Spread is a schedule
The London–New York overlap concentrates the deepest book of the day into roughly four hours. Spreads on the majors compress, depth rises, and the market impact of a given size falls. The Tokyo–London gap is the opposite. None of this requires prediction — it is a property of the calendar.
Where the money is
For a strategy already trading, shifting discretionary execution into the overlap is one of the few genuinely free improvements available: same positions, same signals, lower cost. On a strategy turning over frequently, execution cost is often larger than the signal edge, which makes this the higher-leverage thing to fix first.
The trap
The overlap is also when scheduled data lands, and a tight spread quoted a second before a release is not a spread you will get. Any scheduler that routes to the overlap needs an economic calendar blackout, or it will systematically execute into the worst moments of the best hours.