COINANOMALY
Crypto market data events

When a market departs from its own normal, we keep the record.

Every event is measured against the coin's own history and published with how the data spread in both directions after similar events in the past. No forecasts, no recommendations — only measurement and record.

1792 events recorded · 1763 with the following 24 hours measured · latest Oct 28, 2026 · 18:00 UTC

Event log

In time order, newest first

Events are not ranked or highlighted. Each record is completed automatically with the data of the hours that followed it.

On Oct 27, 2026 BANKUSDT read 3.2×; over the next 24 hours the high was +11.6%, the low −11.0%, the close −9.0%.

# Time UTC Coin Type Trigger Magnitude +24h high +24h low +24h close
1792 RVNUSDT Volume surge Binance 5.3× ≈5× — — —
1791 EDENUSDT Volume surge Binance 5.8× ≈6× — — —
1790 CHIPUSDT Volume surge Binance 5.1× ≈5× — — —
1789 SHELLUSDT Volume surge KuCoin 42.4× ≈25× — — —
1787 TUTUSDT Volume surge Binance 9.5× ≈9× — — —
1786 CELOUSDT Volume surge Binance 7.0× ≈7× — — —
1785 HAEDALUSDT Volume surge Binance 16.3× ≈16× — — —
1788 ARKUSDT Extreme negative funding Binance perp −0.61%/day — — —
1784 INITUSDT Volatility regime Binance spot 3.3× — — —
1783 GMXUSDT Volume surge Binance 5.0× ≈5× — — —

All events →

What we measure

What we measure

The figures below are descriptive statistics computed from past data and are always given for both directions. A figure is shown only if it was measured on the market the event is watched on; where that measurement has not been made yet, this is said instead. Each event shows what was measured; the exact windows and thresholds, which were set by measurement, are not published.

Volume surge

A coin's traded value running far above its own recent normal.

Met 10,175 times in 156,788 measured coin-hours, about one in 15. Measured over 56 days, to Oct 27, 2026. This counts how often the criterion was met; far fewer events are published, because one episode is published once.
How it is worked out
traded value over a recent window
÷ that coin's own median for the same span
Watched on
Binance, KuCoin and Pionex spot, each on its own; an event needs the threshold crossed on one of them

Extreme negative funding

The two sides of a perpetual contract pay each other at intervals, and that payment is called funding. When the contract trades far enough below the spot price, the payment turns around: the selling side pays the buying side. A small gap is ordinary and does not turn it; the rate carries a fixed component that the gap has to clear first.

No single base rate covers this one yet: it is watched on more than one market, and only some of them have been measured. What has been measured is named below.
How it is worked out
funding rate normalised to a day
below a measured line
Measured on
601 Pionex perpetuals, 166 days, tokenised stocks excluded. For Pionex events only; being re-measured for Binance
Watched on
Pionex and Binance perpetuals, each exchange on its own; follow-up data from the same exchange's spot price

Parabolic run

A price multiplying within a few days. Measured close to close, not on the extremes reached inside a day.

Met 34 times in 6,384 measured coin-days, about one in 188. Measured over 56 days, to Oct 27, 2026. This counts how often the criterion was met; far fewer events are published, because one episode is published once.
How it is worked out
close against the lowest close
of the days before it
Watched on
Binance spot; KuCoin spot where Binance does not list the coin

Volatility regime

The range a coin's price travels over three days running several times the range that is normal for that coin.

Met 13,622 times in 147,697 measured coin-hours, about one in 11. Measured over 49 days, to Oct 24, 2026. This counts how often the criterion was met; far fewer events are published, because one episode is published once.
How it is worked out
range over the last 72 hours
÷ that coin’s own median range for 72 hours
Watched on
Binance spot; KuCoin spot where Binance does not list the coin
Method

How we measure

Against its own normal

A coin's volume or price is judged against that coin's own history, not against the market average.

Closes, not wicks

Price events are measured close to close. Extreme prices that appear for a few seconds and vanish do not produce events.

With the base rate

Every statistic is shown beside the same question asked of an ordinary coin-day.

Missing data makes no event

A gap in the record is not a quiet period. If the measuring window is not filled enough, no event is published.

Checkable

Every record is published with its date, exchange, symbol, and the volume or price values it was computed from.

Counted once

The same event is not recorded again on consecutive measurements; each event type has a cooldown.

Important

The information on this site consists only of statistical observations based on past market data; it is not investment advice, investment consulting, or a recommendation to buy or sell.

Rates seen in the past do not indicate future price movements. Crypto assets carry high risk.