The Use of Fuzzy Logic in the Enterprise’s Fundamental Strength Assessment

Abstract

The main aim of the paper is to present a practical application of an original enterprise's fundamental strength assessment fuzzy model, as an alternative to dominant in the literature solutions based on taxonomic methodology. Proposed model is an attempt to combine the fuzzy methodology with financial ratios (both in static and dynamic form) and in contrast to the most popular solutions in this area, it covers wider range of enterprise's performance dimensions, in particular financial condition, financial results, cash flows and operating efficiency, as well as development potential. Such a comprehensive approach to enterprise's fundamental strength assessment makes it more resistant to changes of business conditions, and thus should be more objective in comparison with taxonomic models mostly based on several, often limited in terms of information, financial ratios. In addition, it should be noted, that the use of fuzzy logic allows for significant flexibility in both-the modelcreating step and during its later use. As a result, the proposed solution is relatively easy to adapt to a particular specificity of enterprises and can be further modified/developed. Practical application of the model has been presented in the paper on the example of IT companies listed on the Warsaw Stock Exchange.

Fuzzy Model for Detection of Fraudulent Financial Statements: A Case Study of Lithuanian Micro and Small Enterprises

“…The fuzzy model proposed is intended to be used as a tool for managerial decisionmaking at enterprises. Meanwhile, the models proposed by other researchers are based on the listed enterprises and focus more on support to the investment process (Nawrocki, 2018). The fuzzy model proposed here may help the owners and managers of micro and small enterprises detect fraud of business partners and reduce the business risk.…”

“…The key issue was to identify how the respective researchers use various AI techniques to solve financial problems. Following the systematic literature review, the AI methods used for detection of fraudulent financial statements were identified, such as fuzzy logic (Arora and Saini, 2013;Maltoudoglou et al, 2015;Antonelli et al, 2016;Omar et al, 2017;Nawrocki, 2018;Čičak and Vašiček, 2019;Tiwari et al, 2020), neural networks (Antonelli et al, 2016;Omar et al, 2017;Tiwari et al, 2020), neuro-fuzzy (Arora and Saini, 2013;Antonelli et al, 2016;Rostamy-Malkhalifeh et al, 2021) and decision trees (Arora and Saini, 2013;Chen et al, 2014;Lin et al, 2015;Antonelli et al, 2016;Omar et al, 2017;Zhang, 2020). Hence, previous empirical studies on these AI methods were analysed for finding an appropriate method to us.…”

“…This enables the researchers to model complex systems and use the existing knowledge and experience (Čičak and Vašiček, 2019). In summary, various empirical studies (Huda et al, 2015;Nawrocki, 2018;Čičak and Vašiček, 2019) apply a logical scheme to the financial research based on fuzzy logic. When deciding on the software to be used for fuzzy logic application, certain researchers (Maltoudoglou et al, 2015;Čičak and Vašiček, 2019;Rostamy-Malkhalifeh et al, 2021) tend to opt for MATLAB™ software as it contains several tools for applying of fuzzy logic (Čičak and Vašiček, 2019).…”

“…The Mamdani method seems to be more appropriate and flexible when used for the intuitive models based on human reasoning (Čičak and Vašiček, 2019). Therefore, the Mamdani method is preferred in financial research (Nawrocki, 2018;Čičak and Vašiček, 2019).…”

“…In practice, the fuzzy models usually operate with the IF-THEN-type rules when the input and output variables are specified (Nawrocki, 2018;Čičak and Vašiček, 2019). The output values are based on possible combinations of these weighted rules (Čičak and Vašiček, 2019).…”