A Rule-Based Sell System for Swing Trades
Three Rules, One Daily Routine, No Exceptions
A Rule-Based Sell System for Swing Trades
by Sean Sharpe
A complete sell discipline for catalyst, episodic-pivot, and momentum swing trades. Three rules, one daily routine, no exceptions. Built to remove all real-time emotional decision-making from the exit side of the trade.
Why You Need This
The entry side of any system is the part traders obsess over — chart patterns, catalysts, setups, scanners, watchlists. But every experienced trader will tell you the same thing: the exit is harder than the entry.
This is the failure-mode side of the trade. The list of ways a winning trade gets ruined on the way out is long:
Selling on an intraday wiggle that means nothing
Selling out of fear instead of signal
Over-trimming on emotion (40% off when the rule said 20%)
Cutting winners short because they “looked toppy”
Holding losers because the position “used to be a winner”
Switching to a faster timeframe to justify a sell the daily chart doesn’t support
The point of a rule-based system is to make the exit decision not a decision at all. Pre-decide it once, write it down, follow it.
The Three Core Rules
Rule 1 — Round Number Trim — Take 20% Off
At entry, set a Good-Til-Cancelled limit sell for 20% of the position at the next round number that is at least 30% above your entry price.
The round-number ladder:
$2 → $5 → $10 → $20 → $50 → $100 → $150 → $200 → $300 → $500
Use this table to find the right level for your entry:
Entry $1.00 – $1.53 → first trim at $2
Entry $1.54 – $3.84 → first trim at $5
Entry $3.85 – $7.69 → first trim at $10
Entry $7.70 – $15.38 → first trim at $20
Entry $15.39 – $38.46 → first trim at $50
Entry $38.47 – $76.92 → first trim at $100
Entry $76.93 – $115.38 → first trim at $150
Entry $115.39 – $153.84 → first trim at $200
Entry $153.85 – $230.76 → first trim at $300
Entry $230.77 – $384.61 → first trim at $500
If no qualifying round number exists (entry above $384.61, or no level ≥30% above), skip Rule 1. The next rule becomes your first trim trigger.
Above the table: the ladder simply gets wider as price scales — $500, $1000, $1500, etc. Same logic, same 30% buffer test, just bigger steps.
Why round numbers? They act as natural liquidity magnets and psychological pivots. Retail traders place orders around them. Algorithmic systems target them. The 30% buffer ensures you’re not just selling at a number — you’re selling at a number that represents real progress.
Worked example — BAND (Bandwidth) daily: entry just above $32 on the breakout (yellow up-arrow). The next qualifying round number is $50, which sits ~56% above entry — well clear of the 30% minimum buffer. The GTC limit set at $50 fires automatically when the stock tags the level on first contact. The 20% trim is locked in without a real-time decision; the remaining 80% continues to run under Rules 2 and 3.
Rule 2 — 4EMA-of-Highs Trim — Take 20% Off
After the position has been above the 4EMA of highs for at least 3 days, a daily close below the 4EMA triggers a 20% trim, executed 5 minutes before the close.
Critical detail on re-trims: the 3-day-above qualification resets after every trim. After a trim fires, the position must first close back above the 4EMA for 3 days again, then close below it, before the next 20% trim is triggered. This prevents you from getting whipsawed by multiple trims in a choppy stretch where price keeps tagging the line.
Key details:
All trims are 20% of what remains in the position at the time of the trim, not 20% of original size. The position decays smoothly rather than dropping on a fixed schedule, which leaves a meaningful runner intact for the eventual 20 EMA exit.
The rule does not trigger in the first 3 days of the trade. Early choppy action gets ignored.
After a trim, the 3-day-above clock resets — wait for the position to re-establish itself above the line before another trim is eligible.
The close happens 5 minutes before the bell so you’re acting on a near-final daily print, not intraday noise.
2-Day Cooldown Between Trims (Rule 1 ↔ Rule 2): if the round-number trim has fired within the last 2 days, skip the 4EMA trim even if it triggers. The same applies in reverse — if a 4EMA trim has fired within the last 2 days, skip the round-number trim if the GTC limit hits during that window. Two trims clustered inside two days is the system double-counting the same structural move. Let one trim represent the event; the next trim activates after the cooldown clears.
