Does the fair value gap strategy work? A backtest on twenty assets after costs
A weekly fair value gap, a liquidity sweep on the four-hour chart and a break of structure to enter took 301 trades on 20 assets in 2022–2025. It won 26.6% of them with a target three times the stop, just above the 25.0% needed to break even. Before commission and swap it made +0.06 R per trade; after them, −0.00 R.
What a fair value gap is
A fair value gap (FVG) is a three-candle pattern: the wick of the first candle and the wick of the third do not overlap, leaving a price range the middle candle crossed without trading back. Smart Money Concepts traders expect price to return to that range and react from it. In our engine the gap exists only after the third candle has closed, so a backtest never sees it earlier than a live trader could.
The strategy we tested
- Zone: a fair value gap on the weekly chart.
- Confirmation: price sweeps a four-hour fractal high or low inside the zone.
- Entry: a break of structure on the four-hour chart in the direction of the weekly gap, stop behind the sweep, target three times the stop.
The result
| Value | |
|---|---|
| Trades | 301 |
| Win rate | 26.6% |
| Win rate needed at 1:3 | 25.0% |
| Per trade before commission and swap | +0.06 R |
| Per trade after all costs | −0.00 R |
| Assets with a profit after costs | 11 of 20 |
The idea is not empty: before commission and swap the trades made a small profit, and the win rate sits above break-even. But the edge is about the size of the costs, and after them nothing is left. About half the assets ended in profit and half in a loss, which is what a strategy with no edge after costs looks like.
What could change the result
- A weekly gap gives few trades: about 15 per asset over four years. That is too few to tell a small edge from luck on one asset.
- Costs per trade are lower with a wider stop, so the higher timeframe helps; the commission and swap of holding for several days still take the profit.
- Test the same rules on the assets you trade, with your broker's costs, before relying on them.
How we measured
- Engine backtest on the production server, each asset separately, period 2022-01-01 to 2026-01-01, starting balance $10,000, fixed risk of 1% per trade.
- Commission and swap measured for one commission-charging broker; entries and exits at bid and ask prices, so the spread is inside the result before commission and swap.
- Win rate = take-profits ÷ (take-profits + stop-losses). Results in R, where one R is the amount risked.
Backtests run on October 24, 2026.
Questions
Is a fair value gap the same as an imbalance?
Yes, traders use both names for the same three-candle gap.
Why test on twenty assets?
A result on one asset over four years rests on a handful of trades. Many assets show whether the rules work in general or only where they happened to fit.
For education, not investment advice. Past results, including backtests, do not guarantee future results.
← All articles