The Asian Session Range: Why London Sweeps It Before the Real Move

October 8, 2026 · 8 min read · ChartWise Team

Every trading day opens with a box nobody meant to draw. Between midnight and the London open, gold and the majors grind sideways through hours with no news and barely any volume. Then London comes in and the first real move, more often than not, goes straight through one side of that box. I measured the last two weeks across gold, EUR/USD, GBP/USD, and USD/JPY: 32 of 35 sessions took out the Asian high or low before anything else happened. The box looks like support and resistance on the chart. It functions as an order pool, and the first hours of European trade are when the pool gets collected.

What the Asian Range Actually Is

Tokyo runs the quietest hours on the chart. Local desks work 9 to 5, which is midnight to 8 in the morning UTC, and the flow through those hours is corporate and mean-reverting. Price oscillates, the range compresses, and by the time Tokyo winds down you have a clean high and low, usually tapped two or three times each. That is the Asian range.

The window matters less than consistency. A practical version for gold and the majors is 00:00 to 08:00 UTC. The ICT tradition marks 19:00 to midnight New York time, which overlaps but comes out shorter. Broker charts stamp their own server time, often UTC+2 or UTC+3, so convert the window once and never move it. One calendar note: Europe and the US switch daylight saving about a week apart each spring and fall, so the box shifts by an hour twice a year if you think in New York time. Recheck the conversion when the clocks change, then forget about it.

Here is the part that matters. Those two lines are shared by everyone who marks the box, on every timeframe, in every timezone. Anyone shorted overnight parks the protective buy stop just above the high. Anyone planning the London breakout parks a buy stop just above it too. The other side mirrors it with sell stops under the low. By 7 in the morning UTC there is a measurable pile of resting orders a few ticks beyond each line, and thousands of traders built it without coordinating, because the lines are obvious and stops go where the level is.

The Numbers, Measured

I pulled hourly bars for gold futures and the three most traded pairs and marked the box the same way every day: high and low between 00:00 and 08:00 UTC, then the first bar after 07:00 UTC that traded beyond either side. The sample runs September 25 through October 7, thirty-five instrument-days. Two weeks is a snapshot rather than a study. The snapshot agrees with what session traders have watched for years, and you can rerun it on any window you like.

Combined: 32 of 35 days, 91 percent, and 27 of those 32 hits landed before 10:00 UTC. The box itself held about half of the day’s full range on the majors and near 60 percent on gold, which means the overnight hours compress roughly half of everything the instrument will do into two lines you can see before breakfast.

Why London Runs It First

The mechanical answer is that the pool is where the size goes. A fund that wants to buy a few thousand gold contracts has a problem: market-buying that size lifts every offer in the book and gives away the edge before the position is on. Resting stops solve it. Trigger the buy stops above the box high and thousands of market orders fire upward at once, and the buyer absorbs all of them in seconds at prices that were, moments earlier, the crowd’s protection.

The timing answers a volume question. Tokyo hands the instrument over thin. London arrives with the day’s first serious inventory needs and the day’s first real volatility, and volatility plus a visible pool equals the morning’s first stop. There is no villain in this loop, and there doesn’t need to be one. The cluster is public information. My stop sitting 5 pips under an equal low is one row in a book next to thousands of identical rows, and nobody has to know my name for running all of them to be worth a few million dollars.

This is also why the raw breakout is the trap side of the box. The first orders through the Asian high are, in order: the shorts’ stops firing as market buys, the breakout buyers chasing as market buys, and the large seller who engineered the sequence, filling them all. A poke that runs 20 pips past the high and rolls over did not fail. The filling was the point, and the filling is done.

What Happens After the Sweep

Two days dominate. The rotation day: price sweeps a side, finds no follow-through, and spends the rest of the session traveling back through the box, often to the far side. The trend day: the sweep is the launchpad, the break holds, and the instrument closes the session outside the box entirely. My two weeks split 13 rotation days against 19 trend days, and the tell arrives early. The first hour after the sweep either closes back inside the box, which sets up the rotation read, or holds outside and passes a retest, which sets up the trend read.

The retrace numbers back this up. After the first sweep, price came back through the box at least partially 29 times out of 32, and the median trip back covered 44 to 58 percent of the box height on the FX pairs. Which means the move off the sweep is usually tradeable in both directions: the fade gets a real leg, and the trend entry gets its retest. The one read with no edge is the first tick. Market-ordering the break as it happens buys the top of the pool at the exact moment the pool is being spent.

