Swing Trading Stock Screener: Filter Settings for Pullbacks, Bases and Breakouts (2026 Guide)

Swing Trading Stock Screener: Filter Settings for Pullbacks, Bases and Breakouts (2026 Guide)

The base filters every swing screen needs, four setup screens with exact settings, ATR percent for sizing, and a worked scan that ends in one order.

A swing trading stock screener has one job: cut a market of several thousand stocks down to the ten or fifteen charts worth reading tonight. The screens that do it well share a base layer (liquidity, a price floor, a volatility band, no earnings inside your holding window) and then add three or four filters that describe one specific setup. This guide gives you that base layer, four setup screens with the settings to start from, a way to measure volatility that works across a $12 stock and a $400 one, and a worked scan that ends in a single order with a stop and a share count.

A screen does not find trades. It finds candidates, and most of them will be wrong when you open the chart. Treat every list as a reading assignment, and remember that a stock passing your filters tells you nothing about whether the next move is up.

What a swing screener is for

Swing trades are read on the daily chart and held for a few days to a few weeks, so the screen runs on finished daily candles. That changes what matters. Real-time data, the thing day traders pay for, barely matters here: you run the screen after the close, when the day's high, low and volume are final. The Finviz screener help page lists stock quotes as delayed by one minute, which is irrelevant to a scan you run at 7 p.m.

A wide funnel with hundreds of small stock charts entering the top and a handful of charts coming out the narrow bottom onto a trader's desk
The screener's job is the funnel: thousands of tickers in, a short reading list out, and the chart work starts after that.

What matters is how fast the screen gets you to a list you can read. A screen that returns 300 names has failed as badly as one that returns zero, because nobody reads 300 daily charts carefully in an evening. Aim for five to twenty results. If you get more, tighten one filter. If you get fewer for several nights running, the market may simply not be offering that setup, which is information too.

Swing screening also runs once, then ages. A day-trading scanner fires alerts all session and the list changes every minute. A swing screen gives you tonight's list, you set orders for tomorrow, and the list is stale by the next close. Our day trading scanner setup guide covers the intraday version, where relative volume and gaps drive everything.

The base layer every swing screen needs

Before any setup filter, the base layer removes stocks you should not swing trade at all. These filters stay the same across every screen you build. Save them once and start each new screen from them.

Price floor. Set a minimum share price, $10 being a common starting point. Below that, the universe fills with stocks whose percentage moves are large and whose spreads eat a real part of the expected move. Raise the floor if your account is large enough to buy round lots of higher-priced names. Lower it only if you understand what you are giving up.

Average volume. Finviz defines its Average Volume filter as the average daily shares traded over the recent three-month period. A floor of 500,000 shares is a reasonable start for a small account; a million is safer. The point is the exit. You need to be able to sell your position at a price close to the last trade on the day the setup fails, and that day usually has worse liquidity than the day you bought. Our explainer on what liquidity in trading actually means shows why a thin stock's chart can look clean right up until you try to leave.

Market cap. A floor around small-cap and above removes most of the shells and promotions. This overlaps with the volume floor, so treat it as a second check.

Earnings date. Finviz describes its Earnings Date filter as the company's nearest earnings-report date. Exclude stocks reporting inside your holding window. If you hold for up to ten trading days, exclude anything reporting in the next two weeks. An earnings report can move a stock further overnight than your whole planned trade, in either direction, and no chart setup predicts which.

A volatility band. Every setup screen works better when you remove stocks that barely move and stocks that move so much your stop has to sit miles away. The next section covers how to set that band properly, because the obvious tool for it, dollar ATR, gives misleading results across a screen.

Optional but useful: require the stock to be optionable. Optionable stocks skew toward names with enough interest that the chart reflects real participation. It is a filter, not a requirement.

