Template-type: ReDIF-Article 1.0
Author-Name: Anna Burova
Author-Email: burovaab@mail.cbr.ru
Author-Workplace-Name: Bank of Russia
Author-Name: Henry Penikas
Author-Email: penikasgi@mail.cbr.ru
Author-Workplace-Name: Bank of Russia, Higher School of Economics, Lebedev Physics Institute
Author-Name: Svetlana Popova
Author-Email: popovasv@mail.cbr.ru
Author-Workplace-Name: Bank of Russia
Title: Probability of Default Model to Estimate Ex Ante Credit Risk
Abstract: A genuine measure of ex ante credit risk links borrower's financial position with the odds of default. Comprehension of a borrower's financial position is proxied by the derivatives of its filled financial statements, i.e. financial ratios. We identify statistically significant relationships between shortlisted financial ratios and subsequent default events and develop a probability of default (PD) model that assesses the likelihood of a borrower going into delinquency at a one year horizon. We compare the PD model constructed against alternative measures of ex ante credit risk that are widely used in related literature on bank risk taking, i.e. credit quality groups (prudential reserve ratios) assigned to creditors by banks and the credit spreads in interest rates. We find that the PD model predicts default events more accurately at a horizon of one year compared to prudential reserve rates. We conclude that the measure of ex ante credit risk developed is feasible for estimating risk-taking behaviour by banks and analysing shifts in portfolio composition.
Classification-JEL: E44, E51, E52, E58, G21, G28
Keywords: ex ante probability of default, corporate credit, credit registry, probability of default model, credit quality groups, credit spreads
Journal: Russian Journal of Money and Finance
Pages: 49-72
Volume: 80
Issue: 3
Year: 2021
Month: September
DOI: 10.31477/rjmf.202103.49
File-URL: https://rjmf.econs.online/upload/iblock/e9b/Probability_of_Default_Model.pdf
Handle: RePEc:bkr:journl:v:80:y:2021:i:3:p:49-72