Technical Analysis
Glassnode reports Book Churn metric for ten-minute intervals
Glassnode has published Book Churn, a metric that counts how often an exchange's order-book price levels are revised in each ten-minute interval and scores those intervals versus the instrument's trailing thirty days. According to Glassnode, the metric is available in Glassnode Studio, was tested on Binance BTCUSDT and ETHUSDT back to 2020, and a reading two standard deviations above its norm is followed by an hour of above-median volatility about 63% of the time versus a 50% baseline.
What Glassnode published
Glassnode published a new metric called Book Churn. According to Glassnode, the metric counts how often the price levels of an order book are revised in each ten-minute interval and scores each interval against the same instrument's trailing thirty days at the same hour of day.
Glassnode says the score is available in Glassnode Studio. Glassnode tested the metric as a volatility indicator on Binance BTCUSDT and ETHUSDT perpetuals back to 2020, in-sample, against controls matched on year, trailing volatility and time of day.
| Median ten-minute level updates | BTCUSDT 607,000; ETHUSDT 548,000 | Median ten-minute level updates, Glassnode |
|---|---|---|
| Share of scored intervals with zscore ≥ 2 | 4.3% / 4.6% | BTCUSDT / ETHUSDT, Glassnode |
Each figure as the source states it.
What the metric measures and its limits
Glassnode defines Book Churn as the number of rows an exchange's depth feed publishes per ten-minute interval, counting one row per revised price level. According to Glassnode, trades count toward the total because a fill changes the size at a level.
Glassnode lists three limits. One feed message can carry many level updates and multiple changes to the same price within a message collapse into a single row. Absolute counts depend on the exchange's feed format and update frequency and therefore require recalibration across exchanges. Glassnode excludes rows that restate full order-book snapshots sent on connect or after a resync, so intervals during feed reconnects can read as low activity.
Empirical results and examples
Glassnode reports raw counts are large: a median ten-minute interval carries about 607,000 level updates on BTCUSDT and 548,000 on ETHUSDT. According to Glassnode's backtest, a zscore reading two standard deviations above its norm is followed by an hour of above-median volatility about 63% of the time, against a 50% baseline. Glassnode also reports a reading from 24 hours earlier gets 59%.
Glassnode says most of the edge reflects volatile periods that last for days, and that churn moves with the price move and does not lead it. Glassnode presents an example (BTCUSDT on 27 April 2026, 03:00 to 08:00 UTC) in which three ten-minute intervals carried 1.8 to 2.1 million updates each, up from 0.6 to 0.9 million in the hour before. In that example the scores were 4.34, 3.83 and 3.41. Glassnode gives the corresponding depth and spread in that example: depth within 100 basis points was $377 million to $412 million against $400 million to $464 million in the hour before and then rose to $565 million; the spread stayed at about one tick, between 0.013 and 0.016 basis points.
Relation to depth, spread and how the score is built
Glassnode shows depth and spread capture some but not all of the churn signal. On BTCUSDT in typical months of 2025 and 2026, combined bid and ask depth within 100 basis points was about 1.9x as large in the 95th-percentile interval as in the 5th-percentile interval, while the update count ratio was about 3.9x. Glassnode reports that depth within 25 and 100 basis points and spread together explain 31% of the variation in the score on BTCUSDT and 25% on ETHUSDT, and that the share has fallen since 2023.
Glassnode defines the zscore as: zscore = ( ln(1 + updates) − mean ) / standard deviation. According to Glassnode, the mean and standard deviation come from the instrument's ten-minute intervals at the same hour of day over the previous thirty days (up to 180 intervals). Glassnode says the API returns two fields: zscore, the standardized score, and ratio, the update count divided by the average count of the intervals in the previous thirty days across all hours that had any updates. Glassnode notes ratio is not adjusted for hour of day. Glassnode also notes that for the interval starting at 14:30, nothing in the value comes from after 14:40.
Source: Glassnode Research — Book Churn: How Fast the Order Book Is Being Rewritten