۱- Introduction

۲- Methodology

۳- The effect of dimensionality reduction on stock selection with cluster analysis in different market situations

۴- A stock-selection rotation strategy based on the effect of dimensionality reduction

۵- Conclusions

References

Abstract

Dimensionality reduction is inevitable in stock selection with cluster analysis. Considering relations among dimensionality reduction, noise trading, and market situations, we empirically investigate the effect of dimensionality-reduction methods–principal component analysis, stacked autoencoder, and stacked restricted Boltzmann machine–on stock selection with cluster analysis in different market situations. Based on the index fluctuation, the market is divided into sideways and trend situations. For the CSI 100 and Nikkei 225 constituent stocks, experimental results show that: (1) In sideways situations, dimensionality reduction hardly improves the performance of stock selection with cluster analysis; (2) the advantage of dimensionality reduction is mainly reflected in trend situations, but whether it is in an up or down trend depends on the market analyzed. More importantly, according to the above findings and assuming that the dimensionality-reduction effect will continue, we propose a rotation strategy with and without dimensionality reduction. The results of experiments show that the proposed rotation strategy outperforms the stock market indices as well as the stock-selection strategies based on dimensionality reduction and cluster analysis. These findings offer practical insights into how dimensionality reduction can be efficiently used for stock selection.

Introduction

Stock selection is a crucial issue in investment management, which determines the return of stock investments (Markowitz, 1952; Ren et al., 2017). There are various stock-selection strategies, including multi-factor models (Carvalho et al., 2010; Fama and French, 2018), momentum and contrarian strategies (Grinblatt et al., 1995; Cooper et al., 2004), style rotation strategies (Lucas et al., 2002; Ahmed et al., 2002), volatility strategies (Chong and Phillips, 2012; Hsu and Li, 2013), and behavior biases strategies (Huang et al., 2011). Among these strategies, multi-factor models are the most studied, mainly including the Fama-French three-factor model (Fama and French, 1992), the Fama-French five-factor model (Fama and French, 2017), factor models based on investor attention (Li and Yu, 2012), and factor models based on fundamental and technical analysis (Peachavanish, 2016). Investors can use these models to analyze stock characteristics from different perspectives. If stock characteristics last for a period, investors would obtain a higher benefit from analyzing stock characteristics than from random selection.