The outbreak of the novel coronavirus (COVID-19) will go down in history as a major example of a neglected risk. The topic “infectious diseases” was ranked tenth in terms of impact in the World Economic Forum’s Global Risk Report 2020 (WEF 2020, published on 15 January 2020), but was considered quite unlikely. The attention of corporate decision makers and politicians was mainly focused on traditional sources of business risk and pressing environmental issues. Only a few weeks later, however, their attention shifted dramatically.
According to the World Health Organization (WHO), by 10 March 2020, COVID-19 had led to more than 110,000 confirmed infections and 4,015 deaths in 110 countries – and the numbers are increasing rapidly. Huge disruptions to personal lives are taking place, including curfews (or situations resembling curfews) for a great many individuals. Beyond the immediate tragedies of death and disease, indirect effects through fear are taking hold of an uncountable number of people around the world.
In light of the massive impacts of the coronavirus on public physical and psychological health, the economic and financial impacts may seem secondary. However, the economic effects are potentially going to be of first-order importance. Economists are beginning to consider these consequences. For example, the recent Vox eBook (Baldwin and Weder di Mauro 2020) collects a number of mostly qualitative viewpoints in various domains of the economy. McKibbin and Fernando (2020) simulate the impacts of different scenarios on macroeconomic outcomes in a general equilibrium model. Earlier work also provides valuable insights. For example, Adda (2006) analyses outbreaks of a number of viral diseases in France and evaluates the (sometimes subtle) effects of policies such as school closures and the closure of public transportation networks. Ultimately, predictions are difficult because the spread of the disease, the policy responses, and individual behaviour are unknown.
An important tool for understanding the consequences of events like the emergence of COVID-19 is to consider asset price changes. These price changes capture current expectations. Thus, the researcher need not trace all the future changes to cash flows and discount rates separately (Schwert 1981). Effectively, asset markets provide ongoing, high-stakes surveys regarding future expected outcomes.
Incubation, outbreak, and fever
In new paper (Ramelli and Wagner 2020), we provide a first examination of the stock price reactions to the outbreak of COVID-19. We investigate three periods, which we label Incubation (Thursday, 2 January 2020 to Friday, 17 January), Outbreak (Monday, 20 January to Friday, 21 February), and Fever (Monday, 24 February to at least Friday, 6 March).1
- Incubation: On 31 December 2019, cases of pneumonia detected in Wuhan, China, were first reported to WHO and on 1 January 2020, Chinese health authorities closed the Huanan Seafood Wholesale Market after it was discovered that wild animals sold there may be the source of the virus. The first trading day after these events was 2 January 2020.
- Outbreak: On 20 January, Chinese health authorities confirmed human-to-human transmission of the coronavirus, and WHO issued the first situation report on the outbreak.
- Fever: On Sunday 23 February, Italy placed almost 50,000 people in strict lockdown in Lombardy (one of the most populated and productive regions in Europe) in an attempt to control the outbreak after registering its first deaths from COVID-19 on Saturday, 22 February.
The events initiating the Outbreak and Fever periods markedly changed the attention of market participants. Figure 1 shows that in earnings conference calls, corporate managers and analysts only starting paying attention to the novel coronavirus after 20 January. The first international conference call discussing either of the keywords ‘‘coronavirus’’, ‘‘covid-19”, ‘‘2019-ncov’’, or ‘‘sars-cov-2’’ took place on 22 January. The fraction of firms discussing these topics increased markedly over time, to around 30% at the end of the Outbreak period. When the Fever period began, that fraction increased to approximately 50%.
The global Google search intensity on coronavirus increased massively after 20 January. It subsided somewhat after its interim peak at the end of January, but when the Fever period started, the search intensity spiked.
Figure 1 Global attention to coronavirus
International stock returns by industry
We begin our analyses with an overview of industry-level returns in China, the US, Europe, and Asia ex China, throughout the full months of January and February 2020 (the latest data available to us for countries other than the US). While discussions in the media have generally used raw returns, we also compute CAPM-adjusted returns – that is, returns adjusted for a firm’s exposure to the overall market.
Figure 2 plots the industry averages of cumulative returns for China (top panel) and the US (bottom panel). Energy, Retailing, and Transportation were losers both in China and the US. Healthcare gained substantially in both (and other) countries. There are also differences across regions. For example, the Semiconductor sector gained sharply in China, but lost in the US. Utilities lost in China, but gained strongly in the US.
