Does London break gold's overnight range? 130 days counted

The overnight range here is gold's high and low from the start of its trading day to 08:00 UK: from the 23:00 reopening after the daily break, or from the Sunday 23:00 open on a Monday. London is 08:00 to 17:00 UK the same day, so both stretches are nine hours long. For each of the 130 days we checked whether London traded above the overnight high, below the overnight low, both or neither, and how things stood at noon.

London against the overnight range

Counting the 15 days that went through both, London went above the overnight high on 66 days, 50.8%, and below the overnight low on the same 66. The two sides came out level in this sample, 51 days each. No day went through both sides before noon; all 15 of those reached the second side in the afternoon.

At noon the overnight range had held on 48 days, 36.9%. By 17:00 that was down to 13 days, 10.0%.

The overnight range against the London range

The overnight range had a median of 537.4 pips, about $54 of price, and the London session 607.6 pips, about $61, so London covered a little more ground in the same nine hours. The averages were 617.1 pips overnight and 687.8 in London. The widest London session covered 1,560.4 pips.

The widest overnight range, 2,471.4 pips on 2 April, was also that whole day's range: gold set both its high and its low that day before London opened. Gold's volatility by hour shows where the day's high and low fell across all 130 days; the London morning, 08:00 to 13:00, set the high on only 11 of them.

Against the previous day's high and low

Over a whole trading day, gold went above the previous day's high only on 44 of 128 days (34.4%), below the previous low only on 55 (43.0%), through both on 14 (10.9%), and stayed inside the previous day's range on 15 (11.7%). So it traded beyond the previous day's high or low on 113 of 128 days, 88.3%.

The previous low gave way on more days than the previous high: 69 (53.9%) against 58 (45.3%). Six months is one sample, and that balance can tip the other way in another period. There are 128 days rather than 130 because the first full day, 27 March, and the Monday after Good Friday, 6 April, have no full trading day before them.

What the counts mean for a signal trader

This is not a strategy, and the counts do not test one. We counted whether the price reached a level, not what it did next: a break of the overnight high says nothing here about whether gold kept going or turned back. What the counts do show is that the overnight high and low were not walls: each gave way on about half the days, and a stop just beyond either was within reach of an ordinary London session.

Gold also traded beyond the previous day's high or low on 113 of 128 days. So one of these levels gave way on most days in this sample, and a trade with its stop just beyond one should be sized on the assumption that the stop is hit; the gold lot size calculator sizes a trade from its stop. If you want the previous day's levels on your chart, the pivot point calculator works from the previous day's high, low and close. Our results page lists each closed signal with its entry, stop, targets and outcome, so you can see where stops sat on the days they were hit.

Where the numbers come from

One-minute prices: GetData free sample, MIT licence, 26 March to 25 September 2026, grouped into 130 full trading days of 17:00 to 17:00 New York time, each labelled by the date it ends. Overnight runs from the start of that trading day to 08:00 UK, London from 08:00 to 17:00 UK, and the noon count uses 08:00 to 12:00. UK time was British Summer Time from 29 March.

The first minute after each weekend, daily break or other pause of more than five minutes is left out of highs and lows, because its opening price is carried over from before the pause. Ranges are high minus low, typical values are medians, and a gold pip is $0.10 of price. The method is on the markets page, and gold's other figures on XAU/USD.