Coronavirus - Sector Implications
9 March 2020
COVID-19 is rapidly moving from containment to delay in Western Europe and North America. The authorities are paying the penalty for half-hearted containment and tracking in comparison to China and South Korea. As a result, we are likely to see more widespread infection per head of population, as well as longer to return to normal. On a more positive note, the economic impact is likely to be materially less than in China. We set out in this note some assumptions around spread and the impact on our sectors and companies, as well as considering whether there are longer-term implications.
Italy – into lockdown
The initial response to quarantine just ten small towns has clearly failed and the government yesterday decided to extend the quarantine to the whole of Lombardy and 14 other provinces, which encompasses 16m people, until 3 April. This is a material change of approach and will clearly have major economic impacts as schools, gyms, cinemas, museums and ski resorts will be closed, sporting events cancelled and funerals not allowed. Bars and restaurants can be open between 8:00am and 6:00pm, but customers will need to sit one metre apart. Similarly, other commercial activities will need to have a distance of one metre between people, albeit we do not yet have clarity on the impact on offices and factories. As WHO Director Tedros put it: “They are making genuine sacrifices.”
In this piece
we discussed the exposure of the companies we follow to Northern Italy. We will update this as we have more information on the business impacts. Elsewhere in Italy, cinemas, theatres, schools and museums will be closed and funerals suspended.
So why are the authorities in Italy taking this step? The data for Italy is more concerning than in other regions, with 5% of cases resulting in deaths. Furthermore, 55% of the active cases are hospitalised and 11% are in intensive care. Among the closed cases, 72% have recovered and 28% have died. Of those that have died, 74% are over the age of 70 and 48% over the age of 80. It is likely that the number of cases is more widespread than the statistics show; however, this data is worrying enough to have driven this – albeit belated – response from the authorities.
Will it be successful?
The signs are not promising given that the response to the lockdown has been a sense of confusion as to what is allowed. There has been a rush of people leaving the area, which will inevitably spread the virus further.
Why does Italy appear to be more affected?
It is too early to have a clear view, but the age of population and close family units are likely to be contributing factors. Italy has the highest proportion of people over 80 in the EU at 6.7%, followed by Greece at 6.5% and Spain at 6.0%. France and Germany are at 5.8% and 5.7% respectively, with the UK 4.7%, and Ireland the lowest at 3.1%. Also, the data thus far shows that the majority of cases are transmitted within the home. Thus, having the majority of elderly in the family home or looked after in the community has created greater risk of infection. Furthermore, although Italians live longer than many other nations, they also have a greater number of elderly people with severe difficulties than other countries in the EU, with 59% of those over 80 having at least one severe chronic disease (source ISTAT).
Situation per country
We show below the number of cases and deaths, and as a proportion of population. We also give a view on where we are on the stages of the spread of the virus and the authorities’ responses.
Taking Italy as an example, the death rate has now risen to 5% of confirmed cases. Our view is that the spread has moved from the early stage to aggressive spread, hence being in the mid-stage. Having initially made a very limited response, the authorities are now escalating action to reduce the spread, albeit that it’s probably too late to prevent the spread throughout Italy.
Although these numbers look concerning, it is important to note that South Korea, which has tested 7% of its population, has found only 0.01% positive for the virus with 0.0001% dying.
A large number of countries are now reporting a growing number of cases and no or relatively few deaths. However, the outbreak is still in its early stages in most of these countries, and the majority of reported cases are from people who have travelled abroad. These individuals are likely to be younger and of more robust health than the average population. Moreover, the level of testing is typically limited to those with symptoms, so we should be wary of extrapolating early results. So far in the UK, 23,500 people have been tested and the latest estimate for tests in the US is 5,861 (according to the CDC on Saturday), although two tests are done per person.
Our central case
As a comparison, in the US it is estimated that 5-20% of people have the flu each year, with 3-11% showing symptoms (Source: CDC). For the 2018/19 season CDC estimates that:
• 35.5m (10.7% of population) had symptomatic illness;
• 16.5m (5.0% of population and 46.5% of symptomatic) made a medical visit;
• 490,000 (0.1% of population and 1.4% of symptomatic) went to hospital; and
• 34,157 (0.01% of population and 0.1% of symptomatic) died.
Over the last ten years the estimate for deaths has ranged from 12,000 to 61,000. Clearly these numbers are helped by the immunity built up from previous episodes and a sizeable proportion (37%) of adults having the flu jab.
