I get asked about fair value gaps more than almost anything else in price action, usually by someone who has watched a few videos and is now drawing rectangles on every chart they open. The concept is simple enough that it gets over-applied fast, and once your chart is covered in twenty gaps, none of them mean anything. So let me try to define it tightly, explain why the thing even works, and then get to the part that actually matters, which is telling apart the gaps worth waiting for from the ones that are just noise wearing a costume.
What a fair value gap actually is
A fair value gap is a three-candle pattern. You look at any candle, call it the middle one, and then check the candle before it and the candle after it. For a bullish gap, if the high of the first candle sits below the low of the third candle, the empty space in between is the imbalance, carved out by a strong middle candle that ran through it. For a bearish gap it is the mirror image, the space between the low of the first candle and the high of the third. Nobody traded in that pocket.
The important word is overlap. On a normal, balanced move, consecutive candles share a lot of price range, so buyers and sellers both got filled across most of that band. A fair value gap is the footprint of a move so aggressive that price skipped a whole slice of that band. One side just steamrolled through. That untraded pocket is what you are marking, and it is why the pattern is sometimes called an imbalance. The market spent no time being fair to both sides there.
One thing that trips people up. The gap is not the middle candle itself. It is the space between the wicks of the outer two candles. Draw your rectangle from the high of candle one to the low of candle three for a bullish gap, and that box is the whole thing.
Why price tends to come back to fill it
Nobody can prove the mechanism, and I would be suspicious of anyone who tells you they can. But the order-flow story is reasonable and matches what you see often enough to be useful. When price rips through a zone that fast, it usually means a pile of resting orders got taken out at once, or a big participant lifted everything in front of them. That leaves unfilled interest sitting behind the move, and people who got filled at bad prices now looking for a place to add or get out closer to break-even.
So the pull back into the gap is really price going to find liquidity it skipped. It is efficiency-seeking behavior. A market that moved too fast in one direction often drifts back to let the other side transact before it decides what to do next. This is why I treat a fair value gap as a magnet, not a wall. It attracts price back toward it. Whether it then acts as support or resistance once price arrives is a separate question, and plenty of the time the gap just gets filled and price keeps going the original way.
Which gaps are worth trading
Most gaps you can see are cosmetic, which is the part the tutorials skip. A gap that forms during quiet chop tends to get filled almost immediately and tells you nothing. Here is roughly how I sort them.
- Context. Did the gap form during a real displacement, a candle that broke structure or took out a prior high or low? A gap that comes with a break of structure is worth marking. A gap in the middle of a range is usually just breathing.
- Size relative to the instrument. A gap that is a tiny fraction of the average candle range is not a level, it is a rounding error. I want a gap that is a meaningful chunk of recent range, something price would actually have to work to fill.
- Whether it is already mitigated. If price has already traded back through most of the gap, the imbalance is mostly gone. A fresh, untouched gap is worth more than one that has been revisited twice already.
- Confluence. A gap sitting on top of a prior swing point, a round number, or a higher-timeframe level is far more interesting than one floating in empty space.
If a gap fails all four of these, I do not draw it. That single discipline cleans up a chart more than any indicator.
Timeframe, entries, and knowing when you are wrong
Fair value gaps exist on every timeframe, and lower timeframes produce a flood of them, most of which are garbage. My rule of thumb is to find the gap on a higher timeframe first, the one that matches your holding period, and only then drop down to time an entry. If you are swing trading, the daily or four-hour gap is the object of interest. If you are intraday, maybe the fifteen-minute. The lower you go, the more gaps you get and the less each one means, so let the higher timeframe pick the level and the lower timeframe pick the trigger.
On the entry itself, the mistake is treating the gap as a single line. It is a zone, and price rarely respects the exact edge. A few models I have seen work:
- Full fill. Wait for price to trade all the way through the gap to the far edge, then look for a reaction. Fewer entries, but the ones you get are cleaner because the imbalance is resolved.
- Partial fill, or the consequent encroachment idea. Enter around the midpoint of the gap rather than the edge. The logic is that the fifty percent level of an imbalance often does the job of the whole zone. You get a better price than waiting for a full fill, at the cost of more trades that never fill and leave you behind.
- Edge tap with lower-timeframe confirmation. Mark the near edge, and when price touches it, drop to a lower timeframe and wait for a shift there before entering. This keeps you from catching a knife that blows straight through.
None of these is correct in the abstract. They trade off precision against how many entries you actually get. Backtest whichever one you pick on your own instrument before you trust it, because the fill behavior of gaps is very different between, say, a large-cap equity and a thin altcoin.
You also need a rule for when the trade is wrong. The clean invalidation rule is that a bullish gap is dead if price closes decisively below it, and a bearish gap is dead if price closes decisively above it. Closing through is the key, not just wicking through. A wick that pierces the gap and closes back inside is normal and often the exact spot you wanted to be a buyer. A full-body close on the other side means the imbalance was not defended and your reason for the trade is gone. Get out.
The failure mode to internalize is the gap that gets filled and keeps going. Not every imbalance is a reversal point. Plenty of them are just continuation, where price dips back to fill the untraded pocket and then resumes the original move. If you treat every gap as a guaranteed bounce, you will keep fading strong trends and wondering why. The gap tells you where price is likely to revisit. It does not tell you what happens after. That second question is answered by structure and by which way the higher timeframe is leaning, not by the gap itself.
As for how many actually fill, be careful with the round numbers people throw around. Fill rates depend enormously on the instrument, the timeframe, and how you defined the gap in the first place. A small gap on a liquid instrument fills a large fraction of the time, often quickly. A large gap that formed on a genuine structural break fills less reliably and can stay open for a long stretch, which is precisely why the open ones are worth watching. Rather than trust a stat from a video, measure it yourself on the thing you trade. When I want to see how gaps have behaved across a lot of tickers at once, I lean on the backtesting tools inside Blockcircle, but a spreadsheet and a hundred hand-marked examples will teach you the same lesson.
Mark them tightly, respect the four filters, size the entry to how much precision you need, and let a body close on the wrong side kick you out. Do that consistently and fair value gaps become a useful piece of the read instead of another set of rectangles cluttering the screen.