As the world ramps up its response to the coronavirus crisis, with lockdowns and business closures gathering pace, Insight has estimated the potential peak-to-trough impact to global GDP. Our main views are:
- GDP contractions around 15% peak-to-trough are likely in most economies, with around 3 months in lockdown, followed by a relatively fast recovery, which should produce calendar year 2020 GDP average around -5% in most jurisdictions
- Fiscal and monetary policy responses will be key, but business closures may limit the impact
- Closely watch unemployment, supply chain constraints and behavioral changes for clues on the strength of the eventual rebound
How deep is the initial impact of lockdown?
There is clearly no certainty over this, and indeed we may never know precisely given the difficulties statistical agencies are having of collecting data during lockdown, but we previously looked at a sectoral breakdown to give a broad estimate, with travel and leisure seeing the most severe impacts.
The varying nature of each nation’s service sector and trade balances will likely lead to somewhat different economic outcomes. However, we expect double-digit declines in GDP peak-to-trough, with the US falling slightly less due to a smaller restaurant sector and trade deficit. Due to unknowns around the duration and effectiveness of containment measures, there is much higher uncertainty to these forecasts than during normal economic times.
This led to overall impact around -15% or -12.5% peak-to-trough GDP for most major economies. Note: we are deliberately rounding these numbers to an approximate 2.5%, there is no implication of precision here.
Figure 1: Insight’s peak-to-trough GDP forecasts
|Region||Estimated peak-to-trough GDP Impact|
Source: Insight Investment estimates, March 2020.
Travel and leisure sectors could take -80% hits
We have put together our projections by applying sector-level impact estimates to each nation’s economic composition. We believe sectors tied to travel and leisure will be particularly hard hit (Figure 2). Elsewhere, growing restrictions on nonessential work will likely mean most sectors will face some negative impact, particularly where activity cannot be conducted remotely.
We expect sectors such as construction and manufacturing will face both labour constraints and potential supply chain disruptions, although we believe the impact will be less severe declines than with restaurants, for example.
Figure 2: A sector-by-sector impact of the crisis
|GDP sector||Peak to trough impact|
|Agriculture, forest, and fishing||-10%|
|Wholesale and retail trade||-21%|
|Accommodation and food service activities||-80%|
|Information and communication||-10%|
|Financial and insurance activities||-10%|
|Real estate activities||-16%|
|Human health and social work activities||0%|
|Professional, scientific, and technical activities, administrative and support service activities||-10%|
Arts, entertainment and recreation, other service activities
Source: Insight Investment estimates, March 2020.
Our views on the hardest-hit sectors are largely supported by high frequency data. For example, OpenTable suggests that sit-down restaurant activity has fallen by more than 90% in the United States, Germany, and Ireland. It also expects more regions to follow suit.
Similarly, US cinema box office receipts in the week ended March 19 were $58m, the lowest weekly level since at least 1998 and down 70% from the average week in 2019. As more theatres are forced to close worldwide, receipts would be expected to fall further.
At this early stage, we need to emphasise these forecasts have unavoidably wide margins of error as we cannot yet know how successful each nation’s lockdown strategies will be or how long the authorities will pursue them.
Figure 3: The hardest-hit sectors are following similar paths across countries
How long will the lockdown last?
Given the unprecedented nature of the current situation, it is impossible to predict with any degree of certainty when lockdowns could begin to be lifted. However, we can look to the first country to be affected by the outbreak, China, to provide an indication of how events could progress. There, we have seen less affected parts of the country (ex-Hubei) starting to reopen after around 6 weeks of lockdown, with the epicenter, Hubei, due to start to open up after around 10-11 weeks. However, we have seen one region (Jia county in Henan), which has re-entered lockdown following a re-escalation of cases; this situation needs monitored closely.
Officially, lockdowns in most Western countries will be reviewed later this month, but we believe extensions are likely, and a lockdown duration of 2 to 3 months may be more realistic. Some experts and officials are discussing how restrictions may need to be in place, at least partially, for up to 6 months. Obviously, exiting lockdown too soon risks a re-escalation of cases, which could result in a W-shaped GDP profile. Antibody testing may soon enable some people who have been proven to have recovered from the virus to go back to normal life, but a vaccine that would safely enable a much broader reopening may not be available in sufficient scale until next year.
How quickly does recovery occur?
There are two sides to this question. Firstly, how much damage is done to the economy in the short term? And secondly, how much policy stimulus has been put in place to boost the recovery when it comes?
