USD/JPY: how much does it move, when, and what moves it?
The only pair in the series with two peak hours: at 00:00 UTC it delivers 96% of its best hour. A 112-pip median, measured on 521 days of our own data.
Of the six pairs we measured, USD/JPY is the only one with two peak hours instead of one. At 00:00 UTC, when Tokyo opens, it moves almost as much as it does during the London–New York overlap. No other pair in the series does that: at that hour every one of them is asleep.
That is its character, and it changes entirely what time of day makes sense to trade it.
How the numbers were built. Hourly bars from January 2021 to September 2026. The headline figures use the last two years, 521 Monday-to-Friday days, with Sundays excluded because they are half a session. Every hour below is UTC.
How much does USD/JPY move in a day?
The median daily range is 112 pips, the highest of the five dollar pairs we measured. Mind the unit: a yen pip is the second decimal, not the fourth — it is explained in what a pip is and how to calculate its value.
| Kind of day | Range |
|---|---|
| Very quiet (10th percentile) | 53 pips |
| Quiet (25th percentile) | 78 pips |
| Typical (median) | 112 pips |
| Busy (75th percentile) | 152 pips |
| Very busy (90th percentile) | 207 pips |
59% of days clear 100 pips. For EUR/USD that happens 18% of the time.
| Pair | Median daily range | Days under 40 pips |
|---|---|---|
| USD/CHF | 53 pips | 24% |
| USD/CAD | 59 pips | 17% |
| EUR/USD | 64 pips | 15% |
| GBP/USD | 80 pips | 5% |
| USD/JPY | 112 pips | 4% |
⚠️ That table compares pips, not money, and here it matters more than anywhere else: a yen pip is worth considerably less than a euro pip, so 112 USD/JPY pips are not double the risk of 64 EUR/USD pips. To convert it into currency, use the position size calculator.
What hours is it genuinely alive?
This is where the pair parts company with all the others:
| Window (UTC) | Median range per hour |
|---|---|
| 23:00–01:00 · Tokyo opens | 18 to 28 pips |
| 02:00–05:00 · rest of Asia | 16 to 19 pips |
| 06:00–08:00 · London opens | 23 to 25 pips |
| 09:00–11:00 · European morning | 17 to 20 pips |
| 12:00–16:00 · London–New York overlap | 20 to 29 pips |
| 17:00–21:00 · New York afternoon | 12 to 16 pips |
Its best hour is still 14:00 UTC, at 29.0 pips. But 00:00 UTC delivers 27.7, which is 96% of that. For comparison: at that same hour EUR/USD manages 9.5 pips, half of its own maximum. Why the yen behaves this way in those hours is in the yen’s most violent moves.
From which follows the second consequence, and almost nobody connects it to the first: the London–New York overlap holds only 53% of its daily range, the lowest share of the six pairs. Not because the overlap matters less, but because USD/JPY has a second engine and the others do not. The hour map is in forex trading sessions and in session overlaps; live, on the session clock.
What moves USD/JPY?
- US data. In the hour non-farm payrolls are released, its range is 2.32 times its range for that same hour on an ordinary day. The control is EUR/GBP, which carries no dollar and stays at 1.30.
- The distance between the Federal Reserve (the Fed) and the Bank of Japan (BoJ), for years the widest gap between major central banks. It is the pair's background engine.
- The threat of intervention. This is the variable USD/JPY has and the others do not: when the yen weakens far enough, Japan's Ministry of Finance can step in and buy it. What that is and what it does is in what currency intervention is, and a concrete case in USD/JPY near 162.
With a release coming, the news risk calculator turns that multiplier into position size.
What does not move it, despite appearances?
It is not the reverse of EUR/USD. Tempting to assume, because the dollar sits on opposite sides of the two pairs. We measured it: the correlation of their daily returns is −0.60. Real, but it leaves 64% of the movement unexplained — far too much to treat one as the other's mirror. The pair that genuinely mirrors EUR/USD is USD/CHF at −0.83.
And the pair that resembles it most carries no dollar. GBP/JPY correlates +0.70 with USD/JPY, nearly three times its correlation with GBP/USD (+0.25): it is a yen pair with a sterling name, and trading the two together doubles one bet.
And the 00:00 UTC peak is not "the Asian session" in the abstract: the rest of Asia, 02:00 to 05:00, is the flattest stretch of its day. It is the Tokyo open specifically, not Asian hours in general.
Is it a good pair to start on?
It is the least forgiving of a badly measured stop. With a median of 112 pips and a 90th percentile of 207, the distance between a normal day and a busy one is almost 100 pips: sizing by eye is expensive here. Use ATR (Average True Range), which adjusts to the weeks actually in progress, and measure the spread against the range you expect.
In exchange it offers something none of the other five does: if your hours are Asian, this is your pair. Not by elimination, but because at 00:00 UTC it delivers 96% of what it delivers in its best hour. EUR/USD, at that hour, barely notices.
What these numbers do not tell you
- The median is not your day. It describes the centre of a distribution, not what happens tomorrow.
- No spread and no slippage are inside it. Our history stores prices, not costs.
- The granularity is one hour, so a five-minute spike is invisible. In a pair with interventions, that understates the worst of it.
- It is a single feed — our broker's, normalised to UTC. Another would give similar numbers, not identical ones.
- This is not a forecast. It describes how the pair has behaved; the past shapes the expectation, never the outcome.
