The ability to detect market reversals in a timely manner has become a key task for traders and analysts. Inverse-Logic Spectral Analysis (ILSA) takes it to the next level by overcoming limitations of traditional methods. It was developed by the speaker for the purpose of cycle detection in a volatile environment and is currently pending patent (No GB1709967.2).

ILSA operates with input data of any length, even shorter than the cycle itself. It detects market cycles earlier than conventional methods (based on Fourier or Hilbert transform), and generates in-phase functions, which can used as oscillators or trading signals with virtually no lag.

Practical results demonstrate fast and accurate cycle detection and show cyclical market reversals forecasted months before they occur. The performance of a simple trading model using ILSA for daily GBP/USD trading is presented for a 27-year period, showing a robust P/L growth with the Sharpe ratio of 2.5.