What is the funding rate historical memory?
The live table shows the funding rate on five exchanges right now. This page adds what no free service offers: an own historical series of the multi-exchange AVERAGE. Every hour we store each exchange's annualized rate and its per-asset average; with that series we compute the current funding's historical percentile, how many consecutive days it has kept the same sign, and how far exchanges diverge from each other.
Why is this data hard to find?
Each exchange publishes, at best, its own history. The average across several exchanges must be built by sampling all of them at once, continuously, and storing the result: if nobody captured it at the time, it cannot be rebuilt later. Services that offer it charge for it. Ours starts the day capture was enabled and grows automatically every hour.
How the indicators are computed
We work with rates annualized using each exchange's real interval (8h, 4h or 1h), which makes them comparable. The base series is the daily averages of the multi-exchange mean. The percentile is the share of those days with an average at or below today's (window: up to 365 days). The streak counts consecutive days, including today, with the same sign. The spread is the current gap between the highest and lowest annualized rate among exchanges with data.
Frequently asked questions
What does the funding rate historical percentile mean?
It is an arithmetic fact: the percentage of recorded days in our series whose daily average was at or below the current rate. A 95th percentile means today's rate is higher than on 95% of observed days. It implies no future price direction.
What is a negative funding streak?
The number of consecutive days, including today, in which the multi-exchange daily average has been negative. It describes the persistence of the data, just like counting consecutive rainy days; it is not a buy or sell signal.
Why do you compare annualized rates?
Because exchanges pay funding at different intervals (8 hours, 4 hours, 1 hour). Annualizing with each one's real interval puts all rates in the same unit, making the average and the cross-exchange spread correct.
Since when is there data, and why are some days missing?
The series starts on the date shown under the table, when capture was enabled, and grows hourly. If the server had no traffic and no cron during some hour, that gap is not backfilled: we prefer to show only actually observed data.
Do you store any visitor data?
No. The database only stores public funding rates from exchanges. No visitor preference or data reaches the server.