Why the 4EMA of highs? The 4-period exponential moving average of the daily highs is a fast-response line that follows strong trends closely without being whipsawed by single bars. When price closes below it after riding it for days, the trend has materially slowed — but hasn’t necessarily broken. That’s exactly the right condition for a partial trim, not a full exit.
Worked example — DELL daily, with the 4EMA of highs plotted: breakout entry from a long sideways base around $115 (yellow up-arrow ~$140). The inner fast-tracking line is the 4 EMA of highs; it hugs the daily highs much more tightly than the 20 EMA underneath. The pale candles early in the advance are “Slingshot” bars — bars where price closed above the 4 EMA of highs, the entry pattern this trim rule is built from. The trim rule is the inverse of the Slingshot: Slingshot = close above (a strength signal); the trim trigger = close below (loss of that strength). Once the position has been above the line for 3+ days, each subsequent close-below fires a 20% trim, with the 2-day cooldown preventing clustering against the round-number trim.
Rule 3 — 20 EMA Final Stop — Exit All
If price is below the 20 EMA at the daily close (checked 5 minutes before the bell), exit the entire remaining position before the close.
This is non-negotiable.
No second chances.
No “let’s see if it bounces tomorrow.”
No switching to a faster timeframe to justify holding.
Why the 20 EMA? It’s the standard medium-term trend filter on the daily chart. A close below it isn’t a wiggle; it’s a structural break of the multi-week trend. The position no longer meets the conditions that justified holding it. Exit.
The 20 EMA rule defeats the single most expensive failure mode in swing trading: letting a winner round-trip into a loser because the position “used to be a winner.” When the trend is broken, the trade is over. Exit. Find a new setup.
Worked example #1 — CLPT (ClearPoint Neuro), the clean preventative exit: breakout from a long downtrend base (yellow entry arrow ~$15), quick parabolic run to ~$22, then a single decisive close below the 20 EMA near the high. The rule fires once, you exit the remaining position 5 minutes before the close, and the name drifts lower out of your account. The full exit at the structural break is the one trigger that turns a winner into a closed winner instead of letting it round-trip.
Worked example #2 — QURE (UniQure), damage limitation when a gap beats the rule: breakout entry (~$25), strong run through $30, and the 4 EMA trim rule fires three times at the top as the trend chops. By the time the devastating overnight gap-down hits the right edge of the chart, the position is already ~49% of original size (the geometric decay of three trims: 1000 → 800 → 640 → 512). The gap bar itself is a long-lower-wick recovery candle — by holding to the EOD check rather than panic-selling the open, the remaining 512 shares get a much better fill, and the 20 EMA close-below exit closes the trade. Two rules combine here:
Trim discipline before the event — Rule 2 already pulled meaningful size off three times on the structural slowdown, before anyone knew a gap was coming.
Hold-to-EOD on gaps — the open print was much worse than the close; obeying the rule recovered a substantial portion of the gap-bar damage.
This is what the system is designed to do on trades that don’t end cleanly. Not every winner exits at the high — but every winner exits with the runner reduced and the EOD rule preventing open-print panic.
Choosing The Right Timeframe
The default timeframe is daily. The three rules above all execute on the daily chart for a standard swing position with an expected hold of roughly 3 days to several weeks.
But the mechanics of the system are timeframe-agnostic — the same three rules work on any timeframe. The only thing that changes is which timeframe you read the chart on. Match the timeframe to the expected hold duration.
The four timeframe variants:
Nanocap parabolic (1-day +500%+ moves) — hold 1-2 days — 10-minute chart
In-between momentum — hold ~3 days — 30-min or 1-hour chart
Standard swing (catalyst, EP, breakout) — hold 3 days to several weeks — Daily chart (default)
Secular catalyst / massive weekly base breakout — hold months to a year+ — Weekly chart
Everything else stays identical. The 4EMA-of-highs becomes the 4EMA-of-highs on the relevant timeframe. The 20 EMA stop becomes the 20 EMA on the relevant timeframe. The 3-day-above qualification becomes 3-period-above.
