Failures in France: the slowdown hides a still high level

A calmer pace, a still fragile situation

Altares published on October 13, 2026 its assessment of French companies in difficulty for the third quarter. The group records 14,652 procedures between July and September, a 0.5 % increase over the previous year. Over twelve months, its count approaches 72,000 companies. The message is therefore double: the acceleration of failures has slowed sharply, while the volume remains exceptionally high. A less rapid rise does not mean a return to a normal situation.

Altares highlights the persistent vulnerability of small firms and young companies. Activities dependent on household purchases remain under pressure, despite improvements in several sectors and regions. This snapshot concerns the entire economic fabric. It does not, on its own, allow measurement of the difficulties of French crypto firms nor attribute any possible fragility to a single cause such as regulation, credit, or the price of Bitcoin.

Why the figures can differ across publications

Another official series, published by the Bank of France on October 9, counted 70,766 failures over twelve months up to the end of August 2026. This total was very close to that of July, revised to 70,692. The central bank also described a stabilization at a high level. It mentioned a deteriorated economic climate and successive shocks that have weakened the finances of certain companies, while reminding that the total population of companies continues to grow.

These numbers should not be subtracted to invent a spike in a single month. The reporting dates and accounting scopes differ. The Bank of France includes judicial reorganizations and judicial liquidations associated with a cessation of payments; its definition notably excludes safeguards. The Altares table also presents this latter category. Records can also be revised when judgments are transmitted later.

Correctly comparing the series therefore requires keeping the same definition and the same period. An annual total adds twelve months of events, whereas a quarterly figure covers only three months. It does not describe the number of companies that close each day, nor their exact economic weight. Two close publications can support a general diagnosis without producing strictly identical totals.

A procedure does not tell the whole story of a company

Judicial reorganization aims to organise the continuation of activity, preserve jobs and handle debts when the company is in cessation of payments. An observation period allows examination of possibilities for continuation or takeover. Liquidation corresponds to a situation where recovery appears clearly impossible. Presenting all cases as definitive closures that have already occurred would erase this essential distinction between the different judicial outcomes.

For the digital economy, our reading is that these data mainly illuminate the environment of clients, suppliers and investors. A company may have promising technology while lacking cash to meet its deadlines. Conversely, a high national indicator does not prove that a particular platform is insolvent. Such a claim requires its own accounts, identifiable judicial decisions and precise information on its situation.

The next statistical challenge will be to see whether this stabilization extends into the fourth quarter. Subsequent publications will allow comparison of evolutions on a constant scope, distinguishing procedures and sectors. At this stage, October's data depict a slowdown in the degradation, with an annual stock of difficulties still heavy for companies and their partners.