As both cases and news about the coronavirus are spreading fast across the world, many investors are asking how they can prepare for the impact of a possible wider outbreak. Parallels have already been drawn with the SARS outbreak of 2002-03. So, with history appearing to repeat itself as this latest epidemic exceeds the last, perhaps designing a few historical stress tests may help investors “immunize” their portfolios.
As with most viral infections, this may just be a case of giving it time and lots of fluids (liquidity) and be prepared for symptoms to get worse before they get better. But there is also some preventive medicine that one can take, by, for example, overweighting defensive sectors, such as utilities and real estate, and avoiding cyclical sectors such as materials and consumer discretionary.
Looking back at the SARS timeline, the first cases were reported in China on November 16, 2002, but it wasn’t until February 10 that the authorities officially reported them to the WHO. By then investors had already reacted and markets were down in January. The bulk of the SARS-related selling happened in Q1 2003. Markets rebounded strongly from Q2, once the world joined in tackling this global health threat. It wasn’t until May 2004 that the WHO declared China free of new cases of the SARS virus and the pandemic declared over.
Given this timeline, we decided to focus our modeling of the worst-case scenario through a historical stress test using January 2003 and Q1 2003 as guides. We also wanted to see what to expect once the threat was perceived as averted and so ran Q2 2003 on our portfolio as well. Using the Axioma Risk platform, we setup a historical stress test using the STOXX Global 1800 as of December 31, 2019 as our portfolio. We wanted to know how this portfolio would have performed had we held it through January 2003, January through March 2003, and April through June 2003. The tables below report our results broken down by country, both Standalone (i.e., expected returns for each individual country sub-portfolios), as well in terms of their contribution (i.e., each market’s contribution to the total return of the portfolio) to total forecasted return. The values in each column are further color-coded from positive (blue) to negative (red) returns. All results are reported in USD.
In summary, today’s STOXX Global 1800, if held during the SARS outbreak in 2003, would lose 4.5% of its value in the first month and 12.6% in the first three months. Once the outbreak becomes contained and global health authorities begin to lower the threat levels, in the following quarter we could expect a sharp rebound equal to or surpassing the previous quarter’s loss. That is, of course, if history repeats itself.
The detailed table below shows the impact at the country level for a global portfolio. Not surprisingly, the US market, despite not being the worst hit domestically, is the largest contributor in all our scenarios by virtue of its weight in the index. Interestingly, the UK was hit much harder than the US in the two down scenarios, but also recovered more of its Q1 losses in Q2 USD. Also of interest is the fact that traditional low volatility markets, such as Switzerland (CH), the Netherlands (NL), Norway (NO), Sweden (SE), and Finland (FI), all did worse individually than historically more volatile markets, such as France (FR), Italy (IT), Spain (ES), and Germany (DE). Most surprising of all, China, the epicenter of the SARS outbreak, rose in January by 5% before surrendering those gains later in the quarter to end Q1 at -5.6%.
The table below shows the same test results, but broken down by GICS Sectors and sorted by sector weight in the index. Notice that Consumer Discretionary, where most of the tourism industry resides, would be hit quite hard. Same with other cyclical sectors, such as Materials, Financials, and Industrials. Utilities and Real Estate play their traditional defensive roles, i.e., falling, but also rebounding, much less than other sectors.
Historically, natural disasters do not have a lasting impact on stock prices and depending on their source (e.g., weather versus health scare) tend not to alter the fundamentals upon which stock valuations are based. They do have a relative impact on the affected industries, but that, too, tends to be for a limited time. In our scenario, drawing a parallel between the events surrounding the SARS outbreak in 2003 and today’s coronavirus outbreak, we see sharp market falls in the initial quarter, as the virus spreads and the press coverage gets more negative, thereby affecting investor sentiment. But once the threat is contained, it’s back to fundamentals for investors—and those will have changed little, if at all, because of this type of event.
 Note: The dollar depreciated by 3% against both euro and the pound in January 2003, so the losses in Germany and France are pretty much the same magnitude as in the US, while the UK seems to have been hit even harder.
 “sharp” by natural disaster standards, not by US-China trade war standards as was seen in May 2019