Gap fills: how often small-cap gaps give back the move
A gap fill is a return to the prior close. In FennEdge's archive, few large gaps closed below the prior close, more gave back half; an intraday rate awaits a planned FennEdge data study.
Start in plain English
A gap fill is the claim that a stock which opened far above yesterday's close tends to trade back down to that close. It is one of the most repeated lines in small-cap trading, usually phrased as if gaps owe the market a return trip. For a scalper who might fade a gapper, or hold one, the useful question is narrower: how often, by what definition, and by when?
The core idea of this lesson: "gap fill" is three different measurements that people blur together.
- Full fill: price trades back to the prior close (or below it).
- Partial fill: price gives back part of the gap, for example half (the "gap midpoint").
- When: intraday at any point, or by the close. A gap can fill at 10:15 and close green; it can also hold all morning and fill at 15:55.
FennEdge's archive can measure the close-based versions today. Across 671 US common stocks and ADRs that opened at least 20% above the prior close, 62 (9.2%) closed below the prior close, and 208 (31.0%) closed below the gap midpoint (2026-07-01 to 2026-10-02; 66 sessions; 461 tickers; FennEdge archive; as of 2026-10-02; see Small-cap gapper statistics (2026); the 31.0% is a FennEdge computation from the Gapper HOD study data). So in this sample a full fill by the close was the exception, while giving back half the gap by the close happened in roughly three sessions out of ten. How often price touched the prior close during the day is a different number, and FennEdge has not counted it yet (a FennEdge data study of intraday gap fills is planned).
Gap definitions, gap percentage and "gap and go" belong to US Day Trading Fundamentals: Gappers: gap percentage, gap and go, and gap fill.
Two paths from the same start
All prices, sizes and costs in this example are Illustrative. Aldmere Health is fictional. The trade shown is a short-side fade, one common way traders act on a gap-fill idea; short mechanics, locates and SSR are covered in Shorting small caps.
The setup at 09:41 ET. Prior close $1.80. Opened at $2.34, a gap of +30.0%; the gap midpoint is $2.07. The stock pushed to $2.52 at 09:33 and then faded. The 09:41 bar closed at $2.31, the first 1-minute close below the opening price after 09:35. That is this lesson's fade trigger.
The plan, written before 09:41. Short entry with a marketable limit order (floor $2.28), filled at $2.30. Stop at $2.43, three cents above the 09:40–09:41 high of $2.40. Target $2.07, the gap midpoint (a half fill), not the prior close. Time stop 12:00. Planned risk $0.13 per share. Size: $50 risk ÷ $0.13 = 384 shares. Costs: $0.005 per share round trip for commissions and fees plus a $0.02 per share locate fee, $0.025 in total (Illustrative), plus any slippage.
Path AA half fill by midday
Price slides under $2.25 by 09:50, bounces to $2.33 at 10:05 without reaching the stop, and trades $2.06 at 11:12. The $2.07 buy-to-cover limit fills. The stock closes at $1.94, still above the prior close: a half fill, not a full one.
Path BThe gap holds
09:42 opens $2.31 and closes $2.36, back above the open. By 09:46 the stock trades $2.45. The buy stop at $2.43 triggers and fills at $2.46, three cents of slippage. The stock closes at $2.71 without trading below $2.29 again.
Path CA slow grind
Price drifts between $2.15 and $2.36 for two hours without touching the stop or the target. At 12:00 the time stop covers at $2.22.
The review note for all three: "Plan followed; fade trigger as defined; risk 1R; outcome A, B or C." Nothing at 09:41 separated them. The rest of this lesson sets out what the archive can and cannot say about how often each kind of day happens, and which conditions came with more or fewer fills in this sample.
Defining the fill before you count it
| Term | Definition used in this lesson |
|---|---|
| Gap | Regular-session open ÷ prior session close − 1 (split-adjusted) |
| Gap midpoint | Prior close + half of (open − prior close) |
| Full fill by close | Session close below the prior close |
| Half fill by close | Session close below the gap midpoint |
| Intraday full fill | Any regular-session trade at or below the prior close (needs minute data; planned FennEdge data study) |
| Fade trigger (example only) | First 1-minute close below the opening price at or after 09:35 |
Three choices change the count more than any chart reading:
- Which prior close. After a reverse split, the unadjusted prior close can be a tenth of the adjusted one. A gap computed on the wrong close is meaningless. Use split-adjusted prices and check corporate actions for the day.
