Should You Buy That Stock? Five Questions Before You Do
Of every US stock since 1926, 57.4% never beat a Treasury bill. Five questions a banker runs before buying a single stock — and the first is not about the stock.
Should You Buy That Stock? 4% of Companies Made All the Gains (Explained by a Banker)
Prefer to watch? This article is the written companion to the video above.
The one your colleague will not stop talking about. The one that is already up 200%. The one a stranger online is certain about.
Before your thumb reaches the buy button, one fact, and then five questions. The fact comes from a finance professor at Arizona State University, Hendrik Bessembinder, who went through every US common stock since 1926, all 25,967 of them, and asked something that sounds too simple to need asking: did the stock beat a Treasury bill?
Most of them did not.
The fact that sets the odds
Of every common stock that appeared in the CRSP database between July 1926 and December 2016, only 42.6% produced a lifetime buy-and-hold return above one-month Treasury bills. The other 57.4%, slightly more than four out of every seven, did worse than the most boring government paper there is, over their entire listed lives. Fewer than half, 49.5%, even had a positive lifetime return. The median stock lost 2.29%. The single most common lifetime outcome, rounded to the nearest five percent, was a loss of essentially everything.
And yet the market as a whole beat Treasury bills handsomely over those same ninety years. Both things are true at once. How they can both be true is the whole point of this article.
Every US common stock, 1926 to 2016 · nearly $35 trillion of net wealth creation
Four percent of companies made all of it
Companies ranked from best to worst, left to right. The line reaches the ceiling before it has crossed five percent of the chart — and then runs flat, because the other 96% of companies, added together, earned no more than a Treasury bill.
This is why "the market goes up" and "my stock will go up" are different sentences. The market's gain is real. It was produced by a sliver of companies, and a portfolio that holds everything owns that sliver automatically. A single stock has to be one of them.
The roughly 25,300 companies that listed shares in the United States since 1926 created nearly $35 trillion of net shareholder wealth, measured as of December 2016 — market value above what the same money would have earned in Treasury bills. Five of them account for a tenth of that. Ninety of them, a third of one percent, account for more than half. The best-performing 1,092 companies, slightly more than 4% of the total, account for all of it, and the remaining 96%, added together, matched Treasury bills.
So the market’s gain is real, and it came from a sliver. A broad index fund owns that sliver automatically, because it owns everything. A single stock has to be one of them, and the base rate says most are not. The professionals do not manage it either: the SPIVA scorecard has 79% of active US large-cap funds trailing the S&P 500 in 2025, and roughly 92% trailing over twenty years.
None of this says never buy a single stock. I own some. It says that buying one company is not “investing in the market”. It is looking for a needle, on purpose, and the only sane way to do that is with money you can afford to lose, for a reason you can say out loud, in a size that cannot hurt you. Five questions cover those three conditions, and they are asked in a particular order.
Question one: could you lose all of it?
This is the question that is not about the stock, and it is where most buys end.
Everything worth having in a financial life sits on a structure — four layers, if you use the version I use. A buffer of cash for the week that goes wrong. A floor of defensive assets for the year that goes wrong. An engine, the boring diversified machine that funds the life you actually want to live. Single stocks do not belong in any of those three. They belong in the fourth layer, the free hand: the small slice you could lose entirely without anything that matters changing.
So before anything else, are you buying this with free-hand money, or are you reaching into the engine? If it is the engine, stop. The answer is already no, and nothing in the next four questions can turn it into a yes.
I know you have heard that a hundred times. I mean it in the literal sense. If you cannot afford to lose this exact money, do not touch single stocks, not this one and not any. There is no shame in that. The buffer, the floor and the engine get funded first, and the free hand is whatever remains after they are, which for most people, for a good while, is nothing at all.
Question two: can you say why, in one plain sentence?
Not “it is going to explode.” Not “AI is the future.” A reason: “This company earns a little more every year from something its customers cannot easily walk away from.”
If you cannot finish the sentence “I am buying this because…” in words a twelve-year-old would follow, you do not have a thesis. You have a feeling wearing a suit. A feeling can be right, of course, and plenty of them have been.
Question three: what would prove you wrong?
What a feeling cannot do is tell you when it has stopped being right. A reason can, if you give it a second half.
Every real reason can break. So name the break now, in calm weather, while nobody is shouting. “I am wrong if their main product stops growing.” “I am wrong if a rival takes their best customers.” Write it down, one line, next to the reason.
That sentence is doing two jobs. It keeps you honest today, and it is the only reliable instruction you will ever have about when to sell: the day the reason breaks. Not the day the price falls. A share can drop 30% with the reason fully intact, and hit a record with the reason quietly dead. The portfolio-level rules for selling are the previous episode; the long version for a single company, with kill criteria, thesis creep and an autopsy of how slowly a broken thesis announces itself, is its own piece. Both start from the sentence you write here.
