Investigating Long-Term Short Pairing Strategies for Leveraged Exchange-Traded Funds Using Machine Learning Techniques
Journal of Investing https://doi.org/10.3905/joi.2023.1.293 Suggested citation: Khadivar, H., Nikbakht, E., & Walker, T. (2024). Investigating long-term short pairing strategies for leveraged exchange-traded funds using machine learning techniques. Journal of Investing, 33(2), 104–135.
41 Pages Posted: 16 Apr 2024 Last revised: 21 Feb 2024
Date Written: March 31, 2023
Abstract
This study examines the profitability of short-selling strategies of different portfolios with varying combinations of bull and bear Leveraged Exchange-Traded Funds (LETFs). We find that while short-selling the combination of both bull and bear LETFs does not yield significant positive returns compared to the market, short-selling a portfolio with only bear LETFs can significantly outperform the market, especially if the position is established after a period of heightened market volatility. Moreover, using machine learning techniques, we show that as the correlation of LETFs with their underlying index increases, the return from short-selling both bull and bear LETFs decreases. At the same time, an increase in the net asset value (NAV) of bull LETFs results in an increase in the return of short-sold bull LETFs and a decrease in the return of short-sold bear LETFs.
Keywords: Exchange-traded funds, Leveraged ETFs, Short pairing strategy, Machine learning
JEL Classification: G11, G12
Suggested Citation: Suggested Citation