Using Bayesian Order-Flow Counts to Trade Aggressively at High Frequency

Summary

This high-frequency trading note proposes tracking aggressive order flow over a rolling window of ticks. It explains the Bayesian idea as updating beliefs when new evidence arrives, then applies that intuition to counts of buy- and sell-side activity. In its example, a trader counts events across ten ticks and enters long or short when one side reaches a threshold of seven. The note also suggests using different entry and exit thresholds so the strategy is not continuously in the market, with exits guided by the principle of leaving when the market has not confirmed the trade.

The document offers a conceptual example and pseudocode, but no measured results, execution model, transaction-cost analysis, or risk controls. Its order-flow counting logic is underspecified, and the description of genetic algorithms as the usual name for Bayesian algorithms is incorrect. The proposed signal therefore needs a precise definition, statistical testing, and realistic high-frequency execution assumptions before it can support a trading decision.

Key ideas

• The strategy counts aggressive buy- and sell-side order-flow events over a rolling tick window.
• In the example, a ten-tick window and a threshold of seven trigger long or short entries.
• Different entry and exit thresholds can keep the strategy from holding a position continuously.
• The note supplies no performance evidence and leaves execution costs and risk controls unspecified.
• It incorrectly equates Bayesian algorithms with genetic algorithms.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.