Level Scope › Ours, measured
Volatility outlier
A bar far larger than this instrument usually prints at this hour. The hour is the whole point: the same range is unremarkable at the London open and extraordinary at 03:00.
What it looks like on a chart
A mark under the bar. No level.
| Timeframe | Bars of history it reads | Settings | Events, last 30 days |
|---|---|---|---|
| M15 | 2880 | median 15-minute range per UTC hour, built from the last 30 days; threshold 4x | 416 |
| M30 | 2880 | median 15-minute range per UTC hour, built from the last 30 days; threshold 4x | 206 |
| H1 | 2880 | median 15-minute range per UTC hour, built from the last 30 days; threshold 4x | 57 |
| H4 | 2880 | median 15-minute range per UTC hour, built from the last 30 days; threshold 4x | 11 |
Not on D1. It keys its baseline on the UTC hour, which needs more than one bar per hour to mean anything.
What we measured
The denominator is the median bar range per UTC hour per instrument — a table we already publish, and one nobody else carries. The baseline is rebuilt from the chart's own history so the detector cannot drift from a snapshot it does not control.
What it does not claim
The baseline period is part of the definition and it does not converge. Swept over ten days at baselines of 6, 15, 30, 60 and 90 days, the event count went 4/3/4/1/1 on EUR/USD and 15/9/18/29/22 on USD/JPY — in opposite directions, because one instrument's volatility was rising over the sample and the other's falling. There is no lookback at which this settles down, so the one we use is named rather than defended.
Where the definition comes from
ours — median 15-minute range by UTC hour