Rolling 24-Hour Window
A lot of market metrics are based on “the last 24 hours.” The important detail is whether that 24-hour period is fixed to a calendar day or moves continuously.
A rolling 24-hour window moves with time. At any moment, it looks back exactly 24 hours from “now” and includes whatever data falls inside that range.
This is useful in crypto because the market never closes. Rolling windows avoid sudden discontinuities that come from resetting metrics at midnight or at a session open.
Why It Matters
Rolling windows make benchmarks and volume-based metrics more stable and more representative of continuous trading.
How is a rolling 24-hour window different from “today’s” 24 hours?
“Today” is anchored to a calendar boundary, so it resets at midnight in some time zone. A rolling window has no reset point; it just slides forward as time passes. For 24/7 markets, rolling windows are often a better match for how traders interpret “recent activity.”
What data quality issues show up in rolling-window metrics?
Rolling windows can be sensitive to missing data, stale updates, or late-arriving events. If your inputs aren’t time-consistent, the window may include too much of one venue and not enough of another. That’s why many systems pair rolling windows with strict freshness checks and outlier filtering.
Real-World Example
A price benchmark uses the last 24 hours of trading. At 12:00 UTC today, the window covers data from 12:00 UTC yesterday up to 12:00 UTC today.
Rolling 24-Hour Window and Exchange Rates API
Rolling windows are a common ingredient in benchmark pricing. CoinAPI’s Exchange Rates API uses rolling 24-hour logic as part of how it computes VWAP-24H style rates.