Autocorrelation of tick returns at different horizons
At the tick level, returns tend to show negative autocorrelation due to bid-ask bounce. At a few seconds, the sign can flip as momentum from order flow persists. At a few minutes, it depends on the regime. The horizon at which a strategy trades determines whether it fights or follows the structure.
Measuring the autocorrelation of mid-price returns at different sampling intervals, say 1 second, 5 seconds, 30 seconds, and 5 minutes, shows where mean reversion and momentum dominate. The results vary by pair, time of day, and market conditions.
For a scalper, this informs the choice between fade and follow strategies. If 10-second returns show positive autocorrelation during the London open, momentum entries make sense. If they show negative autocorrelation in the Asian session, fading is preferable.
The numbers are small, often between 0.02 and 0.1, but small correlations can be exploited with enough trades if costs are low.
Compute them across weeks of data, and check stability. If the sign changes randomly, there is nothing to trade.
What do the autocorrelations look like at the horizon the strategy uses?
Autocorrelation of tick returns at different horizons
-
Amanhasnoname
- Posts: 2060
- Joined: Fri Oct 30, 2026 9:45 am
Re: Autocorrelation of tick returns at different horizons
The sign at my holding time was the part that surprised me. I sampled mid-price returns on EURUSD for seven weeks of September ticks at one second, eight seconds, and two minutes. At one second the autocorrelation sat around minus zero point zero seven, which is just bid-ask bounce. At eight seconds it flipped to plus zero point zero four in the first ninety minutes of London and stayed near minus zero point zero three through Asia. Two-minute returns were close to zero most of the day, except plus zero point zero six on mornings that were already trending.
Those numbers are small, as the opening post said, but they were stable enough that I stopped using one entry style all day. I fade small bursts in Asia and I follow the first push after London opens, provided the eight-second figure is still positive that morning.
I recompute the figures every Sunday on the previous fortnight rather than trusting a single long sample. The last time the London eight-second reading went flat I sat on my hands for three sessions and avoided a string of follow-through entries that would have been fades in disguise.
Those numbers are small, as the opening post said, but they were stable enough that I stopped using one entry style all day. I fade small bursts in Asia and I follow the first push after London opens, provided the eight-second figure is still positive that morning.
I recompute the figures every Sunday on the previous fortnight rather than trusting a single long sample. The last time the London eight-second reading went flat I sat on my hands for three sessions and avoided a string of follow-through entries that would have been fades in disguise.