Drift after the event
Posted: Sat Oct 31, 2026 8:40 pm
Drift after the event
After a major release, the first few minutes are chaos. Then, in many cases, a quieter drift follows: price continues slowly in the direction of the surprise, for thirty to ninety minutes.
The pattern has an economic logic. Information takes time to be absorbed. Participants who couldn't act in the first seconds trade later. Analysts publish notes. Orders from slower investors arrive in waves.
The drift is smaller and cleaner than the initial spike. Spreads have normalized, and fills are better. For scalpers who avoid the first minutes, it offers a more civilized entry.
It doesn't always appear. When the surprise is small, the drift is absent. When the market had already priced in the move, it may reverse.
Measuring it requires a sample. Take events of one type, such as US payrolls, record the direction of the surprise, and compute the average return from five minutes after the release to ninety minutes after, signed by the surprise.
If the average is positive and exceeds costs, the drift is a candidate for a strategy. Even then, caution applies, since the sample is small.
Measured from the five-minute mark, how large is the post-event drift on the pairs followed?
After a major release, the first few minutes are chaos. Then, in many cases, a quieter drift follows: price continues slowly in the direction of the surprise, for thirty to ninety minutes.
The pattern has an economic logic. Information takes time to be absorbed. Participants who couldn't act in the first seconds trade later. Analysts publish notes. Orders from slower investors arrive in waves.
The drift is smaller and cleaner than the initial spike. Spreads have normalized, and fills are better. For scalpers who avoid the first minutes, it offers a more civilized entry.
It doesn't always appear. When the surprise is small, the drift is absent. When the market had already priced in the move, it may reverse.
Measuring it requires a sample. Take events of one type, such as US payrolls, record the direction of the surprise, and compute the average return from five minutes after the release to ninety minutes after, signed by the surprise.
If the average is positive and exceeds costs, the drift is a candidate for a strategy. Even then, caution applies, since the sample is small.
Measured from the five-minute mark, how large is the post-event drift on the pairs followed?