6 / 7 · 6 min
Time: what in sessions is testable and what was copied from another market
Crypto trades round the clock but not evenly. The daily rhythm exists and can be measured; the "magic windows" from courses cannot.
The exchange works without weekends, but people do not. Activity comes in waves following the working day: the Asian morning, the European day, the American evening. The overlap of Europe and America — roughly 13:00 to 16:00 GMT — is the busiest time of day. This is not theory: the daily turnover profile can be measured on your own history, and it holds steady from week to week.
How large the difference is
Measured on bitcoin: median hourly turnover at 03:00 GMT is around 41 million dollars, at 14:00 around 186 million. That is more than a fourfold difference, and it is a median, not one day's outlier. The practical consequence matters more than the figure itself: a surge "twice the usual" at three in the morning and at two in the afternoon are entirely different events. The first means two people woke up, the second means money arrived.
What of this is testable
Everything you can measure on your own history is testable: when turnover arrives for YOUR instrument, when the range is wider, when sharp moves happen more often. Untestable is everything presented as a universal window accurate to the minute. Such windows came from forex, where sessions have a physical cause — banks opening in London and New York. In crypto there is nothing to open: the exchange never closed.
Copying the forex schedule into crypto as is
The most common mistake in this subject, and it shows in the details. "The London open at 08:00" makes sense where people with money genuinely start work at 08:00. In crypto that hour simply brings more Europeans — gradually, not in a jump. Worse still is daylight saving: London and New York shift their clocks and GMT does not, so a memorised window slides by an hour twice a year. If your rule is tied to an hour, it must be tied to GMT and measured on your own data.
Times that do have a cause
- The start of the exchange day: everyone measures up or down from the daily open, so it works as a level.
- Funding settlement — usually every eight hours. Positions get trimmed before it to avoid paying, and pressure eases after it.
- Macro data releases and central-bank meetings: the time is known in advance and the reaction can be instant.
- The US stock market open — for our stock perpetuals: before it, the contract's price lives without the share itself.
Keeping one position size at any hour
At night the book is thinner: the same money moves price further, and your exit costs more. Slippage on an exit at four in the morning can eat the whole profit of a trade that would have been clean on daytime turnover. The sensible rule is simple: either trade in the hours when the instrument has turnover, or reduce size outside them. The second is more honest: ordinary life rarely coincides with the American session.
Weekends are a separate case
On Saturday and Sunday crypto turnover falls and the spread widens, while moves can be sharper than usual: a thin book is shifted by less money. A separate trap is stock perpetuals: they trade round the clock while the stock exchange itself is closed. Overnight into Monday the contract's price can travel far from the last real quote, and in the morning it snaps back to it. That is not a move in the share, that is a gap being revealed.
It is precisely because of the daily rhythm that our volume surge is computed not against a common bar but against the norm of THE SAME HOUR for that same coin. Otherwise every day at 14:00 would show a surge across the whole market at once and show nothing at night. The feed of impulses and surges sits next to the table and updates in real time.
Impulses and volume surgesYour instrument's daily profile
Take one coin and over the past thirty days compute the median hourly turnover for each of the twenty-four GMT hours. Chart it as bars from largest to smallest. You will get YOUR profile rather than a picture from somebody's course, and you will see three things: which hours make entry cheaper, which hours make a surge genuinely mean an event, and whether your coin has a time of its own that does not match the market's. Coins with an Asian audience often do.
A volume surge four times the usual at 03:00 GMT. Is that a strong signal?
It depends what it was compared with. If "usual" was taken across the whole day, then at three in the morning turnover is already four times below average — and there is no surge at all, you are looking at a daytime norm applied to a night hour. If the comparison is against the norm of THE SAME HOUR, then it is a genuine event: something happened overnight. The same multiplier means opposite things depending on the denominator.
Why does the "London window" work worse in crypto than in forex?
Because in forex there is a physical cause behind it: banks in London open and volume that was not there before comes to the market. In crypto the exchange never closed, and at 08:00 only the mix of participants changes — gradually, over an hour or two. Add daylight saving: London and New York shift twice a year relative to GMT while the memorised window stays put. There is one testable replacement: measure your instrument's daily profile and work from that.