Ask three traders when EUR/USD “starts moving” and you will get three answers. The honest one is that the pair does not have a single character — it has at least four, and they hand over to each other at predictable points in the clock. The most abrupt of those handovers happens around 08:00 in New York, roughly an hour before the US cash open and right on top of the London–New York overlap.
This note is a description of behaviour, not a strategy. The point is to give you a mental model of what the pair is doing at any given hour so that a stop placed at 03:00 New York time and a stop placed at 09:30 are not treated as the same decision.
Four regimes in one trading day
It helps to stop thinking of the day as “the Asian session, the London session and the New York session” and start thinking about who is actually being forced to transact. EUR/USD is the deepest currency pair in the world, so its intraday shape is driven less by speculation than by the schedule of people who have to deal regardless of price: corporate treasurers, index funds rebalancing hedges, and option desks managing expiries.
| Window (New York time) | Typical character | What is driving it |
|---|---|---|
| 19:00 – 03:00 | Narrow, mean-reverting | Thin book, Asian corporate flow, ranges respected |
| 03:00 – 06:00 | First real expansion | London desks arriving, European data, order books rebuilt |
| 06:00 – 08:00 | Positioning drift | European fixes, pre-New York hedging |
| 08:00 – 11:00 | Highest range, worst mean reversion | US data releases, overlap liquidity, option expiries at 10:00 |
| 11:00 – 16:00 | Decay, then the 16:00 fix | London leaves, month-end and index flows |
The 08:00 boundary matters because two things change at once. US macro releases — CPI, payrolls, retail sales, PPI — almost always land at 08:30, and the market starts repositioning for them thirty minutes earlier. At the same time, London is still fully staffed, so the depth of book is at its daily maximum. High liquidity plus a scheduled information event is exactly the combination that produces large moves with small slippage.
Why the same setup behaves differently before and after 08:00
Before 08:00, EUR/USD spends most of its time doing something that looks like range behaviour. Levels get respected. A push into the Asian high often fails on the first attempt. Traders who learn the pair on the European morning tend to build mean-reverting instincts, because for those hours the pair genuinely does mean-revert more often than it trends.
After 08:00, the distribution changes shape. The tails get fatter. The same “failed break” that would have reversed at 05:00 now has a materially higher chance of continuing, because the order flow behind it is different: it is not a desk fading a level, it is a fund adjusting an exposure. A stop-loss that sits ten pips beyond a session high is a sensible risk decision at 05:00 and close to a donation at 08:29.
The level did not stop working. The population of people trading around that level changed.
What to actually measure
If you want to hold an opinion about this rather than borrow mine, the measurements are simple and you can build them in a spreadsheet from any hourly data export.
- Average true range by hour. Bucket the last 250 sessions by New York hour and take the mean high–low range. You are looking for the shape of the curve, not the absolute numbers.
- Continuation rate. For each hour, when price makes a new high of the last four hours, what fraction of the time does it make another new high within the next two? Compare 04:00 with 09:00.
- Range as a share of the day. What percentage of the full daily range is typically completed by 08:00? On EUR/USD it is usually a minority, which is the whole argument for not committing full risk to the European morning.
- Spread by hour. Log your own broker’s quoted spread every fifteen minutes for a fortnight. Your execution cost curve is not the same as the market’s.
The fourth item is the one traders skip and then wonder why a backtest that looked profitable on mid prices does not survive contact with a live account. Cost is a function of the clock too.
The 10:00 option expiry effect
Currency options in the interbank market conventionally cut at 10:00 New York. When a large strike sits nearby — the kind that gets flagged in daily expiry lists — spot has a tendency to be drawn towards it into the cut and then released afterwards. This is not mystical; it is delta hedging by the desks that are short the option, and it decays to nothing the moment the expiry passes.
You do not need to trade the effect to benefit from knowing about it. It explains a large share of the 09:30–10:00 grind that otherwise looks like the market has simply stopped for no reason, and it explains why a breakout that fires at 09:45 disproportionately fails.
Practical consequences
- Size the day, not the hour. If most of the range arrives after 08:00, an entry at 04:00 is committing risk against a distribution that has not started yet.
- Widen stops with the clock, or shrink size. Same risk in currency terms, different distance. The mistake is holding the pip distance constant across a session boundary.
- Do not carry a European-morning mean-reversion trade through 08:30. Either the thesis survives the data or it does not; that is a separate decision from the one you originally made.
- Log the release calendar before, not after. The pair’s character on a CPI day and on an empty Tuesday are different instruments wearing the same ticker.
What this note does not claim
None of the above is predictive. Hourly volatility profiles are stable enough to plan around and unstable enough to embarrass anyone who treats them as a rule. Regimes shift: through a central bank tightening cycle the European morning can carry far more of the daily range than it does in a quiet summer. Re-measure quarterly, and treat any single day that violates the profile as information rather than as a violation.
The value of a session map is not that it tells you what will happen. It is that it stops you from being surprised by things that happen at the same time every day.
