Transcript Slide 1
Credit Risk Assessment of Corporate Sector in Croatia Saša Cerovac, Lana Ivičić Croatian National Bank Financial Stability Department Structure of the presentation Intro – motivation and credit risk assessment framework Data & definitions Migration matrices Logit model Applications and further steps Objective Modeling credit risk of non-financial businesses entities: assessment and predicting of the rating migration probabilities predicting the probability of being in the default state A contribution to the development of the CNB's technical infrastructure designed for the credit risk assessment (Figure 1) Data sources Two primary databases: CNB’s database with prudential information on bank exposures and exposure ratings (quarterly frequency) Financial Agency (FINA): micro data on corporate financial accounts (annual frequency) Data preparation & cleaning (I) Detailed CNB’s database available since June 2006 full coverage of the banks and detailed risk classification Entries for non-residents, non-corporates, non-market based firms, group of activities and unidentified debtors (other debtors and portfolio of small loans) are removed from the population All exposures towards each single debtor are summed according to their ID number and multiple entries are avoided by prioritizing them according to supervisory actions Data preparation & cleaning (II) Exposures towards small debtors – those not exceeding 100,000 kunas (13,500 euros) - are also removed reducing the volatility steaming from group of debtors that have marginal share in total liabilities of the corporate sector Negative values (“overpayments”) were treated as no exposure Sample was stabilized by removal of enterprises entering and/or exiting the database during the period under observation (year, quarter) Combining the CNB’s and FINA’s databases Some further data reductions took place in the modeling phase due to errors and omissions in FINA’s database Merging CNB’s database with annual financial statements of private non-financial companies obtained from FINA reduced sample dataset to 7,719 firms during 2007 and 2008 (covering more than 75% of bank’s exposures towards market-oriented corporates) Final data set: non-balanced panel of 12,462 observations of binary dependent variable – default state. Construction of credit rating (I) The CNB's database provides only information on the risk classification of individual exposures (placements and off-balance sheet liabilities) - no risk classification of debtors AX - standard A90d – standard, but over 90 days overdue B – substandard (over 90 days overdue) C – delinquent (over 365 days overdue) Construction of credit rating (II) The procedure for classifying debtors into distinct risk categories is based on solving a simple optimization problem derived from the risk classification of their total debt to the banking system as a whole Share of exposure of specific risk category C B A90d AX TOTAL Rating of debtor 50% or more 50% or more 50% or more 100% 100% 100% C B A90d Treshold of 50% maximizes AX rated liabilities to AX rated companies and non-AX rated liabilities to the rest of firms. Distribution of rated debtors from June 2006 to December 2008 AX A90d B C 6,0 7,6 2,1 84,3 Definition of default Following the provisions of the Basel Committee on Banking Supervision (Basel II Accord) and applying general definition of default (Official Journal of the European Union, I.177 p. 113) : Default state: ratings A90d, B or C Rating migrations and the probability of default Migration matrix where • Migration frequency: • Discrete multinomial estimator: • Migrations forecast: • Domestic corporate sector: no absorbing state (reversals are possible); k=4 over horizon Unconditional migration matrices 1-Year AX AX A90d B C A90d 95,0 43,0 10,1 1,7 PR PD B 2,0 22,0 1,8 0,1 C 2,7 32,3 81,9 1,3 0,3 2,6 6,1 96,9 Degree of rating stability 1-Quarter AX AX A90d B C A90d 97,5 40,6 6,0 1,5 1,5 43,6 0,9 0,2 B C 0,9 14,9 90,8 0,8 0,1 0,8 2,3 97,5 Note: Initial rating in rows, terminal rating in columns Conditional matrices I Hypothetical distributions of rating upgrades/downgrades Unconditional distribution Default area Conditional distribution 2 probability Conditional distribution 1 0 rating change Quarterly conditional migration matrices II a. Migration matrices