Fair value gaps — the 3-candle imbalance where the first candle’s high and the third candle’s low never overlap — are the building block of half of ICT-style trading. The usual claim: price returns to "rebalance" the gap, and the 50% level (consequent encroachment) is the entry.
We tested that claim mechanically: 12 months of real Binance 1-minute data on BTC and ETH, aggregated to 5-minute candles, entering on the retrace into the freshest unfilled FVG from 10:00 New York onward, with realistic costs and 1% risk per trade. The full, unfiltered results are below.
Verified Result
No edge: net negative after costs across 2 markets.
| Market | TF | Trades | Win | PF | Max DD | Net |
|---|---|---|---|---|---|---|
| ETH | 5m | 1,078 | 31.6% | 0.50 | 95.6% | -95.5% |
| BTC | 5m | 1,154 | 27.3% | 0.34 | 99.3% | -99.3% |
How the SVS 20 breaks down ▾
12 months of real 1-minute data, fees on (0.05%/side), $10k start, 1% risk. How the score works →
The exact rules we tested
- Detect FVGs as 3-candle imbalances (bar A high < bar C low for bullish, inverse for bearish).
- Track the freshest unfilled gap no older than 30 bars.
- From 10:00 to 16:00 New York time: enter when price retraces to the 50% level of that gap.
- Entry at the qualifying 5-minute close; stop beyond the last opposing swing (fractal lookback 3); target 2R; close any open trade at session end.
- Risk 1% of equity per trade; 0.05% commission per side; 10× max notional leverage.
Results
Binance spot 1-minute klines (data-api.binance.vision), aggregated per strategy timeframe · starting balance $10,000 · risk 1%/trade · Commission 0.05% per side; no spread/slippage modeled (BTC/ETH spot spreads are sub-basis-point); position size capped at 10× notional leverage. Generated 2026-06-12 by the Secuora ai-strategy deterministic runner (same engine as the in-app AI backtester).
| Month | Trades | Win rate | Net P&L |
|---|---|---|---|
| 2025-06 | 113 | 22% | −$4,372 |
| 2025-07 | 98 | 23% | −$1,923 |
| 2025-08 | 109 | 25% | −$1,747 |
| 2025-09 | 91 | 25% | −$793 |
| 2025-10 | 99 | 29% | −$414 |
| 2025-11 | 79 | 32% | −$156 |
| 2025-12 | 80 | 30% | −$128 |
| 2026-01 | 101 | 25% | −$180 |
| 2026-02 | 94 | 33% | −$65 |
| 2026-03 | 88 | 33% | −$47 |
| 2026-04 | 105 | 26% | −$66 |
| 2026-05 | 97 | 28% | −$39 |
| Month | Trades | Win rate | Net P&L |
|---|---|---|---|
| 2025-06 | 92 | 28% | −$3,097 |
| 2025-07 | 87 | 32% | −$1,914 |
| 2025-08 | 85 | 33% | −$990 |
| 2025-09 | 76 | 37% | −$582 |
| 2025-10 | 99 | 34% | −$909 |
| 2025-11 | 90 | 31% | −$471 |
| 2025-12 | 102 | 26% | −$704 |
| 2026-01 | 87 | 31% | −$289 |
| 2026-02 | 98 | 34% | −$143 |
| 2026-03 | 84 | 26% | −$134 |
| 2026-04 | 91 | 34% | −$132 |
| 2026-05 | 87 | 33% | −$185 |
Assumptions (how loose terms were pinned down)
- "after 10:00" read as the 10:00–16:00 New York session
- FVG entry at the 50% of the freshest unfilled gap, max 30 bars old
What the test does and does not show
This is the floor for the mechanical rule, not the ceiling for a discretionary FVG trader: no higher-timeframe bias filter, no displacement requirement, no news filter — one rule, every signal taken. That is deliberate. You cannot judge a filter’s value until you know the unfiltered baseline, and the unfiltered baseline is exactly what nobody publishes.
The monthly table matters more than the headline number: gap strategies live and die by regime. Watch how the same rule behaves in trending versus chopping months before you conclude anything.
Methodology, in one paragraph
Data: Binance spot 1-minute klines (data-api.binance.vision), aggregated per strategy timeframe, June 1, 2025 – June 1, 2026 (12 months). Execution: Secuora’s deterministic strategy runner (the same engine behind the in-app AI backtester) — single position at a time, entries at the close of the signal candle, commission 0.05% per side; no spread/slippage modeled (btc/eth spot spreads are sub-basis-point); position size capped at 10× notional leverage, starting balance $10,000, 1% risk per trade. Swings are confirmed fractals with no look-ahead. These are mechanical results: no discretion, every signal taken. Past performance does not predict future results; this is research, not financial advice.
Frequently asked questions
What is a fair value gap (FVG)?
A three-candle pattern where the wicks of the first and third candle never overlap, leaving a price band the market skipped through — bar A’s high below bar C’s low (bullish) or bar A’s low above bar C’s high (bearish). Traders treat a return into that band as an entry zone, commonly at its 50% level.
What win rate does an FVG strategy have?
Our mechanical 12-month test on BTC and ETH (entry at the 50% of the freshest unfilled gap after 10am NY, 2R target, costs on) produced the win rate and profit factor shown in the results table above. Filtered, discretionary FVG trading can differ in both directions — the published number is the unfiltered baseline.
Do fair value gaps actually get filled?
"Most gaps eventually fill" is true but not tradeable by itself — the question is whether entering at the gap with a stop and a 2R target is positive expectancy after costs. That is exactly what this page measures, month by month.
How do I backtest an FVG strategy myself?
On Secuora, type the strategy in plain English into the AI backtester ("after 10am NY, enter on the retrace into the freshest 1-minute FVG…") — it compiles to the same deterministic engine that produced this research — or practice it manually, bar by bar, in the replay terminal.