Uncommon truths: Viruses, nuclear disasters and financial markets

Could the cure for the Coronavirus be worse than the disease? We look at why we tend to overstate the risks of such events and examine the possible economic and financial market effects.

The Coronavirus is spreading but could the cure be worse than the disease? We look at why we tend to overstate the risks of such events and examine the possible economic and financial market effects.

Which of the following would you choose, if forced: A) a 1-in-12,000 risk of losing $250 or B) a 1-in-50,000,000 risk of losing $850,000? Though the expected loss under B is smaller, I suspect you would choose A given the serious consequences of option B going wrong. Similar thinking explains why we worry about viruses such as Coronavirus but not simple flu, which in fact claims more lives (of which, more later).

As of the date of publication, there have been 24,554 reported cases of Novel Coronavirus (2019-nCov) and 492 reported deaths (World Health Organisation (WHO) data). Given that the first cases were reported by China to the WHO on 31 December 2019, the virus would appear to be spreading quickly.

Coronaviruses get their name from their crown like appearance under the microscope. They are not new, with the most recent common ancestor of all coronaviruses placed at around 8000 BC (10,000 years ago)1. Bats and birds are commonly the carriers but these viruses are now common among other species, including humans.

Coronaviruses typically cause mild respiratory infections in humans (they are thought to cause most common colds). Some rarer forms (SARS, MERS and Novel Coronavirus) can cause pneumonia and may be lethal. Human-to-human transmission is thought to be via respiratory droplets (coughing and sneezing). These viruses do not like heat but remain infectious until exposed to 56 degrees C for 90 minutes2.

Figure 1 shows how the outbreak of SARS developed in 2002/03. The first known case was in China on 16 November 2002 but was not identified until much later. In fact, SARS was initially thought to be a problem in Hong Kong and Vietnam, though spreading (especially to Singapore, the US and Canada). However, when China released data to the WHO on 26 March 2003, it became apparent that China accounted for around 60% of global cases and deaths and was in fact the origin of the virus (see the footnote to Figure 1).

The number of SARS cases accelerated over time. As often happens with flu epidemics, the worst of the SARS outbreak occurred during the winter and early-spring period, with a flattening out of cases during Q2. It is commonly said that Novel Coronavirus has a lower fatality rate than SARS (around 2% versus nearly 10%). Strictly speaking this is true but note how the fatality rate of SARS increased over time (from around 3% in the early stages to 10%). The final number of SARS cases is reckoned to be 8096 with 744 deaths.

Though SARS spread to 29 countries, around 87% of the cases and deaths were in China and Hong Kong and only 11 countries recorded deaths (Canada stood out with 47 deaths and a fatality rate of 17%).

Figure 1

Figure 2 shows that Novel Coronavirus has developed far more rapidly than SARS (in terms of cases), though the number of deaths remains lower (for now). This speed of transmission is one of the frightening aspects of this outbreak and has naturally led to concern being expressed in the media and by markets. Governments around the world are also taking preventative measures and China’s economy is in a form of temporary lock-down.

However, it is interesting to ask why such outbreaks seem so frightening, especially when we compare the number of deaths in the last month (less than 500) to the 21,000 people that die in road accidents every month in China (112,000 globally) or the near 40,000 monthly global deaths from seasonal flu (assuming the deaths are spread evenly throughout the year).

Seasonal flu is reckoned to impact 340m-1000m people per year (5-15% of the world’s population), so there is a reasonable chance of contracting it. The number of deaths is reckoned to be 290k-650k per year. That is a lot of deaths but is less than 0.1% of the number of cases. So, the risk of contracting seasonal flu is high but the consequences are limited.

Based on the data available so far, we seem at very little risk of contracting the Novel Coronavirus (less than 0.01% if we assume the 25,000 cases so far is repeated every month throughout the year). Even with a fatality rate as high as SARS (10%), the probability of contracting and dying from this virus would be less than 0.001% (around 30,000 deaths in total). So, there is a small risk of catching this virus but the implications of doing so are quite serious.

Slovic et al3 conducted experiments on the perception of risk. Though “experts” were found to have a decent grasp on the risk posed by various activities, substances and technologies, the same was not true for “lay-people”. In particular, among “lay-people”, groups of students and women (members of the League of Women Voters) rated nuclear power as the most risky (“risk of dying”) out of 30 categories, whereas “experts” ranked it 20th (which is exactly where it was placed based on the researchers’ assessment of the number of deaths in the US). On the other hand, “lay-people” ranked electric power (traditional power generation) as the 18th most risky, whereas “experts” ranked it 9th (it was 5th). By the way, the three most risky categories were smoking, alcoholic beverages and motor vehicles (“experts” correctly identified them but put motor vehicles first).

