Key takeaways
- Across roughly 26,000 US stocks from 1926 to 2016, only 42.6 percent beat one month Treasury bills over their lifetimes.
- The best performing 4 percent of companies account for the entire net gain of the US market over that period. The rest, collectively, matched T bills.
- The median stock that eventually delisted lost about 92 percent.
- Picking a handful of names is a bet that you landed in the thin slice, and the base rate says otherwise.
The finding
Hendrik Bessembinder went through the entire CRSP database of US common stocks, about 26,000 of them, covering 1926 to 2016, and asked a plain question. Over its whole life as a listed company, did this stock beat a one month Treasury bill?
For most of them the answer was no. Slightly more than four in every seven failed to beat the safest asset available.
Where the market's gains actually came from
If most stocks lose to T bills, how does the market return roughly 10 percent a year? Because the distribution is wildly lopsided.
The best performing 4 percent of listed companies, around 1,092 firms, account for the entire net wealth creation of the US stock market since 1926. Everything else, taken together, merely matched Treasury bills. Narrow it further and about 90 companies, roughly one third of one percent, produced more than half the total.
The index does not work because the average stock is good. It works because it guarantees you own the handful that are.
What this means for a concentrated portfolio
Hold ten stocks and you are not holding a small version of the market. You are holding ten draws from a distribution where most outcomes are mediocre and nearly all the return sits in a tail you probably missed.
The median stock that eventually left the exchange lost about 92 percent on the way out. Companies do not usually collapse dramatically. They drift, underperform, get acquired cheaply, or quietly delist.
The honest counterargument
Skew cuts both ways. The same maths that makes concentration likely to disappoint is what makes it occasionally spectacular. Someone owned those 90 companies.
The question is not whether concentration can win. It is whether you have a reason to believe you are that person, and whether the rest of your plan survives if you are not. Size the bet accordingly and keep the core of the portfolio boring.
Sources
- Hendrik Bessembinder, Do Stocks Outperform Treasury Bills?, Journal of Financial Economics, volume 129 issue 3, 2018, pages 440 to 457.
- The 4 percent and 90 company figures are from the same paper. The widely quoted 86 stocks and $16 trillion version comes from the earlier working paper covering 1926 to 2016.
- Long run market return context follows the historical series maintained by Aswath Damodaran at NYU Stern.
Open Forecast. Run this math on your own numbers, with every assumption labeled.
Open in Capital Allocation →