Most volatile crypto hours of the day, in UTC and local time
We measured average hourly moves for 12 crypto markets on Binance and Coinbase. See the busiest hours in UTC, New York, London and Tokyo time.
The hour from 14:00 to 15:00 UTC moved crypto the most in our test. That is 10 to 11 in the morning in New York in summer. For Bitcoin on Binance it averaged a 0.54% move, almost twice the quietest hour, 04:00–05:00 UTC, at 0.30% (2,598 hourly price changes each, September 8, 2019 to October 20, 2026). Nine of the 10 Binance coins we measured peaked in that same hour, and so did Bitcoin and Ethereum on Coinbase.
UTC is the world time standard used to label the hours here. BTC is Bitcoin, ETH is Ethereum, and SOL is Solana. We measured perpetual contracts, or perps: contracts based on a coin's price that have no expiry date, priced in USDT, a token designed to track the US dollar.
For when the daily candle closes, see What time does the crypto daily candle close?. For direction at the US open, see Does Bitcoin dump at the US market open?.
What does an average hourly move measure?
A price bar records the first, highest, lowest and last prices over a set time. The last price is called the close. We used one-hour bars and measured the percentage change from each close to the next.
We then removed the plus or minus sign before taking the average. This is an absolute move: its size without its direction. Rises and falls both add to it.
So “most volatile” here means the largest average hourly move. It is not the gap between an hour's high and low. A price can rise, fall back and finish near where it started, which we count as a small move. To see swings inside an hour, you need smaller bars, such as one-minute bars.
Which UTC hours moved BTC, ETH and SOL the most?
These are Binance USDT perps with one-hour bars, ending October 20, 2026, at 00:00 UTC. Each cell shows the average move size and, in parentheses, the number of hourly price changes.
| Market and observed date range | 04:00–05:00 UTC | 14:00–15:00 UTC | 15:00–16:00 UTC |
|---|---|---|---|
| BTCUSDT: Sep 8, 2019–Oct 20, 2026 | 0.30% (2,598) | 0.54% (2,598) | 0.50% (2,598) |
| ETHUSDT: Nov 27, 2019–Oct 20, 2026 | 0.39% (2,518) | 0.69% (2,519) | 0.65% (2,519) |
| SOLUSDT: Sep 14, 2020–Oct 20, 2026 | 0.65% (2,226) | 1.00% (2,227) | 0.98% (2,227) |
For each of these coins, 14:00 UTC was the largest of all 24 hours, 15:00 UTC came next, and 04:00 UTC was the smallest.
SOL's quietest hour still moved more than BTC's busiest. The coins cover different dates, so part of that gap may come from the periods covered.
What time is that in New York, London and Tokyo?
Local clocks move by an hour for daylight saving time, so one UTC hour lands at a different local time in summer and winter. Tokyo does not change its clocks.
| UTC hour | New York (summer / winter) | London (summer / winter) | Tokyo |
|---|---|---|---|
| 00:00–01:00 (TRX peak) | 20:00 / 19:00 the day before | 01:00 / 00:00 | 09:00 |
| 04:00–05:00 (quietest for BTC, ETH, SOL) | 00:00 / 23:00 the day before | 05:00 / 04:00 | 13:00 |
| 14:00–15:00 (peak for 9 of 10 coins) | 10:00 / 09:00 | 15:00 / 14:00 | 23:00 |
| 15:00–16:00 (second for BTC, ETH, SOL) | 11:00 / 10:00 | 16:00 / 15:00 | 00:00 the next day |
So the busiest hour falls in the New York morning, around when US stock markets open (9:30 New York time). The quietest hour falls in the middle of the night in New York. We did not test whether US trading hours cause the peak; the timing lines up, and that is all.
Midnight UTC also stands out: it ranked in the top three hours for 8 of the 10 Binance coins. That is when a new UTC day starts and daily bars close. Again, we measured when moves happened, not why.
Did the rest of the group follow the same pattern?
The other Binance markets mostly shared that peak. Here are their largest hourly averages, on the same terms.
| Market | History starts | Peak hour, UTC | Average move size | Hourly price changes at peak |
|---|---|---|---|---|
| XRPUSDT | Jan 6, 2020 | 14:00–15:00 | 0.76% | 2,479 |
| DOGEUSDT | Jul 10, 2020 | 14:00–15:00 | 0.90% | 2,293 |
| ADAUSDT | Jan 31, 2020 | 14:00–15:00 | 0.86% | 2,454 |
| AVAXUSDT | Sep 23, 2020 | 14:00–15:00 | 0.99% | 2,218 |
| LINKUSDT | Jan 17, 2020 | 14:00–15:00 | 0.94% | 2,468 |
| BNBUSDT | Feb 10, 2020 | 14:00–15:00 | 0.64% | 2,444 |
| TRXUSDT | Jan 15, 2020 | 00:00–01:00 | 0.56% | 2,469 |
TRX had a different peak hour, but only just. In that same Binance TRXUSDT history, 14:00–15:00 UTC averaged 0.55% across 2,470 hourly price changes, a hair below midnight.
