Ask the desk: AUDUSD into Asia session

[LondonScalper]

I'm planning to add a small AUDUSD book in the Asian session over the winter, and before I spend money learning the hard way I'd like to hear from those of you who already trade it.

My situation: London-based, so the Sydney and Tokyo morning sits at roughly 23:00 to 04:00 for me, depending on the time of year. I'm not willing to trade every night. Two or three nights a week, maybe 23:30 to 02:00, would be my limit.

What I've done so far is watch the pair for about three weeks without trading, and pull six months of M1 data to measure ranges. My rough findings: the first hour after the Sydney open is quiet, then the Australian data releases (usually at 00:30 London in winter, 02:30 in summer) can move it 20 to 40 pips. The Chinese data releases occasionally do more.

My questions for those with experience:

What do you see as the realistic spread on a raw account between 23:00 and 01:00 London? My demo shows 0.3 to 0.6, but demos can flatter.

Do you trade the Australian release itself, or wait for the dust to settle?

How do you deal with the RBA days? Is it worth even trying?

And the big one: how do you keep a London morning routine intact after trading until 02:00? I'm 45, not 25, and I'm wary of what sleep debt does to my judgement at 08:00 the next day.

[Shadow Trader]

On the spread, my experience on raw accounts between 23:00 and 01:00 London is closer to 0.4 to 0.9 on AUDUSD, and it widens around the rollover just before 22:00 and for a while after. I'd check your live account across a week before deciding, because demo quotes in that window are usually tidier than live.

I don't trade the Australian releases themselves. I wait at least five minutes and look for the first pullback. The jobs numbers in particular can move 30 pips and then give half of it back while the spread is still wide.

RBA days I'd skip for the first few months. The statement wording matters more than the decision, and reading it fast while trading isn't easy.

On sleep, I tried something close to your plan for a winter. What worked was making the nights fixed, say Tuesday and Wednesday only, and not trading London on the following morning at all. Trying to do both on the same day lasted about two weeks before my London trades got noticeably worse.

Have you measured how often the 23:30 to 02:00 window gives you more than one clean setup?

[Fairman]

What do you see as the realistic spread on a raw account between 23:00 and 01:00

On the spread question, the thing to watch is that the window starts not long after the daily rollover. Most brokers roll at 17:00 New York, which is 22:00 London for most of the year, and AUDUSD spreads can stay wide for anything from a few minutes to the best part of an hour while liquidity providers come back. In the weeks when the UK and US clocks are out of step, the rollover moves to 21:00 London. By 23:00 it's usually back to normal, but the first part of the window is worth sampling separately. On releases, the RBA decision is at 14:30 Sydney time, which is 03:30 London in the northern winter, after a 23:30 to 02:00 book has closed. China's data, around 09:30 to 10:00 Beijing time, is 01:30 to 02:00 London in winter and does sit inside it.

[Shadow Trader]

Fairman's point on the rollover window is the filter I wish I'd had earlier. I now skip AUDUSD if the spread at 23:30 London is still above 1.0, even when the demo looks fine, because that first half hour after rollover is where thin Sydney liquidity sits.

The other skip is range based. If the first 45 minutes after the Sydney open cover less than about 8 pips of ADR, I treat the night as dead and leave the platform. Those sessions paid almost nothing last winter and still cost spread.

Australian data at 00:30 still gets a look, but only when that ADR filter has already passed. Under 8 pips of early range, even a jobs print tends to whip rather than trend.

China's 01:30 to 02:00 London prints can sit inside your book, so I mark them the night before. I also cut size in half if I'm still in when they land.

LondonScalper, have you tried a hard spread cutoff at the start of the window? Or do you just wait it out until the quotes look normal?

[Shadow Trader]

Watch tick-level quotes for a while and the bid and ask don't always move together. Sometimes the ask ticks up while the bid sits still, widening the spread by a fraction, then it snaps back. That's flicker: rapid, small quote changes that aren't trades.

Most of it is noise from liquidity providers refreshing prices. Different feeds show different amounts of it, which is why the same minute can look different across brokers.

Some scalpers read it as a clue anyway. One-sided repricing, with the ask lifting and the bid late to follow, may show dealers adjusting to something. Others treat it as meaningless. The honest answer is that it depends on the feed, and the evidence is mostly anecdotal.

Where it matters for sure is execution. A market buy during a flicker fills at the momentarily wide ask, so the entry is a fraction worse than the chart implied.

Do you watch the bid and ask lines separately on the chart, and has the gap between them ever warned you off an entry?

[FXS Prop Desk]

Rollover timing is the key detail for AUDUSD into Asia, Fairman. In the weeks when the clocks are out of step, a rule written in London time can land right in the wide-spread window without anyone noticing. A simple safeguard is to make the start of the trading window depend on the spread rather than the clock: entries allowed only once the spread has narrowed below a chosen threshold. That works regardless of the season and also handles the nights when liquidity returns more slowly than usual.