Monitoring Privately-held Firms' Default Risk in Real Time: A Signal-Knowledge Transfer Learning Model
Electronic Access:
Free Download. Use the free Adobe Acrobat Reader to view this PDF file
Disclaimer: IMF Working Papers describe research in progress by the author(s) and are published to elicit comments and to encourage debate. The views expressed in IMF Working Papers are those of the author(s) and do not necessarily represent the views of the IMF, its Executive Board, or IMF management.
Summary:
We develop a mixed-frequency, tree-based, gradient-boosting model designed to assess the default risk of privately held firms in real time. The model uses data from publicly-traded companies to construct a probability of default (PD) function. This function integrates high-frequency, market-based, aggregate distress signals with low-frequency, firm-level financial ratios, and macroeconomic indicators. When provided with private firms' financial ratios, the model, which we name signal-knowledge transfer learning model (SKTL), transfers insights gained from 35 thousand publicly-traded firms to more than 4 million private-held ones and performs well as an ordinal measure of privately-held firms' default risk.
Series:
Working Paper No. 2024/115
Frequency:
regular
English
Publication Date:
June 7, 2024
ISBN/ISSN:
9798400278396/1018-5941
Stock No:
WPIEA2024115
Format:
Paper
Pages:
45
Please address any questions about this title to publications@imf.org