When to use each variant:
10-min for nanocap parabolics: these move 500-800% in a session and reverse just as fast. A daily exit is far too slow — the trade is over before the daily print resolves. The 10-min 20 EMA captures the structural break in real time.
30-min / 1-hour for in-between trades: short-shelf-life catalyst that’s more than a one-session move (a strong reaction expected to ride 1-3 days). Used less often than the standard daily.
Daily for standard swings: the default. Catalyst-driven episodic pivots, multi-day continuation plays, normal breakouts.
Weekly for secular catalysts and perfect weekly base breakouts: a multi-year base breakout with a structural catalyst behind it is fundamentally different from a 3-day pop. The position is held for the structural rerate, not the reaction. Daily rules would shake you out on the first normal weekly pullback. Move the whole system up to the weekly.
The Same Rule Across Four Timeframes
Reading the charts: the white line is the 20 EMA on the relevant timeframe. Every red triangle marks a candle that closed below the 20 EMA — i.e. Rule 3 firing on that bar. The MA Ribbon (20 / 50 / 100 / 200) is plotted for context; the system itself uses only the 20 EMA and the 4 EMA of highs.
10-min — nanocap parabolic (BIRD): gap-and-go from a $2.50 base, opened ~$8, ran to $26.21. The rising 10-min 20 EMA tracked the entire move; every pullback held the ribbon until the first decisive close below fired the exit trigger.
10-min — sequential signals (ORBS): sub-$10 base to $87.55 high. First signal at the blow-off top, second on the structural break back through the EMA. Two clean fires from the same rule on the same chart.
30-min — multi-day catalyst swing (DRUG): biotech catalyst run from ~$1 base to $90+ over several sessions. The 30-min 20 EMA rode underneath the entire trend; the final close-below fires after the swing peak. The honest tradeoff: on a multi-leg parabolic, a faster timeframe will produce mid-trend shake-out signals — which is why the hold-duration estimate dictates the timeframe at entry.
Daily — multi-week position trade (FSLY): long sideways base $7-9, vertical breakout to ~$26. Daily 20 EMA tracked the breakout leg, sell triangles clustered at the top, latest bar broke back to the EMA / prior shelf — textbook daily-variant exit sequence.
The Weekly Variant — Secular Base Breakouts
When the chart shows a multi-year base breakout on the weekly with a structural catalyst behind it, the trade is fundamentally different from a 3-day catalyst pop. The position is held for the structural rerate. Daily rules would shake you out on the first normal weekly pullback. Move the entire system up to the weekly: 4 EMA of weekly highs, 20 EMA on the weekly. Trims and exits happen at weekly closes.
Worked example — PL (Planet Labs) weekly: multi-year base $1.60-$8 from 2022 through 2024, then a clean breakout above the $8-9 horizontal in late 2025. Yellow up-arrows mark the breakout entry and the retest re-add. The trendline-break sell triangle fires once on the first pullback structure, then the trend rides the rising weekly 20 EMA to ~$40. Textbook secular base breakout — the same setup on a daily would have shaken you out repeatedly during normal weekly pullbacks.
Worked example — NVDA weekly, the re-add cycle within a secular trend: the weekly 20 EMA tracked the multi-year uptrend cleanly. Yellow up-arrows mark a catalyst-driven re-add near $48 (early-2024 hyperscaler catalyst) and a second re-add around $95 on a pullback to the rising EMA. Sell triangles fire at the eventual peak and the most recent break. Two trims, two re-adds, same trade — the trims locked R from each leg; the re-adds sized the next leg on new fundamental catalysts.
The Counterfactual — Same Stock, Wrong Timeframe
The most powerful teaching pair for timeframe selection is showing the same stock managed on two timeframes, where one was correct and one would have destroyed the trade.
BE (Bloom Energy) weekly — correct timeframe for the secular breakout: multi-year base $11-$33 from 2020 through mid-2025, breakout above the $33 horizontal on a structural energy-infrastructure catalyst, two yellow up-arrow re-adds, ride to ~$350 with only a handful of trim triggers across the entire run.
