Predicting Bankruptcy through Neural Network: Case of PSX Listed Companies
DOI:
https://doi.org/10.52131/jom.2022.0402.0080Keywords:
Bankruptcy Prediction, Artificial Neural Network, Logistic Regression, ROAUC, PSX, Non-financial companiesAbstract
The paper reconnoiters if logistic regression (LR) and neural network (NN) can estimate bankruptcy for PSX non-financial companies a year ahead of bankruptcy occurrence; particularly it endeavors to explore how exact LR and NN models are? Financial ratios were utilized forecast the bankruptcy in firms. Empirical results demonstrated that both models have capability to predict the event of bankruptcy with NN outperforming LR model. Although both models possess capability to predict bankruptcy, current research demonstrated that use of neural networks (NN) enhances the precision of prediction by being a superior approach over logistic regression method (this is based on accuracy level achieved earlier by NN over LR). These results will cover the literature gap existent in bankruptcy research in Pakistan especially about NN estimation model, proposing an advanced forecasting with precision as proven through figure 4.1.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2022 Javed Iqbal, Furrukh Bashir, Rashid Ahmad, Hina Arshad
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.