Research record · Descriptive
Do Fair Value Gaps Work? 3 Years of Data
WarRoom tested what happens when price first returns to a fair value gap (FVG) on daily and weekly charts, across 512 coins and 3 years, with a sealed holdout. The first touch did not beat comparison periods, and rejections showed no dependable follow-through. FVGs are useful as a map of fast moves, not as a standalone entry.
- About 20% of first visits to a daily FVG (19% weekly) produced a clear rejection within two candles.
- The first-touch rejection rate was below the comparison groups in the holdout.
- After a clear daily rejection, the average two-candle move in the direction of the rejection was −0.34%: no continuation edge.
- Previously tested gaps rejected more often only because they had already survived a test.
What was tested
A fair value gap is the price space a fast move leaves behind between the first and third of three candles. The study replayed WarRoom’s own three-candle FVG definition across 512 coins with at least $1 million of daily trading activity, from June 30, 2023 to June 28, 2026, on daily candles and complete Monday-to-Sunday weeks.
The earlier part of the data was used to build and inspect the study. The later part, from July 1, 2025, was held back and checked once after the rules were fixed.
How often price rejected the gap
A clear rejection means the candle entered the gap and closed back out through the far side in the expected direction.
| FVG state at the visit | Daily | Weekly | How to read it |
|---|---|---|---|
| Untouched (first visit) | 19.8% | 18.9% | Below the comparison groups |
| Touched before, not yet filled | 40.0% | 33.6% | Only because it had already survived a test |
What happened after a rejection
| Timeframe | Average move | 95% uncertainty interval | Reading |
|---|---|---|---|
| Daily | −0.34% | −0.59% to −0.08% | No dependable continuation |
| Weekly | +1.74% | −0.82% to +4.78% | Range includes no continuation |
When daily and weekly gaps pointed in opposite directions, the weekly direction was followed about 57% of the time over the next two daily candles (372 holdout episodes). On the weekly horizon the sample was too small and inconsistent to pick a winner, so WarRoom shows the disagreement instead of resolving it.
Limits of this research
- Daily and weekly candles only: it cannot see the order of events inside a candle, and it does not test 1-minute or 5-minute behaviour or order flow.
- Some coins listed later than others, so the history is not identical for every coin.
- These are descriptions of past behaviour, not a strategy.
Frequently asked questions
Do fair value gaps work?
As a standalone entry, the evidence says no. Across 512 coins and 3 years, a first touch of a daily or weekly FVG did not beat comparison periods, and rejections had no dependable follow-through. FVGs are still useful for seeing where price moved fast and where it is interacting with that structure.
Are older or previously tested FVGs stronger?
Previously tested gaps rejected more often, but only because they had already survived a first test. That is a survivorship effect, not evidence that an old gap is a better starting point. Older gaps saw more movement while active, but not more rejections.
Educational content, not financial advice. Historical results do not guarantee future outcomes.