MEASURED · 7 SEPTEMBER 2026
Do round numbers work as gold support and resistance?
Not measurably. Across 35,002 approaches to price levels in a year of XAUUSD, round levels turned price back about as often as any other price, and the gap is small enough that we can put a number on how small. Every gold analysis page publishes a list of levels, and those lists cluster on round numbers.
The claim: price turns more often at round numbers
Any “XAUUSD support and resistance” page gives you a list of prices. They are stated as fact, they expire within a day, and they cluster heavily on multiples of $50 and $100. The implied claim is that these prices are special, and that price is more likely to turn there than elsewhere.
That claim is testable, and as far as we can find nobody has published the test.
How we measured it
We used 70,952 five-minute XAUUSD bars from a live MetaTrader 5 feed, 2025-07-31 to 2026-08-04, with gold trading between $3,281.59 and $5,595.35.
An approach is the first bar where price trades into a band around a level, having been outside it. From there we look one hour ahead and record one of two outcomes. Price either rejected the level, ending back on the side it came from, or broke it, closing through and staying through. We ignore approaches within 24 bars of the previous one, so price oscillating around a level does not add dozens of correlated readings.
The control decides everything. For every round level we also test two deliberately unremarkable ones, offset 30% and 70% into the same increment: prices nobody draws on a chart. If roundness carries information, round levels must reject more often than these. If they do not, the levels everyone publishes are decoration.
Round levels rejected price about as often as arbitrary ones
| Level size | Approaches | Round rejected | Control rejected | Difference |
|---|---|---|---|---|
| $10 | 20,707 | 54.2% | 53.9% | +0.3 pts |
| $25 | 8,233 | 57.1% | 56.5% | +0.6 pts |
| $50 | 4,016 | 60.9% | 60.5% | +0.4 pts |
| $100 | 2,046 | 69.8% | 69% | +0.8 pts |
| All pooled | 35,002 | 56.6% | 56.1% | +0.45 pts |
Round levels rejected price 56.6% of the time. Levels chosen specifically because they are unremarkable rejected it 56.1% of the time.
The difference is +0.45 percentage points, with a 95% confidence interval running from −0.65 to +1.55. That interval contains zero, on a sample of 35,002 decided approaches. Taken individually, none of the four level sizes produced an edge distinguishable from chance either.
The sample is large enough that a real effect larger than about 1.5 percentage points would have shown up. Roundness carries no measurable information about where gold turns.
One number in that table is misleading, and it is ours
Rejection climbs steadily with level size, from 54.2% at $10 up to 69.8% at $100. That does not mean bigger levels are stronger. It is an artefact of our own method.
The band scales with the increment, so at $100 price must travel $5 to count as a break, against $0.50 at $10. A wider band means more moves fail to clear it, which raises the rejection rate. Round and control levels share the same band at each size, so the comparison within each row is sound. The trend from row to row measures our band width and says nothing about the market. We point it out because someone would otherwise quote it as a finding.
Why round numbers feel like they work
Price reverses at most levels. A 56% rejection rate at an arbitrary price is what an oscillating market does over an hour, wherever the line is drawn. Every level you draw will appear to work slightly more often than it fails.
Round numbers feel different because of memory. A reversal at $2,700 gets annotated. An identical reversal at $2,683.50 does not, because nobody drew a line there. The reversals are equally common, and only the noticing is selective.
We did not test levels taken from market structure, such as a prior swing high, a session extreme or the edge of an overnight range. That is a different question. We have tested session ranges separately and found follow-through of 51.7%, which is also close to a coin flip. What this study rules out is choosing a level because it ends in two zeros.
Set the stop from volatility instead
If the level does not decide the trade, the stop distance does. We measured that separately: ordinary movement hits a $2 stop on gold 81% of the time within an hour, and gold’s median daily range is around $76. A stop placed just beyond a round number, on the theory that the number will hold, sits exactly where noise reaches. The numbers are in the cost of scalping gold.
Replace the level list with a size calculation. Decide the stop from volatility, then size the position so that stop is affordable. The stop loss calculator does it against your own account, and our entry on support and resistance covers levels drawn from market structure instead of arithmetic.
Method, limits, and the data
This is one symbol, one year, one broker feed and one horizon. A different holding period would change the absolute rejection rates: a longer horizon lets more levels break and pushes every number down. It would move round and control levels equally, and that comparison is what the conclusion rests on.
The band is the judgement call. Too tight and genuine touches are missed. Too wide and unrelated price action counts as an approach. We used 5% of the increment either side and applied it identically to both groups. The script is research-round-numbers.py. It runs on the same published bar archive as our other studies, and the full output, break counts included, is in the JSON behind this page.
Underlying data, method and licence: the XAUUSD archive, published under CC BY 4.0. This study has its own record with DOI 10.5281/zenodo.22637578, mirrored on Hugging Face, so the counts behind every figure above are downloadable. If you disagree with the band, the horizon or the control choice, the inputs are public and you can rerun the test with your own.
Common questions
Do round numbers act as support and resistance on gold?
Not measurably. Across 35,002 approaches to price levels on XAUUSD over a year, round numbers turned price back 56.6% of the time and arbitrary non-round levels turned it back 56.1%. The difference of 0.45 percentage points has a 95% confidence interval from −0.65 to +1.55, which contains zero. None of the four level sizes tested showed an edge distinguishable from chance.
What are the current support and resistance levels for XAUUSD?
We do not publish a list. A level quoted today is stale tomorrow, and our measurement suggests the exact number matters less than the industry implies. Price reversed at roughly 56% of approaches to any level, round or not. The useful question is how much room your stop needs, and volatility answers that better than a level list.
Why does price seem to react at round numbers then?
Because price reacts at most levels. Our control levels were prices offset 30% and 70% into each increment, picked because nobody watches them, and they produced 56.1% rejections. Memory does the rest: a reversal at $2,700 gets marked on the chart, and an identical reversal at $2,683.50 does not. The reversals are real. Giving the credit to the round number is the mistake.
Does this mean support and resistance does not work?
No. The study tests one widely published claim: that round numbers are stronger than other prices. Levels drawn from market structure, such as prior swing highs and lows, session extremes and opening ranges, were not measured here and are a different question. What this rules out is picking a level because it ends in two zeros.