Using the April MBS (conducted between April 28th and May 6th), we updated this picture at the end of April. According to the business managers surveyed, the loss of activity in industry and in construction appears to have been less severe in April than in the first weeks of lockdown at the end of March. In market services, there is a dichotomy between business-oriented services, which remained less affected (with the exception of temporary employment), and those more focused on households, whose loss of activity remained almost as high as at the end of March. Overall, whereas a month ago we estimated the loss of GDP for a typical week of lockdown in March to be around -32%, our new estimate for a typical week of lockdown in April is around -27%. These are of course orders of magnitude with a large margin of uncertainty, but they nevertheless suggest smaller GDP losses in April.
As regards the outlook for May and the beginning of the exit from lockdown, the business leaders surveyed expect in most sectors (with the exception of accommodation and food services) a partial recovery of their activity. However, in view of the great uncertainty surrounding the lockdown's length and exit procedures, we are not publishing a GDP forecast for the current quarter this month, contrary to our usual practice.
In the framework of the Eurosystem staff macroeconomic projections for the euro area as a whole, the Bank of France will publish macroeconomic projections for France for the period 2020-2022 during the week of June 8th, together with its next monthly business survey.
The monthly business survey was conducted between 28 April and 6 May and involved a sample of 8,500 companies and institutions.
With a full month of lockdown in April (compared to half a month in March), economic activity reached a particularly low level.
However, companies have adapted and have gradually put in place various measures to protect their employees' health. This made it possible to limit site closures and restart production, albeit at a much lower rate than normal and unevenly across sectors.
There were 5 exceptional closure days, like in March. However, this compares with a total lockdown time of 21 days in April and 10.5 days in March (both excluding weekends and public holidays), i.e. an average opening rate of around 75% in industry (compared with around 50% at the end of March).
However, the number of closing days varied from 1 day in the pharmaceutical industry to 11 days in transport equipment. Sectors with the highest numbers of exceptional closure days were typically those reporting the most negative balances of opinion regarding the change in activity relative to the previous month and the lowest capacity utilisation rates.
For the second month in a row, activity has declined in all industrial sectors; this contraction is less marked overall than in March, given that the length of lockdown was twice as long as in March.1 But this contraction was still very significant in the automotive, plastics processing, equipment manufacturing and metalworking sectors.
If the level of activity in April as a whole had been the same as in the two-week lockdown in March, the drop in activity in April (compared to the previous month) should be of the same order of magnitude as in March. The fact that the balance of opinion on the outlook for activity was less negative in April than in March therefore suggests a smaller decline in activity in April than in the two-week lockdown period in March. The red diamonds on the chart (a "theoretical" extension of the March trend) show this difference.
The overall capacity utilisation rate for manufacturing industry fell from 77% in February to 56% in March, an all-time low for the survey.
Market services activity recorded a further sharp decline, although there was a clear dichotomy between business services (except for temporary employment) and those aimed at households.
The number of exceptional closure days in April averaged 9 days, after 6 days in March, with a very wide range, from 1 day in computer services to 24 days in accommodation and food services.
In services as well, the sectors with the highest number of exceptional closure days were also those reporting the most negative balances of opinion regarding the change in activity relative to the previous month.
Accommodation and food services were particularly affected for the second month in a row, and the drop in temporary work was even greater in April, as a result of the sharp fall in activity recorded in sectors employing a large number of temporary workers (in particular the automotive sector).
Business services (with the exception of temporary employment) remained relatively less affected than services aimed at the public, such as accommodation and food services, which were the most directly and lastingly affected by the lockdown measures.
At the same time, the services that fared best were those that were able to resort to teleworking on a massive scale. In services, 39% of business leaders surveyed reported that they had recourse to teleworking, compared to 16% in industry and 14% in construction. Three service sectors resorted to teleworking more than 70% of the time: publishing, computer services, and legal and accounting activities.
Construction activity further deteriorated in April, but to a lesser extent than in March.
In services, companies reported that their cash positions deteriorated further in April. In industry, cash positions were still degraded, but roughly unchanged from March.
