Adaptive Position Series with Multiscale Trend and Flat Confirmation

Summary

This article describes a self-adapting trading algorithm that analyzes price in variable-sized blocks and opens positions in linked series. Its central addition is a rule for starting another series while an earlier one remains open: the prior series’ working scale must first exceed a threshold, then the candidate scale must show a sufficiently large trend followed by a flat segment. This is meant to distinguish a smaller oscillation within a larger trend from continuation of that same trend.

The confirmation range is adaptive. Its minimum block count is derived from the number of prevailing directional blocks and a chosen flat percentage; its maximum comes from how many smaller blocks fit within recent larger blocks and the unfinished current block. A probability-based table adjusts the flat criterion across different range sizes. The article illustrates the method with a prolonged GBPUSD decline and argues that concurrent series can use smaller-scale movement while a main series waits for a rollback. It also lists unresolved work, including portfolio-based lot balancing, instrument-specific statistics, and improved loss compensation. The excerpt provides no independent performance test for this added mechanism.

Key ideas

• The algorithm analyzes price at multiple block scales, where trends and flat conditions may coexist.
• Additional position series can begin before earlier series close if scale and confirmation conditions are met.
• A trend signal at the candidate scale must be followed by a flat segment to confirm a new series.
• The block range for flat detection adapts to prevailing directional blocks and the structure of recent larger blocks.
• Instrument-specific sizing and statistical adaptation remain open development tasks.

Tags

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