Previous day high and low on gold: what 115 days of data show.
Yesterday’s high and low are the most watched levels on any gold chart. We measured how often they were taken, what happened in the next hour, and how long equal highs and lows survived.
By XAU Setups Desk
Published 3 November 2026
8 min read
Short answer: across 115 trading days in 2026 gold traded below the previous day’s low on 57% of days (65 of 115 days · 95% range 47–65%) and above the previous day’s high on 46% (53 of 115 days · 95% range 37–55%). After the low was taken, price closed back above it within an hour 85% of the time (55 of 65 takes · 95% range 74–91%); after the high was taken, back below within an hour 68% of the time (36 of 53 takes · 95% range 55–79%). Equal highs and lows were even less durable: 81% were taken within 24 hours.
The previous day high (PDH) and previous day low (PDL) are on almost every gold trader’s chart, so stop orders cluster just beyond them. That makes them classic liquidity levels. Instead of repeating the theory, we counted what gold actually did with them.
Previous day high and low: the numbers
Bars show the share; the white whisker is the 95% range. The low/high difference partly reflects the window: gold fell from about 4,630 to about 4,195, which favours lows being taken.
The low was taken more often than the high in this period (65 vs 53 days out of 115). With these sample sizes the two ranges overlap, so treat the gap as a tendency of this downtrend, not a rule.
Both in one day happened on 16 days (16 of 115 days · 95% range 9–21%), the “outside day” where both pools are cleared.
Quick reclaims at the low: 55 of 65 PDL takes (85%) closed back above within an hour, compared with 36 of 53 (68%) at the high.
What it looks like
Illustration, indicative prices. The level starts at the candle that made yesterday’s low. Today a wick trades below it (red ring) and the 15-minute candle closes back above: the take and the close back. In our data that close back came within an hour in 55 of 65 previous-day-low takes (85%, 95% range 74–91%).
Draw the level from the actual candle that made yesterday’s low, not from a rounded number. A few dollars matter: in our Asian-range data the median overshoot before a close back was about $3 (30 pips).
Equal highs and lows
Of 366 equal-high and equal-low pools, 81% were taken within a day. The median wait was about 1 hour 45 minutes. Pools often overlap in time, so the real uncertainty is wider than the whiskers.
Equal highs and lows (two or more swing points at almost the same price) look like strong resistance or support to many traders. In our data they behaved more like magnets: of 366 pools, 298 were taken within 24 hours (298 of 366 pools · 95% range 77–85%) and 328 within 72 hours (328 of 366 pools · 95% range 86–92%), with a median wait of about 1 hour 45 minutes. Many pools overlap in time, so these events are not independent and the real uncertainty is wider than the ranges suggest.
How to use this (education, not a rule)
Treat PDH, PDL and equal highs and lows as places where orders sit, not as walls.
If you place stops, avoid the most obvious spot just beyond these levels, and size the position so a wider stop risks the same amount (see the lot size guide).
In our method a level being taken is where we start watching, never the entry. The sequence is fixed: 1 · sweep of the level; 2 · CHoCH, a candle body closing beyond the external swing in the new direction (a wick does not count); 3 · retest of the broken level or the order block of the CHoCH leg; only then is the idea confirmed, and the order block of the CHoCH leg is the zone to study. No break and retest, no trade. This is a method for study, not a call to buy or sell.
Illustration, indicative prices. The previous day high is taken (green ring). A candle body then closes below the external swing low (blue ring): the CHoCH. Price retests the order block of that leg (gold box, red ring). That sequence is the confirmation. If no body closes beyond the swing, or price never comes back to retest, there is nothing to do.
How we measured it
Data: 15-minute XAU/USD bars from our scanner’s price feed, 115 complete trading days from 30 April to 9 October 2026 (one fewer than the Asian study, because the first day has no previous day). No days were picked or dropped for their result.
Clock: New York time; the gold trading day runs from 18:00 to 17:00 New York. New York was on daylight time (UTC−4) in this period.
Taken: price trades beyond the level during the trading day. Closed back: a 15-minute candle closes back on the original side within four candles (one hour).
Equal highs and lows: two or more swing points within 12 pips ($1.20) of each other, at least one hour apart, with no trade through the level in between.
Ranges: 95% Wilson score intervals, assuming independent events; related events make the true range wider.
Previous day low taken on 57% of days (65 of 115), high on 46% (53 of 115).
After a PDL take, a close back above within an hour came 85% of the time (55 of 65).
81% of 366 equal highs and lows were taken within 24 hours.
Levels show where orders sit. Confirmation needs a CHoCH by body close, then a retest of the order block of that leg.
Frequently asked questions
How often does gold take the previous day high?
In our 115-day sample gold traded above the previous day high on 53 days, 46% (95% range 37–55%).
How often does gold take the previous day low?
On 65 of 115 days, 57% (95% range 47–65%). After doing so it closed back above within an hour on 55 of those 65 days (85%).
Are equal highs and lows support and resistance?
They are better understood as liquidity. 81% of the 366 equal highs and lows in our data were taken within 24 hours.
Should I trade every previous day low sweep?
No. A sweep is only the first step. Our method waits for a CHoCH by candle body close beyond the external swing and a retest of the order block of that leg. No break and retest means no trade. This is education, not a recommendation.
Will these percentages stay the same?
Not necessarily. They describe one period of 2026 in which gold trended down. We re-run the study as data grows and update the date on this page.
Educational chart analysis only · not financial advice. Statistics describe past data from our own feed (30 Apr – 9 Oct 2026) and are not a forecast. Charts marked as illustrations use indicative prices. Trading gold on margin carries a high risk of loss.