In the scope of our survey, with a sample made up essentially of SMEs and mid-sized firms, nearly half of the companies surveyed reported having applied for a State-guaranteed loan and 10% of companies were considering applying in the coming weeks.
In the accommodation and food services sector, almost 60% of the companies reported having applied for a State-guaranteed loan.
In the transport sector, the share of companies expressing an intention to apply for a State-guaranteed loan in the coming weeks was particularly high (20%).
The upturn in activity expected in May (compared to the previous month) is of significant magnitude but would in no way make up for the declines of the previous two months.
In addition, business leaders were specifically asked about their business prospects for the post-11 May period: they were asked to estimate their level of activity in relation to a level deemed normal. In most sectors, they expect an upturn in activity, which would be more pronounced in industry and construction, although activity will not return to a normal level. Overall, the most severely affected sectors during the lockdown (such as the automotive industry or auto repair) would recover the most, although their activity levels would still be far from normal. Some activities would remain virtually at a standstill, such as accommodation and food services.
In our previous publication dated April 8th and based on the March monthly business survey, we estimated the loss of activity for the whole economy to be about -32% over a typical lockdown week in March. This estimate is of the same order of magnitude as those previously provided by INSEE and OFCE. Taking into account the lockdown in the second fortnight of March, we deduced an estimate of the growth rate of GDP in the first quarter of 2020 to be around -6%. In its publication of the preliminary results, INSEE estimated the fall in GDP at -5.8%. These results confirm the overall assessment and the breakdown of sectoral shocks: at end-March, the decline in construction was almost twice as strong as in industry and market services, while agriculture and non-market services remained relatively unscathed.
As explained above, our last survey suggests that the loss in activity was less significant in April than during the lockdown weeks in March. Indeed, the number of exceptional closure days in April remained stable in industry and increased only marginally in market services (see Chart “Average number of exceptional closure days in industry”, “Average number of closure days in services”), while the number of lockdown weeks doubled in April compared to March. This confirms that a large number of companies have been able to integrate health measures into their production process and resume their activity. However, this resumption might have been only partial if the same level of activity per working day as in normal times could not be achieved due to the health measures and distancing requirements.
The capacity utilisation rate in industry averaged 56% in March, after 77 % in February, which represents a capacity utilisation rate of around 35% in the last two weeks of March. It rose to 46% in April, again reflecting a recovery in industrial activity compared to end-March. Survey balances, which measure the monthly change in activity, also improved, showing less negative values in April than in March. With a lockdown period about twice as long in April as in March, this again suggests that the level of activity in April was less deteriorated than in the second half of March. This is particularly the case in industry, where the balance on evolution of deliveries rose to -70 from -95 in March. In services, the improvement was more modest, with the balance on activity climbing to -99 in April from -116 in March.
We used this information at the finest disaggregation level possible to update our estimate of the loss of activity for a typical lockdown week. This loss is estimated for the month of April at approximately -27%, after -32% in the fortnight of lockdown in March. The improvement occurred essentially in manufacturing industry excluding agri-food, coking and refining (-37%, after -48%) and market services (-27%, after -37%). However, the loss of activity in the construction sector was still quite substantial, at around -75%.
If we combine the sectors most exposed to the lockdown measures, i.e. manufacturing industry excluding agri-food, construction, and non-financial and non-real estate market services (which together account for 55% of GDP), the loss of activity represents around 40% of the normal level (compared with 50% in our early April estimate for the second half of March).
This rise in activity in April compared with end-March is also corroborated by alternative high-frequency indicators, such as temperature-adjusted electricity consumption, credit card payments and ATM withdrawals. These data, collected during the month of April, pointed to a partial recovery in consumption and production.
This new assessment points to an improvement of activity compared with our previous estimate for March, as well as with the latest INSEE estimate (-33% on May 7th). This can be explained by the fact that the survey on which our estimate is based was conducted in late April/early May, i.e. towards the end of the lockdown period. However, these differences should be interpreted with caution, given the unusually wide margins of uncertainty surrounding all these estimates.
Updated on: 06/10/2020 13:56