Key takeaways · 6
- Float sets the size of the amplifier; it says nothing about direction. Weinstein stage is the field that supplies the direction, and
screen float<40m stage=2 rs>80asks for both at once. - Weinstein
stageis a numeric field (aliasesws,weinstein) with four values: 1 Basing, 2 Uptrend, 3 Topping, 4 Decline.stage=2keeps only names the terminal reads as being in a confirmed uptrend on the last close. - A breakout flag does not check stage before it fires. On the 2026-07-09
bo float<25mboard, 22 of 50 rows were sitting in Stage 4 decline while only 14 were in Stage 2. Requiringstage=2up front refuses those Stage 4 pops before you ever read them. - Float is a share count (alias
flt), never a percentage;rsis the relative-strength percentile; there is no price or close field anywhere in the terminal. Space-separated predicates AND together, so all three conditions apply at once. - Stage is a classification computed off end-of-day data, and it can change. A Stage 2 tag tonight describes the trend as of the last close. It does not promise the trend holds, and it says nothing about whether the entry is good.
- Thin float plus Stage 2 is the amplifier already pointed up. It is not a signal. Pair it with a
dollarvolfloor and read the base underneath before treating any name as more than a condition on tonight's screen.
Float sets the size, stage sets the direction
The float-and-supply hub makes one point over and over: a thin float amplifies a stock's response to demand, but the amplifier is direction-agnostic. The same lack of shares that produces a fast markup on buying produces a fast markdown on selling. Float tells you how big the move will be relative to the order flow. It does not tell you which way the stock is currently going.
That second question has its own field. Weinstein stage classifies where a name sits in the trend cycle, and stage=2 is the value for a confirmed uptrend. Put the two together and you get a specific, narrow condition: thin supply, and a trend that is already pointed up. screen float<40m stage=2 rs>80 is the command, and it answers a question a raw float screen cannot: not just how loud can this name get, but is it already moving in the direction you would want the amplifier pointed.
What Stage 2 actually means
Stan Weinstein's stage model splits a chart into four phases, and the terminal carries them as a single numeric field, stage, aliased ws or weinstein. Stage 1 is basing, a stock going sideways after a decline, building a floor. Stage 2 is the confirmed uptrend, price advancing above a rising moving average. Stage 3 is topping, the uptrend stalling into a range. Stage 4 is decline, price falling below a falling average. stage=2 filters to the second of those four, the phase where the trend is up and confirmed rather than merely forming.
The number is a label the terminal computes from the last close, not a forecast. A name reads Stage 2 tonight because its recent price and moving-average structure fit the definition of a confirmed uptrend on this close. That can change: a Stage 2 name can roll into Stage 3 over a few weeks, or break straight to Stage 4 on a bad session. The tag describes the trend that exists right now. It is honest about the present and silent about the future, the same as every other reading in the terminal.
Running `screen float<40m stage=2 rs>80`
screen filters the whole universe, not the 100-row cap that leaders applies, so nothing clearing the bar gets dropped before you see it. The three predicates sit side by side and AND together automatically: float<40m for a tradeable float under 40 million shares, stage=2 for a confirmed Weinstein uptrend, and rs>80 for relative strength in the top fifth of the market. A name has to clear all three to appear, thin supply and confirmed trend and strength, not any one of them alone.
Read as one sentence, the command says: of the names lightly enough floated to move fast, show me the ones already trending up and already outperforming. That is a materially different list from a raw leaders rs>90 float<40m board, which ranks by strength and lets any stage through, including names deep in Stage 3 or Stage 4 that still carry a high relative-strength memory from an earlier run. Adding stage=2 is how you insist the trend is current, not historical.
Typed live: the command decoded clause by clause. On the 2026-07-09 breakout board 22 of 50 rows were Stage 4; requiring stage=2 refuses those up front.
