Bankruptcy Prediction of Swedish SMEs and the importance of CEO characteristics

Abstract: This thesis investigates whether an accounting-based bankruptcy prediction model based on the logit analysis developed by Altman and Sabato (2007) can be enhanced by incorporating CEO specific variables. The study concentrates on Swedish non-listed SMEs by reason that the SME sector is an important engine to the growth of the Swedish economy. Furthermore, the applicability of the original model is evaluated in regard to the new setting of Swedish SMEs as well as whether the inclusion of CEO characteristics increases in importance for smaller firms. The data encompasses 15,885 non-listed Swedish limited liability companies, out of which 226 are failed. By validating the models on an out of sample, it is found that the original model using solely financial ratios is applicable to predict bankruptcy among Swedish non-listed SMEs. Moreover, the classification accuracy is improved when adding the CEO specific variables age and gender, given a predefined cut-off rate. However, the inclusion of CEO specific variables does not gain in importance as firm size decreases.