Combining Fourier Cycle Detection with a Spiking Neural Network for Trade Timing

Summary

This article proposes a trading system that pairs a discrete Fourier transform (DFT) with a leaky integrate-and-fire spiking neural network. The DFT examines a rolling series of prices or indicators such as MACD and RSI, selects the strongest frequency component, and represents the estimated cycle phase as a bounded value. The network accumulates weighted input signals over time, with a decay factor reducing its stored potential; it generates a trade signal only when that potential crosses a threshold. The intended use is to identify market transitions while filtering brief price moves and whipsaws.

The article reports an optimized 16-month test with a 90-bar Fourier window, 56 trades, a 62.5% win rate, profit factor of 1.64, and maximum drawdown of 7.57% on a $10,000 account. These are reported results from one test, not evidence of general profitability. The author recommends testing other instruments and longer periods, and suggests exploring adaptive weights, faster Fourier computation, and volatility-scaled decay. The article does not provide enough detail here to independently assess the test design or robustness.

Key ideas

• The DFT decomposes a rolling market series into frequency components and identifies the component with the greatest amplitude.
• A bounded cycle estimate is derived from the dominant component's phase.
• The spiking network accumulates weighted signals while its stored potential decays over time.
• A trade signal occurs when accumulated potential crosses a preset threshold.
• The reported test is limited to one 16-month evaluation and requires broader validation.

Tags

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