Our numbers don't pick stocks that beat the market. Here's what they are for
We tested whether the numbers on our stock pages pick stocks that beat the S&P 500, from three months to ten years. They don't, so far. Our growth estimate is steadier than the one hidden in the price, but stocks the price rates "expensive" against it did not fall behind enough to matter, and no group beat the S&P 500. The pages are for understanding a business, not for timing a buy.
If you were buying the whole company, these numbers would matter a lot: its cash would be yours. If you are trading the shares, they will not tell you where the price goes. Use them to understand what you own and what its price assumes.
Every three months we sorted companies into five groups, from "the price expects less growth than we do" to "the price expects far more." Then we checked each group three years later.
How often each group beat the S&P 500 over three years
A coin flip is 50%. No group got there.
Typical price gain over three years, by group
The "expensive" group did a little worse, the "cheap" group did not do best, and over one year all five groups were within about a point of each other.
If the price expects too much growth, you might think the stock will fall later. Mostly it did not. Those companies grew their sales faster than the rest, and their price-to-sales ratio only shrank a little.
Three years later: sales growth (green) and change in price-to-sales (grey)
The gap closes mostly because the company grows into its price, not because the price drops.
Maybe the price catches up only slowly. So we held each group for five, seven and ten years, with dividends reinvested and companies that went bust counted as total losses.
Our value estimate, held five years: yearly gain of the typical stock in each group
"Cheapest" by our model did no better than "dearest". At seven and ten years it did worse.
The one thing that shows up: companies burning cash. Typical yearly gain, held ten years, by cash yield
The typical cash-burning company barely gained or lost ground, and only 15% of them beat the S&P 500. That is a warning about risk, not a way to pick winners: among companies that make cash, cheap and expensive ended up close.
A fair caveat: these years (2010–2026) were famously bad for "cheap" stocks and great for a few giant companies that carried the S&P 500. Older studies found cheap stocks did win over long periods before then. We cannot test those years with our data, so we do not claim it.
| Idea | What happened |
|---|---|
| Buy what our value estimate calls cheapest | Over a year it finished within 0.7 points of the "dearest" stocks, and came out ahead in only 44% of quarters |
| Bet against fast growers the price loves, on earnings day | They did not fall. On average they rose slightly |
| Find mispriced options | Any edge was eaten by the gap between buying and selling prices |
- Knowing what you would have to believe. If the price assumes 30% growth a year and companies like it have rarely done that, you know the bet you are making.
- Understanding the business: where the cash comes from, what it owes, and whether it is shrinking or growing.
- Our growth estimate is steadier than the one hidden in the price (chart below). That helps you judge the price; it does not tell you where the price will go.
Whose growth number was closer to what companies actually did
For the curious: how the growth number is built and checked
The recipe: 30% from the company's own revenue growth over the past five years, plus 70% from the typical rate for its industry, capped. Recent growth tends to fade toward normal faster than people expect; 30% was the most accurate weight on 2014–2018 and stayed so on later years.
The price's number is the yearly growth that makes today's price add up, assuming a 10% yearly return and a sale at 15 times cash flow in ten years. On average it was not too high or too low; it was scattered, because a price also carries mood, risk and interest rates.
The chances on the pages were built on earlier years and checked on later ones: said 15%, happened 16%; said 25%, happened 25%; said 50%, happened 50%.
Banks and insurers: our earnings estimate was only about as good as the price's, so their pages show no chance figure. REITs and utilities, on dividends, were clearly closer but with too few past cases to confirm the chances.
At market turns the growth edge held for companies in general but shrank to a coin flip for fast growers. Changing the recipe at turns did not help.
SAVNG is an educational research publication, not a Registered Investment Adviser. Everything here is general-circulation information produced by an automated pipeline — the same for every reader — and is not personalized investment advice, an offer, or a recommendation to buy or sell any security. Any performance figure is hypothetical; past performance does not predict future results. Details.
Published by SAVNG Research. Every figure on this page is read from the study data at render time; if the study changes, the page changes with it.