BE daily — the exact same trade on the wrong timeframe: count the red sell triangles. Each one is a 4EMA close-below that would have fired a 20% trim of the remaining position. Over the same run that produced the clean weekly hold above, the daily-variant rules trim the position down on every normal pullback to the daily rising 20 EMA — and the most recent bar is a hard close below the 20 EMA, which on the daily timeframe is a full exit.
Same entry. Same catalyst. Same stock. The weekly variant holds for the structural rerate; the daily variant trims down to a fraction and exits. The rules don’t tell you which timeframe to use — the hold-duration estimate at entry does. The timeframe choice matters more than any other decision in the system.
How To Decide At Entry
Pick the timeframe at entry, not midway through the trade. Switching mid-trade is one of the failure modes the system is designed to defeat (Hard Rule #3).
A reasonable check at entry: “What does this trade need to do for me to consider it a winner?”
Move 50%+ in an hour or two? → 10-min
Move 30-50% over a couple of days? → 30-min / 1-hour
Move 30%+ over a couple of weeks? → Daily
Re-rate from a multi-year base, hold for the new uptrend? → Weekly
The answer determines the timeframe. The timeframe then locks for the duration of the trade.
The Full Sell Sequence
Every winning trade follows this exact sequence. No deviations.
1. ENTRY → set stop + GTC limit sell 20% at the round number. Position remaining: 100%
2. ROUND # HITS → 20% of remaining sold automatically. Position remaining: 80%
3. CLOSE < 4EMA HIGHS (after 3+ days above, ≥2 days since last trim) → sell 20% of remaining, 5 min before close. Position remaining: 64%
4. CLOSE < 4EMA HIGHS (next occurrence, same conditions) → sell 20% of remaining, 5 min before close. Position remaining: ~51%
5. CLOSE < 20 EMA → exit all remaining, 5 min before close. Position remaining: 0%
Trims are 20% of what remains at the time of the trim, not of original size. Starting with 1,000 shares: the round-number trim sells 200 (leaving 800); the next 4EMA close-below sells 160 (leaving 640); the next sells 128 (leaving 512); the 20 EMA exit sells the remaining 512. The position decays geometrically and a substantial runner is always intact for the final 20 EMA stop. No two trims (Rule 1 ↔ Rule 2) within 2 days of each other — the cooldown prevents the system double-counting a single structural move. All discretionary sells happen 5 minutes before the close — no earlier.
The Daily Routine
Same time. Same actions. Every day.
During market hours: Nothing. No sell decisions. No touching orders. (Round-number GTC orders are the only exception — they execute automatically.)
5 min before close: if price is below the 20 EMA → exit the remaining position now.
Then check: has it closed below the 4EMA of highs (after 3+ days above, and ≥2 days since the last trim)? → sell 20% of remaining at 5 min before close.
After the close: log the day’s actions in your journal.
Done: walk away.
The Three Hard Rules
These three sentences are the entire enforcement mechanism. Everything else flows from them.
1. NO selling during market hours — except round-number GTC orders. 2. NO overriding signals. 3. NO switching timeframes.
If you find yourself wanting to sell during market hours (other than a pre-set round-number GTC), you’re not seeing what the daily timeframe will eventually print. The end-of-day check is the only discretionary signal that matters.
If you find yourself wanting to override a rule because “the market looks different today” — that’s exactly the moment the rule exists to defeat.
If you find yourself thinking “the 1-hour chart looks worse than the daily” — that’s the most common rationalization for selling out of position. Stay on the timeframe you picked at entry.
Setting Up A New Position — The Checklist
When opening a new swing position, do these immediately — same trade, same minute as the entry:
1. Set the entry-day stop. The default stop for any new position is the Low of Day of the entry day (LOD). That structural reference is the only stop you need until something changes (a trim fires, or the position runs far enough to be trailed). Pre-set this on entry.
2. Look up the entry price in the Round Number Level Table. Set a GTC limit sell for 20% at the determined round number. It executes automatically when hit.
3. Pick the timeframe based on expected hold duration. It locks for the duration of the trade.
4. Note the date. Day 0 + 3 periods = the earliest possible 4EMA trigger date. Until then, only Rule 1 (round number) and Rule 3 (20 EMA stop) are live.