Two days from the sample make the point better than any rule could. October 5, gold swept the $4,194.90 box high at 08:00 UTC, extended just $4 beyond it, and closed the session back inside the box. Pure collection, no trend. October 6, the sweep was a $3 poke at the same hour, and price ran to $4,202.80 by 11:00 UTC, closing more than $13 above the box high. Nearly identical sweeps, opposite outcomes. The hourly closes sorted them inside the first two hours: a close back inside the box meant rotation, closes holding outside meant trend. October 7 swept the low and never came back either, finishing more than $60 under the box, which is the same acceptance read pointing the other way.

Trading Around the Box

Mark it before London does

Both sides of the box go on the chart before 07:00 UTC, along with the prior day’s high and low. Round numbers sitting at or just beyond a box edge go on the map too, because a big figure one poke past the box high stacks two pools in the same place and deepens the sweep. The stop-hunting guide covers the full cluster map; the box is its time-anchored subset.

Two reads, two trades

The sweep-and-fail: price takes a side by at least the median overshoot, an M15 close comes back inside, and the trade targets the box mid first and the far side second, with the stop beyond the sweep extreme. The acceptance trade: an hourly close holds outside, the retest of the level holds, and the target is the next session extreme or the prior day’s level. If you only ever take one, take the first. The median overshoot gives it a defined trigger and a defined invalidation, and the base rates say it shows up on about nine days in ten.

The sizing math

EUR/USD, box 34 pips, median overshoot 21. A stop 10 pips below the box low dies on a normal day. Surviving the median tag takes structure plus roughly 25 to 30 pips, which at $10 a pip is $250 to $300 per standard lot. Risking $150 a trade on that placement means 0.5 to 0.6 lots. The trader with the 10-pip stop runs 1.5 lots for the same $150 and donates it on the median morning. The dollars were identical. The outcomes were not.

Gold scales it up. Median box $47, median overshoot $33, and the contract is 100 ounces, so $1 of price is $100 a lot. The median, meaningless tag is $3,300 of adverse travel on a single lot. On a $50k account risking 1 percent, that’s a $500 budget trying to survive a $3,300 wave with the stop parked where the crowd parks it. Size the position to the allowance, or the allowance gets tested by your broker instead of by your plan.

News days

Red-folder mornings sweep both sides. CPI and NFP tag the Asian range and the London morning extremes within the same hour, and the second sweep is more often the real move than the first. On those days the box is a map of where the whipsaw lives, and the correct position size is smaller or zero.

Frequently Asked Questions

What time window is the Asian session range?

For gold and the majors, 00:00 to 08:00 UTC is the practical box; Tokyo’s session runs 9 to 5 local time. The ICT convention marks 19:00 to 00:00 New York time, which overlaps but comes out shorter. Broker server time is often UTC+2 or UTC+3, so convert the window once, keep it fixed, and recheck the conversion at the daylight-saving switches, which Europe and the US do about a week apart.

Should I trade the Asian range breakout or the sweep?

The measured base rates favor waiting. 29 of 32 recent sessions returned through the box at least partially after the initial sweep, and only 13 closed back inside by the New York close, so most first breaks are part of the collection rather than the trend. A break that holds outside through a full hourly close and then survives a retest is the version with follow-through, and it was the minority day. If you trade only the rejection side, the median overshoot, 19 to 34 pips on the majors and about $33 on gold, defines both your trigger and your invalidation.

Does this work on gold?

The mechanics are identical and the magnitudes are the point. Gold’s median box ran $47 over the sample against daily ranges of $50 to $150, and the median overshoot was $33, which is $3,300 of adverse travel on one standard 100-ounce lot. Pip-sized gold habits left over from $20-a-day markets are arithmetic from a different instrument.

What if both sides of the box get swept?

It happens on news days and on genuinely two-way sessions, and the second sweep is more often the real move than the first. Treat the first tag as information about position size rather than direction, and let the hourly behavior after each sweep argue the case: rejection back inside is the rotation read, holding outside with a retest is the trend read.

The Short Version

• The Asian box is a pair of order pools, one resting above the overnight high, one below the overnight low. Everyone builds them, nobody coordinates, and both facts matter.
• 32 of 35 recent sessions across gold and three majors swept an Asian extreme first, and 27 of those hits came inside London’s first two hours.
• The median first poke ran 19 to 34 pips on the majors and $33 on gold, which is bigger than most traders’ "safe" cushions.
• 29 of 32 came back through the box at least partly; 13 of 32 closed back inside. A close back inside sets up the rotation read, and closes holding outside with a retest set up the trend read.
• Stops go past the sweep extreme with size cut to match, so the dollar risk you started with is the dollar risk you keep.