Range expansion after narrow-range days

[Shadow Trader]

A narrow-range day is one with an unusually small high-to-low range, compared with the recent average. The NR7, the narrowest range of the last seven days, is a well-known version. The idea is that compression is followed by expansion, and that the next day may bring a bigger move.

The statistic is easy to test. For each day, check whether it's the narrowest in a lookback window, then measure the next day's range relative to average. Results in many markets show a mild tendency for ranges to expand after compression, which fits the observation that volatility clusters and mean-reverts.

Direction is a different matter. Compression doesn't say which way price will break. Strategies typically place orders on both sides of the range and let the market choose, accepting whipsaw risk.

For scalpers, the idea provides a day-level filter: on days following a narrow range, expect more activity and adjust size and targets accordingly. It doesn't give entries by itself.

Combining the signal with a time-of-day filter often improves it. Expansion after a narrow day tends to show up at the major session opens, so ignoring the quiet hours concentrates the strategy on the periods when the compression is most likely to resolve.

Also consider the scheduled calendar. A narrow range before a big data release or central bank meeting is a very different thing from a narrow range in an event-free week. The first is waiting for a catalyst that is already on the schedule, while the second may just be a dull week.

Has the narrow-range tendency shown up in the pairs traded, and was it strong enough to affect the plan for the next day?

[FXS London Scalper]

A narrow-range day is one with an unusually small high-to-low range, compared with the recent average.

NR7 on EURUSD pairs nicely with the calendar. A narrow day before ECB or US payrolls is compression with a known catalyst. A narrow day before an empty calendar often just produces another narrow day. Splitting NR7 results by whether high-impact data follows would probably sharpen the idea a lot.