Long-term dependency between sovereign bonds and sectoral indices of India : evidence using Hurst exponent and wavelet analysis

Purpose: The purpose of this study is to provide a new way to optimize a portfolio and to show that combining the Hurst exponent and wavelet analysis may help to increase portfolio returns. Design/methodology/approach: The authors use the Hurst exponent and wavelet analysis to study the long-term dependencies between sovereign bonds and sectoral indices of India. The authors further construct and evaluate the performance of three portfolios constructed on the basis of Hurst standard deviation (SD) – global minimum variance (GMV), most diversified portfolio (MDP) and equal risk contribution (ERC). Findings: The authors find that an ERC portfolio generates positive superior return as compared other two. Since our sample includes periods of two crisis – post-2007 financial crisis and the ongoing pandemic, this study reveals that combining government bond with equities and gold provides a higher returns when the portfolios are constructed using the risk exposures of each asset in the overall portfolio risk. Practical implications: The findings provide guidance to portfolio managers by helping them to select assets using the Hurst approach and wavelet analysis thereby increasing the portfolio returns. Originality/value: In this study, the authors use a combination of Hurst exponent and wavelet analysis to understand the long-term dependencies among various assets and provide a new methodology to optimize a portfolio. As far as the authors’ knowledge, no study in the past has attempted to provide a joint framework for portfolio optimization and therefore this study is the first to apply this methodology.