Intra and Inter Sectoral Risk Spread and Portfolio Risk Management: Case of S&P 500

Abstract

Abstract A critical issue of diversification in portfolio management is the intra and inter-sectoral spread of risk. The aim of this study is to capture potential intra and inter sectoral risk spread. In this regard, the Bayesian Neural Networks (BNN) model was involved, the method being applied for a portfolio of 12 shares from the American index S&P 500, on the period January 1st, 2011 – January 28th, 2023. The expected shortfall was involved as a risk estimation measure and the Tabu Search learning algorithm of BNN was employed. The robustness of the results was tested at three significance thresholds namely, 0.85, 0.90 and 0.95. The differences of the results were highlighted on two subsamples, from January 1st, 2011 – December 31st, 2019, and January 1st, 2020 – January 28th, 2023. Evidence of intra and inter sectoral contagion was found on the full sample period, however during the financial turmoil period represented by the last sub-sample, the results display that the linkages between different sectors weakened to certain extents. The key contribution stands in the significant implications for portfolio risk management. The results highlight and strengthen the importance of building a portfolio based on an accurate selection of assets that are not inter-linked.