01

The range before the breakout

This reading uses the OANDA XAUUSD H1 file supplied with the app project, from 30 June to 6 August 2026. A sharp decline reached 3,942.125 on 30 June; price then rallied to 4,202.660 on 6 July and returned to the 3,960 area. For this example, 4,202.660 is the observed range ceiling and 3,942.125–3,960.320 is the lower support area. Those levels are measured from completed bars. “Accumulation” remains a hypothesis because price alone cannot establish who bought or why.

02

An early entry from the support test

On 17 July, the 15:00 H1 bar fell to 3,964.155 and closed at 3,966.320. The next bar had a wide range: it made a 3,959.800 low, then closed at 3,997.380. A faster rule uses that completed close as the long reference, before the range breakout; a more cautious rule waits for the 17:00 close at 4,001.375 for follow-through. These are price-based rules to replay, not fully confirmed Wyckoff signals: traded volume is absent and the test bar did not narrow clearly. Neither is a Spring, since the low did not undercut the earlier 3,942.125 campaign low.

03

Invalidation and first objective

The support-test idea is wrong if price breaks below 3,959.800 after the entry; a practical stop needs a pre-set buffer for spread and execution costs, so 3,959.800 is the invalidation reference, not a guaranteed fill. The first objective is the opposing range edge at 4,202.660. Price first touched it on 5 August at 15:00 (high 4,213.540), while that candle closed back below at 4,184.025. A limit exit at the boundary would have been reachable before the 16:00 close at 4,206.180 confirmed a breakout. From the 3,997.380 reference entry to the range edge is 205.280 points; to the test low is 37.580 points, about 5.46R before costs and any stop buffer.

04

What this example says about Wyckoff

Here the useful part is not a label on the later breakout: the range gives a place to watch, the return to support gives an earlier candidate entry, and the test low gives a price that cancels the idea. That made the method actionable before the move to the range ceiling. The trade worked in this one historical sample; one winner cannot show that the reading has an edge. To test that, apply the same definitions to every qualifying range, including failed tests, and compare with a simple support-bounce rule under the same costs.

05

What the file can and cannot confirm

The H1 file contains OHLC and tick volume, but its VOL column is zero. This article uses only price; it does not treat tick volume as total traded gold or claim effort-versus-result confirmation. The bars can show a sharp decline, a rebound, repeated support tests and a later break. They cannot prove accumulation, identify a Composite Man or guarantee another range will behave the same way.