How to Screen for Stocks Worth Trading

By Silviu Vasilescu · Updated 1 August 2026 · 9 min read

There are thousands of listed stocks and you can follow perhaps thirty properly. Screening is how you get from the first number to the second — not by finding the "best" stocks, which nobody can do reliably, but by discarding the ones your method cannot trade well and keeping a manageable list of the ones it can.

The purpose is elimination, not discovery. That reframing solves the most common screening problem, which is a screen returning four hundred results that nobody looks at. A screen you cannot act on is the same as no screen.

1. Three layers, in order

A screen that works has three stages, and the order matters because each one is cheaper than the next.

Tradeability comes first: can you get in and out of this stock at a sensible price? This is a permanent filter, set once, and it should remove the large majority of the market before you think about charts at all. Condition comes second: is the stock in a state your method wants — trending, strong relative to its peers, in the right kind of sector? This changes slowly, over weeks. Setup comes last: is there something to do right now? This changes daily and is the only layer that needs looking at often.

Getting the order wrong is what produces exhaustion. Scanning for setups across the whole market every evening means most of what you find will fail the tradeability test anyway, and you will have spent the evening finding out.

2. Layer one: can you trade it at all?

Liquidity is the first and least negotiable filter. Use average dollar volume — average daily share volume multiplied by price — rather than share count, because a million shares of a $2 stock and a million shares of a $200 stock are not comparable positions. Set the floor so your intended position is a small fraction of what trades in a normal day. If you need a meaningful share of the day's volume to get filled, your own order will move the price against you both on the way in and, worse, on the way out.

Illiquidity is expensive in a way that is easy to overlook because it does not appear as a commission. It shows up as a wide bid-ask spread you pay on every entry and exit, and as slippage when a stop is triggered and there is nobody there. A method with a +0.3R edge can be turned negative by trading a universe where the spread costs 0.4R.

Price floor. Very low-priced stocks tend to have proportionally wider spreads and more erratic behaviour. A minimum price is a crude filter, but it removes a category of stock where technical analysis works poorly because the participants are different.

Structural exclusions. Decide once whether you will trade around scheduled earnings dates, and whether you will hold leveraged or inverse exchange-traded products at all. Both introduce behaviour your chart-based rules do not model — a gap through your stop in the first case, and decay that has nothing to do with the chart in the second.

3. Volatility: enough movement to pay for the risk

Volatility is not the same as risk, and screening it out entirely is a mistake. A stock that moves 0.4% a day cannot produce a 3R winner in a two-week hold, because there is not enough movement available. Your target and your stop both have to fit inside what the stock actually does.

Average true range as a percentage of price is the practical measure — a normalised figure you can compare across stocks of any price. Set a floor so there is enough range for your targets, and a ceiling so stops do not have to be so wide that your position size becomes trivial.

The right band depends entirely on your holding period. A trader holding for weeks needs less daily range than one holding for hours, because the move accumulates over more days. This is one of several places where screening criteria follow from your system's timeframe rather than from a universal rule.

4. Layer two: relative strength and trend

Relative strength means comparing a stock's performance with the market's over the same period. A stock up 8% while the index is up 2% is behaving differently from one up 8% while the index is up 10%, even though the absolute return is identical. The first is attracting money; the second is being carried.

This matters because relative strength has some persistence — the same trend-following effect, applied to comparison rather than to price alone. It is a weak effect, as all of them are, but it is one of the more durable ones and it is cheap to screen for: rank your universe by return over three, six and twelve months, and look at the top decile.

Combine it with a simple trend filter. Price above a rising long-term moving average is the standard version and it is adequate — its value is not precision but consistency, keeping you out of the falling stocks that look cheap and continue to fall.

If you trade mean reversion rather than trend, this layer inverts: you want stocks that are stretched away from their average, in names whose long-term structure is intact. The principle is unchanged — the condition layer selects for the state your edge needs — but the criteria are the opposite ones, which is worth stating because most screening advice silently assumes trend-following.

What a screen is for: putting a stock on your list while it is quiet, so the decision is already prepared when it moves.

5. Sector and market context

Stocks move in groups. Money rotates between sectors, and a stock in a strong sector has a tailwind that has nothing to do with the company — which is useful when you are with it and expensive when you are not. Checking which sectors are leading takes a minute and improves the quality of a watchlist more than most refinements to the individual filters.

The broad market matters in the same way. Breakouts follow through more often when the index is advancing, and fail more often when it is not, because the same background flows that lift the index lift its components. Some systems formalise this as a market filter — take long setups only while the index is above its own long-term average — which reduces the number of trades and tends to remove a disproportionate share of the losing ones.

Sector strength also warns you about a risk your position sizing will otherwise miss. If your screen keeps returning the same industry, your watchlist is concentrated, and taking five of those setups is closer to one large position than five small ones — the correlation problem.

6. Layer three: the setup itself

Only now do you look for something to do. On a list of thirty or forty stocks that have already passed the first two layers, this is quick, and it is the part that can be done by eye without becoming a nightly ordeal.

What you are looking for depends on your method — a tightening base, a pullback to a rising average, a pattern nearing completion. What matters is that the criteria are written down, so that "it looks good" is never the reason a stock advances to your watchlist.

Two mechanical filters earn their place at this stage. A relative volume flag surfaces stocks trading well above their own average, which is where the news and the institutional interest is. And a range contraction filter — recent range narrow compared with its own history — finds stocks coiling before a move, which is exactly the condition that precedes a tradeable expansion.

7. A weekly routine that fits in an hour

Screening fails from being too laborious to sustain far more often than from being too simple. A workable rhythm:

• Once, at the start: define the tradeability filters — liquidity floor, price floor, volatility band, exclusions. Save the screen. Do not revisit it monthly.
• Weekly, 30 minutes: run the condition layer. Rank by relative strength, apply the trend filter, glance at sector leadership. Produce a watchlist of twenty to forty names. Remove the ones that no longer qualify.
• Daily, 10 minutes: look only at the watchlist for setups approaching a trigger. Write the trigger, stop and size for tomorrow's candidates before the market opens, so the decision is made when you are calm.
• Monthly, 20 minutes: check whether the watchlist actually produced your trades. If most of your positions came from somewhere else, the screen is decoration and either it or your behaviour needs to change.

That last check is the one people skip, and it is diagnostic. A screen that does not feed your actual trading is telling you something true about your process.

8. Screening mistakes worth avoiding

Too many criteria. Each filter you add removes candidates, and past a certain point you are selecting for the peculiar rather than the good. A screen returning three results has usually been over-specified rather than finely tuned.

Changing the screen after a losing trade. The screen did not cause the loss; losses are the cost of doing business. Adjusting filters in response to individual outcomes is curve fitting performed one trade at a time.

Screening for what already happened. A filter for "up 40% in a month" finds moves that are over. Useful screens find conditions that precede moves — contraction, strength relative to peers, a base forming — not the moves themselves.

Trusting the data without looking. Screeners carry stale figures, mishandle corporate actions and report volume oddly around splits. Before trading a name you have never traded, look at the chart and the spread yourself. Two minutes of checking prevents the occasional expensive surprise.

This guide is educational material about how markets and trading methods work. It is not financial advice and not a recommendation to buy or sell any security. Trading involves the risk of losing money, including more than you intended if you use leverage.