conditional on economic activity b. Migration matrices conditional on economic cycle Industry AX AX A90d B C A90d 97,5 34,6 5,3 1,2 Acceleration phase B 1,5 48,2 0,6 0,3 C 0,9 16,4 91,9 0,8 0,2 0,8 2,3 97,7 AX AX A90d B C A90d 97,2 45,2 6,1 2,3 B 1,7 40,2 1,0 0,2 C 0,9 13,9 90,3 0,8 0,2 0,7 2,6 96,7 Construction Retardation phase AX AX A90d B C A90d 97,5 46,5 8,7 1,7 B 1,5 40,8 1,5 0,0 C 0,9 12,1 87,1 1,3 0,1 0,6 2,8 97,0 AX AX A90d B C A90d 97,8 36,1 5,9 0,7 1,3 47,1 0,9 0,2 B C 0,8 16,0 91,2 0,9 0,1 0,9 1,9 98,2 Non-financial services AX AX A90d B C A90d 97,5 40,9 5,6 1,6 1,5 42,8 0,9 0,2 B C 0,8 15,4 91,4 0,7 0,1 0,9 2,1 97,5 Note: a. Initial rating in rows, terminal rating in columns b. Differences in migration frequencies that are statistically significant (5% level) in relation to the parameters of unconditional matrix are in italic[4]. [4] The t-statistics is derived from binominal standard error. Empirical regularities Probability of default (reversal) in correlation with credit rating 120 Historical evolution of PDs across sectors Distribution of debtors according to their rating Empirical probability of default (1-Y Matrix) 4,0 Empirical probability of default (1-Q Matrix) 100 3,0 Default rates, % 80 % 60 2,0 40 1,0 Industry Construction 20 Non-financial services q4/2008 q3/2008 q2/2008 q1/2008 q4/2007 q3/2007 C q2/2007 B q1/2007 A90d q4/2006 AX q3/2006 0,0 0 One-year forecasts Annual forecast of migartion probabilities Annual Forecast based on 1-Y Migration Matrix AX AX A90d B C A90d 91,4 53,6 18,7 3,4 B 2,4 6,3 2,1 0,2 C 5,4 34,8 68,0 2,4 0,8 5,2 11,0 94,0 Annual Forecast based on 1-Q Migration Matrix AX AX A90d B C A90d 93,1 69,6 21,8 6,2 B 2,5 5,4 1,6 0,4 C 3,9 22,0 68,9 2,9 Note: Initial rating in rows, terminal rating in columns 0,5 2,8 7,8 90,5 Modeling default state Multivariate logit regression Binary dependent variable yi,t explained by the set of factors X The probability that a company defaults is Using the logit function: Share of firms in default across sectors 0,20 0,18 2007 0,16 2008 0,14 0,12 0,10 0,08 0,06 0,04 0,02 0,00 Agriculture and manufacturing Construction and real estate Non-financial service Total Selection of explanatory variables Initial set: Financial ratios: liquidity (16), solvency (23), activity (12), efficiency (7), profitability (27) and investment indicators (1) Size variables Sectoral dummies Time lag: t-1 Correction of outliers: winsorization Selection of explanatory variables Univariate analysis Mean equality test Graphical analysis: scatterplots Univariate logit models: ROC 3 6 5 / a c co u n ts re c e iv a b le tu rn o ve r 20 15 10 5 (sa le s + d e p re cia tio n ) / to ta l a ss e ts C a s h to to ta l a ss e ts 5 4 3 2 1 0 S h a re h o ld e rs ' e q u ity to to ta l a ss e ts Boxplots 6 1.2 -1 0 25 0 0 1 Default 1.0 0.8 0.6 0.4 0.2 0.0 1 -0.2 0 Default 1 Default 6 5 4 3 2 1 0 0 1 Default Scatterplots Shareholders' equity to total assets 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 -0.05 average defualt rate Average default rate Cash to total assets 0.30 0.25 0.20 0.15 0.10 0.05 0.00 0 0.05 0.1 0.15 0 0.2 0.3 0.4 Percentile range avergae Percentile range average 365 / accounts receivable turnover Sales + depreciation to total assets 0.80 Average default rate Average default rate 0.1 0.60 0.40 0.20 0.00 0.50 0.40 0.30 0.20 0.10 0.00 0 0.2 0.4 Percentile range average 0.6 0 2 4 Percentile range average 6 ROC The predictive power of a discrete-choice model is measured through its: Sensibility (fraction of true positives): the probability of correctly classifying an individual whose observed situation is “default” Specificity (fraction of true negatives): the probability of correctly classifying an individual whose observed situation is “no default” ROC curves in univariate analysis Profitability indicators seem to have highest univariate classification ability: AUCs ranging from 0.69 to 0.75 Among liquidity indicators, the best performing is the ratio of cash to total assets Funding structure appears to be a good individual predictor of default too: ratios of equity capital to total assets and to total liabilities reach AUC values of above 0.70 Multivariate models Intermediate choice: 28 financial ratios Numerous models including different groups of variables were tested Final multivariate model was chosen among