Building Trading Signals from Cycles and Finite-Difference Oscillators
Summary
The article develops trading signals from cyclical price models expressed with finite differences rather than directly with sine and cosine functions. In its basic harmonic oscillator, a parameter tunes sensitivity to a cycle period, and sign changes in a difference equation trigger position reversals. It then extends the construction to damped cycles, combinations of oscillations intended to approximate the first five Elliott waves, and generalized higher-order difference models with weighted terms. Further variants introduce nonlinear transformations, including sign, quadratic, Van der Pol, Duffing, and logarithmic forms, as well as a non-autonomous oscillator driven by external inputs.
The author illustrates the models with reported EA balance curves and says testing used EURUSD on an hourly chart over the stated 2024 period. The article argues that parameter combinations can produce stable results, but provides no detailed benchmark, transaction-cost analysis, out-of-sample comparison, or statistical significance assessment in the supplied text. Higher-order patterns may occur less often, and the many adjustable coefficients make selection and overfitting important concerns.
Key ideas
- A harmonic oscillator can be represented with finite differences, with sign changes used to generate trading entries and exits.
- Damping adds a parameter that controls how quickly modeled oscillations fade.
- Combining oscillations or weighting higher-order differences can represent more complex price patterns.
- Nonlinear oscillator variants aim to respond differently to sharp moves and changing market behavior.
- The reported EURUSD hourly test and balance illustrations do not establish robustness after costs or out of sample.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.