Incorporating Graph Permanent in Forecasting Financial Performance

Forecasting financial performance metrics plays a pivotal role in guiding decision-making processes for the management, investors, creditors, regulators and other stakeholders of every business entity. Multiple linear regression is a widely used popular method to forecast financial performance metrics. But multiple linear regression fails to capture the interrelationships and correlations among various factors influencing financial performance metrics. It’s a notable limitation of multiple linear regression. In this research paper, we have proposed a graph permanent based technique named networked regression that integrates graph theory with regression analysis to model the dynamic relationships among the explanatory variables and to offer a more comprehensive understanding of their collective impact on the response variable. We demonstrate that networked regression, when combined with multiple linear regression, can offer more comprehensive insights and predictive capability. Thus by integrating graph theoretic concepts in regression analysis, our proposed approach contributes to advancing the field of financial performance forecasting in dynamic financial environments. In this research work, we have conducted experiments on the financial datasets of various companies like PepsiCo, P&G and Nestlé.