Novel Portfolio Construction Based on Traditional Stock Index
Conghao Jin
2024
Abstract
Novel portfolio construction with great performances under controllable risks is always pursued by the finance industry. This study explores a novel approach to portfolio investment based on an active management strategy, centered on uncovering stocks that have yet to receive sufficient market attention despite their robust fundamentals, or individual stocks whose prices have deviated from their intrinsic values due to market contingencies, resulting in abnormal fluctuations. During the research process, a stock pool comprising the constituents of the CSI 1000 Index was first constructed. These stocks, though relatively small in market capitalization, exhibit good liquidity and receive limited market attention, providing an ideal environment for the application of portfolio strategies. The research findings reveal that all three machine learning algorithms, i.e., Gradient Boosting Decision Tree (GBDT), Support Vector Machine (SVM), and Random Forest (RF), achieve stable excess returns on the CSI 1000 Index. Among them, GBDT performs best in terms of overall returns, followed by SVM and RF in third place. From a risk control perspective, RF exhibits the lowest maximum peak-to-trough decline, indicating strong risk resilience, while GBDT and SVM follow closely. These results creatively integrate machine learning algorithms from the field of artificial intelligence into portfolio construction strategies, with a focus on the CSI 1000 Index, aiming to enhance investment efficiency through technological means.
DownloadPaper Citation
in Harvard Style
Jin C. (2024). Novel Portfolio Construction Based on Traditional Stock Index. In Proceedings of the 1st International Conference on E-commerce and Artificial Intelligence - Volume 1: ECAI; ISBN 978-989-758-726-9, SciTePress, pages 134-142. DOI: 10.5220/0013208000004568
in Bibtex Style
@conference{ecai24,
author={Conghao Jin},
title={Novel Portfolio Construction Based on Traditional Stock Index},
booktitle={Proceedings of the 1st International Conference on E-commerce and Artificial Intelligence - Volume 1: ECAI},
year={2024},
pages={134-142},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013208000004568},
isbn={978-989-758-726-9},
}
in EndNote Style
TY - CONF
JO - Proceedings of the 1st International Conference on E-commerce and Artificial Intelligence - Volume 1: ECAI
TI - Novel Portfolio Construction Based on Traditional Stock Index
SN - 978-989-758-726-9
AU - Jin C.
PY - 2024
SP - 134
EP - 142
DO - 10.5220/0013208000004568
PB - SciTePress