No trade: the band stated at 00:25 UTC broke to the upside on the 18:00 UTC close

1What the market did

At issue BTCUSD had closed at 83,536.4, halfway between the two firmest zones of the hourly structure: below, the zone of lows between 83,304.1 and 83,730.4; above, the zone of highs between 84,278.9 and 84,443.6.

On 30 September, in the 13:00 UTC bar, price had swept the highs at 84,959.9 with a sweep of 1.28 times the ATR, the deepest in the window, and had printed the period high at 85,588.9. It then fell 2,284 points to 83,304.1, bounced to 84,415 without clearing the equal highs at 84,443.6, and turned back down. Volatility was wide: ATR percentile at 78 on the hourly chart.

2What was there and what was not

This reading had only the hourly chart to work with, 238 bars from 21 to 30 September. What was there and what was not:

• Economic calendar. None supplied.
• Macro data. None. No rates, flows, positioning or volume: the reading rests on structure and liquidity only, and says so.
• Structure. Hourly only: without the four-hour, daily and weekly charts, agreement across timeframes could not be assessed. Since 23 September every structure break had been classed drift or indeterminate; the only displacement-quality break in the window, on 23 September, was bearish and stale.
• Liquidity. A pool of three equal lows at 82,825.2, below the zone of lows; two equal highs at 84,443.6, at the top of the zone of highs.

3Two readings of the same facts

The upside version: the hourly chart was labelled as an uptrend, with higher lows at 82,895.8 and 83,304.1 and an upside structure break at 83,795.2. But that break was drift quality, and price had already slipped back below its level.

The downside version: the deepest sweep in the window had marked the cycle high, followed by a 2,284-point fall and a lower high at 84,415, the usual shape of distribution. But price was already inside the zone of lows, with the pool at 82,825.2 about 700 points away.

Neither version was firm enough to decide on. Drift and indeterminate breaks in alternating directions are the shape of a market rotating inside a range, and wide volatility describes broad swings rather than a coil about to break.

4Where the levels are

Every level in this table comes from the run cited at the foot of the document; the outcome below was checked on a public series.

5Why the engine stands aside, and what would change the picture

The engine concluded that the structure gave no edge either up or down, and stood aside. It wrote instead where it expected price to stay: inside the 83,304.1–84,443.6 band for the 24 hours after issue. It also stated its doubt: with wide volatility and a liquidity pool nearby, a move out of the band was far from ruled out, but nothing said which side.

The picture changes if either condition is met:

• an hourly bar closes below 83,304.1, the bottom of the band: a move out to the downside, with the pool of lows at 82,825.2 as the next level;
• an hourly bar closes above 84,443.6, the top of the band: a move out to the upside, the lower-high sequence is voided and the next level is the zone of highs between 85,236.4 and 85,588.9.

Either way the market would show itself directional rather than range-bound, and the decision to stand aside would need reading again.

6What happened next

Written afterwards, at 21:40 UTC on the same day, on Binance BTCUSDT public hourly bars, brought onto this reading’s scale with the −10.4 basis-point divergence measured at issue (see “Data and limits”).

The band did not hold: it broke to the upside. The hourly bar that closed at 18:00 UTC closed at 84,819.7, 376.1 points above the 84,443.6 top: the second of the two conditions stated at issue. Until then price had stayed inside the band on a closing basis for more than seventeen hours. It came close twice: the lowest close, 83,343.2 at 08:00 UTC, stayed 39.1 points above the floor; and between 15:00 and 17:00 UTC price went above the top during the bar, up to 84,655.9, without closing above it. The conditions were written on the hourly close, so those excursions do not count. The next level named by the reading, the zone between 85,236.4 and 85,588.9, had not been reached at the time of the check: the high after the break was 85,185.

In the outcome count this reading goes among the no-trade readings, with the outcome “out of the band”: the engine named no trade, so there is no result in R to report.

What this does not show. One reading checked on public data is no evidence about the next one. The band and the two conditions above were recorded, with a timestamp, before the outcome was known; conditions and outcome are published together. The outcome can be checked independently on public data. The recording time is not something this page can prove and has to be taken on trust: that is why the conditions are published in full rather than summarised once the outcome is known.