best performing combinations of 3, 4, 5 and 6 explanatory variables + economic activity dummy Best performing competing models Indicator C Sector Liquidity Financial leverage Activity Profit Size Construction and real estate dummy Cash to short-term liabilities Cash to total assets Shareholders' equity to total assets Shareholders' equity to total liabilities After tax profit + depreciation to debt/365 365 / accounts receivable turnover EBIT to total liabilities Model 3_1 4.41 (0.22) -0.45 (0.06) -0.29 (0.01) -0.23 (0.01) Model 4_1 -0.41 (0.17) -0.26 (0.07) Model 5_1 -0.30 (0.22) -0.24 (0.07) Model 6_1 -0.17 (0.22) -0.28 (0.07) Model 6_4 -0.06 (0.22) -0.30 (0.07) -0.67 (0.04) -0.67 (0.04) -1.87 (0.19) -0.63 (0.04) -1.96 (0.19) -0.65 (0.04) -2.17 (0.20) -0.27 (0.01) 0.10 (0.01) Sales + depreciation to total assets Sales -0.75 (0.04) R2 AUC % of correct 0 % of correct 1 % of total correct 0.18 0.79 71.57 73.21 71.80 0.11 (0.01) -0.17 (0.01) -0.51 (0.05) 0.19 0.79 72.37 71.20 72.22 -0.01 (0.00) 0.19 0.79 71.29 72.99 71.51 0.09 (0.01) -0.14 (0.01) -0.37 (0.05) -0.01 (0.00) 0.20 0.80 74.89 71.20 74.41 -0.04 (0.00) 0.09 (0.01) -0.41 (0.05) -0.01 (0.00) 0.20 0.80 75.90 69.50 75.05 Marginal effects at the means of independent variables Variable Coefficient Marginal effect (M) 1 Std.dev.*M Constant -0,17 Construction and real estate dummy -0,28 -0,02 -0,009 Cash to total assets -0,63 -0,048 -0,020 Equity to assets -1,96 -0,149 -0,032 365/accounts receivable turnover 0,09 0,007 0,003 EBIT to liabilities -0,14 -0,011 -0,011 Sales + depreciation to total assets -0,37 -0,028 -0,029 Sales -0,01 -0,001 -0,0004 Kernel density estimate of default probabilities distribution for defaulted and non-defaulted companies 1000 800 700 Nodefault=1 Nodefault=1 Nodefault=0 600 500 100 Nodefault=0 10 400 1 300 200 0.1 100 0 0.01 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Cross-border lending effects on credit risk distribution "In the presence of the effective credit limits, foreign banks help arrange direct cross-border borrowing for their clients, typically for the most creditworthy large corporates, leaving the Croatian banks mostly with customers with no other sources of financing.” IMF (2008): Republic of Croatia: Financial System Stability Assessment—Update Model application I (debt) Cumulative distribution of debt according to the origin of a creditor a. Cumulative distribution of debt, 2007 b. Cumulative distribution of debt, 2002 100,0 100,0 90,0 90,0 80,0 80,0 70,0 70,0 60,0 60,0 50,0 50,0 40,0 40,0 Domestic creditors only 30,0 Domestic creditors only 30,0 Dominantly domestic creditors Dominantly domestic creditors Dominantly foreign creditors 20,0 Dominantly foreign creditors 20,0 10,0 10,0 0,0 0,0 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 Percentiles 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 Percentiles Model application II (debtors) Cumulative distribution of debt according to the origin of a creditor c. Cumulative distribution of debtors, 2007 d. Cumulative distribution of debtors, 2002 100,0 100,0 90,0 90,0 80,0 80,0 70,0 70,0 60,0 60,0 50,0 50,0 40,0 40,0 Domestic creditors only 30,0 Dominantly domestic creditors Dominantly foreign creditors 20,0 Domestic creditors only 30,0 Dominantly domestic creditors Dominantly foreign creditors 20,0 10,0 10,0 0,0 0,0 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 Percentiles 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 Percentiles Further steps Refinements of the approach: Searching for alternative definitions of default Applying alternative estimators and modeling conditionality of ratings dynamics Examining alternatives for the selection of explanatory variables Correcting for selection bias using multinomial logit Modeling the event of default (PD) Modeling the event of reversal (PR) Improving explanatory power using macroeconomic variables (contingent on longer data series) Model applications: Forecasts of EAD Stress-testing of the corporate sector Credit risk assessment in the Croatian National Bank Macro approach Sensitivity of NPLR's Macroeconomi c risk model Corporate credit risk models Sectoral approach Households credit risk Bank failiure model EWS CAMELS downgarde model Linear probability model (LOGIT) Migration matrices Credit deafult Sensitivity of financial margin Capital adequacy