Portfolio construction and weight optimisation using principal component analysis
by Kartikay Laddha; Vidhi Kapoor; Siba Panda
International Journal of Forensic Software Engineering (IJFSE), Vol. 1, No. 4, 2022

Abstract: Principal component analysis helps in constructing statistical risk factors by explaining the variance inherited from the dataset. It transforms the dataset into new independent components which are uncorrelated and are hence used to optimise portfolios for earning higher returns and more financial control. It helps uncover the underlying drivers hidden in the data by summarising huge feature sets using a few components. In this paper, the method of PCA is used to derive the principal components that capture most of the market volatility and are used to define the cash allocation strategy which helps outperforming the sectoral benchmark index (NIFTY IT) in terms of returns and Sharpe ratio. Our optimised portfolio provides a risk to reward ratio of 1.0605 in comparison to 0.4583 provided by the NIFTY IT index as the weights allocation in our portfolio is explained by the captured variability, which was obtained using the PCA technique. This methodology helps in creating portfolios with reduced dimensionality and variability of the dataset used for practical applications.

Online publication date: Tue, 05-Jul-2022

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Forensic Software Engineering (IJFSE):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com