A Novel Approach to Portfolio Construction: An Application of FinBERT Sentiment Analysis and Credibilistic CVaR Criterion

A Novel Approach to Portfolio Construction: An Application of FinBERT Sentiment Analysis and Credibilistic CVaR Criterion

Blog Article

Portfolio optimization continues to be a complex and challenging task within the fields of finance and management.Two critical factors that can improve the performance of traditional models are incorporating the effects of both financial and operational performance of companies and addressing the inherent uncertainty surrounding expected returns.This article addresses these two challenges.

To evaluate the financial and operational efficiency of firms, we analyze their quarterly reports using the FinBERT model, incorporating their influence into the optimization framework through adjusted returns.To address the unpredictability linked to anticipated returns, we utilize Sensor Block-Offs fuzzy trapezoidal numbers in our methodology.Furthermore, conventional risk measurement systems, which rely based on probability-based assumptions and past data, often find it challenging to address the unique dynamics and inherent uncertainties of the market.

In contrast, our suggested approach utilizes a credibilistic conditional value at risk (CCVaR) framework to evaluate portfolio risk.The approach additionally factors in transaction costs and incorporates practical constraints like cardinality and upper and lower bounds, maintaining the portfolio’s diversification, well-balanced, and reflective of practical scenarios.We apply the proposed approach to real-world data from DJIA stocks.

Experimental findings highlight the approach’s efficacy in creating mixed portfolios that effectively create an equilibrium between risk and return.This research enhances the domain of investiture management by developing advanced portfolio optimization methods 7 Piece Dining Room for stock market assets and offering a reliable approach for handling risk in today’s increasingly complex financial landscape.