Does the same hour lead on another exchange?
Yes, in the two Coinbase markets we measured. These are spot markets: you buy the coin itself, priced in US dollars. Their hourly history starts earlier, July 20, 2015 for BTC-USD and May 18, 2016 for ETH-USD, and ends October 20, 2026, at 00:00 UTC.
| Coinbase spot market | Peak hour, UTC | Average at peak | Quietest hour, UTC | Average at quietest |
|---|---|---|---|---|
| BTC-USD | 14:00–15:00 | 0.54% (4,107) | 05:00–06:00 | 0.35% (4,104) |
| ETH-USD | 14:00–15:00 | 0.75% (3,799) | 04:00–05:00 | 0.48% (3,793) |
A different exchange, a different kind of market and several more years of history gave the same busiest hour. So the 14:00 UTC peak is not a quirk of one exchange. Bitcoin's peak average even matched: 0.54% on both.
Does the peak hour explain why prices moved?
No. We grouped price changes by hour, but we did not measure what caused them. The tables also mix weekdays, weekends and every year in each history, so they do not show whether the same hour stayed on top the whole time.
Is the busiest hour the best time to trade?
This study cannot answer that. Larger moves can go in either direction, and removing the sign means these averages cannot say how often prices rose or fell. An average also hides spread: a few very large moves can pull it up.
For the day-of-week question, see crypto weekend volatility. For what followed a sharp fall, see Bitcoin after big one-hour drops.
How can you use this in your own research?
Judge a move against the normal size for that hour. A 0.54% Bitcoin hour was average at 14:00 UTC in this history, but nearly twice the 04:00 UTC average. If you test a rule that reacts to "big" moves, try measuring "big" against the same hour's average for the same coin and exchange, and see whether your results change. That is a research check, not a reason to trade at any hour.
How we measured this
We used public one-hour price bars for the named Binance perpetual contracts and Coinbase spot markets over the dates shown in the tables. We measured the percentage change from each close to the next, removed its sign, and averaged the results by the UTC opening hour in which the move occurred. The tables show the number of hourly changes used; we left missing prices empty and skipped the first close because it had no earlier close to compare. We kept partial years, and each coin covers a different period. We did not subtract trading costs or test any trade. The coins are selected ones still trading, and we did not check whether the rankings hold in a later period.
Try this in Comparand
With Comparand connected to your AI app, ask: “Get completed hourly price bars for Binance BTCUSDT perpetuals between my chosen UTC start and end times. Use the closes to calculate the average move size by UTC opening hour, ignoring direction. Show the dates and the count for each hour.”
What comes back:
- venue_bars returns the first, highest, lowest and last prices, plus trading volume, for each completed hour. It is free.
- Your AI app works out the hourly averages; the tool returns prices, not a ranking.
- Each answer carries source details so you can check the dates.
To look inside an hour on Hyperliquid, ask: “Check which archive days are complete. Then show BTC's one-minute price bars between my chosen UTC start and end times within a covered day.”
What comes back:
- hl_coverage (free) shows which days are complete.
- hl_bars returns BTC one-minute bars, so you can see swings inside the hour. It is included on plans, or $0.005 per call with pay-as-you-go credits.
- Those bars describe Hyperliquid, so keep them apart from the Binance and Coinbase results above.
The docs explain access and tool settings. You can also try 3 coins free to explore Comparand's past-chart search. Those are past cases, not a forecast.
The short version
- On Binance, 14:00–15:00 UTC (the New York morning) had the largest average move for 9 of 10 coins; for BTC it was 0.54% versus 0.30% at 04:00 UTC.
- Coinbase BTC-USD and ETH-USD, with history back to 2015 and 2016, peaked in the same hour.
- TRX's largest average came at midnight UTC, with only a small lead.
- Move size depends on the coin as well as the hour.
- Each average mixes changing market conditions over the dates shown.
- Larger average moves do not tell you whether prices will rise or fall.
Questions people ask
What is the most volatile hour for Bitcoin?
For Binance BTCUSDT perpetuals, contracts with no expiry date, the largest average hourly move was 0.54% at 14:00–15:00 UTC. We used 2,598 hourly price changes from September 8, 2019, to October 20, 2026. We counted move size whether the price rose or fell.
What time is crypto most volatile in New York time?
In our Binance test, the busiest hour for 9 of 10 coins was 14:00–15:00 UTC. That is 10 to 11 a.m. in New York during summer time and 9 to 10 a.m. in winter. Coinbase BTC-USD and ETH-USD peaked in the same hour. We measured when moves happened, not why.
Do all crypto coins peak at the same hour?
No. BTC, ETH and SOL peaked at the same UTC hour in our Binance study, but TRX peaked at midnight UTC. The largest average was not at the same hour for every coin.
Does a volatile hour mean price is likely to rise?
No. We measured the size of each move after removing its direction. A larger average can include both rises and falls.
Does this measure the full price range within each hour?
No. We measured the change from the previous hourly closing price to the next. A price could rise and fall within the hour but finish near where it started.