Measuring volatility with ATR as a percent of price

Average true range (ATR) is the standard volatility measure for swing trading. StockCharts' ChartSchool credits J. Welles Wilder with developing ATR, introduced in his 1978 book New Concepts in Technical Trading Systems. True range is the largest of three numbers: today's high minus today's low, the distance from today's high to yesterday's close, and the distance from today's low to yesterday's close. Those last two capture gaps that a simple high-low range would miss. Finviz's ATR filter uses a 14-day average of true range.

The problem in a screener is that ATR is measured in dollars. A $400 stock with a $6 ATR and a $20 stock with a $1.20 ATR look like very different animals, but the first moves 1.5% on a normal day and the second moves 6%. ChartSchool makes this exact point: because ATR is absolute, lower-priced stocks show lower ATR values, so the raw numbers cannot be compared across securities. Its fix is to divide ATR by a price average and multiply by 100.

That percentage, ATR divided by price, is what your volatility band should use. A working band for most swing setups is roughly 2% to 5%:

  • Below about 2%, the stock moves so little that a normal swing target takes weeks, and costs eat a bigger share of the move.
  • Above about 5%, a stop placed below real structure sits far enough away that your share count gets small, and a single bad open can cost a multiple of your plan.

Finviz does not offer ATR percent as a ready filter, so you approximate it. One way is to use the Volatility filter (it measures average daily range as a percentage over the week or month) and pick the band nearest 2% to 5%. Another is to add an ATR column to your results view, export, and divide by price yourself. TradingView and thinkorswim both let you build the ratio directly in a custom column or scan formula.

ATR percent also feeds straight into position size. If your stop sits at 1.5 times ATR below entry and you risk a fixed $250 per trade, a more volatile stock gets fewer shares automatically:

Same 250 dollar risk, same 50 dollar stock: higher ATR percent means a wider stop and a smaller position.

The arithmetic: a 3% ATR on a $50 stock is $1.50, a 1.5 ATR stop is $2.25 away, and $250 divided by $2.25 is 111 shares. The risk management guide covers how to pick the dollar risk in the first place. The screener's role is to keep the list inside a band where that math produces a position you are willing to hold.

Screen 1: pullbacks inside an uptrend

The pullback is the most common swing setup and the easiest one to screen for, because each part of it maps to a filter. The full entry and exit rules are in our swing trading strategies guide; here is how to find the candidates.

Start from the base layer, then add:

  1. Price above the 50-day simple moving average. This is the trend filter. Finviz has separate filters for the 20, 50 and 200-day SMAs.
  2. 50-day SMA above the 200-day SMA. A second trend check that removes stocks bouncing inside a longer downtrend.
  3. Price near the 20-day SMA. Pick the option that puts price within a few percent above the 20-day, or just below it. This is where the pullback has come to rest.
  4. RSI(14) in the middle of its range, roughly 40 to 55. Finviz's help page describes RSI(14) as comparing upward and downward close-to-close moves to flag overbought and oversold. An uptrend's pullback rarely reaches the textbook oversold level of 30. If you demand 30, you will screen out almost every healthy pullback and catch trends that are breaking.
  5. One-week performance negative. This confirms the stock has actually pulled back in the last few sessions.

Relative volume below 1 is a useful extra. A pullback on below-average volume suggests profit-taking. A pullback on heavy volume suggests real selling, and that one deserves more caution.

What this screen cannot tell you is whether the pullback has finished. That is what the trigger is for: you set a buy stop above the high of the last pullback day and let the stock prove the turn. Many nights this screen returns names whose pullback continues another three days. The order never fills, and that is the screen working.

Screen 2: tight bases near highs

The base breakout catches stocks that have stopped going down after a run and are compressing just under resistance. The screen looks for that compression.

From the base layer, add:

  1. Price within 5% to 10% of its 52-week high. Finviz's High/Low filter covers this. Stocks near highs have no overhead supply from recent buyers waiting to get even.
  2. Price above the 50-day SMA. Same trend filter as before.
  3. Low weekly volatility. Use the Volatility filter set to the lowest band for the week. A tight base shows up as a narrow daily range.
  4. Relative volume below 1. Volume drying up inside the base is the classic sign that sellers have stepped back.

