Newton Polynomial Interpolation for Trading Signals and Trade Management

Summary

The article explains Newton interpolation through divided differences and basis terms, then describes a scalable procedure for deriving polynomial coefficients from sampled points and evaluating the polynomial at a new input. It illustrates the method with price observations and proposes using it in an Expert Advisor as a signal, trailing stop, and money management component. The approach is attractive for its transparent equation and modest storage and computation needs compared with neural networks.

The article discusses application examples and testing, but the supplied text does not provide complete performance statistics. It emphasizes material limitations: polynomial interpolation can fit noise in the sample, leading to poor out-of-sample behavior, and financial series can change regime and exhibit abrupt moves that smooth polynomial curves may not capture. The author suggests pairing the method with other indicators to filter noise and notes that the trailing and money management uses require further testing. The method should therefore be treated as an exploratory modeling tool rather than established evidence of a trading edge.

Key ideas

• Newton interpolation builds a polynomial from unique input values using successive divided differences.
• The resulting coefficients can be evaluated at a new input to produce a forecast.
• The article applies the method as a proposed trading signal, trailing stop, and money management component.
• Fitting every sampled fluctuation can make the polynomial capture noise and weaken out-of-sample performance.
• Changing market properties and sharp price moves are difficult for smooth polynomial curves to represent.

Tags

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