Key takeaways · 6
- Coil is a 0-to-100 tightness score, where 100 is the tightest.
screen coil>90 | sort coilranks the whole universe by how compressed each name's recent daily range is and floats the tightest to the top. - On the 2026-07-08 close,
coil>90returned 153 names. Many are tight because they are dead: low relative strength, basing after a decline, going nowhere. Tightness is a precondition, not an edge on its own. - Adding one filter fixes it.
screen coil>90 rs>80 | sort coilkeeps only the coils that sit under strength, the handful like LPRO, RLYB, DOC, ROKU, and GBTG rather than the low-RS names that are quiet because nobody wants them. - Coil is the smaller of two readings. One is relative: today's 4-day average of true range over price, scored as an inverted percentile against the stock's prior 60 sessions. The other is absolute: a 2% daily range maps to 100 and 8% maps to 0. The minimum of the two means a name has to read tight both ways.
- The score needs about 65 bars of history to compute, so a freshly listed name reads blank rather than falsely tight. Coil is the absolute-range cousin of ADR and the numeric footprint of a VCP's successive contractions.
- Coil reports a condition, not a direction. A tightly wound stock can break either way, and it tends to resolve in the direction of the trend it sits inside, so pairing it with an RS or stage filter is what keeps the read useful. Every reading is end-of-day and point-in-time.
What "coiled tightest" means
To find the stocks that are wound tightest right now, rank the whole universe by how much each name's daily range has compressed and read the top of the list. In the tickerstance terminal that is one command, screen coil>90 | sort coil, and on the 2026-07-08 close it returned 153 names. The catch is that a raw tightness screen mixes two very different kinds of stock: the ones coiling under strength and the ones that are simply dead.
The idea behind coiling is old. Big, clean moves tend to come out of quiet. A stock that swings 6% a day rarely launches a tidy breakout from that noise; the launches come after the swings contract and the chart goes still. Traders picture a spring winding up: the tighter the range gets, the less room the stock has to keep doing nothing, and the more a decisive move has to resolve the tension. Minervini built a whole pattern around it, the VCP, where each pullback in a base is shallower than the last until the range is a coil.
The problem is that tightness is hard to eyeball across thousands of charts, and a chart that looks tight is not always tight in a way that matters. A sleepy small-cap that never moves looks just as quiet as a leader resting after a run. To sort one from the other you need the tightness as a number you can rank, and then a second filter for whether the tight name is worth watching at all.
Coil, made into a number
Coil is that number. It scores how tight a stock's recent range is on a 0-to-100 scale, where 100 is the tightest and 0 is loose. The raw ingredient is the 4-day average of true range divided by price, so the reading is a normalised range rather than a dollar figure. Dividing by price matters: it stops the score from reading a cheap, drifting stock as tight just because its dollar swings are small.
The score is the smaller of two legs. The first is relative: it takes today's normalised range and scores it as an inverted percentile against the same stock's prior 60 sessions, so 100 means the range is tighter than almost any recent day for that name. The second is absolute: a fixed map where a 2% average daily range or less scores 100 and 8% or more scores 0. Taking the minimum of the two forces a high coil to mean tight both ways, tight for this stock and tight on the tape. A wild name that has merely calmed down from its own chaos fails the absolute leg and never scores high.
The calculation needs about 65 trading days of history to run, so a stock fresh from its IPO reads blank rather than falsely tight. Everything after that is ranking: screen coil>90 keeps every name scoring above 90, and | sort coil orders them tightest first.
Coil is the minimum of a relative and an absolute tightness reading, so a name has to read tight both for itself and on the tape to score high.
The 153 tightest names
Type screen coil>90 | sort coil and the terminal filters the universe to the 153 names whose range has compressed into the top decile of tightness, then sorts them so the tightest sit at the top. Read the columns across: RS is the 0-to-99 relative-strength rank, ADR is the average daily range in percent, and the stage tag says whether the name is basing, trending, or breaking down.
The screenshot is where the lesson jumps out. The tightest rows are a mixed bag. AHRT sits at the top at RS 81 with an ADR of 1.9%, genuinely quiet. But right alongside it are ATGL at RS 11, CIG at RS 21, and BGSF at RS 28 and down 30% from its 52-week high. Those are tight, but they are tight the way a parked car is quiet. Further down the same list are LPRO at RS 98 in a Stage 2 uptrend, RLYB at RS 96, ROKU at RS 88, and GBTG at RS 94. Same tightness score, completely different setups.
screen coil>90 | sort coil on the 2026-07-08 close. 153 names cleared coil 90; the list mixes strong coils like LPRO and RLYB with low-RS names that are tight because they are dead.The trap: tight because dead
This is the thing to internalise about a pure tightness screen: coil measures how tight, not whether the tightness means anything. A stock can wind into a narrow range because buyers and sellers are in a standoff at the top of a strong base, or because it is a forgotten Stage 1 basing name that nobody is trading, or because it is deep in a Stage 4 decline and the selling has simply exhausted itself. All three read as high coil. Only the first is coiling in the sense a momentum trader means.
The ADR column is the quiet confirmation. Most of the low-RS names on the list carry a small average daily range not because they just contracted, but because they are always this quiet. Their coil is high in absolute terms and unremarkable in relative terms. Tightness, on its own, is a precondition for a clean move, not evidence that a move is coming. Screening for it and stopping there hands you a watchlist that is mostly noise.
