Narrow it, and you are betting against it.

What this means at Closelooknet

This is the intellectual spine of the barbell we keep coming back to: own the broad haystack for the base of a book, and treat every narrow, thematic or single-name tilt as a deliberate, sized decision that has to clear a high bar — because the base rate is against it. Diversifying into a theme is not lowering risk; it is concentrating it while feeling safer.

It also reframes, honestly, what our own indices are. The Rubin, HALO, AW40 and agentic families are not passive haystacks; they are curated attempts to sit on the winners' side of the skew. Bessembinder's work is the reason we build them with explicit methodology and hold them with humility — as an investment diary of our own reasoning, never as a promise to beat the market or as advice to anyone else. The skew is exactly what makes that hard, and pretending otherwise would be the dishonest move.

And the skew is not abstract this week — it is live on the tape. As we lay out in the current Global edition, the “international rally” is an ex-US benchmark whose entire return is carried by two semiconductor economies, with twenty of twenty-six regions lagging the index they compose. That is Bessembinder's four-percent, drawn on a world map: a few winners making the number, the median destroying relative wealth. The return skew is the single most useful lens for reading it.

The discipline

The whole of it reduces to two of Bessembinder's own sentences. On the odds:

“We find that the percentage of bootstrapped mutual fund portfolios that outperform the SPY during the full thirty-year sample decreases from 47.5% at the monthly horizon to 5.5% at the 30-year horizon.”

And on the price of ignoring them:

“We tabulate an aggregate wealth loss of $1.02 trillion to mutual fund investors over our 30-year sample, when opportunity costs are based on beta-adjusted SPY returns.”

Own the haystack. Be very sure before you go looking for the needle — and be honest that looking is what you are doing.