Highlighting weak signals statistics on the short-term credit risk of SMEs

Scientific project

The aim is to determine variables present in weak signals allowing to assess the financial wellbeing and the risk associated to companies in the Partners perimeter. More precisely, ACOFI Gestion and Euler Hermes wish to access new analytic tools in order to assess the predictive default probability over a set horizon (for instance a year). Even though the analysis could cover big companies, it focuses on small and medium sized companies named SME. These companies are not listed on the stock exchange or when they are, it is hardly useful because of a lack of liquidity and transactions, and they are generally not given a rate by rating agencies. Hence, common models assessing default probability generally do not apply to SMEs. Yet, SMEs play a central and increasing role in the world’s economy. About 97% of registered companies in OCDE countries are SMEs. In 2011, SMEs represented 99% of European companies and 2/3 of European employment, the net contribution to employment growth being about 58%.


The partners wish to use flexible statistics analysis tools that allow to obtain a forecast of the default probability without having to assume a structural financial model below. The idea is to get reliable predictions, easily up-dated. According to the needs the analysis should cover various sectors, use new available data sources, standards (company’s financial data, macroeconomics data…) or non-standards if needed. The analysis should also allow the creation of risk-evolution scenarios connected to a specific SME or a group of SMEs following the evolution of the default probability predictors.

Scientific officer

Valentin PATILEA
Valentin PATILEA

Economic Partners