Support and resistance on gold: real levels against random levels, 2008 to 2026

The short answer

On 18 years of gold, the levels traders draw (yesterday's high and low, last week's high and low, recent swing points) are not magic bounce lines on their own: price turned at them about as often as at lines drawn close to them. Their value is as places to watch, targets and stop-loss zones. And one finding stands out: last week's high and low were broken through clearly more often than nearby lines, so they attract breakouts.

The idea

Traders draw lines at prices where the market turned before: yesterday's high and low, last week's high and low, the last swing high and swing low. The belief is that the price will "bounce" off these lines again. These are called support (below the price) and resistance (above the price).

How we tested it

We took 18 years of gold prices (2008 to 2026) and found every time the price reached one of these lines for the first time. Then we checked: did the price turn back (by half of a normal day's move) before it broke through the line (by half a day's move)?

Then the important part: we did the same with fake lines. We took each real line and moved it a random small distance up or down. These fake lines mean nothing. If the real lines matter, the price must turn at them more often than at the fake ones.

What happened

How often the price turned back at the line:

• Yesterday's high: real line 44.7%, fake line 44.6%
• Yesterday's low: real 47.5%, fake 48.4%
• Last week's high: real 41.2%, fake 45.4%
• Last week's low: real 41.8%, fake 46.9%
• Last swing high: real 42.9%, fake 44.6%
• Last swing low: real 47.1%, fake 49.4%

Yesterday's levels and the swing points behaved much like the lines next to them. Last week's high and low stand out: the price broke through them more often than through nearby lines, and that difference is too big to be luck. We also checked Bitcoin (2018 to 2026) and saw the same: yesterday's high 47.3% against 47.2% for the fake line, last week's low 42.2% against 49.0%.

What it means for you

• Use a line as a place to watch for a setup, together with a confirmation, not as a reason to buy or sell by itself
• Last week's high and low were broken more often than respected. On gold, a break of them deserved more respect than a bounce
• What protects you is your exit plan: a stop loss at the price where your idea is clearly wrong, and a trade size that fits your risk
• An obvious line is where everybody puts their stop loss. Yours does not have to sit one point behind it

This test only checks lines by a fixed rule. An experienced trader may see more on a chart than a simple rule can, and that part we cannot test.

Want to run this test on your own chart? Our free indicator ICT Liquidity Levels Honest Stats draws the previous day and week highs and lows and the Asian range, and shows next to them how often price turned there compared with random lines: https://www.mql5.com/en/market/product/201465 (MT4 version: https://www.mql5.com/en/market/product/201466)

If you trade from lines on the chart, the free TrueRisk Trade Panel lets you drag your stop-loss line and works out the lot size from your risk % while you drag: https://www.mql5.com/en/market/product/200808

Details for those who want them

• Data: XAUUSD hourly from Dukascopy, 2008 to 2026, and BTC 2018 to 2026
• Only the first touch of each level counts; on gold there were 4,616 first touches of yesterday's high
• Fake lines: the real line moved by a random 0.2 to 0.5 of the average daily range

Have a strategy, an indicator or an EA you want tested like this? Write it in the comments.

If one of our free indicators or tools helped you, a short review on its Market page helps other traders find it.

Next week: a strategy that did hold up on data it was not picked on, buying dips in stock indices.