Why a breakout flag skips this check
The low-float breakouts screen, bo float<25m, is the clearest illustration of what stage=2 buys you, because it deliberately does the opposite. A breakout flag checks one thing: did the stock close up at least 4% on heavier volume today. It does not look at the chart underneath the print, so a name can fire the flag while sitting in the middle of a long decline. On the 2026-07-09 board, that is exactly what happened: 22 of the 50 breakout rows, nearly half, were in Stage 4 decline, against only 14 in Stage 2. The loudest single mover on that board carried a relative strength of 21 and a close 94% below its 52-week high, a violent bounce inside a downtrend, not a new leader.
screen float<40m stage=2 rs>80 refuses those rows before you ever read them. Where the breakout scan hands you a mixed board and asks you to sort the Stage 2 uptrends from the Stage 4 bounces by eye, the stage predicate does the sorting up front. That is not better or worse than the breakout scan, it is a different job: the breakout flag catches today's thrust regardless of trend, and the stage screen catches the trend regardless of today. Running both, and knowing which one you are looking at, is the point.
What the screen still leaves to you
A stage label narrows the field; it does not finish the read. Two names can both read Stage 2 and sit in very different places: one just crossed into a fresh uptrend out of a long base, the other is extended far above its moving average and closer to the end of the move than the start. The stage field cannot tell those apart, and neither can the RS number next to it. That distinction lives on the chart, in the base underneath the trend, not in the screen.
The liquidity caveat that runs through this whole cluster applies here too. A thin-float Stage 2 name can still trade almost nothing, and a confirmed uptrend on a stock nobody can actually buy in size is a chart, not an opportunity. Stack a dollarvol>10m floor on the command, screen float<40m stage=2 rs>80 dollarvol>10m, and the names that clear it are thin, trending, strong, and tradeable. Everything the screen reports is end-of-day and point-in-time: it tells you the float is thin and the stage is up. What you do with that is yours to decide.
Where it sits in the float family
This screen is the float family's answer to the direction question. screen shares<30m coil>85 rs>80 finds thin supply gone quiet, compressing before it moves. bo float<25m finds thin supply firing on a single day, stage unchecked. screen float<40m stage=2 rs>80 finds thin supply already in a confirmed uptrend, the amplifier pointed up rather than coiled or merely twitching. Three frames on the same thin-supply idea, at three different moments in the cycle.
Open the terminal and run screen float<40m stage=2 rs>80 against tonight's close. It ships with tickerstance Pro at $28 a month, grandfathered, so the price you join at is the price you keep. The screen reports a condition, thin float in a confirmed uptrend, and stops there. The trade is yours.
Frequently asked questions
What does `screen float<40m stage=2 rs>80` find?
It returns names with a tradeable float under 40 million shares that are also in a confirmed Weinstein Stage 2 uptrend and ranking above 80 in relative strength. The three predicates AND together, so a name has to clear all three: thin supply, an up trend, and strength. It is the float family's way of screening for the amplifier already pointed up rather than merely thin.
What are the four Weinstein stages in the terminal?
stage is a numeric field (aliases ws, weinstein) with four values: 1 Basing, 2 Uptrend, 3 Topping, 4 Decline. stage=2 filters to the confirmed uptrend. The tag is computed from the last close and can change as a stock moves from one phase to the next, so it describes the present trend rather than predicting the next one.
How is this different from the low-float breakouts screen?
bo float<25m fires on a single day, a close up 4% or more on heavier volume, without checking what stage the stock is in. On the 2026-07-09 board, 22 of its 50 rows were in Stage 4 decline. screen float<40m stage=2 rs>80 requires a confirmed uptrend up front, so it refuses those Stage 4 pops before you read them. One catches today's thrust; the other catches the trend.
Does a Stage 2 tag mean a stock is a good buy?
No. Stage 2 means the terminal reads the stock as being in a confirmed uptrend on the last close, nothing more. A name can be early in a fresh Stage 2 out of a base or extended far above its moving average near the end of the move, and the stage label cannot tell those apart. It is a condition, not advice, and the read of the base underneath it is still yours.
Should I add a liquidity filter to this screen?
Yes, the same as any float screen. A thin-float Stage 2 name can still trade almost nothing, which makes a clean uptrend untradeable in size. Add a dollar-volume floor: screen float<40m stage=2 rs>80 dollarvol>10m keeps only the names that are thin, trending, strong, and actually liquid enough to act on.
Can stage change after I run the screen?
Yes. Stage is recomputed every close from a stock's price and moving-average structure, so a name that reads Stage 2 tonight can roll into Stage 3 over a few weeks or break to Stage 4 on a bad session. Re-running the screen against each new close is how the list stays current; the tag is always a description of the last close, never a lock on the future.