Trend Smoothing and Statistical Criteria for Trading Signals
Summary
The article surveys methods for smoothing price series and identifying whether observed movement departs from randomness. It discusses moving-average variants intended to reduce lag, then describes Abbe variance comparisons, signs of first differences, Kendall’s statistic, the Wald–Wolfowitz rank test, and Foster–Stewart record counts. Some measures indicate trend presence, while others also estimate direction or strength.
The proposed trading use treats extreme criterion readings as overbought or oversold conditions and looks for a possible reversal after a strong trend. Additional filters, including Abbe trend detection and average price speed, are suggested to qualify signals; Foster–Stewart is presented as a combined signal and filter. The article reports illustrative indicator and strategy applications, but the supplied text omits part of the strategy discussion and gives no complete quantitative performance evidence. It cautions that criteria may generate false signals, need separate exit rules, and can depend on sample length, making smoothing potentially necessary.
Key ideas
- Moving-average variants aim to reduce lag while retaining some noise suppression.
- Trend criteria test whether a price sequence appears nonrandom, and their outputs differ in whether they show direction.
- The article proposes using extreme criterion values as reversal signals and other criteria as filters.
- False signals, sensitivity to sample length, and the need for explicit exit rules limit the approach.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.