Test of Elliot Wave theory by time series segmentation algorithms

Abstract

Discovering the patterns of stock prices in a volatile and rapidly-changing market environment is a challenging problem. Elliot Wave theory, a form of technical analysis, attempts to investigate market price movements. The purpose of this study is to test the principles of Elliot Wave theory on Istanbul Stock Exchange(ISE)-National 100 index. Two time series segmentation algorithms, top-down algorithm and a modified version of bottom-up algorithm, are applied. Standard (compulsory) rules and rules expressed as the golden ratio of Fibonacci are formulated by statistical hypotheses. The existence of Elliot Wave pattern in ISE -National 100 index is tested. Hypothesis tests show that Elliot patterns exist in ISE-National 100 index, stating that Istanbul Stock Exchange is not an efficient market.