A gap is the cleanest setup a screen produces. The stock opens somewhere other than where it closed,
the distance is visible, and there is a well known rule attached to it: gaps fill. That rule is the
reason gap trades are usually entered, and it is the part worth checking before anything else.
We measured 144,286 gaps of 2% or more across 181 sessions, from 2,173,336 symbol-days. What follows
is what that measurement changes about how a gap trade should be set up, and where the money actually
goes once it is on.
1. Fix what fill means before you trade one
Gaps fill is unfalsifiable as usually stated, because no horizon is attached. Given enough time a
stock touches almost any price. Attach a horizon and the claim becomes answerable, and considerably
less flattering.
Our definition, and the one you should hold anyone else to: the previous close is touched again, the
session low reaching it on an up gap or the session high on a down gap, within a stated number of
sessions.
That definition is already generous to the rule. A price touched once intraday and immediately
rejected counts as filled and would not have been a workable exit. Keep that in mind for every number
below, because they are upper bounds on anything tradable.
2. Same-day fill is the minority at every gap size
| Up gap | Events | Filled same day |
|---|
| 2-5% | 55,923 | 39.2% |
| 5-10% | 13,691 | 28.8% |
| 10-20% | 4,067 | 22.4% |
| 20%+ | 1,774 | 12.3% |
| Down gap | Events | Filled same day |
|---|
| 2-5% | 53,483 | 42.0% |
| 5-10% | 11,088 | 29.6% |
| 10-20% | 3,202 | 17.1% |
| 20%+ | 1,058 | 3.9% |
Even the smallest gaps here, a median of 2.8%, close the distance the same session under half the
time. By 20% the up gaps are at one in eight and the down gaps at one in twenty-five.
If your plan is to fade a gap and be flat by the close, the base rate you are working against is
somewhere between roughly a third and one in twenty-five depending on what you picked. That is the
number to size against, not the chart.
3. Direction matters, and not the way it feels
The intuition is that a stock opening far above its close is expensive and should give it back, while
one opening far below is cheap and should bounce. The measurement runs the other way.
| 20%+ gaps | Same day | Within 60 sessions |
|---|
| Up | 12.3% | 62.9% |
| Down | 3.9% | 37.9% |
A large up gap is 1.7 times more likely to fill within three months than a large down gap, and
roughly three times more likely on the day. The median of that up bucket opened 31.3% above the
previous close; the down bucket opened 27.4% below.
The reason is structural. A large down gap is usually the market repricing something permanent, and a
company that is now worth less has no obligation to revisit its old price. A large up gap is more
often enthusiasm, and enthusiasm decays.
The practical form of this: treat a large down gap as a repricing until something tells you
otherwise, and do not buy it because it looks cheap against yesterday.
4. Why the rule survives anyway
Because it is close to true about the gaps most people see. Gaps of 2-5% are 109,406 of the 144,286
events, 75.8% of them, and those fill 87.5% and 90.4% of the time within three months.
If your experience of gaps is the ordinary ones, the rule looks excellent, and you will carry it
intact into the rare cases where it is worst. The rule is not so much wrong as quoted at the wrong end
of the distribution.
Full fill curves by horizon are in
do gaps always fill.
5. The horizon you can actually hold changes the answer
Small gaps do most of their filling early. For 2-5% up gaps the cumulative rate runs 39.2% on the
day, 69.2% by five sessions, 82.4% by twenty and 87.5% by sixty, so the move from twenty to sixty
sessions adds about five points.
Large gaps are the opposite. A 20%+ down gap is at 3.9% same day, 15.8% by five sessions, 27.0% by
twenty and 37.9% by sixty, and that curve has not flattened at the end of the window. Those are
filling over months, not days, and most of them had not filled at all.
So the horizon in your plan has to be a horizon you can hold. A day trader quoting a sixty-session
fill rate is quoting a number from a different strategy.
6. The exit is where gap trades actually die
A gap trade is entered near the open on a name that just moved, which is the most expensive
combination in US equities. Three measured costs apply directly.
| Cost | What we measured |
|---|
| Crossing the spread | A $5,000 market buy in a $5-20 stock pays about 38bp one way |
| Depth at the quote | The quoted price supports $1,155 to $5,150 depending on price band |
| Stops on a falling stock | On falls of 1.00% or more the spread averaged 2.11%, 4.2 times the 0.50% on falls up to 0.25% |
That last row is the one that ruins fade trades on down gaps. The spread a falling stock quotes is
not a constant; it widens with the move that triggers your stop, so the exit is worst exactly when you
need it. Detail is in
stop-loss slippage
and the depth numbers are in
order book depth measured.
Size follows from the depth figure rather than from your account. If the quote supports $1,155 and
you send $25,000, you are not getting the quoted price. A $25,000 order could not be filled inside ten
levels 28.7% of the time.
7. Do not size the gap from premarket prices
The gap you see before the bell is drawn from prints that may not represent a market you could have traded.
The median symbol-day did 0.65% of its dollar volume before the open, across 33 of the 330 premarket
minutes, and per traded minute the regular session is 10.2 times denser.
A premarket quote on a thin name is a number on a screen, not a price with size behind it. Measure
the gap from the official open, and read
premarket trading measured
before building anything that triggers before the bell.
8. A checklist before you take a gap trade
Write the horizon down first. Same day, five sessions or sixty are three different strategies with
three different base rates, and the fill statistic only means something once the horizon is fixed.
Bucket the gap by size and direction, then look up the base rate for that bucket rather than the
headline. A 2-5% up gap and a 20%+ down gap have almost nothing in common.
Price the exit before the entry. Use the spread you will actually pay on the way out, not the one
quoted while the stock is calm, and check that your intended size fits inside the depth at the quote.
Assume a touch is not a fill. If your plan needs to sell at the previous close, require the level to
hold rather than be tagged, and expect that to cost you some of the fills counted here.
Check that the gap is a gap. A large overnight move in an unadjusted feed can be a corporate action
rather than a price change, and those do not fill because nothing moved.
What this guide does not give you
It does not give you an edge, and the base rates above are not one. They are the denominator any gap
strategy has to beat after costs, and on most buckets that is a demanding bar rather than an
opportunity.
It also says nothing about the path. A gap that eventually fills may have moved violently against you
first, and a fill rate measured on daily highs and lows cannot see that.