Top gainers by percent or by dollar volume: the same day, two lists that share 9 names out of 50

Every platform has a "Top Gainers" list. Almost none of them tell you what "top" means, and it turns out to matter more than any other choice in the screen. We built both versions from the same universe on the same days, changing nothing except the sort order. Across 40 sessions the two top-50 lists shared a median of 9 names. Two random draws from the same pool would have shared 7.

1. What we built

For each session we assembled one qualifying universe: US common stock, split-adjusted, ETFs and ETNs removed, priced between $0.50 and $100, up 3% or more on the day, with at least $1 million of dollar turnover. A median of 414 stocks qualified per session. Then we ranked that same set two ways and took the top 50 of each: List A — by percent change. The biggest movers. List B — by dollar turnover. The most heavily traded of the movers. Both are defensible readings of "top gainers". Both are in wide use. The base universe, the date, the filters and the corrections are identical — the only difference between A and B is the column they are sorted on. Split adjustment and the ETF exclusion are not optional. Without the first, a reverse split reads as a huge gain; without the second, leveraged ETFs crowd both lists. Skipping either produces rankings that are wrong before the sort even happens.

2. They share nine names out of fifty

Eighty percent of each list is absent from the other. On the day they diverged most, the two lists a trader would call "today's top gainers" had 2 names in common.

3. That is barely better than chance

Nine out of fifty sounds low, but low compared to what? The honest test is what you would get by drawing two random samples of 50 from the same pool. The pool size changes daily, so we computed the expectation for each session from that session's own pool and averaged it: two independent draws of 50 share 7.0 names. We observed 9. So the ranked lists agree about 1.29 times as much as coin-flipping would. On **7 of the 40 sessions** the overlap was at or below the random expectation outright. That is the finding. These are not two views of the same phenomenon with different emphasis. As sets of stocks, they are close to unrelated.

4. They contain different kinds of stock

The reason the overlap is so low shows up immediately in what each list holds. Ranking by percent selects small, cheap and violent. Ranking by turnover selects large, expensive and liquid. The typical name in list B trades 16.6 times the dollar volume of the typical name in list A, and rose about a third as much. Neither is a distortion. They are answers to different questions: *what moved most* and *where the money went*. The problem is that both get printed under the same heading.

5. Which one your screener is giving you

You can work it out in about a minute, without documentation. Sort the output by price and look at the cheapest name. If almost nothing is under $5, you are looking at a turnover-ranked or size-filtered list. Look at the median gain. Around 5% suggests turnover ranking. Around 15% or more suggests percent ranking. Look at the turnover column, if there is one. A list where the typical name trades hundreds of millions of dollars is not the biggest-movers list. We have written before about a widely used preset that turns out to [exclude almost every real mover](/research/yahoo-finance-top-gainers-missed-all-15-of-the-day-s-biggest-movers-we) through undisclosed size filters. This is the same class of problem one level down: even with no hidden filter at all, the sort order alone decides most of what you see.

6. Which one you want

That depends on what you are going to do, and it is worth deciding deliberately rather than inheriting it. If your position is small relative to the book, ranking by percent surfaces the names that actually moved, and the liquidity floor is what keeps them tradeable. If your position is large, turnover ranking finds names that can absorb it. A 40% move you cannot get filled in is not an opportunity. If you use both, use both lists. They are cheap to compute from one universe, and running only one means never seeing four fifths of the other. Our own daily scan publishes the turnover-ranked version, which is a choice we made for a specific reason and one you should not simply adopt. The filters behind it are written up in how to find stocks to day trade.

7. What we did not measure

Which list makes money. We measured composition, not returns. Whether percent-ranked or turnover-ranked candidates perform better is a strategy question and nothing here answers it. Other rankings. Relative volume would be a third list, and we did not build it. There is no reason to think it lands near either of these. Our filter choices. The $1M turnover floor and the $0.50 to $100 price band shape both lists. A different floor changes the composition of A far more than B, because A is the list living near the boundary. Regime. Forty sessions in one stretch of 2026. The size of the gap should be expected to move with how much small-cap activity there is in a given period.