If nothing could ever prove you wrong, you are not an investor. You are a fan, and fans hold all the way down.
Question four: what do you know that the price does not?
This is the one almost nobody asks, and it is the uncomfortable one.
That company everybody is excited about — the excitement is already in the price. Thousands of professionals with better data than you have already bought and sold it; that is what set the number on your screen. So a great company is not the same thing as a great investment. A wonderful business at too high a price is a bad buy. An ordinary business, cheap enough, can be a good one. How “cheap enough” is actually measured, and why the P/E ratio is the wrong tool for it, is its own article.
You do not need inside information. You do need an honest answer to one question: why would this price be wrong, when so many people who do this for a living have already had their say? “It is a good company” is not an answer. Everyone already knows it is a good company, which is precisely why it is expensive.
Question five: zero tomorrow — your life, or your week?
Go back to the fact at the top. With a single stock, going to zero is not a freak accident; it is the most common outcome in the data. Position size is therefore not a detail to settle after the decision. It is most of the decision.
One position goes to zero · the rest of the portfolio earns an illustrative 7% a year
Lose a position worth half your portfolio, and the rest needs ten years to catch up
Time until the portfolio is back where it started, by the size of the position that was lost.
The first two rows cost you months. The last one costs you a decade. Nothing about the company changed between them — only how much of your money you let it hold. That is the whole of question five.
The rule is the same one that runs through stop-loss placement and portfolio governance on this site: small enough that a total loss is annoying, never devastating. A common institutional version is to risk no more than one or two percent of the portfolio on any single position, and to keep any single holding below about five percent unless it is the core index fund. If this stock going to zero would wreck your year, the position is too big, whatever the story.
How it works in real life
Your colleague will not stop talking about a stock. Run it.
Run the tip through it · in order · most buys stop at the first row
The five-question filter
Could you lose all of it?
This money sits in the free hand — not the buffer, not the floor, not the engine.
Stop here. Nothing that follows applies to money you cannot afford to lose.
Can you say why, in one plain sentence?
"I am buying this because…" — finished in words a twelve-year-old would follow, with no hype in it.
You have a feeling wearing a suit, not a thesis.
What would prove you wrong?
"I am wrong if…" — one concrete event, written down now, in calm weather.
If nothing could prove you wrong, you are a fan. Fans hold all the way down.
What do you know that the price does not?
Why would this price be wrong, given that professionals with better data have already had their say?
"It is a good company" is not an answer. Everyone knows. That is why it is expensive.
Zero tomorrow — your life, or your week?
Sized so that a total loss is annoying, never devastating.
Too big, whatever the story. Shrink it or skip it.
Five honest yeses: buy it calmly, and enjoy it. One honest no, anywhere in the list, is a finished decision — not a missed opportunity.
Is this free-hand money? If it is your rent or your retirement, done, do not. Can you say why in one plain sentence, and what would prove you wrong — and would you actually sell then? Then the two that are about the price and the size: what do you know that the market does not, and is it small enough to lose completely and shrug?
Five honest yeses: buy it calmly, and enjoy it. One honest no, anywhere in the list, and you have just saved yourself from a decision you would have made with your thumb instead of your head.
The series, in one idea
That is the last of eight decisions, and there was one idea underneath all of them: the calm choice was never the complicated one. It was the one you had already thought through before the pressure arrived.
The eight were how much cash to hold, whether to pay off debt or invest, whether to rent or buy, whether to invest all at once or over time, how much risk to carry, how many funds to own, when to sell, and now whether to buy at all. Not one of them needed a hot take. They needed a framework and a few honest minutes before the moment, and you have those now.
One precision, because a knowledgeable reader will notice it: the video calls the man behind the opening fact “a banker”. He is not. He is a professor of finance, and the paper is his. The series voice took a liberty there; the article does not.
Keep going
Finished reading? Three places to go next
- Start Money Decisions from the beginningHow Big Should Your Emergency Fund Really Be? Forget 'three to six months of expenses.' Size your emergency fund on your fixed costs, not total spending — times how fragile your income is. Here's your number.
- Read nextWhy Are Interest Rates So High in 2026? What It Means for You The 10-year Treasury hit its highest close since 2002 in October 2026. What's behind it, and what it means for your mortgage, savings and bonds.
- On YouTubeSubscribe on YouTube (opens YouTube in a new tab)New videos every Tuesday: one money question per episode, checked against the data.
