A fair value gap is three candles. That’s the whole definition before anyone dresses it up. Candle one leaves a wick. Candle two moves hard in one direction. Candle three leaves a wick that doesn’t overlap candle one’s. The space between those two wicks never traded. Nobody bought there, nobody sold there, and some price-action traders treat that untouched space as a magnet — an “inefficiency” price is statistically more likely to come back and fill. I mark it the same way I mark everything else on a blank chart. Then I wait.
The three candles, with real numbers
Say a stock is trading around 84 on a five-minute chart. Candle one prints a high of 84.10. Candle two is the move — a strong green candle that opens at 84.15 and closes at 85.40, barely pausing. Candle three opens near 85.30 and has a low of 84.90. Now look at the two wicks that matter: candle one’s high sits at 84.10, candle three’s low sits at 84.90. Nothing in between — from 84.10 up to 84.90 — ever traded. No buyer and no seller met inside that band on any of the three candles. That’s the gap. Eighty cents of untraded space, left behind by a move that outran the orders sitting under it.
That’s the mechanical test, and it’s the only one that matters: does candle one’s wick stop before candle three’s wick starts. If yes, there’s a gap. If the wicks overlap even slightly, there isn’t one, no matter how fast candle two moved. Speed makes the gap likely. Only the wick positions confirm it.
Why the gap forms in the first place
A normal candle-to-candle sequence has overlap — candle two’s range touches candle one’s, and candle three’s touches candle two’s, because trading is continuous and price moves through levels in small steps with orders on both sides at every tick. This kind of gap means that continuity broke for one candle. Buyers hit market orders faster than sellers could show up to take the other side, or the reverse on the way down. The imprint of that imbalance is the gap itself — a stretch of price nobody had to defend or contest, because nobody was there.
This is also why the gap tends to show up around news, earnings reactions, or the open of a session, when order flow arrives in a burst instead of a steady stream. A quiet, grinding chart rarely produces one. A chart that just got hit with a wave of one-sided orders produces them constantly.
Fair value gap or just a level with new packaging
Here’s my honest take, and it’s not going to make the concept sound more exotic than it is. I’ve been marking places price left behind since before I’d ever heard the term “fair value gap.” I called it a gap-fill zone. Some people called it an imbalance. Smart money concepts traders gave it the FVG label and a three-candle rulebook, which is genuinely useful — it’s more precise than “price moved fast here” — but the underlying idea isn’t new. It’s the same principle behind every level I’ve ever drawn: mark the place where price didn’t do normal business, and pay attention if it comes back.
A support line marks a spot where buyers showed up before. A supply zone marks a spot where sellers showed up before. This one marks a spot where nobody showed up at all. Different mechanism, same category of tool — a record of something the chart did, not a prediction of something it’s going to do. I don’t treat FVG as a separate system bolted onto my chart. I treat it as one more way a level gets created, alongside old highs, old lows, and reversal candles I’ve already been marking for years.
What the “must fill” claim gets wrong
The part of fair value gap trading I don’t buy is the certainty some accounts sell with it — the idea that price is somehow obligated to return and fill every gap because the market demands “efficiency.” Price isn’t obligated to do anything. Some gaps fill within hours. Some sit untouched for months. Some never fill at all because the stock trends straight through them and never looks back. Treating it as a guaranteed round trip is the same mistake as treating a support line as a guaranteed bounce — it’s a place worth watching, not a promise.
What I actually do with a fair value gap is mark the edges — candle one’s wick, candle three’s wick — the same two clicks I’d use on any zone. Then I wait and watch how price behaves if it comes back into that space, instead of assuming the return itself is the trade.
A trade where the gap was the whole setup
AMD, hourly chart, after a strong earnings reaction. The stock gapped from around 168 to 176 inside a two-hour window, and inside that move sat a clean three-candle gap between 170.40 and 172.10 — candle one’s high at 170.40, candle three’s low at 172.10, nothing traded in between. I marked it and left the chart alone.
Four sessions later, price pulled back. It came down through 176, through 174, and slowed right at 172.10 — the bottom edge of that gap. It didn’t blow through it. It sat there for about ninety minutes, printed two small-bodied candles, then turned back up. I got long near 172.30 with a stop under 171.90, just below the gap’s lower edge. Price ran back to 177.80 over the next three days.
The gap didn’t predict that bounce. It told me where to watch closely, the same way any level would have. If price had sliced straight through 172.10 without slowing, I’d have taken the small loss on the stop and moved on — I’ve had that happen with these gaps just as often as I’ve had it happen with lines. The setup wasn’t “buy the gap because gaps fill.” The setup was “price is approaching a spot where it left something behind — watch the close, wait for confirmation, size the trade like the level might not hold.”
I’ve had the reverse happen too, on a different name a few months earlier. The gap held on the first touch, failed on the second visit a week later, and price tore through it without slowing. Two touches, same drawn box, two different outcomes. That’s not the tool failing. That’s the tool doing exactly what a level does — describing a place, not a certainty.
How I actually mark one
On the chart, I’m not running a scanner or an indicator to find these. I look for the same thing every time: a candle with real range, followed by two candles whose wicks don’t overlap it. When I see one, I draw a box from the top wick to the bottom wick, same as I’d box a supply zone off a reversal candle. No fixed width, no rounding to a clean number — the candles set the edges, not me.
Once it’s marked, it sits on the chart like every other level. I don’t chase price into it, and I don’t assume a touch means a bounce is coming. I wait for the candle to actually close in a way that shows rejection — a small body, a long wick back out of the gap — before I do anything. That’s the same patience I’d apply to a supply zone or an old swing high. The name on the box doesn’t change how carefully I wait. A gap that gets touched and sliced through cleanly isn’t a failed trade idea. It’s information that the imbalance wasn’t strong enough to matter this time, and the chart moves on.
Where this leaves me
I use fair value gaps now. I didn’t for years, mostly out of stubbornness about a term that sounded more complicated than it needed to be. Once I sat down and worked through what the three candles were actually showing, it turned out to be a level like any other level — just one built from absence instead of a wick that got rejected. Same rule applies as everything else on a blank chart: mark it, wait, and let the close tell you what happened. Not the label.
I trade with a blank chart, levels built from wherever price left something behind — old highs, reversal candles, fair value gaps — and nothing else in Static, the free daily chart room run by Draw Lines Make Money. If marking a gap instead of chasing it makes sense to you, you can sit in and watch how it’s done live:
Join the free Static chart room →
Disclosure: that’s an affiliate link — I may earn a commission if you join a paid tier later, at no extra cost to you. The free room is free.