September 10, 2020

Fathom’s Recovery Watch 09.09.2020

by Fathom Consulting.

Subscribe to Fathom’s regular Recovery Watch newsletters for the latest insights into the impacts of COVID-19.


  • UK case numbers rose sharply over the weekend; while measured infection rates are some way from their peak, true infection rates, although inherently unknowable, are likely to be orders of magnitude below levels seen in March and April
  • India has overtaken Brazil as the country with second most confirmed cases of COVID-19 globally
  • Government policy is key to the economic outlook; further full lockdowns appear unlikely, but a withdrawal of fiscal support poses risks

We are currently in the process of putting together our Global Economics and Markets Outlook for Q4. The juicy bits are reserved for our clients and will be presented to them in the coming weeks. Today’s Recovery Watch takes a closer look at infection rates in a handful of key countries and considers these numbers within the context of changes to testing capacity. (If you are not a client and would like more information on our services please contact

Refresh this chart in your browser | Edit the chart in Datastream

The chart above shows official confirmed COVID-19 per capita case numbers in a handful of countries — selected because they have all had notable outbreaks of the disease. Different stories emerge: in some, cases are rising, but are still well below previous peaks (UK and Italy); in India, they are relatively low on a per capita basis, but rising and hitting new highs; they are high, but falling, in the US and Brazil; they continue to rise and hit new highs in Spain.

These data suggest that countries are at different stages of the outbreak, but also that none of the countries in question can confidently claim that the virus is under ‘control’. But the trajectories of infections in the chart above are somewhat misleading since they do not account for changes in testing capacity, which, in most countries has increased significantly since the start of the outbreak. It would be wrong, therefore, to infer too much from the trajectories in the chart above alone. In the UK, for example, more than ten times the number of daily tests are now being carried out compared with April. Back in April tests were reserved almost exclusively for those who were seriously ill to the point of requiring hospital admission. Now, in principle, they are available to anyone with symptoms. In India, the number of daily tests is nearly 80 times higher now than it was in April. Admittedly, the number of tests carried out is not the same thing as the testing capacity, but it is a reasonable proxy.

It is impossible to know with certainty how many people in the UK had COVID-19 in the early stages of the pandemic, but the actual numbers were certainly far higher than the official confirmed cases due to testing capacity constraints back then. Anecdotal evidence supports this point — the author of today’s Recovery Watch, for example, personally knows several people who experienced strong COVID-like symptoms back in March and April, but who were unable to get tested. While some testing capacity constraints still exist today, most people that want to get tested in the UK today can do so. Academic studies also suggest that official COVID cases were massively understated earlier this year. For example, according to estimates by the COVID Symptom Study website, the number of live symptomatic COVID cases in the UK peaked on 1 April at just over 2 million, while they estimate that the number of current live infections (of those symptomatic) is less than 31k — on this metric, active infections were more than 50 times higher in April than they are now. Official figures for this metric are not published, but the difference between new infections today and new infections back in April (a reasonable proxy for live infections), is less than two times.

The chart below compares new infection figures today with those in April, as well as adjusting the April figures by a factor proportional to the change in testing since then. Admittedly, this is a crude adjustment, and it is possible that those who got tested in April were more likely to have had the disease than those who get tested today (if so, the adjusted case numbers for April would be inflated in the chart below). But the adjusted April figures are much lower now in the US, Spain, Italy and the UK than they were in April. Interestingly, the current rates of new infections in Brazil and India are still higher than they were in April, which most probably reflects that the peak in these two countries happened later than in Europe and the US and that they are still at relatively bad stages of their outbreaks.

Refresh this chart in your browser | Edit the chart in Datastream

The data from the COVID Symptom Study Survey suggest that the adjusted figures for April in the chart above may even still be downplaying the rate of infection in that month, at least in the UK. For example, they estimated that there was a total of around 2.1 million live infections on 1 April. Since the average infection period is believed to be between one and two weeks, this would be consistent with new real symptomatic cases peaking between 150k and 300k per day in early April. Official figures, by contrast, show that COVID-19 infections in the UK peaked at 6.2k on 1 May (the peak in the first half of April was 5.9k). In other words, according to this study, new COVID infections were being underreported by a magnitude of between 25 and 50 times. COVID-related fatalities might lag new case numbers by a couple of weeks, but the trajectories illustrated by the chart below tell a similar story to the testing-adjusted COVID numbers above.

Refresh this chart in your browser | Edit the chart in Datastream

This is not intended to downplay the severity of the disease or the recent increase in cases in Europe, but it is intended to show that infection rates are almost certainly significantly lower today than the official non-testing-capacity figures suggest. The bottom line is that the shrinking pool of people that the virus can infect in the places that were worst hit in the early stages of the pandemic (like New York — see chart below — and London — see suggested reading list), coupled with behavioural changes more generally, seem to be keeping infection rates down. Whatever the reason, it seems unlikely that, even though figures are rising again in some places, full-scale lockdowns will be needed again. Indeed, some governments have said that they will not do it (France), while the UK Prime Minister has said that another full lockdown would be a “nuclear option”. And while the rise in cases in Europe is a cause for concern, we think that case numbers are likely to be brought down without drastic changes to social or economic behaviour being required.

Refresh this chart in your browser | Edit the chart in Datastream

Finally, the short-term economic outlook will depend heavily on governments’ policy responses. Not just in terms of influencing people’s behaviour but also the level of stimulus they chose to provide going forward. A swift withdrawal, prompting a large pickup in unemployment is, of course, a risk. But while the UK government has spoken about ending its furlough scheme in October, the opposition are calling for an extension and it would not be surprising if some Tory MPs start to follow suit as that deadline approaches. Germany has extended its furlough scheme until the end of 2021. The US meanwhile, has enacted large stimulus programmes in the wake of the pandemic, which saw personal income spike. Congress is working on a fresh large stimulus package. These are some of the reasons that we remain in the camp of a V-shaped recovery and which we will discuss with clients in the upcoming forecast round.

Refresh this chart in your browser | Edit the chart in Datastream

Interesting reading



Financial time series database which allows you to identify and examine trends, generate and test ideas and develop view points on the market.

Refinitiv offers the world’s most comprehensive historical database for numerical macroeconomic and cross-asset financial data which started in the 1950s and has grown into an indispensable resource for financial professionals. Find out more.

Article Topics
Article Keywords ,

Get In Touch


We have updated our Privacy Statement. Before you continue, please read our new Privacy Statement and familiarize yourself with the terms.×