We backtested the fair value gap honestly — it beat buy & hold on only 10 of 47 famous assets
A fair value gap is the best-known Smart Money Concept: a three-candle imbalance where price jumps and leaves a gap between the first candle's high and the third candle's low that it may later trade back into. It is defined purely by closed-bar price, with nothing to redraw — the first SMC construct we've backtested honestly (others are on the way; see below). We ran it the same way as everything else here: the gap is read only from bars that have closed, every fill is charged fees and slippage, and the result is always measured against simply buying and holding. No favourable window, no deleted losers. Here is the whole glass — every famous asset, every winner and every loser.
So — does the fair value gap work?
On the honest evidence: mostly, no. Across the 47 famous assets with enough trades and history to judge, the fair value gap beat buy & hold on 10 of them — 21%. A handful of real winners exist — but across dozens of assets a few beats are expected by chance alone, so the base rate puts the burden of proof on the gap, and one screenshot of a gap that filled perfectly is not that proof. That is below the 24% we measure across every strategy on every asset — so the gap does worse than a rule picked at random, not better. The typical fair-value-gap run didn't just lose narrowly either: its median result came in at -71 pts versus buy & hold over the window.
These are famous names that still trade today. Assets that delisted, went bankrupt or (for crypto) died or were rug-pulled aren't in the set (survivorship bias) — and that history isn't available from the free public data this tool runs on — so this rate isn't a representative base rate for every asset that has ever traded.
Where it shone, where it died
The most useful thing about the whole glass isn't the average — it's the spread. The same fair-value-gap rule, unchanged, swings from a clear beat to a rout depending purely on which asset it is pointed at.
Ethereum — the fair value gap beat buy & hold by +206 pts here. Where the timing added value it usually did so by sitting out a decline the buy-and-holder rode down, then stepping back in after the drop — as much a statement about that asset's fall as about the gap.
Nvidia — the fair value gap came in -898 pts behind buy & hold. On a strong, sustained trend price keeps leaving gaps and running on without returning to fill them, so the rule sits out the move — the worst thing to do to a trending asset.
That inconsistency is the honest answer to “does the fair value gap work”: it isn't a property of the gap, it's a property of the asset and the regime. Returning to a gap can help on something that ranges and hurt badly on something that trends — and you can't know which future you're in from the past.
Why the fair value gap looks like it always gets respected
- • You are shown the one gap that filled and bounced perfectly. Picking the chart after you already know the gap was respected is hindsight, not evidence — the whole glass above is what you get when you can't choose.
- • The filled-and-failed gaps get deleted. Smart-money charts quietly stop drawing a gap once price trades through it, so the eye only ever meets the ones that held — a survivorship illusion. We count every gap the rule ever formed, winners and losers alike.
- • The costs are usually hidden. Returning to every gap trades a lot; fees and slippage on each entry and exit quietly turn many “respected” gaps negative — the friction was always real, most demos just leave it out.
Which Smart Money Concepts we backtest — and which we decline, and why
Everyone draws fair-value gaps; we honestly backtest the one closed-bar-safe FVG rule — costs on, versus buy & hold, whole glass — and decline the Smart Money Concepts that repaint or are defined by hindsight.
The fair value gap is one idea inside Smart Money Concepts. We ship the ones a machine can read the same way every time from closed bars, and we decline the ones whose meaning is added by eye after the move. Here is where each one lands — graded by whether it can ever be shown wrong — and the honest, testable thing we run in its place.
- • Break of structure → the Donchian breakout: a close above the highest high of the last N closed bars. The same idea — price clearing a prior extreme — with a level that never moves after the fact.
- • Liquidity sweep → the false break: a new N-bar high, then a close back below it. The observable half of a “sweep” — the failed push — without the claim about resting orders no price feed can show.
- • Premium and discount → channel position: whether price sits in the upper or lower half of an N-bar range, measured from bounds the machine sets, not lines drawn by hand.
The inversion fair value gap, the balanced price range, the opening-gap imbalance, and the mechanical order block anchored to the gap-creating move (the last opposite-colour candle before it) are all defined by closed-bar price too. They are on the list to build and backtest the same honest way — not declined, just not here yet.
- • The popular order block is chosen by eye after the move, and two chartists mark two different candles — there is no single rule the data can settle, so a backtest of it would really be a backtest of one person's eye.
- • A liquidity sweep rests on resting orders a price feed never shows, and the label is only pinned on once price turns back — a reading added after the fact.
- • A break of structure is drawn from a swing pivot that later bars can still shift, so the level moves as the chart updates — a redraw, or a lag, depending on when you look.
- • A change of character is a break of structure plus a call on which way the trend was already running — the second step is a judgement, not a measurement.
- • Hand-drawn premium and discount hang off a range whose top and bottom are placed by eye and shift as the chart is redrawn — move the anchor and the label moves with it.
- • Inducement is a move said to trap traders before the “real” one — but it is only ever named after the real move has happened, so nothing could ever show it wrong. An idea that cannot be wrong cannot be tested.
This line can move. As more of these ideas turn out to have a closed-bar definition a machine can read the same way twice, they move from the declined column into the built one — this is never a promise to add no further Smart Money Concept.
Every fair-value-gap backtest is public — winners and losers, no cherry-picking. Read the whole fair value gap scoreboard, then run it on the asset you care about and see honestly which side of the line it lands on.
Figures are computed live from simulated backtests over roughly the last five years and update as new bars close. “Fair value gap” here is the standard bullish three-candle imbalance, entered on a return to the gap after price first leaves it and exited when price closes back above the gap or trades through its lower edge; other definitions behave differently, which is the point — you can test them yourself. Hypothetical / simulated results — Past performance is not a reliable indicator of future results. This is an educational, analytical tool — not investment advice, and not a recommendation to buy or sell anything. You make all decisions and execute on your own broker.