Detecting Market Cycles Using Inverse-Logic Spectral Analysis

Abstract

The author describes a method of cycle detection using inverse-logic spectral analysis. It allows for fast detection of market cycles using any amount of input data, which can be smaller than the cycle length. The proposed method copes well with volatile input data and is more suitable for financial markets than traditional methods based on the Fourier or Hilbert transform methods. It detects dominant cycles earlier and generates in-phase functions, which can be used as oscillators or trading signals, with virtually no phase lag.