How to find stocks to day trade: what to screen for, and what most lists hide

Most advice on finding stocks to day trade starts with a screener preset and a rule of thumb. The problem with presets is that they carry filters you did not choose and cannot see, and on any given day those filters remove most of what you were looking for. This guide is the opposite approach: start from the whole market, decide each filter yourself, and know what each one costs you. Every number below is measured, not assumed.

1. Start from the whole market, not from a list

On 2026-08-20 there were 5,077 US common stocks trading. A widely used Top Gainers preset showed 78 of them that day, and none of the 15 biggest movers were in it. Two corrections are mandatory before you rank anything, and skipping either produces a list that is mostly garbage. Adjust for splits. Daily bars from most sources are unadjusted. A 20-to-1 reverse split reads as a 1,900% gain. Without this correction the top of your gainers list is fake every single day. Drop anything that is not common stock. Leveraged ETFs move 10% on an ordinary day by design. If you do not exclude them, they crowd out the real movers permanently. Neither correction is optional and neither is difficult. They are just usually skipped.

2. The filters that actually matter

There are only three that do real work, and each has a cost you should know. Price floor. Excludes stocks below some level. The honest reason is not quality, it is mechanics: on a $2 stock the minimum tick of one cent is already 50 basis points, so the spread cannot get tight. A $5 floor is common. It is also the single filter that removes the most day-trading candidates. Liquidity floor. The one filter nobody should skip. It should be measured in dollars traded, not share count, because a million shares of a $0.10 stock is $100,000 and a million shares of a $50 stock is $50 million. Those are not the same market. Market cap floor. This is the filter that quietly does the most damage to a day trading screen, because the stocks that move 40% in a session are almost never large companies. On 2026-08-20 a stock closed at $30.00, up 44.5%, on $119.9 million of turnover — and it was missing from the preset list. It cleared any price filter. Size was the only thing excluding it. If you take one thing from this guide: **a market cap filter and a day trading screen are working against each other.** If a preset has one and does not tell you, its list is not the list you think.

3. What a real day of movers looks like

Filter that same day to stocks that rose 10% or more on at least $5 million of turnover — a liquidity floor low enough to include small names but high enough that a normal position is not moving the price by itself. That produced 27 stocks. Between them they traded $1.42 billion. That is the realistic candidate pool for one session. Not three, not three hundred. A couple of dozen, of which you will trade a handful. For reference, here is what our own scan uses, stated openly because that is the whole point of this guide: On 2026-08-20 those rules captured 23 of the 27. We missed four: three priced under $0.50, and one that traded fewer than 500,000 shares. 16 of the 23 were priced under $5 — the part a large-company preset cannot reach. We are telling you our filters and what they cost us. Ask the same of any list you use.

4. Your list will not stay still

This is the part that surprises people who build a watchlist at 09:30 and work from it all day. We recorded which symbols a 30-slot, change-ranked universe actually held at every instant across ten sessions. A median of 140 symbols passed through those 30 slots in a single session — a turnover of 4.7 times. The quietest session still cycled 3.6 times. Of the symbols present at 09:35: Half your 09:35 list is gone by 10:00. A morning watchlist is not a plan for the day, it is a snapshot with a short shelf life. Whatever you build, it needs to re-rank during the session, not once at the open.

5. Rank, do not just filter

Filtering tells you which stocks qualify. Ranking decides which few you actually watch, and the choice of ranking changes the list completely. By percent change gives you the biggest movers, which skews small and cheap. By dollar turnover gives you the most tradeable movers, which skews larger. Our own gainers list is ranked this way, and the effect is visible: on a recent session the cheapest name in our top 50 gainers was $8.64, even though the price floor is $0.50. Nothing filtered the cheap names out. They simply did not rank. By relative volume — today's volume against a normal day — finds stocks doing something unusual rather than something large. One warning: many sources compute a 10-day average that **includes today**, which drags the average up and understates the ratio badly on exactly the days you care about. Check before trusting it, or compute it yourself excluding the current session. Pick the ranking that matches what you trade. Just be aware that "top gainers" and "most tradeable" are different lists.

6. A checklist you can run

Whatever tool you use, these are the questions to answer before you trust its output. Does it adjust for splits? If a reverse split shows as a huge gain, everything downstream is wrong. Does it exclude ETFs? If leveraged ETFs appear in your gainers, it does not. Is there a market cap filter you did not set? This is the one that is usually hidden and does the most damage. Is there a price floor you did not set? Check the cheapest name in the output. If nothing is under $5, there is a floor whether or not it was disclosed. Is the liquidity floor in dollars or shares? Shares are close to meaningless across price bands. Does it re-rank during the session? If it runs once at the open, two thirds of it is stale by the close. You can answer every one of these in a couple of minutes by looking at the output rather than the documentation: sort by price and look at the minimum, sort by market cap and look at the minimum, and check whether anything in the list has moved since you loaded it. If a list cannot survive that inspection, do not build a strategy on top of it.