Working Paper Series no. 823. Taking advantage of detailed firm-level data on VAT returns, we estimate the monthly impact of the Covid-19 crisis on the turnover of more than 645,000 French firms. Our approach, based on a micro-simulation model, is innovative in a triple way. Firstly, we quantify the activity loss with respect to a counterfactual situation in which the crisis would not have hit. Secondly, we estimate this shock at the firm level, enabling a thorough analysis of activity loss heterogeneity throughout the crisis. In particular, we shade light on the dispersion of the shock both within and between industries. We show that the industry the firm operates in explains up to 48% of the monthly activity shocks’ variance weighted by employment, a much larger share than in a normal year. Finally, we leverage our monthly firm-level data on sales to show how corporate activity has evolved along four distinct trajectories throughout 2020. The main determinant of belonging to a given profile of activity is the firm industry – defined at a very granular level. Conditional on industry, the activity trajectory is also correlated with the ability to adapt some firms have demonstrated during the crisis in terms of organization and production.
The outbreak of the COVID-19 pandemic in 2020 and the restrictive sanitary measures taken to contain it had a major impact on the activity of French companies. Using detailed VAT returns data for a sample of more than 645,000 companies, we can estimate corporate sales on a monthly basis throughout 2020. Our analysis innovates in a three ways. First, we evaluate the impact of the crisis by estimating activity shocks measured with respect to a simulated counterfactual situation. Then, we estimate this shock at firm level, allowing a thorough analysis of the heterogeneity of activity loss throughout the crisis. Finally, we leverage our monthly firm-level data on sales to show how corporate activity has evolved along four distinct trajectories throughout 2020.
Our results show that the first lockdown created a generalized shock whose magnitude varied across sectors, while the shock induced by the second lockdown period was weaker and more localized. We show that during the crisis the industry that the firm operates in explains up to 48% of the monthly activity shocks’ variance weighted by employment, a much larger share than in a normal year.
The dynamic analysis of business losses reveals four profiles of activity trajectory in 2020, each one corresponding to different levels of corporate resilience in the crisis. The main determinant of belonging to a given profile of activity is the firm industry – defined at a very granular level. Conditional on industry, belonging to a high type profile of activity trajectory is also correlated with the greater ability to adapt some firms have demonstrated during the crisis in terms of organization and production. On the contrary being an exporter before the crisis is correlated with belonging to a low type profile of trajectory of activity in 2020.