Keeping the coils that sit under strength
The fix is one more predicate. screen coil>90 rs>80 | sort coil keeps only the tight names that also carry real relative strength, and the 153 names collapse to a handful worth charting: LPRO, RLYB, DOC, ROKU, GBTG and their neighbours. In terminal grammar the two conditions simply sit next to each other and both have to be true, so coil>90 rs>80 reads as tight and strong. Swap rs>80 for a stage or uptrend condition if you would rather define strength by trend structure than by rank.
What you are doing is layering the two questions coil cannot answer alone. Coil is the how-tight. RS or stage is the is-it-worth-watching. A coil that sits under strength is the setup the VCP describes: a leader that ran, then pulled into a tightening base, storing energy for the next leg. A coil with no strength under it is just a stock at rest. The same screen works inside a single group with stocks <group> coil>90, so you can ask which names in a rotating industry are winding up rather than scanning the whole tape.
Coil has a close relative in ADR, the average daily range. ADR is the absolute-range cousin, the raw percentage a stock travels on a normal day; coil turns that same normalised range into a ranked, two-legged tightness score. When a name's ADR keeps shrinking week over week, its coil climbs. That is the numeric version of the successive contractions you would otherwise look for by eye in a base.
A coil is a condition, not a direction
Coil is computed from the close and stamped with it; the terminal never claims to know the intraday range. A tight reading after Friday's close describes the range through Friday, not what happens when the stock opens Monday.
A high coil reports stored energy, not direction. A tightly wound stock can break either way, and plenty of tight bases resolve downward or just keep basing. Coil tells you where a move is more likely to be clean if it comes, not that it is coming. It tends to resolve in the direction of the trend the coil sits inside, so a coil under an uptrend and a coil under a downtrend are not the same read, and the RS or stage filter is what tells them apart.
And the terminal reports conditions and leaves the trade to you. "153 names cleared coil 90, and a handful of them also clear RS 80" is a fact about the tape on one date. Whether any of those names belongs in your book depends on your strategy, your timeframe, and your risk, and the terminal does not editorialise about it. Remember too that any name with fewer than about 65 bars of history reads blank rather than tight, so a thin screen right after a wave of IPOs is the data being honest, not the market going quiet.
Where the coil screen lives
The coil screen, the strength filter, and the per-group drill are all in the tickerstance terminal, where a typed line puts a question to the tape. The interactive terminal is a Pro feature at $28 a month, grandfathered, and everything it computes is end-of-day and point-in-time honest, with no "here is what to buy" layered on top. It reports where the tension is; the trade is yours.
Open the terminal and type screen coil>90 | sort coil to see tonight's tightest names, then add rs>80 to keep the coils that sit under strength. The tightest chart is only worth your attention when something strong is doing the winding.
Frequently asked questions
What does it mean for a stock to be coiled?
A coiled stock is one whose daily price range has contracted into a narrow band, so the chart goes quiet after a stretch of wider swings. Traders picture a spring winding up: the tighter the range, the less room the stock has to keep doing nothing, and the more a decisive move has to resolve the tension. Minervini's VCP formalises the idea as a base where each pullback is shallower than the last.
How is the coil score calculated?
Coil is a 0-to-100 score where 100 is the tightest. It starts from the 4-day average of true range divided by price, a normalised range rather than a dollar figure. The final score is the smaller of two legs: a relative leg that scores today's range as an inverted percentile against the stock's prior 60 sessions, and an absolute leg that maps a 2% average daily range to 100 and 8% to 0. Taking the minimum means a name has to read tight both for itself and on the tape. The calculation needs about 65 bars of history.
How do I screen for the tightest stocks?
In the tickerstance terminal, screen coil>90 | sort coil filters the universe to names scoring above 90 on the coil tightness score and orders them tightest first. On the 2026-07-08 close that returned 153 names. To narrow it to the tight names that are also strong, add a relative-strength filter: screen coil>90 rs>80 | sort coil.
Why are so many coiled stocks weak or low relative strength?
Because tightness measures how compressed a range is, not why. A stock can coil at the top of a strong base, or sit quiet in a forgotten Stage 1 base, or go still deep in a Stage 4 decline once the selling exhausts itself. All three read as high coil. On the 2026-07-08 screen, most of the 153 names were low-RS basing names that are tight because they are dead, not because they are winding up under strength.
How do I find coiled stocks that are actually strong?
Layer a strength filter onto the tightness screen. screen coil>90 rs>80 | sort coil keeps only names that are both tight and carrying real relative strength, which on the 2026-07-08 close left a handful like LPRO, RLYB, DOC, ROKU, and GBTG instead of 153. You can swap the RS filter for a stage or uptrend condition if you prefer to define strength by trend structure.
Is the coil score the same as ADR or a VCP?
They are related but not the same. ADR, the average daily range, is the raw percentage a stock travels on a normal day; coil turns that normalised range into a ranked, two-legged tightness score from 0 to 100. A VCP is a chart pattern of successive shrinking contractions inside a base, and a rising coil is the numeric footprint of that same contraction. Coil is the metric, VCP is the pattern, and ADR is the underlying range coil is built from.
Does a high coil score predict which way a stock will break?
No. Coil reports a condition, stored energy, not a direction. A tightly wound stock can break either way, and many tight bases resolve downward or keep basing. Coil tends to resolve in the direction of the trend it sits inside, so pairing it with a relative-strength or stage filter is what tells a coil under strength apart from a coil under a decline. It is end-of-day and never a trade instruction.