Figure 2 Stock returns by industry in China and the US, January to February 2020
Next, we consider when these returns were realised. For this question, we study reactions in the US, which initially was not directly affected.2 Figure 3 plots the CAPM-adjusted returns over time for a selected subset of industries. Interestingly, of the total run-up (loss) that US firms in Healthcare and Utilities (Energy and Transportation) experienced in the Incubation and Outbreak periods, about one third occurred already in the Incubation period. And also strikingly, in the Fever phase there was temporarily strong reversal in the relative returns (as investors apparently sold all stocks), followed by a whipsaw pattern (as in the aggregate market).
Figure 3 Relative winning and losing US industries, 2 January to 6 March 2020
Firm-level stock returns: The roles of trade and debt
Next, we quantify the effects of various firms-specific factors.
First, we employ data on US firms’ international exposure from Hoberg and Moon (2017). These authors analyse 10-Ks for annually updated firm disclosures regarding their international activities, counting the number of times each country is mentioned, distinguishing between input reliance and exposure to the export market.
We find that US firms with any Chinese exposure experienced 7.1% lower CAPM-adjusted returns from 2 January to 6 March than comparable firms (5.3% when controlling for industry fixed effects). More mentions of either export or supply chain exposure to China resulted in substantially lower cumulative abnormal returns (CARs) over the whole period. A one standard deviation higher export (input) exposure is associated with 1.91% (1.80%) lower CARs.
These considerations hold also when looking at the stock price effects of firms’ share of foreign revenues (not necessarily from China), indicating a general pessimism of financial markets regarding the disruptive impact of COVID-19 on global trade.
Figure 4 plots the sensitivity over time of CAPM-adjusted returns per one-time mention of China in a firm’s disclosure of its international activities. Interestingly, as for Chinese supply chain exposure, about half of the overall effect until the end of the Outbreak period was actually realised in the first part of January. Thus, it is conceivable that sophisticated investors had already started pricing in the concerns about supply chain disruptions in the Incubation period. It is notable that these asset-price changes occurred at a time when firms exporting to China in particular should, in principle, have done comparatively well, given the arguably good news contained in the ‘Phase 1’ trade agreement between China and the US signed on 15 January.
Figure 4 Stock prices and exposure to China, 2 January to 6 March 2020
In the Fever period, the market entered a feverish whipsaw pattern of aggregate returns, and so did the value of firms’ international exposure, even though little fundamental information of different signs appears to have arrived each day. Herd behaviour appears to have gripped investors.
Finally, during the Incubation and Outbreak periods, corporate leverage does not explain much of the cross-section of stocks. However, in the Fever phase concerns about corporate debt (leverage) started to play an important role. The observed de-risking of investors suggests that market participants are increasingly worried that the COVID-19 crisis may turn into a financial crisis more broadly.
Overall, our study illustrates how markets are adjusting to the rapid emergence of a previously neglected risk. These early results suggest that the market fairly quickly began to respond to concerns about the possible economic consequences of the novel coronavirus. The reaction happened initially in quite an orderly fashion, focusing on international trade. In the most recent period (late February and early March), large price moves of the aggregate market occurred. But behind these feverish price moves, some patterns emerge. In particular, the cross-section of stocks reveals that investors started to become concerned about potential amplifications of the COVID-19 shock through financial channels
Our work looks at stock price effects, which capture the expectations of market participants regarding future economic consequences. Other research should examine the realised consequences of the novel coronavirus. The combination of this research will inform policymakers, investors, and businesses in their responses to the emergency (and hopefully help prepare for future emergencies).
Adda, J (2016), “Economic activity and the spread of viral diseases: Evidence from high frequency data”, The Quarterly Journal of Economics 131: 891–941.
Baldwin, R, and B Weder di Mauro (eds) (2020), Economics in the Time of COVID-19, a VoxEU.org. eBook, CEPR Press.
McKibbin, W and R Fernando (2020), “The global macroeconomic impacts of COVID-19: Seven scenarios”, working paper.
Ramelli, S and A F Wagner (2020), “Feverish Stock Price Reactions to COVID-19”, Working Paper.
Shiller, R J (1981), “Do stock prices move too much to be justified by subsequent changes in dividends?”, American Economic Review 71: 421-236.
WEF – World Economic Forum (2020), The Global Risks Report 2020.
1 For a full timeline see https://en.wikipedia.org/wiki/Timeline_of_the_2019-20_coronavirus_ outbreak.
2 The number of cases in the US, while relatively limited at the time of writing this column, is poised to rise rapidly, as the disease spreads and as systematic testing begins.