The top of the hour has its own gravity
Many algorithms and fund flows are scheduled on the hour or half hour. Fixes, rebalancing trades, and time-weighted execution programs all tend to cluster at round clock times, and the market shows small bursts of activity at those moments.
On a one second chart, a spike in tick volume appears at the :00 and :30 marks, with sometimes a small price jerk. It is rarely large, but it can be enough to disturb a tight scalp.
The effect is strongest in the busiest sessions and at specific hours like 10am New York, 4pm London, and the opening of major markets.
A scalper holding a position through those moments faces slightly more volatility than the average second. Tight stops get tested, and entries placed seconds before the hour can be caught in the burst.
A habit worth testing is avoiding fresh entries in the 20 seconds before the hour and half hour, and seeing whether the results change. Some find a small but consistent improvement.
One small test: bucket the one second returns by seconds past the hour, and compare the average absolute move in the first five seconds of each minute with the rest of the minute. A visible hump near :00 and :30 confirms the effect on the broker's own feed.
Does the strategy show any difference in performance around round clock times?
The top of the hour has its own gravity
The top of the hour has its own gravity
It’s Fairman 
-
Amanhasnoname
- Posts: 2060
- Joined: Fri Oct 30, 2026 9:45 am
Re: The top of the hour has its own gravity
The round clock effect is real, and the bucket test you describe is about the cleanest way to see it, because it uses the broker's own feed rather than assumptions. It is worth separating the half hours from the full hours in that test as well, since the full hour usually carries more weight, and splitting by session so that the 4pm London fix does not swamp what happens during quieter hours.
One pitfall is treating every hour the same. A lot of the bursts come from scheduled data releases that happen to fall at :00 or :30, such as 8:30 and 10:00 New York, so the average can look stronger than the underlying flow really is. Excluding news days, or tagging them, shows whether the effect survives on ordinary days where only the execution programs are running.
If it does survive, the practical response is not only skipping entries just before the hour. Positions already open can be managed too, for example by having the stop slightly wider than the typical burst size or by taking partial profit before the mark. Comparing the results of those options on past trades should show which one actually helps rather than just feeling safer.
One pitfall is treating every hour the same. A lot of the bursts come from scheduled data releases that happen to fall at :00 or :30, such as 8:30 and 10:00 New York, so the average can look stronger than the underlying flow really is. Excluding news days, or tagging them, shows whether the effect survives on ordinary days where only the execution programs are running.
If it does survive, the practical response is not only skipping entries just before the hour. Positions already open can be managed too, for example by having the stop slightly wider than the typical burst size or by taking partial profit before the mark. Comparing the results of those options on past trades should show which one actually helps rather than just feeling safer.