Open AccessBUSINESS PERFORMANCE AND FINANCIAL HEALTH ASSESSMENT THROUGH ARTIFICIAL INTELLIGENCE
Author(s) -
Tomáš Krulický,
Jakub Horák
Publication year - 2021
Publication title -
ekonomicko-manažérske spektrum/economic and managerial spectrum
Language(s) - English
Resource type - Journals
eISSN - 2585-7258
pISSN - 1337-0839
DOI - 10.26552/ems.2021.2.38-51
Subject(s) - competitor analysis , self organizing map , shareholder , business , asset (computer security) , marketing , finance , cluster analysis , computer science , corporate governance , artificial intelligence , computer security
Research background: Globalisation and the development of technology introduce new requirements for effective business management. Every business must constantly adapt to the environment, analyse and know its competitors and its customers’ requirements, and meet other stakeholders’ commitments. An unsuccessful business will go into liquidation. The intention of any business should be not only to avoid this situation, but to thrive and prosper and create value for its shareholders. Purpose of the paper: The aim of this study is to propose an appropriate tool for cluster analysis and determine the ability of a business to survive a potential financial distress. Methods: Details from financial statements of construction companies operating in the period 2015-2019 in the Czech Republic are analysed. Attention is mainly directed to items that represent the capital and asset structures of a company, liquid assets, and the ability to generate sales and profit. Artificial neural networks in the form of Kohonen networks are used for the purpose of cluster analysis. Financial analysis is used to examine the underlying dataset as well as for a detailed analysis of selected clusters, i.e. the contribution margin and ratio indicators. Findings & Value added: The basic analysis clearly shows that companies in liquidation attempt to reduce the value of inventories and engage additional foreign capital with a view to survival, while there is a certain solidarity between companies’ key persons. Cluster analysis using Kohonen networks is quite successful. The present methodology and approach can still be applied to the design of an enterprise decision support tool. Further research may study whether the representation of businesses in the different clusters will change over time, or whether the development of the construction industry can indeed be predicted based on an analysis of the leaders.