Nonlinear Price Indicators Using Medians and Ordinal Patterns
Summary
The article explains nonlinear transformations for financial time series and develops indicators based on order statistics. It contrasts logarithmic transformations, which turn arithmetic operations into geometric ones, with methods that sort observations to calculate medians, modes, and range midpoints. Examples include Bayesian smoothing toward selected reference values, a median of medians as a triangular-window analogue, and a pseudo-median built from local highs and lows.
It also applies the ideas to trend strength, RSI, Ichimoku, and candlestick-pattern analysis. For trend strength, sorted prices are weighted to form bounds representing idealized rising and falling sequences; their separation is presented as a volatility clue. The pattern method ranks selected candle prices and uses their ordering as a pattern signature. These are indicator construction proposals, not evidence of profitable trading: the article offers visual examples and implementation concepts but no systematic performance evaluation, and notes that some measures are spike-sensitive or unstable.
Key ideas
- Medians and modes are nonlinear measures of central tendency that can complement moving averages.
- Bayesian smoothing blends observed prices with selected reference values to make median-based series more responsive.
- A median of medians offers a nonlinear analogue of a triangular moving-average window.
- Sorted prices can define bounds for assessing trend strength and may indicate volatility through their spread.
- Ordinal rankings of selected candle prices can encode candlestick sequences as pattern signatures.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.