The absence of an opening bell is described as the market never sleeping. What actually happens is that participation follows the working hours of three regions, and the hours when all three are asleep produce a book that behaves differently in ways that cost money if you have not noticed.
A market with no closing time gives up the clearest structure a trader can have: a moment when everybody is present and a moment when nobody is. What replaces it is not uniformity but a softer version of the same pattern, driven by the fact that the humans and the desks providing liquidity keep ordinary working hours in ordinary time zones.
That produces a daily rhythm with three phases, corresponding roughly to the Asian, European and American working days, plus the gaps between them. The rhythm is not a schedule anybody publishes and it shifts with daylight saving changes and with where activity happens to be concentrated in a given period, but it is visible in any measure of participation.
The practical importance is not that some hours are better for making money. It is that the same order, placed at the same size on the same asset, has a different cost at different hours, and that difference is knowable in advance and entirely under your control.
There is no bell, but the people providing liquidity keep working hours. The rhythm survives the absence of a schedule.
Four things move together and they all move in the same direction. The spread between the best bid and the best offer widens when fewer participants are quoting. The depth behind those prices thins, meaning a given order size consumes more levels. Volume falls, so a large order is a larger share of what is trading. And the rate at which a consumed level refills slows down.
The fourth is the one people miss and it is arguably the most important. In an active hour, the level you consume is replaced within moments by somebody else's order, so the price you moved comes back. In a quiet hour, it may not be replaced for a while, which means your own order has moved the price and left it moved.
Those four together mean that the cost of taking liquidity is materially higher at some hours than others, for reasons that have nothing to do with what the price is doing and everything to do with who is awake.
The transitions between regional working days behave differently from either the busy or the quiet periods. During a handover, one group of participants is closing positions they do not want to carry while another is opening, and the two flows are not the same flow, which produces activity without necessarily producing direction.
The European and American overlap is generally the deepest window of the day, because two regions are working simultaneously rather than one handing to another. That is where a large order has the best chance of being absorbed without moving the price, and where the difference between a well-executed and a badly-executed entry is smallest.
The handover into and out of the Asian day, by contrast, tends to be thinner, and the transition can produce moves that look significant and reverse when the next region arrives. Interpreting a move made in a thin handover as information about direction is a common and expensive mistake.
The hours when all three regions are outside working time are the ones worth understanding in detail, because that is where the differences are largest. The book is thinner, quoted by fewer participants, and a larger proportion of what remains is automated rather than discretionary.
That last point matters more than the thinness. Automated liquidity is provided under rules, and one of those rules is generally to withdraw when volatility rises above a threshold. So the quiet-hours book is not merely thinner, it is thinner in a way that gets thinner still exactly when something is happening, which is the opposite of what you would want.
The practical consequence is that a move beginning in the quiet hours can travel much further per unit of order flow than the same move would in an active hour. It is not that the news is bigger; it is that there is less standing in the way, and what was standing there steps back.
A stop placed at a distance that is comfortable during active hours can be reached during quiet hours by a move that would not have reached it at any other time. That is not bad luck and it is not manipulation; it is the same distance measured against a different depth.
There is a second effect on top of the first. A stop that triggers becomes a market order, and a market order in a thin book fills worse than the same order in a deep one. So the quiet hours both make the stop more likely to trigger and make the fill worse when it does, and the two effects compound rather than adding.
None of this argues for removing stops, which trades a known cost for an unbounded one. It argues for setting the distance with the thinnest book in mind rather than the one you were watching when you placed the order, and for knowing that a position held across the quiet hours is a position with a wider effective risk than the same position held during the day.
The quiet hours make a stop more likely to trigger and make the fill worse when it does. The two effects compound.
Everything above applies more strongly across the weekend, when all three regions are out for two consecutive days. The book is at its thinnest, the professional desks are least engaged, and the connection to conventional markets is severed because those markets are closed.
