High Technology ETF Forecasting: Application of Grey Relational Analysis and Artificial Neural Networks

Abstract

This study employs the grey relational analysis model and provides robust identification of the S&P 500 stock index as having the greatest influence on exchange-traded funds (ETFs). The subsequent influencing factors are the volatility index (VIX), commodity research bureau (CRB) index, Brent crude oil index, put-call ratio, and trade index (TRIN). Our results show that the back propagation network model outperforms the recurrent neural network model in predicting both high technology and non-high technology ETFs. The low grey relational grade (GRG) variables (i.e., put-call ratio, TRIN and crude oil index) have greater influence than the group of high GRG variables (i.e., S&P 500 stock index, VIX, and CRB index) and the group of all variables in high technology ETFs, while on non-high technology ETFs, the all variables group showed stronger influence.

Keywords: high technology and non-high technology ETFs; grey relational analysis; artificial neural network

JEL Classification: E27, F47, G17