Group sparse enhanced indexation model with adaptive beta value. Issue 10 (3rd October 2022)

• Record Type: Journal Article
• Authors: Xu, Fengmin; Ma, Jieao; Lu, Haibing
• Is Part Of: Quantitative finance. Volume 22:Issue 10(2022)
• Page Start: 1905
• Page End: 1926
• Publication Date: 2022-10-03
• Subjects: Enhanced indexation -- Group sparse -- Beta value -- Data-driven learning; C51 -- C53 -- C61 -- G11; Finance -- Periodicals; Business mathematics -- Periodicals; Finance -- Mathematical models -- Periodicals; Investments -- Mathematics -- Periodicals; Economics -- Periodicals; Finances -- Modèles mathématiques -- Périodiques; 332.015118
• DOI: 10.1080/14697688.2022.2092542
• Languages: English
• ISSNs: 1469-7688

Abstract

Abstract : Enhanced indexing, which has been used by professional portfolio managers for decades, is a portfolio management strategy that attempts to increase returns by building a portfolio around core, index-like positions and adding tactical tilts toward specific styles or individual stocks. This paper proposes an improved enhanced indexation model by considering the systematic risk, measured by Beta value, and the industry rotation phenomenon. The systematic risk is the risk related to the stock market as a whole and can be reasonably controlled to improve portfolio performance, by actively tracking and forecasting the market trend. Sector rotation refers to the investment strategy of taking money that's invested in one stock market industry and moving it to another, by taking advantage of the historical performances of specific industries during different phases of the cycle. In specific, our model aims to find a small set of industries that is mostly likely to thrive in the anticipated future, which is mathematically realized by dividing stocks into industries and minimizing their L 2, 1 norm. To evaluate our strategy, we conducted extensive numerical experiments against some major world indices, e.g. CSI 300, S&P 500, FTSE 100 and Nikkei 225. The experimental result shows that our approach can generate sparse portfolios with excellent out-of-sample excess returns and high robustness after deducting transaction costs.