Furthermore, lay-people’s perception of the high risk of nuclear power was not based upon their estimate of the number of deaths in an average year (their mean estimate was 20 versus the researchers’ estimate of 100). Rather, it came from their belief about what would happen in a “disastrous” year. Based on the answers given, more than 40% of participants implicitly believed there would be more than 10,000 deaths in a disastrous nuclear year and more than 25% expected 100,000 or more fatalities (in other work, Slovic found respondents expected a serious nuclear accident to result in hundreds of thousands or even millions of deaths). Slovic compared this with the “worst-case” estimate of 3,300 prompt fatalities from the maximum credible nuclear accident, with the odds against such an accident being 3,000,000:1 (US Nuclear Regulatory Commission 1975).

Figure 2

Slovic concluded that three factors form our perception of risk: dread, familiarity and the number of people exposed. This framework seems ideal for explaining the reaction to the Novel Coronavirus. As the name suggests it is new, hence the lack of familiarity (fear of the unknown). Second, the fact that it can be fatal instils dread. Finally, in theory we could all be exposed, which raises the potential number of deaths. The Spanish flu pandemic of 1918-20 is a great example of the latter point: it is thought that 500m contracted that strain of flu (one-third of the world’s population) and that 50m-100m died (10%-20% of the number of cases and 3%-7% of the world’s population).

Another element of unfamiliarity about the current outbreak is that it is occurring in China. Many in the West seem suspicious of anything Chinese. This became clear during the 2015 stock market sell-off in China, when a small devaluation of the currency sent global markets into a tail-spin. Given our lack of knowledge of such a vast country, we tend to make up what we don’t know and assume the worst. The current reaction of other countries suggests that China is effectively being put into quarantine (though to be fair it has shut down much of its own economy).

This attitude towards China is interesting when compared to what happened during the swine flu outbreak in 2009 (swine flu is the same A(H1N1) virus as Spanish flu). In the first month that swine flu data was reported, the US, Canada and Mexico accounted for 92% of all cases (the US alone accounted for 52%). By mid-2010 the worldwide death toll had risen to 18,449, with around 46% of them in the Americas (these were laboratory confirmed deaths but the USCenter for Disease Control and Prevention puts the actual number of deaths at 151,700 to 575,500)4.

Figure 3 suggests the path of the Novel Coronavirus is so far like the early stages of swine flu (though the number of deaths is higher). However, there was no suggestion of putting the US and its neighbours into quarantine in mid-2009. Is the dread and lack of familiarity with both the virus and China causing an overreaction on the part of non-Chinese governments? Perhaps in the modern world they cannot risk being seen to do nothing, especially when others are acting.

This, along with the self-imposed lock-down within China (perhaps due to mistakes made with SARS), risks making the cure worse than the disease, with unnecessary loss of income and welfare. According to the US National Academy of Medicine5, the average direct global economic cost of infectious diseases is around US$60bn per year. However, using a more complete approach, Fan et al.6 estimate the annual cost to be US$490bn, which is around 0.6% of global GDP. This cost is incurred every year, on average.

That average cost does not cover what happens during pandemics, when global costs are likely to be accentuated by the closure of travel and trading links. The World Bank estimated in 2014 that a pandemic on the scale of the 1918 Spanish flu would incur costs of around 5% of global GDP7. WHO data suggests the Ebola outbreak in Guinea, Liberia and Sierra Leone caused 28,616 cases (around 0.1% of the population) and 11,310 deaths (a fatality rate of 40%). The World Bank estimates those countries suffered a cumulative economic loss of at least 10% of GDP8.

Figure 3

The Spanish flu and Ebola examples are unlikely to provide good templates for the current outbreak in China (the fatality rate looks nowhere near as high as for Ebola and the way we deal with pandemics has evolved massively since 1918). At the same time, the economic damage caused to China could be greater than during the 2002/3 SARS outbreak. Though it is hard to know what might have happened otherwise, World Bank China GDP growth rates for 2001 (8.3%), 2002 (9.1%) and 2003 (10.1%) are not suggestive of a big issue. Also, to the extent that China is impacted, the knock-on effect on the rest of the world may now be greater as World Bank data suggests the Chinese economy in 2018 represented 15.8% (18.6% PPP basis) of global GDP versus 4.2% in 2002 (8.5%).

So, how bad could the economic effect be? Obviously, if the whole Chinese economy were shut down for a long period, the effect would be catastrophic at both the local and global level. Though we don’t know how long this will last, there are some points of comfort: first, the lock-down started during the Chinese New Year period when the economy would have been largely closed for a week anyway; second, Hubei province (where most of the cases are occurring) accounted for only 4.4% of China’s GDP in 2018 (according to China National Bureau of Statistics data); third, according to China’s National Health Commission, 80% of deaths are in the over-60 age range and 75% had some underlying condition (which suggests much less risk for the general population); fourth, the China lock-down suggests radical action to stop the spread of the virus (as opposed to what happened in 2002/3) and, fifth, if the disruption is short, the loss of economic activity can be made up in subsequent periods.