Short-term damage, or hysteresis, is principally due to an increase in unemployment that leads to a loss of consumption, in addition to business failures that mean that people do not get re-employed quickly afterwards. US monthly payrolls for March at -701k is clearly the worst level since the Global Financial Crisis, but weekly initial jobless claims at 6.6mln (w/c March 30) following 3.3mln the previous week (w/c March 23) is really breaking all records for pace. That suggests nearly 10 million Americans lost their jobs in a fortnight.
Figure 4: US weekly initial jobless claims spike
Source: Insight Investment estimates, April 2020.
We are starting to get unprecedented unemployment surges from countries around the world.
Figure 5: Norway unemployment begins to climb
Source: Insight Investment estimates, April 2020. Note: The above will also have been impacted by the oil price fall.
In parts of Europe, including the UK, governments are taking the approach pioneered previously by Germany, of subsidizing wages while workers are furloughed.
Figure 6: Germany’s wage subsidization spikes
Source: Insight Investment estimates, April 2020.
Note: The last data point is inferred from the number of companies applying (470,000) times the series average of 12 employees per company. Recent numbers of employees per company have been around 23, so this could be significantly higher.
It remains to be seen how long-lasting these effects are. Some sectors in some economies like the US, or the Norwegian oil sector, are historically good at making redundancies and then rehiring the same people over a temporary slowdown. But it seems intuitive that provided the firms themselves still exist, employees who have been furloughed rather than made redundant would find it easier to restart work once a recovery starts.
Comprehensive data on business failures has yet to be released. Government policy is clearly pivoting towards supporting businesses with government guaranteed loans, etc. However, it remains to be seen how successful this will prove to be.
Policy response will be substantial
Monetary policy will likely be highly accommodative for over 12 months while fiscal policy will need to be loose to ensure firms and individuals have the necessary liquidity to withstand the downturn.
Figure 7: The scale of the fiscal and monetary response announced so far
Source: Insight Investment estimates, April 2020. Note: US quantitative easing (QE) is “unlimited”. In Japan Yield Curve Control means QE is already variable.
These are large numbers, and would normally lead to a strong recovery in Q4 or into 2021.
As can be seen, central banks are doing what they were created to do – which is nothing to do with inflation – but instead to be a financer of last resort for a government that finds itself in need of dramatic fiscal expansion. These levels of QE (although they do include corporate QE purchases as well) will ensure that government bond yields do not rise significantly as a result of increased supply.
We are back to the era of big government we last saw during and after WW2, with financial repression. It’s not so much a magic money tree as a magic money forest. This will ultimately have some long-term effects that we will be looking into further in due course.
Should authorities be able to contain the virus, this stimulus will likely set the scene for growth to run well-above trend. In most regions, we expect it will take more than 12 months to return to pre-virus levels of activity.
Putting it all together
There are obviously plenty of uncertainties still, but if we take our estimate of around -15% peak to trough impact, a base case of around 3 months in lockdown, followed by a relatively fast recovery then we should get calendar year 2020 GDP average around -5% in most jurisdictions.
The greatest sensitivity is to the length of the lockdown. Extending out to 6 months would give calendar year averages of between -7.5% and -10%. For comparison, in the GFC 2009 calendar year GDP was -2.5% in the US, -4.5% in Euro area, -5.4% in Japan and -4.2% in the UK.
Fiscal expansions are likely to reach or exceed the levels of 2009. Only in Emerging Europe do we see lower levels. This will push government debt/GDP levels to above those of 2009 in most countries.
Quantitative easing, or other monetary policies such as yield curve control, will keep yields low or lower, at least for this year. Longer term, we may see a similar outcome to the US financial repression that followed WW2.
Three things to watch for the eventual rebound
1) Unemployment: While helicopter money combined with standard safety net benefits can help displaced workers, they will need to return to the labor market to maintain consumption eventually.
But how quickly workers are rehired at restaurants, hotels or (for example) airports, is a key question. The greater the initial rise in unemployment or corporate bankruptcy filings, the greater the risk of lingering unemployment is to the economy.
2) Supply chain constraints: These will likely delay the return to full capacity. In China this has been a multi-week process even as virus restrictions are removed. In all likelihood, not all countries will remove restrictions concurrently. With the global nature of the supply chain, production may not return as quickly as demand, prolonging the recovering in output.
3) Behavioral changes: The key question here is how long will it take (if ever) for consumers to be as willing to eat out, go to large sporting events, fly and / or travel as before? These preferences can be quickly tracked via real-time data like film tickets, restaurant reservations, as well as information on weekly oil demand and number of flights. But the potential longer-term psychological behavioral impact of this event also needs to be considered. Will businesses and individuals want to maintain less leverage or hold more liquidity, which could increase the savings rate, lower consumption, or defer business investment? These changes may take more time to grow clear.
Please note: any forecasts or opinions are Insight Investment's own as of April 6, 2020 and subject to change without notice.