- Close vs. touch. A stock that trades $0.01 under the prior close at 10:20 and closes 25% green counts as an intraday fill and as a close-basis hold. Both statements are true; they answer different questions.
- Which session. A premarket dip below the prior close before 09:30 is not a regular-session fill. Write down whether premarket trades count.
Odds box
Read the two boxes together. The measured box describes days, judged at the close. The Illustrative box describes trades, judged by a plan with a stop. A 31% half-fill rate by the close does not mean a fade entered at 09:41 reaches the midpoint 31% of the time: some of those days squeezed through a stop first and faded later, and some faded before 09:41 and left no trigger. Only an intraday study with stated entries can connect them, which is why that study is planned.
What shifts the odds
These splits come from the same 671 observations (FennEdge computation from the Gapper HOD study data; 2026-07-01 to 2026-10-02; FennEdge archive; as of 2026-10-02). They are associations in one sample of 66 sessions, not triggers.
| Condition | Full fill by close | Half fill by close |
|---|---|---|
| Gap 20%–50% (n = 459) | 50 (10.9%) | 147 (32.0%) |
| Gap 50%–100% (n = 129) | 10 (7.8%) | 38 (29.5%) |
| Gap 100% or more (n = 83) | 2 (2.4%) | 23 (27.7%) |
| Open under $1 (n = 97) | 6 (6.2%) | 34 (35.1%) |
| Open $1–$5 (n = 259) | 22 (8.5%) | 77 (29.7%) |
| Open $5 or more (n = 315) | 34 (10.8%) | 97 (30.8%) |
- Gap size. Full fills by the close were less frequent as the gap grew in this sample, which is partly arithmetic: a 120% gap has much further to travel than a 22% gap. Half fills varied less across gap sizes.
- Price tier. Differences between price tiers were small relative to their intervals; with 97 sub-$1 observations, a gap of a few points is within noise.
- Month. Full fills by the close: July 26 of 196 (13.3%), August 21 of 270 (7.8%), September 14 of 192 (7.3%). October's 13 observations (2 sessions; 1 full fill) are too few to report as a month, which is why the three months add up to 658 of the 671 and 61 of the 62 fills. Regime shifts week to week; see Hot and cold small-cap markets.
- When the day's high printed. Of gappers whose 04:00–20:00 high printed in premarket, 45 of 356 (12.6%) closed below the prior close; of those whose high printed in the regular session, 17 of 260 (6.5%). The other 55 printed their high after hours (16:00–19:59) and account for none of the 62. This split uses hindsight: you only know the premarket high was the day's high after the close.
- Hypotheses to record, not findings: catalyst type, float, an offering or filing during the session, SSR status, and how far price had already faded before your trigger. The archive does not yet split by these.
Observe vs. infer
Back at 09:41 in the Aldmere Health example (fictional; figures Illustrative):
| What the screen shows (observed) | What a trader might infer | Why it is not established |
|---|---|---|
| The 09:41 bar closed at $2.31, below the $2.34 open | "The gap is failing" | In the Gapper HOD sample (2026-07-01 to 2026-10-02; FennEdge archive; as of 2026-10-02), 438 of 671 gappers closed below their open, yet only 62 closed below the prior close |
| Price is $0.51 above the prior close | "Gaps fill, so it has $0.51 to fall" | In the same sample, 62 of 671 closed below the prior close; a full fill by the close was the minority outcome |
| Volume fell from 820,000 at 09:30 to 270,000 at 09:40 | "Buyers are exhausted" | Volume after the opening minutes is commonly lower; a decline does not identify who trades next |
| Quote at 09:41:59: bid $2.30 × ask $2.31 | "Easy to cover later" | The cover happens at a future spread, often wider during a squeeze |
Cost of being wrong
Using the Aldmere Health fade at 384 shares (Illustrative):
| Outcome | Exit | R after costs | Dollars after costs |
|---|---|---|---|
| Path A: half-fill target | $2.07 | +1.58R | +$78.72 |
| Path B: stop, 3¢ slippage | $2.46 | −1.42R | −$71.04 |
| Path C: time stop | $2.22 | +0.42R | +$21.12 |
| Path B with a halt-up and reopen at $2.70 | $2.70 | −3.27R | −$163.20 |
Fees and locate total $9.60 per round trip at this size. The last row is the case a stop cannot handle: if the stock pauses under LULD while moving up and reopens higher, the buy stop fills at the first available price (LULD, Trading halts). A short-side fade has no natural ceiling on that row. Size for it.