That last point produces the most distinctive weekend behaviour. Digital assets trade continuously, but the equities, currencies and rates they are correlated with do not, so anything that happens on a Saturday cannot be arbitraged against a related instrument until Monday. Moves can therefore run further and hold longer than the same news would produce midweek.
The corresponding effect at the reopening is worth expecting. When conventional markets return, positions that were established without a reference point get re-evaluated against one, and the adjustment can be abrupt. Neither the weekend move nor its Monday adjustment is a signal about anything; both are consequences of who was available to trade.
Everything described here is checkable on your own data and worth checking rather than accepting. Record the spread and the depth on the assets you trade at intervals through several complete days, and the shape appears immediately if it is there, in your assets, at the sizes you use.
The measurement matters because the pattern is not identical everywhere. An asset whose holder base is concentrated in one region has a rhythm shaped by that region rather than by the global three-phase pattern, and an asset traded predominantly by automated participants has a flatter profile than one traded predominantly by people.
This is also where a general claim becomes a specific number. Knowing that the book is thinner at night is not actionable; knowing that on the assets you trade the spread is a certain multiple wider between certain hours is actionable, because it tells you what a given execution will cost and whether it is worth waiting.
For anybody with discretion about when to execute, the answer is to place discretionary orders in the deep hours and avoid taking liquidity in the thin ones. That is not a trading edge, it is a cost reduction, and it is available to everybody without any view on direction whatsoever.
For anybody without that discretion, because their strategy fires when it fires, the answer is to model the cost honestly by hour rather than using an average. A strategy whose backtest applied a single spread assumption across all hours will underperform if it happens to trade disproportionately in the thin ones, and the difference is attributable to nothing more exotic than the clock.
And for positions held across the quiet hours, the answer is to size and to place stops with the thinner book in mind. That is a different distance from the one that felt right at midday, and choosing it deliberately is cheaper than discovering it.
It is tempting to convert all of this into a claim that certain hours are better for trading, and that claim is not supported. Lower cost is not the same as higher return, because the reason costs are low in the deep hours is that many participants are present, and many participants present is also the condition under which any edge is competed away fastest.
What is supported is narrower and more useful: execution cost varies by hour in a measurable way, that variation is a real component of returns, and it is one of the few components a trader controls completely. Reducing a known cost is not a prediction and does not require one.
The distinction matters because the two claims lead to different behaviour. One leads to waiting for a good hour to trade, which is a directional bet dressed up as discipline. The other leads to executing more cheaply whenever you were going to execute anyway, which is arithmetic.
Yes, though softer than a market with an opening bell. Participation follows the working hours of three regions, producing a daily rhythm nobody publishes. It shifts with daylight saving and with where activity is concentrated, but it is visible in any measure of participation.
Four things moving together: the spread widens, the depth behind it thins, volume falls so a given order is a larger share of trading, and consumed levels refill more slowly. The fourth matters most, because it means your own order can move the price and leave it moved.
Generally the overlap between the European and American working days, because two regions are active simultaneously rather than one handing to another. That is where a large order has the best chance of being absorbed without moving the price.
Because a larger proportion of what remains is automated, and automated liquidity generally withdraws when volatility rises above a threshold. So the book is thinner in a way that gets thinner still exactly when something is happening.
Yes, in two compounding ways. A distance comfortable during active hours can be reached in quiet hours by a smaller move, and a triggered stop becomes a market order that fills worse in a thin book. Set the distance with the thinnest book in mind, not the one you were watching.
The same effects amplified, plus one distinctive one: correlated conventional markets are closed, so nothing that happens can be arbitraged against a related instrument until Monday. Moves run further and hold longer, and the Monday re-evaluation can be abrupt.
No, measure it. Record spread and depth on your own assets at intervals through several complete days. An asset whose holders are concentrated in one region has a rhythm shaped by that region, and one traded mostly by automated participants has a flatter profile.
That claim is not supported. Low cost is not high return: costs are low when many participants are present, which is also when any edge is competed away fastest. What is supported is that execution cost varies measurably by hour, and that is a cost you control without predicting anything.