Changes in Aggregate Ponzi Interests and Commercial Banks'Non-Performing Loan Forecasting
XIE Deren,SHI Xuezhi
DOI: https://doi.org/10.2139/ssrn.4976544
2024-01-01
SSRN Electronic Journal
Abstract:The emerging literature suggests that the accounting information disclosed by micro-enterprises is useful for macro forecasting.Previous studies have mainly analyzed the relevance of aggregate accounting information for predicting macroeconomic indicators,such as future GDP growth,job creation and destruction,monetary policy,and inflation(Konchitchki and Patatoukas,2014a;Crawley,2015;Luo et al.,2016;Shivakumar and Urcan,2017;Rouxelin et al.,2018;Ma and Zhang,2020;Ye et al.,2020;Hann et al.,2021).However,little is known about the usefulness of aggregate accounting information in predicting meso-level indicators,especially the credit risk of the regional banking industry. The non-performing loan ratio is a leading indicator of an impending banking crisis(Reinhart and Rogoff,2011)and can effectively reflect a bank's risk status(Fang,2015).Thus,this paper investigates the predictive value of changes in aggregate Ponzi Interests for the growth of non-performing loans in commercial banks.This study uses data from publicly listed companies to construct a measure of firm Ponzi Interests,considering the primary repayment source and subsequently aggregating this measure to the provincial level.We measure Ponzi Interests by tracking the financial sources of firms'interest payments over an extended period and separating the component of total cumulative interest payments covered by financing cash inflows. Consistent with our conjecture,we find that changes in aggregate Ponzi Interests effectively predict the subsequent growth of non-performing loans in commercial banks.We observe a significantly positive association between changes in aggregate Ponzi Interests and the subsequent growth of non-performing loans in commercial banks at the provincial level.This relationship is attributed to the fact that our Ponzi Interests indicator reflects the reliability of a firm's primary repayment source.Firms with a deepening degree of Ponzi Interests will have a higher operating risk and likelihood of experiencing financial distress.Furthermore,the increase in non-performing loans in commercial banks is primarily driven by firms experiencing a deterioration in their financing capacity and asset profitability as well as firms lacking affiliations with banks.Moreover,the findings of a cross-sectional analysis reveal that the forecasting power of the aggregate Ponzi Interests indicator is more pronounced in provinces where Ponzi financing sustainability is poorer and banks have less incentive and ability to conceal non-performing loans. Our paper contributes to the literature in three aspects.First,it introduces an innovative approach by constructing a regional aggregate Ponzi Interests indicator based on the primary repayment source of firms.In addition,this study establishes the relationship between changes in aggregate Ponzi Interests and the subsequent growth of non-performing loans in commercial banks.By delving into enterprise cash flow and the ability to pay interest,our paper extends the research on the meso and macro predictive value of accounting information(Konchitchki and Patatoukas,2014a;Crawley,2015;Luo et al.,2016;Shivakumar and Urcan,2017;Rouxelin et al.,2018;Ma and Zhang,2020;Ye et al.,2020;Hann et al.,2021). Second,this paper enhances the literature on the prediction of firms'financial distress and default risk.Previous studies in this field have mainly focused on the predictive value of financial indicators(Altman,1968),short-term cashflow(Aziz et al.,1988),and market return(Beaver,1966)information.Our study provides evidence that the Ponzi Interests indicator based on long-term cash flow information can be effectively used to predict corporate default risk. This paper complements the literature on bank risk-taking determinants.We examine how accounting information about the reliability of a firm's primary repayment source influences a bank's passive risk-taking,thereby affecting changes in non-performing loans.Given that the literature has mainly focused on the effects of macroeconomic conditions and bank characteristics on banks'passive risk-taking,this paper provides additional insights from the perspective of firms acting as loan demanders. This paper has crucial policy implications and practical relevance.Our results suggest that aggregate Ponzi Interests derived from micro-enterprises'accounting data contain valuable information regarding future credit risk and expected credit losses associated with bank credit assets.These findings imply that regulators should consider the Ponzi Interests indicator to enhance their credit supervision.In addition,commercial banks should incorporate the Ponzi Interests indicator in their specific loan approval procedures and expected credit loss models.