In this paper, we propose a novel approach in stock clustering with the purpose of the construction of a portfolio optimization strategy. The idea is to exploit hierarchical Neural Network Principal Component Analysis and Adaptive LASSO in combination with the Arbitrage Pricing Theory in order to group stocks whose returns are affected by the same risk factors, and then eliminate such dependence through an appropriately constructed portfolio. We test our proposal on the Italian stock market.
A Machine Learning Approach in Stock Risk Management / Cuomo, Salvatore; Gatta, Federico; Giampaolo, Fabio; Iorio, Carmela; Piccialli, Francesco. - (2021), pp. 308-311. (Intervento presentato al convegno 13th Scientific Meeting of the Classification and Data Analysis Group (CLADAG 2021)) [10.36253/978-88-5518-340-6].
A Machine Learning Approach in Stock Risk Management
Salvatore Cuomo
;Federico Gatta;Fabio Giampaolo;Carmela Iorio;Francesco Piccialli
2021
Abstract
In this paper, we propose a novel approach in stock clustering with the purpose of the construction of a portfolio optimization strategy. The idea is to exploit hierarchical Neural Network Principal Component Analysis and Adaptive LASSO in combination with the Arbitrage Pricing Theory in order to group stocks whose returns are affected by the same risk factors, and then eliminate such dependence through an appropriately constructed portfolio. We test our proposal on the Italian stock market.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
