Skip to content
GetProfitable
Search
Wiki

Gap-Fill Swing

Multi-day version of the gap fill: fade non-catalyst gaps in liquid names or indices that fail to extend within the first two days, targeting the pre-gap close over the following week.

What it is

The gap-fill swing is a slower cousin of the intraday gap fill. Rather than fading a gap in the first hour, it waits for evidence over one or two days that the gap has no follow-through, then enters a position aimed at the pre-gap close over the following 3 to 10 sessions. It applies to gaps caused by sympathy moves, sector news, analyst notes and other soft catalysts, not to earnings gaps, which belong to post-earnings-drift.

The logic

A gap that does not extend within two days has usually exhausted its cause. The buyers who reacted to the note or the sector move are done; the stock now has a pocket of thin volume between the gap and the pre-gap price, and holders who bought at the top of the gap are sitting on small losses with no new information to justify holding. Gaps fill because those holders sell, and because there is little natural demand in the gap's price zone until the pre-gap level, where the previous buyers live.

On the other side are traders who bought the gap as a breakout and are hoping for continuation, and the analysts' clients who bought the upgrade. When the catalyst is soft, they are the supply.

Setup rules

  • Market: stocks with average daily dollar volume above $30 million, sector ETFs, index ETFs. No small caps; their gaps are too often real.
  • Timeframe: daily for the setup; intraday only for execution.
  • Gap conditions: an up gap of 2 to 6 percent (for shorts) not caused by earnings, guidance, M&A or FDA-type binary news; a catalyst that is either absent or soft (upgrade, sympathy, index inclusion rumour).
  • Failure conditions: two full sessions after the gap without a close above the gap day's high; the second day closes in the lower half of the gap day's range; volume on days 1 and 2 declining.
  • Trend context: the stock is not in a strong uptrend (below a 3-month high) and the gap did not break a multi-month base. Fading gaps out of bases loses.
  • Short constraints: shares are borrowable at a reasonable rate; short-interest is below 10 percent of float to avoid squeezes.

Entry, stop, target

Short on day 3's open, or on a break of day 2's low. Stop above the gap day's high plus 0.5 ATR. Target is the pre-gap close (a full fill); a conservative version takes half at the gap's midpoint. Time stop of 10 sessions.

Item Level Notes
Pre-gap close 100.00
Gap-day range 103.50 to 105.00 Gap 4 percent, closed 104.20
Day 2 close 103.80 Lower half, volume down
Entry 103.40 Break of day 2 low
Stop 105.90 Gap high plus 0.5 ATR, risk 2.50
Target 1 102.00 Gap midpoint, reward 1.40, 0.6R
Target 2 100.20 Full fill, reward 3.20, 1.3R

The R:R is unattractive by design; the setup is a high-win-rate trade, and it depends entirely on the classification of the gap being right. If the win rate in your own testing is not above 60 percent, the setup does not pay.

Position sizing and risk

Shorting into a gap carries squeeze risk; risk no more than 0.5 percent of equity per trade using /tools/position-size, and use a hard stop order rather than a mental one. Long versions (fading gap-downs) are easier to execute and carry gap-down-again risk instead. Read /learn/risk-management on asymmetric outcomes before running a short-biased book of these; a single squeeze can erase ten fills.

What breaks it

  • Misread catalysts. A "soft" upgrade that is actually the start of a re-rating. The 2-day failure filter helps but does not eliminate this.
  • Squeezes. A short into a stock with hidden demand can gap higher a second time. The stop is a real order, not a plan.
  • Trending markets. In a strong bull market up-gaps in strong names extend more often than they fill; restrict to gaps in stocks that are not at highs.
  • Costs. Borrow fees, the spread and the small target combine to make this a low-margin trade. The full-fill target is needed to make it work, and it is the target reached least often.
  • Edge decay. Gap statistics in index ETFs have shifted as more systematic capital fades gaps; the fill rate at 5 days is lower than older studies claim.

How to test it

Scan for every gap in your universe over 5 years, manually tag the catalyst type (this is the slow, unavoidable step), and record the fill status at 2, 5 and 10 days. That gives you fill rates by catalyst type before any entry rule. Then apply the 2-day failure filter and your entry and exit rules with costs including borrow. If the fill rate for soft-catalyst gaps that failed the 2-day test is not at least 15 points higher than for all gaps, the filter is not adding value. Minimum 300 tagged gaps. Paper-trade for a quarter, since the setup is quiet outside of busy news periods.

Variations

  • Gap-down long version for index ETFs and quality large caps that gap down on sector news.
  • Options version: a short-dated put debit spread instead of a short stock position to cap squeeze risk; see credit-spread-program for structuring logic.
  • Intraday version: see gap-and-go-gap-fill.

Further reading

gap, mean-reversion, short-selling, short-interest, short-squeeze, float, atr, win-rate, expectancy, bull-trap.

Related playbooks: gap-and-go-gap-fill, post-earnings-drift, range-mean-reversion-20ma, afternoon-reversal

See it drawn

Original diagrams for the ideas on this page. Illustrative, not real market data.

Risk and reward on one tradeA price scale showing an entry with a stop two points below and a target six points above, so the reward band is three times the risk band.PRICETARGET 106.00ENTRY 100.00STOP 98.00REWARDRISK6.00 pointsthree times the risk2.00 pointsthe most you loserisk : reward = 1 : 3
Risk and reward on one trade. One trade on a price scale: the entry sits 2.00 points above the stop and 6.00 points below the target, so the shaded reward band is three times the risk band. The ratio compares what is lost if the stop is hit with what is gained if the target is reached.
Bollinger bands squeezing and then expandingA price line between three curves: an average in the middle and a band above and below it that pinch together in the centre of the chart and then spread apart as the price runs higher.PRICE WITH BOLLINGER BANDS (20, 2)SQUEEZEupper bandpricemiddle band20-day averagelower bandbands widen asvolatility risesIllustrative prices. The bands sit two standard deviations from the average.
Bollinger bands: squeeze and expansion. The middle line is a 20-day average and the outer bands sit a set number of standard deviations away, so they measure how far price has recently been straying. When moves are small the bands pinch together; when moves grow they spread apart.

Educational only, not advice. Spotted an error? Post in Site Feedback.