Modeling Financial Distress Risks for SMEs Based on Random Effects Logistic Model
LIANG Qi,GUO Xin-wei,SHI Ning
DOI: https://doi.org/10.3969/j.issn.1004-6062.2014.03.016
2014-01-01
Abstract:Small-and-Medium Enterprises( SMEs) play an indispensable role in increasing employment,encouraging innovation,and promoting economic prosperity.However,appropriate models of measuring credit default risks for SMEs are not available to commercial banks,thus leading to restricting loans for SMEs.Although many domestic researchers have been modeling default prediction,their models mainly focus on large enterprises and are ineffective when applied to SMEs.Logisic model is currently the most popular method in modeling credit default risk.However,its ability to predict out-of-sample default events is limited because of its poor performance in controlling those factors in the model.To solve this problem,we employ a random effects logistic model to improve the estimation accuracy of parameters in logistic model.Besides,most domestic researches consider only financial ratios,but ignore non-financial factors such as corporate governance factors.Some researchers argue that corporate governance may affect the quality of financial information and the effectiveness of measuring credit default risk.In this study,we use financial distress events( Special Treatment events) as proxy indictors for credit default,and employ both financial ratios and corporate governance factors to model a random effects logistic model of credit default for SMEs.In Section I,we introduce the principle and estimation procedure of random effects logistic based on cross sectional data.Under the assumption that the default probability follows Beta distribution,a likelihood function of random effects logistic is derived and the maximum likelihood estimation( MLE) is employed to obtain parameters.In Section Ⅱ,we discuss data source and variables application.In Section Ⅲ,we establish an empirical work under the model specification of random effects logistic model,and compare its prediction power with logistic model.The results show that regardless whether corporate governance indictors are included the random effects logistic model performs better than the logistic model in both discrimination power within sample and prediction power out-of-sample.Robustness test under different critical values further suggests that random effects logistic model has a better prediction power and a higher stability than logistic model.Additionally,we find the prediction power is also improved when model specification considers corporate government indicators,which may eliminate the financial information distortion caused by potential earnings management behaviors.In sum,logistic model can control factors out of model specification due to random effect.Compared with logistic model,the parameter estimation is more stable and the prediction power is higher in random effects logistic model.Prediction power will achieve the highest when both financial ratios and corporate governance indictors are included in random effects logistic model,which also indicates that financial distress risk of SMEs depends on both financial and corporate governance conditions.Corporate governance information should not be ignored when modeling financial distress prediction model for SMEs.Although our study is based only on a limited number of SMEs,the findings can help model financial distress risk and measure credit risk for unlisted SMEs.Based on the estimation of the default probability with the internal rating-based approach( IRB) recommended by Basel Ⅱ,our findings may help commercial banks evaluate credit default risks of SMEs and promote commercial banks to augment loans for SMEs,and thus alleviating SME's financing problems.