Weak Signals" system to identify companies in difficulty

Over the years, numerous administrative data have been added to the Signaux faibles application, created in 2016. By cross-referencing these various data, such as unpaid social security contributions, requests for short-time working or downsizing, Bercy's algorithm should enable companies in difficulty to be detected more quickly.

A system "unique in the world", according to Bercy

With the gradual withdrawal of various public support measures for businesses, many of them are likely to find themselves in serious difficulty, particularly in the tourism, automotive and aeronautical sectors.

If they are to continue to benefit from certain types of aid, it is essential that the authorities are able to identify them. With this in mind, French Finance Minister Bruno Le Maire mentioned the existence and usefulness of the Signaux faibles application when announcing a new business support plan.

This tool is not new, since its birth actually dates back to 2016. However, over the past 5 years, it has been enhanced with numerous tax and financial data, initially on a local scale, in Bourgogne-Franche-Comté, then on a national scale from 2019.

The different warning signals

Before the onset of the health crisis, the data used to identify companies in difficulty were relatively "classic": they included debts to URSSAF, and requests for short-time working.

However, the crisis and the various measures put in place to support businesses have changed the situation. Short-time working has been used on a massive scale, and business leaders have made extensive use of the opportunities for deferring or waiving contributions.

New data has been added to enable the application to identify risk situations more effectively. For example, it now includes monthly VAT payments, as well as many other items supplied by :

  • URSSAF,
  • Direction générale des entreprises,
  • Banque de France,
  • or the Direction générale des finances publiques.

Data supplied by the French Ministry of Labor, and even by suppliers themselves in the case of payment terms, will be gradually integrated into the application.

From all this data, which is made available to a large number of public agents, the algorithm draws up two lists:

  1. a red list of all companies considered to be in difficulty,
  2. a yellow list for the others.

Only data from companies with at least 10 employees is taken into account, as the algorithm gives more random results below this threshold.

As soon as a company enters the red list, a public official is responsible for contacting it and directing it to the departmental crisis exit committees, for appropriate support. The Ministry of the Economy has announced its intention to pay particular attention to the situation of companies whose debt is increasing, even as their cash flow continues to decline. According to the Banque de France, these currently represent between 5 and 8% of French companies.