Charting: Chaos theory in disguise?

Abstract

The primary hypothesis of this article is that technical modeling methods may represent crude but useful ways of exploring nonlinear qualities in data. More specifically, it is proposed that graphical technical analysis may be restated in terms of attractors and strange attractors. Further, it is suggested that technical analysis methods may allow prediction on systems of higher dimension than nonlinear methods do at this time. An objective algorithm that identifies technical patterns is applied to highdimension nonlinear data, and provides support for the primary hypothesis. This article is distinguished from previous studies in that (a) a specific linkage/equivalence between technical analysis and nonlinear forecasting is proposed, and (b) statistically significant evidence in support of this specific linkage/equivalence is presented. If this link is valid, it has important implications for the future study of technical and nonlinear analysis, which should be studied jointly, and also suggests that other disciplines applying nonlinear analysis might benefit from applying technical methods.