According to the WHO the spread of the coronavirus is broadly similar to seasonal flu and well below mumps or measles. So, for the purposes of this note we are assuming that the spread in the UK (and similarly in the majority of Western Europe) and North America reaches 0.1-5.0% of the population, with a death rate of 0.5-1.0%. Clearly these are merely ‘finger in the air’ at this stage, but it seems reasonable to assume that South Korea’s broader testing provides an insight to the ultimate death rate.
For the UK, this would represent 70,000-3.5m people infected and deaths of 335-34,000. Clearly this is purely a guess in order to provide a framework for discussion rather than a clear analysis of the likely outcome. This is also on the assumption that a large number of people have mild symptoms and self-quarantine/don’t know and so don’t become part of the statistics.
Furthermore, we expect the level to be materially lower than normal flu, given that with flu large numbers of people continue their normal lives rather than quarantine themselves (willingly or unwillingly). To put this in context, a severe winter flu season (eg, 2014/15) resulted in 28,330 deaths, with a low of 1,692 reported for 2018/19. This compares to 599,000 deaths in the UK in 2018.
We assume that the virus spreads rapidly and peaks by the end of April, but that it will take around three months to fade away. There is no clear evidence that warm weather will stem the progression of the virus, but that does seem a reasonable assumption. Also, if the virus spreads quickly then people are far more likely to take precautions (handwashing, self-quarantining, avoiding crowded areas and working from home).
If this does play out, we would expect to see a gathering impact in the coming weeks and a return to normal in June. We expect economic activity to be affected during this period, but expect businesses and people to aim to keep economically active. We currently do not see widespread closures of businesses (as seen in China) as we expect the focus to be on delay rather than containment. Assuming the data is correct, China’s actions have resulted in containment of the virus and most businesses expect to be operating normally by the end of March. This means there will have been a severe impact for just two months. In the UK, we may have a longer, but shallower, experience.
It is too early to have a clear view on the longer-term impact, but we don’t believe that everything will be as before.
There is a risk that the virus re-emerges and that we get further outbreaks beyond our central case. We are confident that a vaccine will be found, but it will clearly need to be appropriately tested to ensure efficacy and safety and so it is likely to be at least a year for a viable vaccine to be available. There is also the risk that the virus mutates. This occurred with Spanish Flu (which infected 27% of the world population), which returned in different forms in the following two years after the initial outbreak in 1918. Furthermore, the mutant forms were more aggressive and caused higher death rates.
Given that there is no evidence to suggest there will be mutations, our central case is that this does not occur and that the global economy will be in recovery mode in H2 and going into next year. Central banks are looking to soften the impact with rate cuts, and there will be a requirement for governments to take action to ensure short-term cash flow issues do not result in a deepening of the economic impact. We expect further action to be taken, for example, the use of helicopter money as in Hong Kong, a reduction in consumption taxes (VAT) or deferral of tax payments.
In the longer term we expect:
• Higher government deficits
• Lower long-term interest rates
• Faster shift to online
• Increased spend on healthcare (with greater contingency planning)
• Weaker consumer expenditure, but marginal given employers are likely to retain and pay workers and the self-employed will keep working
• Reduction in cruises (publicity could not have been worse)
• Increase in home working as businesses/individuals are better set up at home and home working is seen as more acceptable
• Greater focus on supply chain (companies will need to have contingency plans)
• Reduced concentration of supply chain (ie, less exposure to China)
• Pension fund deficits are likely to worsen, unless mortality rates are materially higher than our assumptions
Sector & company thoughts
As above, our assumptions are:
• 70,000 to 3.5m people have the virus, although most will be unrecorded.
• A number of schools will close.
• The impact will have largely ended by June.
• Business decisions are deferred due to travel disruption, inability to complete due diligence, lack of key staff, etc.
• There will be business failures, which will increase unemployment in the short term.
• China returns to normal working practices by end of March. Supply chain issues are likely to persist through April given the level of disruption. Consumer spending in China is unlikely to recover before H2 given the level of disruption and loss of income for some people.
• In general, the well-positioned, well-run, well-financed businesses will come through this with their position enhanced and with greater opportunities for growth.
• Risk assets (eg, equities) remain under severe pressure in the short term, but should bounce back as the virus reduces and markets look to economic recovery. Ratings could move higher given that listed companies tend to be better financed and have access to equity capital to fund growth. Also, the WACC will reduce with the reduction of long-term interest rates and equities will look attractive given the gap between the higher dividend yields and bond yields.