Which trading session actually performs best?

Across 766 published trades, entries taken during London hours returned +16.1 pips each. New York returned +8.1 and Asia +7.8. On the same system, with the same rules, the London session was worth roughly twice as much per trade as either of the others.

Session by session

Sessions are defined by the UTC hour the signal fired: Asia 00:00–06:59, London 07:00–12:59, New York 13:00–20:59.

1. London — 341 trades. 54.5% reached a take-profit, 24.6% closed at a breakeven exit, 20.8% stopped out. +16.1 pips per trade.
2. New York — 287 trades. 43.2% take-profit, 27.2% breakeven, 29.6% stopped out. +8.1 pips per trade.
3. Asia — 121 trades. 47.9% take-profit, 19.8% breakeven, 32.2% stopped out. +7.8 pips per trade.

London wins on both halves of the equation: it reaches targets more often and stops out less often. New York is the interesting case — it produces plenty of trades, but a materially higher share of them stop out, and more of the rest end up as breakeven exits rather than reaching a target.

The sample we will not draw a conclusion from

There is a fourth window — 21:00 to 23:59 UTC, after the New York close. It looks terrible: 52.9% stopped out and barely a pip per trade.

It is 17 trades. That is not enough to conclude anything. 17 trades could look like that by chance on a system with no time-of-day effect whatsoever, and any article telling you otherwise is selling you something. We restrict new entries in that window for reasons of liquidity rather than because of this number, and we are reporting it here only so the totals reconcile.

The same caution applies in weaker form to Asia at 121 trades. London at 341 and New York at 287 are the two numbers worth acting on.

What to do with this

If you cannot take every signal, take the London ones. That is the honest version of the advice, and it follows from the largest and cleanest sample in the table.

It is worth being clear about why this is plausible rather than just true-in-our-data. London hours carry the heaviest volume in the currency market, and the European open produces directional moves that a trend-following system can hold. Thin conditions produce moves that start and fail. That is a mechanism, not just a correlation — but five months is five months, and we will revisit it with more trades.

Check this yourself

Every trade is published with the exact UTC timestamp it fired at, so any of the above can be recomputed from our record rather than taken on trust.

Browse the full record, or read why we separate take-profits from breakeven exits in every figure above.

Figures recomputed from the live record on every page load — 766 closed trades since 2026-04-20, current as of 14 September 2026. Past performance does not indicate future results.