Glassnode introduces Book Churn metric to track how fast order books get rewritten
Glassnode has introduced a new metric called Book Churn. It measures something most order-book data ignores: how often the price levels in an order book get revised.The analytics firm describes Book Churn first and foremost as a gauge of order-book activity. Its research also tests whether that activity can serve as a volatility signal for crypto perpetual futures, and the early results are interesting enough to take seriously.Technically, the metric tallies every revision to price levels on an exchange's depth feed. That feed is the stream of updates showing bids and asks being added, changed, or pulled.The volume of those revisions is substantial. For the BTCUSDT perpetual, Glassnode's research puts the typical figure at around 607,000 updates per interval.Glassnode publishes the metric at ten-minute resolution. Each reading is scored against the same instrument's trailing thirty days, at the same hour of day.That last detail matters more than it sounds. Crypto trades around the clock, but activity is not evenly spread. A quiet Asian morning and a busy US afternoon look nothing alike, so comparing each hour only to its own history keeps the metric from flagging routine daily rhythms as anomalies.The result is expressed as a z-score. The metric is currently applied to perpetual futures on Binance, specifically the BTCUSDT and ETHUSDT contracts.The headline finding concerns what happens when churn spikes. On Binance BTCUSDT and ETHUSDT, Glassnode's research found that a z-score of 2 or higher preceded above-median volatility in the following hour approximately 63% of the time.By definition, above-median volatility shows up in about half of all hours. So a 63% hit rate is an edge over a coin flip, not a crystal ball.The research also indicates the signal works best during persistent volatile conditions, when turbulence tends to cluster. The z-score's reliability improves in evaluations from 2022 onwards.Glassnode reports that depth and spread metrics explain only 25-31% of the variation in Book Churn scores. That share, the firm says, has declined from earlier years.The research adds that the metric offers information beyond conventional price ranges in around half of the episodes analyzed.Glassnode built its reputation on on-chain analytics, the study of what happens on the blockchain itself. Order-book data lives somewhere else entirely, inside exchange matching engines.The firm has been expanding into that territory throughout 2026. It began with spot order-book metrics in June 2026, then followed with perpetual futures metrics and Book Churn in October 2026, with the metric launching on October 9, 2026.Glassnode released its comprehensive study on the metric, which it also refers to as Perpetual Orderbook Churn, on November 18, 2026.The most immediate use case is risk management. A churn spike could work as an early warning that the order book is being reshuffled aggressively, before that shows up in price candles.The caveats deserve equal billing, though. A 63% hit rate means the signal misses a meaningful share of the time.The scope is also narrow for now. The published research covers two Binance perpetual contracts. Whether the relationship holds on other venues, smaller assets, or spot markets remains an open question as Glassnode's order-book coverage expands.