How many trades before this means anything
The archive figure of 9.2% rests on 671 observations (2026-07-01 to 2026-10-02; FennEdge archive; as of 2026-10-02), and its interval is still about 7%–12% (FennEdge computation). Your own log is much smaller. If you count fills on 10 sessions with about 10 gappers each and see 9 full fills by the close out of 100 (Illustrative counts), a 95% Wilson interval runs from about 5% to 16%: consistent with the archive, and also with rates nearly twice as high. Observations from the same session are not independent (a hot day lifts many gappers at once), so 100 gappers from 10 sessions carry less information than the count suggests. How many trades before your numbers mean anything has the method.
What this lesson does not claim
- It does not claim gaps fill, or that fading gaps is profitable or unprofitable. The fade split is Illustrative until the FennEdge data study of intraday gap fills is published.
- The measured rates are close-based. They say nothing about whether price touched the prior close during the day.
- The sample covers 66 sessions from July to early October 2026. Repeat tickers (multi-day runners) count once per session.
- The archive does not control for float, news type or dilution filings.
- Aldmere Health, its bars, quote, fees and locate costs are fictional and Illustrative.
Go deeper
- FennEdge first.
- Top Gappers Calendar lists the day's top gappers with their opening gap and day change (close vs. prior close) in percent. Guests, Free and Plus see the latest session at default filters; full history and filters need Pro or Max (tiers as of 2026-10-05). It is historical daily data, not real-time.
- Top Gappers daily recaps are public pages of the top 10 per completed session, from 2026-10-01; they show no high or low columns, so they support close-based counts only.
- Gapper HOD Timeline shows when each gapper printed its high (guests, Free and Plus: latest completed session; earlier sessions need Pro or Max). The Gapper HOD research article publishes the same 671-observation sample.
- In-Play Days shows 1-minute charts of past active days, where you can check intraday touches by hand. Without Max, including as a guest, you can open each ticker's most recent active day in full; earlier days and the two-year history need Max.
- FennEdge's article Gap up and gap down covers gap basics.
- Gap-down fills. The mirror case (a gap down that trades back up to the prior close) is a separate count; the archive's gapper list holds gap-ups only.
- Split-adjusted prior close. Reverse splits are common among small caps; an unadjusted close makes the gap and the midpoint wrong.
- Look-ahead in your own tests. Filtering by "the high printed in premarket" uses information you only have after the close; Backtesting intraday ideas covers this.
- Related lessons: Why small caps behave differently, Opening range breakouts, Multi-day runners.
Exercises
Six exercises follow this lesson: three concept checks, one calculation (a gap midpoint after a reverse split), one scenario question on reading the measured splits, and one practice task. Exercises, answers and progress are available after signing in.
Practice task: count fills on 10 sessions. Use the Top Gappers Calendar (history needs Pro or Max; with guest, Free or Plus access, record the latest session each day for 10 sessions) or the public daily recaps (top 10 per session, from 2026-10-01). For every listed gapper with a gap of 20% or more, record the opening gap and the day change (close vs. prior close) in percent, then classify: full fill by close (day change below 0%), half fill by close (day change below half the opening gap), or neither. Write the definitions first and log every listed gapper, not only the ones that look interesting. Use the FennEdge practice log: download the practice log (CSV), with a column guide. These are observations, not trades: set mode to replay, write fill=full, fill=half or fill=none plus the gap and day-change percentages in conditions, and leave the fill, R and path columns blank. The exercise checklist tracks each step; your log stays on your own device. Keep it separate from the Trading Journal, which holds only real broker-synced trades.
Primary sources
- FINRA. Trading Halts, Delays and Suspensions.
- Limit Up-Limit Down Plan. LULD Plan website.
- SEC. Key Points About Regulation SHO.
- FINRA. Frequent Intraday Trading: Understanding the Basics.
Sources checked 2026-10-05. Aldmere Health and all prices, volumes, costs and the fade outcome split are fictional and Illustrative. Measured figures come from the FennEdge archive as stated. This lesson is education, not a recommendation to buy or sell any security. Past patterns do not indicate future results.