Predicting Herding Trend by Fuzzy Logic

Abstract:

In this article we focus on the predicting of herding behavior towards the Toronto Stock exchange index (TSX60) using a fuzzy logic. We use monthly data from January 2000 to December 2006. Our observations are divided into two subsets: the first is used for estimation and the second for the forecasting. Despite the estimated model produced an inexact answer, with a probabilistic output, it is possible to implement an investment strategy, using the herding index as a proxy for investors, which outperforms a buy-and-hold one.