The London and New York Overlap as a Session
For the major pairs, the hours when London and New York are both dealing are the busiest of the day by a comfortable margin. Two of the largest centres are active at once, books are deepest, spreads narrowest, and moves more likely to continue rather than stall. If the currency day has a main session, this is it, and it is a reasonable candidate to build a range around.
Why It Concentrates
The overlap is not busy because of the clock. It is busy because the participants who move currency in size are at their desks simultaneously in both centres, and because a great deal of business that has to be transacted between European and American counterparties naturally lands there.
The practical consequences are the ones that matter to a breakout rule. Deeper books mean orders fill closer to their price, which reduces the execution cost that erodes short holding periods. More participants mean a level broken is a level many people saw break, which is the mechanism by which breakouts get followed at all.
Anchoring the Range to It
Taking the start of the overlap as the open, and measuring a range from there, produces something structurally similar to an index opening range: a burst of activity from a fresh set of participants, forming levels under good liquidity, followed by a session in which those levels can be tested.
It has an advantage over the London open considered alone, which is that the London session has already been running for hours by the time New York arrives. Direction is often established, the tone of the day exists, and the range formed at the overlap sits inside a context rather than in a vacuum. Whether that helps depends on the rule. A break in the direction of an established move is a different proposition from a break with nothing behind it.
You Are Not Starting From Nothing
The important difference from an exchange open is that the market has been trading all day. The high and low of the London session already exist, and they are levels a great many participants are watching. A range measured from the overlap start can sit entirely inside them, in which case its edges are not the significant nearby levels.
This is worth checking before treating a break of the overlap range as meaningful. If the range high sits a short distance below the London high, a break of it runs immediately into the level that actually matters, and the trade has very little room before meeting resistance established under conditions just as good. The overlap range is not the only structure on the chart, and pretending the session started fresh discards information that is sitting right there.
Releases Land in the Same Window
American scheduled data is published in the early part of the overlap with some regularity. That is not incidental, it is a defining feature of the window, and it means a range built at the overlap start is frequently a range built either just before or straight through a release.
A range formed through a release is a range formed by a single repricing event, and its edges are the extremes of a spike rather than levels anyone defended. Breaking them says almost nothing. Knowing what is scheduled and when, and either delaying the range or standing aside, is not an optional refinement here. It is the difference between measuring a session and measuring an announcement.
What Happens Later
The second half of the overlap behaves differently from the first. As London approaches its close, European participation withdraws, and the market that remains is thinner even though New York is still fully active. Moves initiated late in the window have less behind them and are more prone to stalling or reversing.
For a rule that enters on a break and holds for part of the session, this argues for taking entries early in the window rather than throughout it. A time cutoff after which no new positions are opened is a crude device, but it addresses a real change in conditions rather than an imagined one, and the change is visible in the participation itself rather than inferred from results.