Download

Download Full Text (982 KB)

Description

Financial distress and bankruptcy in construction firms can have significant economic consequences, making early detection of financial risk an important research problem. This project investigates whether financial ratios can be used to predict corporate bankruptcy up to two years in advance. The study focuses on publicly traded construction companies and analyzes financial statement data collected from SEC filings. Key financial variables include revenue, earnings before interest and taxes (EBIT), net income, total assets, total liabilities, current assets, current liabilities, inventory, shareholder equity, and debt obligations.

Using these variables, several financial ratios are calculated to evaluate company performance and financial stability, such as working capital to total assets, EBIT to total assets, net income to total assets, and debt to total assets. These ratios are then used to compute a bankruptcy prediction metric based on the Altman Z-Score framework. The dataset includes both companies that later filed for bankruptcy and companies that remained operational, allowing for comparison between financially distressed and healthy firms.

The objective of the project is to build a predictive model that can identify financial warning signs prior to bankruptcy. By analyzing patterns in the financial ratios and comparing Z-Scores across firms, the research evaluates whether the model can successfully classify companies into distress or non-distress categories. The results contribute to understanding how financial indicators can support early risk detection in the construction industry.

Publication Date

4-30-2026

Financial Data to Decision: Bankruptcy Prediction Using Machine Learning vs Altman Z-Score

Share

COinS