This screen often returns stocks that look flat and boring. Flat and boring is the target: you want a range narrow enough that a stop below the base's low sits close to a buy stop above its high. Draw both lines on every result. If the distance between them is wider than about 1.5 times ATR, the base is too loose for this setup.

Finviz's Pattern filter can help here: it tags charts with shapes such as horizontal channels and wedges. Treat those tags as a hint. An automated pattern label is a starting point for your own read of the chart, and you should check every tagged chart yourself. Our guide to price action trading covers how to read the base itself: where the real support is, and whether the highs are flat or drifting lower.

Screen 3: momentum bursts

The momentum burst is a short, sharp move, often off a news day, that swing traders join on the first or second day. This screen is the opposite of the base screen: it wants motion.

From the base layer, add:

  1. Change today above 4% (or the nearest option).
  2. Relative volume above 2. The move needs real participation behind it.
  3. Price above the 50-day SMA. A burst inside a downtrend is usually a short squeeze or a bounce, a different trade.
  4. Optional: a gap filter set to gap up. This isolates moves that started at the open.

Two cautions belong with this screen. First, it runs late. By the close, the burst is a day old and the easy part may be over. The entry plan is usually a small first-hour pullback or a break of today's high, with a stop under today's low or under the gap. Second, the earnings filter matters more here than anywhere else, because many bursts are earnings reactions. A stock that just reported is outside its next earnings window, which is fine, but check that the burst is not a single-day reaction already fading into the close. A long upper wick on the screened day is the warning.

Screen 4: support bounces and failed breakdowns

These two setups are harder to screen for, because what defines them is a level, and filters do not see levels. A screen can still narrow the field.

For support bounces in a range, start from the base layer and look for stocks with a flat 50-day SMA (price crossing it often), RSI(14) between 30 and 45, and one-month performance near zero. That combination tends to surface stocks moving sideways and sitting at the lower half of their range. Then you open each chart and find the support level by hand: the price where the stock turned up at least twice in the last few months.

For failed breakdowns, the screen is a recent new low followed by a reclaim. Some platforms let you express "made a 20-day low within the last three days and closed back above it". Finviz cannot express this directly. The closest approximation is price within a few percent of the 20-day low with a positive change today and relative volume above 1. Expect most results to be stocks that are still breaking down. The ones worth reading show a long lower wick through the prior low and a close back above it.

Both screens return more noise than the first three. That is normal. The setup is in the chart, and the screen only gets you to the right neighborhood.

Building the same screens in TradingView and thinkorswim

Finviz is the reference here because its filters are named plainly and its free tier covers everything above. The same logic transfers to other tools, with different strengths.

TradingView's screener sits next to the charts, so you can click a result and see the daily chart with your own drawings and indicators already on it. Its filters include moving-average relationships, RSI, volume, and performance. Our walkthrough of the TradingView stock scanner covers setting it up and when its Pine-based screener is worth the extra effort.

Thinkorswim's Stock Hacker, inside the Schwab platform, lets you write custom scan conditions in thinkScript. Custom scripts make ATR percent, "closed above the 20-day low after making it", and other conditions Finviz cannot express possible. The cost is a learning curve and a desktop platform.

Whichever tool you use, write the screen down in plain words first: "liquid stocks above $10, above a rising 50-day, resting at the 20-day, RSI 40 to 55, down this week, no earnings for two weeks". Then translate it. One Reddit user who built their own tool described the frustration well: going through more than 25 filter options, dropdowns, and tabs to find the right combination. A screen you can state in one sentence is easier to rebuild, easier to test, and harder to fiddle with after a bad week.

The earnings filter and overnight gaps

Swing trading means holding through nights, and nights are where the largest losses come from. The screener is your first defense, because the earnings filter removes the single most predictable gap: a scheduled report.

A stock chart with a wide empty gap between one evening's close and the next morning's open, a desk calendar with one date circled beside it
Earnings reports cause the biggest scheduled gaps, which is why the earnings filter sits in the base layer of every swing screen.

