Does a Volatile Day Follow a Volatile Day?
https://fxabsolute.com/stats/GBPUSD.json.
Across nine pairs
| Pair | Wide threshold | After a wide day | vs base rate | Samples |
|---|---|---|---|---|
| USD/JPY | 146.9 | 47.4% | +22.4 | 331 |
| EUR/JPY | 149.1 | 46.5% | +21.5 | 329 |
| GBP/JPY | 188.3 | 45.6% | +20.6 | 338 |
| EUR/GBP | 53.1 | 45.4% | +20.4 | 335 |
| USD/CAD | 97.7 | 44% | +19.0 | 339 |
| AUD/CAD | 75 | 42.7% | +17.7 | 330 |
| EUR/USD | 93.3 | 41.8% | +16.8 | 335 |
| AUD/USD | 79.6 | 40.7% | +15.7 | 334 |
| GBP/USD | 115.6 | 38.9% | +13.9 | 332 |
Every pair sits above 25%. That consistency is the finding — it is not one instrument's quirk, and it is the reason a fixed stop distance behaves differently depending on what yesterday did.
Why the base rate matters
"Thirty-nine percent of days after a wide day are also wide" sounds like an edge until you ask what share of all days are wide. It is 25%, because wide is defined as the top quarter. The finding is the 13.9-point difference, not the 38.9%.
Any statistic about market behaviour that arrives without its null is doing the same trick. It is worth asking for one every time.
What follows from it
A stop sized on a long-run average is too tight after a wide day and too loose after a quiet one, and the same is true of a target. That is an argument for sizing off recent range rather than a fixed number — not an argument for trading more after a busy day.
FXAbsolute free backtester
Test the finding on the chart—not just on the page
Open FXAbsolute and replay GBP/USD candle by candle. Turn the observation into explicit entry, stop, target and exit rules, then measure the result yourself.
Questions
Does forex volatility cluster?
Yes. Across all nine pairs measured, a top-quartile range day is more likely than chance to be followed by another — 38.9% against a 25% base rate on GBP/USD.
What counts as a volatile day here?
A day in the top quarter of that instrument's own daily range distribution. For GBP/USD that is 115.6 pips or more.
Is this a trading edge?
It is a sizing input, not a signal. It says the width of tomorrow is partly predictable; it says nothing about direction, which is the part that decides whether a trade makes money.