Automatic identification and evaluation of Fibonacci retracements: Empirical evidence from three equity markets

Abstract

[EN] We examine the empirical performance of using Fibonacci retracements as a tool in technical analysis, in three main equity markets. To this end, we propose a novel algorithmic approach that not only identifies Fibonacci Support and Resistance levels but also constructs zones around them, in an attempt to address the inherent subjectivity associated with drawing and assessing these levels on a chart. Among our main findings we report a positive relationship between the width of the Fibonacci zone and the probability of identifying a price bounce. We show that although this finding does not necessarily imply a profitable strategy, it may potentially explain why Fibonacci retracements are widely used by practitioners. Our novel algorithmic approach can also be used to rigorously assess other variations of Support and Resistance levels.