It cannot remove the unscheduled gaps: a guidance cut, a downgrade, a sector headline, a market-wide selloff. When a stock opens below your stop, the stop does not protect you at the stop price. The SEC's investor guidance states it plainly: when the stop price is reached, a stop order becomes a market order. After an overnight gap, the first price available is the opening price, wherever that is. A stop-limit order avoids filling at a bad price, but it may not fill at all, which leaves you holding a stock that is falling.

So the screen has a second job besides finding setups: it should keep your list inside a volatility band where a plausible gap is survivable. A stock with a 6% ATR can gap 10% on ordinary news. A stock with a 2.5% ATR usually gaps less, although no filter rules out a large gap on any stock. Size every position as if the stop might fill worse than planned. If a gap of twice your stop distance would hurt more than you can accept, the position is too big.

A worked example: from a full market to one order

Here is a full scan run on a hypothetical evening. The ticker and prices are illustrative, chosen to make the arithmetic clear; the process is the real one.

The screen. The pullback screen from above: price over $10, average volume over 500,000, price above the 50-day SMA, 50-day above 200-day, price within a few percent of the 20-day, RSI(14) 40 to 55, one-week performance negative, no earnings in the next two weeks. It returns 23 stocks.

The first pass. Open each daily chart for about a minute. Twelve go immediately: some have a 50-day that is flat even though price sits above it, some have pulled back on heavy volume, a few have a gap in the last week that left the chart messy. That leaves eleven.

The second pass. For each of the eleven, mark the pullback low and the high of the last pullback day, and check that the distance between them is no wider than about 1.5 times ATR. Five fail; their pullbacks are too deep for a tight stop. Six remain. Check the earnings date by hand on each, because a company can announce a date after the screener's data was last updated. Two now report within ten days. Four remain.

A trader at a kitchen table in the evening with a laptop showing a screener results table and a notepad listing three sketched chart setups
The evening scan: run the saved screen after the close, read every result, and leave with a short list and orders set for the morning.

The pick. One stock stands out: a $48 stock with a 14-day ATR of $1.40 (2.9% of price, inside the band). It ran from $41 to $53.10 over five weeks, then pulled back four days on falling volume to $46.20, right at its rising 20-day SMA. The last pullback day's high is $48.55.

The order.

  • Buy stop at $48.60, just above the last pullback day's high.
  • Stop loss at $46.05, just under the pullback low.
  • Risk per share: $48.60 minus $46.05 is $2.55, which is 1.8 times ATR. Slightly wide, still acceptable for this setup.
  • Fixed risk of $250 divided by $2.55 is 98 shares, a position of about $4,763.
  • First target at the prior high of $53.10: $4.50 above entry, or about 1.8 times the risk.

What can go wrong. If the stock never trades above $48.60, the order never fills and you lose nothing. If it fills and drifts down to $46.05 during the session, you lose about $250 plus commissions and slippage. If it fills and then opens at $44 two days later after a sector headline, the stop becomes a market order at the open: $48.60 minus $44 is $4.60 per share, times 98 shares, about $451. That is 1.8 times the planned risk on a stock with no earnings on the calendar. The plan holds up only if you can absorb that outcome.

Four candidates, one order. On a different night the same screen might produce zero orders. Some nights it should.

Saving, scheduling and keeping a screen honest

The Finviz help page notes that combinations of filters can be saved as a preset and opened from the screener menu later. Save each setup screen separately, and give each a name that says what it is for: "Pullback, liquid, no earnings 2w". The base layer goes into every one.

Run the screens at the same time each day, after the close and before you plan tomorrow's orders. Consistency matters more than the exact settings. Run the pullback screen some nights and the momentum screen on others, depending on mood, and you cannot tell which one works for you.

Keep a simple log: date, screen, number of results, how many you read, which ones you ordered, and how each trade ended. After a few months you will have real answers. Does the pullback screen produce more filled trades than the base screen? Do the results you skipped tend to work anyway, which would mean your chart reading is too strict? Without the log, you only remember the trades that hurt.