Primary sources (4)
- 01Do Stocks Outperform Treasury Bills? — Journal of Financial Economics, Vol. 129, No. 3 (2018), pp. 440–457. Figures cited here from the author's May 2018 version: 25,967 CRSP common stocks, July 1926 to December 2016; 42.6% with lifetime buy-and-hold returns above one-month Treasury bills, 57.4% below; 49.5% with a positive lifetime return; median lifetime return −2.29%; nearly $35 trillion of net wealth creation, of which five firms account for 10%, the top 90 for over half and the top 1,092 (slightly more than 4%) for all; median listed life 90 months; only 47.8% of monthly stock returns exceed the one-month Treasury return — Hendrik Bessembinder, W. P. Carey School of Business, Arizona State University
- 02Do Stocks Outperform Treasury Bills? — SSRN abstract page for the same paper — SSRN, abstract 2900447
- 03Long-Term Shareholder Returns: Evidence from 64,000 Global Stocks — the global follow-up: 55.2% of US and 57.4% of non-US stocks underperformed one-month US Treasury bills from 1991 to 2020, and the top 2.4% of firms accounted for all $75.7 trillion of net global wealth creation — Bessembinder, Chen, Choi and Wei, Financial Analysts Journal, 2023, as summarised on the Arizona State University research page
- 04SPIVA U.S. Scorecard, year-end 2025 — 79% of active large-cap US equity funds underperformed the S&P 500 in 2025; over twenty years roughly 92% of domestic US funds underperformed their benchmarks — S&P Dow Jones Indices
Questions people actually ask
Should I buy this stock?
Run it through five questions, in order, and stop at the first honest no. One: is this money you could lose completely — does it sit in the free hand, outside your emergency buffer, your defensive floor and your long-term engine? Two: can you say why you are buying it in one plain sentence, with no hype? Three: what specific event would prove you wrong, written down now? Four: what do you know that the price does not already reflect, given that professionals have already traded on everything public? Five: if it went to zero tomorrow, would it change your life or just your week? Five honest yeses mean buy calmly. One honest no is a finished decision, not a missed opportunity.
What percentage of stocks beat Treasury bills?
Fewer than half. Hendrik Bessembinder's study 'Do Stocks Outperform Treasury Bills?' (Journal of Financial Economics, 2018) examined all 25,967 US common stocks in the CRSP database from July 1926 to December 2016. Only 42.6% had lifetime buy-and-hold returns above one-month Treasury bills; 57.4% — slightly more than four out of every seven — did not. Only 49.5% had a positive lifetime return at all, and the median lifetime return was −2.29%. The stock market as a whole beat Treasury bills comfortably; the average stock did not.
How many stocks account for the stock market's gains?
About four percent. In Bessembinder's 2018 study, the roughly 25,300 companies listed in the US since 1926 created nearly $35 trillion of net shareholder wealth — market value above what the same money would have earned in one-month Treasury bills — measured as of December 2016. The 1,092 best-performing companies, slightly more than 4% of the total, account for all of it. Just five firms (Exxon Mobil, Apple, Microsoft, General Electric and IBM) account for 10%, and the top 90 companies for over half. The remaining 96% of companies, added together, matched Treasury bills.
Is it worth buying individual stocks at all?
It can be, provided you know which game you are playing. Owning the whole market through a broad index fund captures the small group of companies that produce the market's gains automatically; owning a single stock requires that company to be one of them, and the base rate says most are not. That does not make single stocks a mistake — it makes them a bet, which belongs only in the part of your money set aside for bets, sized so that a total loss is annoying rather than devastating, and bought for a reason you can write in one sentence.
What is an investment thesis and how do I write one?
An investment thesis is the specific reason you own a company, stated in one plain sentence that a twelve-year-old could follow — for example, 'this company earns a little more every year from something its customers cannot easily walk away from.' 'It is going to explode' or 'AI is the future' are not theses; they are feelings. A real thesis has a second half: the concrete event that would prove it wrong. Writing both down before you buy is what later tells you when to sell, because the day the reason breaks is the day the position has lost its job.
How much of my portfolio should be in one stock?
Small enough that a total loss would be annoying, not devastating. A common institutional rule of thumb is to risk no more than 1–2% of the portfolio on any single position, and to keep any single holding below about 5% except the core index fund. The arithmetic is the reason: if the rest of the portfolio earns an illustrative 7% a year, a 2% position that goes to zero is earned back in about four months, a 10% position in about a year and a half, and a 50% position in about ten years. Position size decides the damage; the company merely decides whether it happens.
Why is a good company not always a good investment?
Because the price already contains the good news. Thousands of professionals with better data have already bought and sold the company's shares, and the number on your screen is the result of that. A wonderful business at too high a price is a poor buy; an ordinary business, cheap enough, can be a good one. The question is therefore never 'is this a good company' — everyone knows it is, which is exactly why it is expensive — but 'what do I know that the price does not already reflect', and 'it is a good company' is not an answer to that.
Is a stock that has already risen 200% too late to buy?
The rise itself tells you nothing either way. What matters is whether the reasons for the rise are already fully priced, which is question four of the filter: what do you know that the price does not? If your only information is that the price went up and that other people are excited, you know nothing the market does not, and you are buying the excitement. If you can state a specific reason the price is still wrong, and what would prove that reason false, the past return is irrelevant — the position is judged on the five questions like any other.