5. Write the position into your journal with entry price, position size, timeframe, stop level, and round-number trim target — so when the close arrives, the day’s check is mechanical.
After the first trim: consider moving the stop to break-even. This is the standard move once the position has paid for itself via the round-number trim — but it’s not absolute. On some setups the original LOD stop is left in place to give the position more room. Use judgement. The point of moving to B/E is that a winning trade should never become a loser; the point of sometimes leaving the stop wider is that the best trades never come anywhere near the original stop in the first place.
Edge Cases
The three core rules cover ~90% of what happens in a trade. The cases below are the recurring scenarios the rules don’t address by themselves — handle them the same way every time so they don’t become discretionary doors.
Gap-Downs
If the stock opens below the 20 EMA on a gap-down, do not exit at the open. Wait for the close. The rule is end-of-day driven for a reason — opening prints are noise-heavy and frequently reverse during the session. Let the daily print resolve, then act on the 5-min-before-close check exactly as written. (QURE’s lower-wick recovery candle above is the canonical example.)
Earnings Mid-Trade
Earnings reports are the single largest overnight-gap risk a swing position carries. The system handles them with a discretionary trim, not a hard rule:
Trim down to roughly a 5% position size going into the print, in most cases.
The actual trim depends on the quality of the setup into earnings, situational awareness (broader tape, prior reactions in the sector), and how many trims have already fired from Rules 1 and 2.
The point: never carry a full-size position through a binary event. Take the reaction-side exposure off; keep a runner.
This is one of the few places in the system where judgement is explicitly allowed. Document the trim size and reasoning so the discretion stays honest and reviewable.
Re-Entry After A 20 EMA Exit
Once the 20 EMA close-below has triggered a full exit, the trade is over. No re-entry unless a fresh fundamental catalyst AND a fresh technical setup are both present — exactly the conditions that justified the original entry. A reclaim of the EMA on its own is not a re-entry signal; a new catalyst on its own without a clean technical setup is not either. Both signals must be there. If they are, that’s a new trade with a fresh entry, fresh stop, and fresh round-number lookup. If either is missing, walk away. There are always other setups.
The Best Trades Never Come Near The Stop
A note on calibration: in real swing-trading practice, the trades that turn into the meaningful winners almost never come anywhere near the original entry-day LOD stop. They run from entry. The stop exists to define risk before the trade is taken — not because most winners test it. If a position is repeatedly poking at the original stop, that’s information: the setup probably isn’t working, and the “give it more room” instinct is exactly the wrong response. Honour the stop; the next setup is around the corner.
What Belongs Outside The System
The Sell System operates inside a trade that has already passed your entry rules. It governs how a trade is managed, not whether it should be entered.
A 20% trim at the round number = rule-based management.
A 20% trim on a 4EMA close-below = rule-based management.
The 20 EMA exit = rule-based management.
Anything outside this system that closes a position — a discretionary trim because “the market looks toppy,” a panic exit because “things feel off,” a target hit that isn’t on a round number — is a rule violation, regardless of how it works out in P&L terms. Track those separately. The point of the system is to make those discretionary exits stand out as the anomalies they are.
A Final Note
This system isn’t designed to maximise upside on any single trade. It’s designed to make sure your winners actually become winners on the way out, and your losers don’t get a second life because you got attached to them.
Every rule trades a small amount of theoretical upside for a much larger amount of behavioural reliability. The 30% buffer on the round-number trim means you sometimes leave a few points on the table. The 4EMA trim sometimes fires early on a healthy pullback. The 20 EMA exit sometimes sells at the bottom of a wick before a reversal.
That’s the cost. It’s worth it. The cost of not having a system is far higher — it’s the gap between what your trades actually produced and what they would have produced if you’d stayed disciplined.
The only way to find out is to run it for a quarter, log every trade against the system, and see what your actual execution-vs-rule gap looks like. That gap is your real edge to recover.
Very nice write-up.
Sean, it looks like Rule #1 does not apply to weekly structural breakouts (like your NVDA example) since no trim was taken at $100. Am I correct?