Change one filter at a time. Tighten RSI and add a volume filter in the same week, and if results improve you will not know which change mattered. Most screens need a small number of adjustments early and then very few.

Be honest about what a screen's history shows. A screen that found good candidates during a strong uptrend may produce almost nothing in a choppy or falling market. The screen is telling you the conditions changed. Forcing it to return results by loosening filters puts you in trades the setup was never meant for.

Common mistakes

  • Screening for oversold in an uptrend. RSI below 30 catches trends that are breaking. For pullbacks in a healthy uptrend, 40 to 55 is the usual zone.
  • Skipping the earnings filter. One report can undo weeks of careful trades. Check the date on the chart too, because screener data can lag a new announcement.
  • Using dollar ATR across the whole list. Divide by price. A $3 ATR is quiet on a $300 stock and violent on a $30 one.
  • Treating the result list as a buy list. Every name still needs a chart read, a trigger, a stop and a size. Most results fail one of those.
  • Loosening filters on slow nights. Few results usually means few setups. A loose screen fills the list with weaker charts.
  • Adding filters until the screen fits last month. A screen tuned to past winners keeps finding last month's market. Keep it simple enough to describe in one sentence.
  • Ignoring liquidity on the way out. A stock that trades thinly can be easy to buy and hard to sell on the day it breaks. The volume floor exists for the exit.

Where Quant AI fits

A screener and Quant AI do different halves of the same evening. The screener reads numbers across thousands of stocks and hands you a list. Quant AI reads one chart: take a screenshot of any result's daily chart, and it marks the trend, the support and resistance levels it finds, and the setup it sees, in a few seconds. That speeds up the first and second pass in the worked example, where most of the time goes.

It does not scan the market, it does not know a company's earnings date unless it shows on your chart, and it does not decide whether a setup is worth your money. The trigger, the stop, the size, and the decision to take a gap risk stay with you.

Frequently asked questions

What is the best stock screener for swing trading?

For most people, Finviz's free screener covers every filter in this guide, and it runs fine on end-of-day data. TradingView suits you if you already chart there and want results next to your drawings. Thinkorswim's Stock Hacker is the choice when you need custom conditions such as ATR percent or a reclaim of a recent low. Our comparison of the best free stock screener apps goes through the trade-offs.

What settings should I use on Finviz for swing trading?

Start with the base layer: price over $10, average volume over 500,000, a market cap floor, and no earnings inside your holding window. Then add the filters for one setup, for example price above the 50-day SMA, near the 20-day SMA, RSI(14) between 40 and 55, and one-week performance negative for a pullback. Save it as a preset and run it after each close.

How many stocks should a swing screen return?

Five to twenty is a workable range. More than that and you skim instead of reading; fewer for several nights in a row may mean the market is not offering that setup right now. Tighten or loosen one filter at a time to get there.

What RSI setting is best for a swing trading screener?

Keep the standard 14-period RSI and change the range you filter for. For pullbacks in an uptrend, 40 to 55 catches most healthy rests. For range bounces, 30 to 45. Screening for RSI under 30 mostly returns stocks in real trouble.

Should I add fundamental filters to a swing screen?

They can help as a quality floor, for example positive earnings growth, but a swing trade lasting a week mostly depends on the chart and the market. Fundamentals matter most around earnings, which the earnings filter already handles. If you add them, add one or two, and expect fewer results.

What are the top stocks for swing trading right now?

No fixed list stays useful. The point of a screener is that the answer changes every night as charts set up and fall apart. A list published by someone else was built for their screen, their timing and their risk, and is usually stale by the time you read it.

Can a screener tell me when to buy?

No. It tells you which charts meet your conditions as of the last close. The entry comes from the chart: a trigger price, set in advance, that the stock has to trade through. Many screened stocks never trigger, and that protects you from setups that did not finish forming.

Do I need real-time data for a swing screener?

Not for the screen itself. You run it on completed daily candles after the close. Real-time data helps during the session if you want to watch a trigger fill, but the screening work does not need it.