So, hopefully, there will be a short, sharp downward shock to the Chinese economy during Q1, with recovery during Q2 and H2. We believe the best test of this hypothesis will be for how long daily cases and deaths continue to rise. Figure 4 shows that SARS deaths progressed in an exponential fashion until peaking and then declined rapidly (the same applies to the number of cases). We would hope to see a peak soon but cases and deaths continue to accelerate for now. Also, in terms of putting matters into perspective, during the month that 400-500 people have died from Coronavirus, there are likely to have been 7,000-15,000 deaths in China from seasonal flu9.

If the above is correct, we expect little noticeable effect on China’s full year economic data and even less on global aggregates. If so, we believe financial markets will eventually be driven by other factors (global growth trends and politics in the US, for example).

However, we also believe the virus data will get worse before it gets better and would not be surprised to see further bouts of panic as cases and deaths accelerate and spread geographically. Normally, we would expect to see improvement as the weather improves during Q2 but the extreme measures taken by China could advance the peak in cases and deaths. Figure 4 shows that SARS deaths peaked at the end of April 2003, which was when the Hong Kong stock market bottomed. It would be convenient if the two were linked but we are not sure: global markets were weak at that time for other reasons (the end of the dotcom bubble and the US invasion of Iraq) and no similar correlation with global markets is apparent during the swine flu outbreak (which was more centred on the US).


Figure 4

We started the year believing that stock markets were stretched and in need of a consolidation phase. It strikes us that Coronavirus has given the pretext and we think it could yet push markets lower. However, and unless we are very much mistaken, we believe the economic impact of this virus will be short lived and barely noticeable by the end of the year. We therefore intend to use any future market weakness to add to our equity allocations (we are currently underweight equities – see our Model asset allocation in Figure 8). We would expect the biggest rebounds in the assets that have suffered the most over recent weeks, especially those in the emerging markets.

As said by a US president from a different era: the only thing to fear is fear itself.


Unless stated otherwise, all data as of 5 February 2020. Paul Jackson is Global Head of Asset Allocation Research with the Global Market Strategy Office. András Vig is Multi-Asset Strategist with the Global Market Strategy Office.


^1 Source Wertheim J.O., Chu D.K.W., Peiris J.S.M., Kosakovsky Pond S.L., Poon L.L.M.: A case for the ancient origin of coronaviruses. Journal of Virology. June 2013

^2 Source Duan S.M., Zhao X.S, Wen R.F., Huang J.J., Pi G.H., Zhang S.X., Han J., Bi S.L., Ruan L., Dong X.P.: Stability of SARS coronavirus in human specimens and environment and its sensitivity to heating and UV radiation. Journal of Biomedical and Environmental Sciences. Sep 2003

^3 Source Slovic P., Fischhoff B., Lichtenstein S.: Facts and fears: Societal perceptions of risk, in NA – Advances in Consumer Research Volume 08, eds. Monroe K.B., Abor A., MI: Association for Consumer Research, Pages; 497-502. 1981

^4 Source Dawood F.S., Luliano A.D., Reed C., Meltzer M.I., Shay D.K., Cheng P-Y et al.: Estimated global mortality associated with the first 12 months of 2009 pandemic influenza A H1N1 virus circulation: a modelling study. The Lancet Journal. June 2012

^5 Source The Neglected Dimension of Global Security: A Framework to Counter Infectious Disease Crises, the US National Academy of Medicine's Commission on A Global Health Risk Framework for the Future. Journal of Clinical Psychology. 2016; 33: 1125-1127

^6 Source Fan V, Summers L, Jamison D: The inclusive cost of pandemic influenza risk. The National Bureau of Economic Research, Cambridge, MA2016

^7 Source Jonas O.: Pandemic Risk. World Bank. 2014

^8 Source The economic impact of the 2014 Ebola epidemic: Short- and medium-term estimates for West Africa. World Bank, Washington, DC2014

^9 Source China does not publish flu statistics in the same way as other countries (it only counts deaths directly caused by flu and not those caused by complications such as respiratory problems). This range of estimates assumes that the total number of annual deaths is spread over either 6 or 12 months, using the 88,100 annual deaths estimated by Li Li, Yunning Liu, Peng Wu, Zhibin Peng, Xiling Wang, Tao Chen et al.: Influenza-associated excess respiratory mortality in China, 2010–15: a population-based study. The Lancet Journal. September 2019.

^10 Source Sands P., El Turabi A., Saynisch P.A., Dzau P.V.: Assessment of economic vulnerability to infectious disease crises. The Lancet Journal. May 2016