The expectation of the potential of textual data for predicting corporate bankruptcies has grown in recent years. Various studies suggest that factors, e.g. readability and the choice of words, are suitable for differentiating between solvent and ongoing bankrupt companies. We, therefore, derive three hypotheses for demarcation criteria, as well as one hypothesis for language development, from existing research and test them using a dataset consisting of 31 solvent and 31 bankrupt German companies. We show that the influence and usefulness of such parameters are strongly affected by circumstances, e.g. text size but also the time distance to the occurrence of bankruptcy. We thus argue that the differentiated consideration of companies with respect to comparison groups leads to a significantly improved benefit of considering language concerning bankruptcy prediction. Moreover, we show that the analysis of language development is suitable to identify clues about the future financial situation of companies.
Nießner, Tobias; Wiederspan, Olga; and Schumann, Matthias, "Consideration of the Use of Language in Corporate Bankruptcy Prediction: A data analysis on German Companies" (2022). PACIS 2022 Proceedings. 122.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.