Why I only trade the London-NY overlap on AUDUSD.

I trade one window. London-New York overlap. Roughly 8pm to midnight my local time in Sydney, which is 8am to noon UTC, which is 3am to 7am New York. Outside that window I am almost always not at the chart. Inside that window I do almost all of my work for the week. The reason is structural and the structure is worth writing down.

A four-hour window is when the two largest forex sessions are simultaneously open. London is the largest single session in forex by volume, accounting for roughly forty percent of all spot turnover. New York is the second largest, at around twenty percent. The four hours where both are open are the four hours where actual price discovery happens. Spreads tighten because liquidity is deep. Stops are less likely to be cleaned out by thin-book moves. Order flow has direction because the institutions on both sides of the Atlantic are simultaneously active.

Outside the overlap, the math changes. The Asian session, in isolation, has roughly half the participation of London. The London-only window before New York opens has decent volume but is more directionally noisy because it is dominated by European positioning ahead of US data. The New York-only window after London closes is dominated by position-squaring and end-of-day flow, which is informative but rarely sets up the kind of clean continuations my system is built around.

The pure session-by-session analysis on AUDUSD specifically is even starker. AUD has its biggest moves during the overlap because the AU central bank statements that drive AUDUSD repricing are released during the Asian session, but the price action that resolves those repricings happens hours later when London and New York both react. A morning Reserve Bank of Australia statement in Sydney sets a level. The actual move into or away from that level usually plays out at the London open and continues into the overlap.

Spread is the second reason. AUDUSD's median spread on a typical broker during the overlap is around 0.6 to 0.8 pips. During the dead Asian-only afternoon, the same pair trades closer to 1.5 to 2.0 pips, sometimes worse. Two-pip spreads on a forty-pip move are five percent of the trade. Compounded across hundreds of trades, the spread cost outside the overlap eats a serious chunk of edge that simply does not exist for trades taken during it.

Slippage is the third reason. Stops outside the overlap fill worse on average because the order book is thinner. A hundred-pip stop fills at minus one hundred and three on a quiet Asian afternoon and fills at minus one hundred and one during the overlap. The difference is the depth of resting orders against you. Three pips of average extra slippage on losers, multiplied by the loss frequency, is again a real cost that disappears for window-only trading.

The other reason I trade one window is energy. Trading is a decision job. Decision quality erodes with hours at the chart. A four-hour focused window where I am rested, present, and watching one or two pairs is the highest quality work I do. Trying to extend the window to six or eight hours produces lower quality decisions in the bonus hours, and the lower quality decisions cost more than the marginal trades they generate.

The objection I hear most is that limiting yourself to one window is leaving money on the table. Some traders make money in the Asian session. Some traders specialise in New York close. The math on those traders shows that what they have done is built a different system, optimised for the specific conditions of those windows. They are not just taking the same setups during off-hours. They are running a different rule set. If you are using the same rule set across all sessions, the rule set is implicitly tuned for one of them, and trading the others is running an untested system on real money.

My system is tuned for the overlap. The continuations I take resolve quickly because liquidity is deep. The pullbacks I enter on are clean because the order flow is directional. The stops I set hold because the move is moving for a reason. None of those properties apply during the Asian session, when continuations stall and pullbacks chop and stops get picked off by thin-book scans. So I do not trade then.

The discipline cost is small. I open the chart at 8pm, I trade until midnight, and I close the laptop. The opportunity cost of skipping every other window is the trades I would have lost taking them, plus the time I would have lost waiting for them, plus the energy I would have spent watching for them. Subtracting those costs leaves the four-hour window I actually trade, which has done all the work for the year.

If you are running a system that is breaking down on average and you do not know why, look at when your trades happen. Filter your last hundred trades by hour of the day. The window that pays will be visible. The windows that bleed will be visible. Most retail traders are paying for a small edge during one window and then giving it back during the others.