When does the day's high and low actually form?

Every platform in this category sells this report — high and low by session, high and low by weekday — and it is the clearest case we have found of a true number that means nothing. We ran it on 73,001 trading days across 19 markets. The first answer we got was wrong, and the way we found out is the whole article.

The answer that looks like a finding

Cut each day at midnight UTC, record which hour held the day's high, and the opening hour wins in 15 of 19 markets. That is a real count off real data. It would make a confident chart. It is also an artefact of two things that have nothing to do with the market.

The control

Ours keeps every bar in its own hour — so the day's volatility profile is exactly as it was — and randomises only its direction: each bar's move is flipped about its own open with probability one half, the path is rebuilt, and the extreme is relocated. Anything that survives is not "the London hour is big", because the London hour is still just as big in every one of the 30 draws.

Against that control, the opening hour is the busiest in 17 of 19 markets — more than we actually observe. The famous finding is not merely unsupported. It is slightly weaker than chance.

The test that settles it: move the boundary

If the concentration belongs to the start of the day rather than to midnight, then moving the boundary has to move it too. So we re-ran everything on a day beginning at 22:00 UTC — the FX rollover, and the boundary your broker actually uses.

What actually survives

104 of 912 hour-tests clear our significance bar against 2.5 expected — 41.6× — so there is real structure in where the extreme lands. But the biggest single block of it is a negative, and it is about the hours this site already tells people to avoid: 46 of those 104 flags sit at 21:00–23:00 UTC and say the extreme lands there less often than the control expects, against 2 saying more. The rollover does not make new highs and lows. It is thin, it is the widest spread of the day, and it does not even produce the extremes.

And the weekly version is mostly a trend

Counted raw, the week's high lands on Fri in 15 of 19 markets and the week's low on Mon in 15. Against the same control only 6 of 19 clear even a weak bar on the Friday high — led by gold and the US indices, which is what an asset that rose across the sample is obliged to do. The correlation between a market's own trend and its Friday-high strength is 0.372: positive, and too weak to carry the story alone. ⚠ And nineteen markets agreeing is not nineteen observations — US30 and SPX500 correlate at 0.94, so a head-count here overstates its own evidence.

What we would actually say

• Check what your boundary is doing before you believe a time-of-day result. This one flipped every market's answer.
• Uniform is almost never the right null for a "when" question. Extremes concentrate at the ends of a path whether or not anything is happening.
• The overlap and the US close are real. The rollover is real too, in the opposite direction.
• It tells you when, never where. A session that produces the high on 13% of days produces it somewhere else on the other 87%, and none of this says anything about the level.

Full method and every market's table in section 19 of the evidence page. The data is free to download if you want to redo it.

Other reports this category sells, tested

Every claim we have tested, including the ones that held, is on the reports index.