Taking the Twists into Account: Predicting Firm Bankruptcy Risk with Splines of Financial Ratios

Abstract

Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. The Working Paper series presents reports on matters in the sphere of activities of the Riksbank that are considered to be of interest to a wider public. The papers are to be regarded as reports on ongoing studies and the authors will be pleased to receive comments. Abstract We demonstrate improvements in predictive power when introducing spline functions to take account of highly non-linear relationships between …rm failure and earnings, leverage, and liquidity in a logistic bankruptcy model. Our results show that modeling excessive non-linearities yields substantially improved bankruptcy predictions, on the order of 70 to 90 percent, compared with a standard logistic model. The spline model provides several important and surprising insights into non-monotonic bankruptcy relationships. We …nd that low-leveraged and highly pro…table …rms are riskier than given by a standard model. These features are remarkably stable over time, suggesting that they are of a structural nature. Terms of use: Documents in

Trade Credit and the Propagation of Corporate Failure: An Empirical Analysis

“…Our baseline model incorporates creditor characteristics that have been shown to be important determinants of firm failure, such as capital structure, cash and liquid asset holdings, profitability, size, and age (see, e.g., Shumway (2001), Campbell, Hilscher, and Szilagyi (2008), Jacobson, Lindé, and Roszbach (2013), Giordani, Jacobson, von Schedvin, and Villani (2014)). Financial ratios in accounting data are typically, for a given firm, highly persistent over time, which may introduce bias in variance estimates.…”

“…18 See Giordani et al (2014) for a detailed overview of the applied interpolation procedure. The shares of shorter (less than 12 months) and longer (more than 12 months) statement periods are both around 5 percent.…”

“…To control for firm-specific failure risk, V i t includes data on firms' capital structure, cash and liquid asset holdings, profitability, size, and age. 24 These are variables that are documented as important determinants of firm failure (see, e.g., Shumway (2001), Campbell, Hilscher, and Szilagyi (2008), Jacobson, Lindé, and Roszbach (2013), Giordani et al (2014)). Table III shows results for Equation (2) where the TDF variable is included to control for exposures to trade debtor failures, estimated for the baseline (Panel A) and extended (Panel B) sample periods.…”

“…(2) where the Claims/Assets variable is included to measure the exposure of trade debtor failures) for the period 2007-2011. Column (I) reports results from a spline version of the debtor-fixed effects model reported in Column (V), Table V. In the additive spline model, we include knots at the 20th, 40th, 60th, and 80th percentiles of each control variable (the model corresponds to a LPM version of Equation (5) in Giordani et al (2014)). Column (II) reports results for a debtor-fixed effects model including additional control variables related to the asset turnover ratio and net working capital.…”

Generalized smooth finite mixtures

“…The scatter for profit is especially peculiar, suggesting quite different relations between leverage and profits when profits are positive (strong negative relation) from when profits are negative (constant or slightly positive relation). Strong non-linearities seem to be a quite general feature of balance sheet data, and there have been a few recent attempts with nonlinear/nonparametric models for these data, see Giordani et al (2011) for an additive spline approach and Bastos and Ramalho (2010) for an application with regression trees.…”