Are emerging markets the “next shoe to drop”, so to speak, in the COVID-19 pandemic?
JPMorgan’s Marko Kolanovic doesn’t think so. Or, at the least, he says there are potentially important mitigating factors to consider when it comes to postulating a severe outbreak in the developing world.
Clearly, there have been some EM “hot spots”. South Korea and China are emerging markets, after all. Their outbreaks are largely under control now. Turkey is potentially worrisome, as is Brazil. But, when it comes to which countries have suffered the most, developed markets have obviously been hardest hit.
That’s a simplistic visual, but Kolanovic notes that “if one looks at COVID-19 deaths per million of population, within the top 20 countries, there is only one emerging market — Iran”.
Iran is a special situation. Most obviously, it’s cut off from dollar funding and although Donald Trump repeatedly suggested the crushing sanctions imposed by Steve Mnuchin would not hinder the regime in Tehran from obtaining international aid and humanitarian support, those sanctions (along with recent events, including the accidental downing of a passenger flight during retaliatory strikes against US targets in Iraq earlier this year) certainly did not help. Simply put: You don’t want to be a pariah state in a pandemic. In addition, Kolanovic writes that “Iran has a relatively cold climate and relatively older population for an emerging market country”.
Looking at the list of countries with the highest number of deaths per million, Marko notes that the next country which comes up is Ecuador, which he observes is “the only country in the developing world across South America, Asia, and Africa that does not have a universal Bacillus Calmette-GuÃ©rin vaccination policy”.
Why does that matter? Well, as it turns out, there’s a statistically significant relationship between those vaccinations and reduced coronavirus infection rates and mortalities. This dynamic has been documented at length in a variety of places, including a feature piece in The New York Times this month which begins with these passages:
A vaccine that was developed a hundred years ago to fight the tuberculosis scourge in Europe is now being tested against the coronavirus by scientists eager to find a quick way to protect health care workers, among others.
The Bacillus Calmette-Guerin vaccine is still widely used in the developing world, where scientists have found that it does more than prevent TB. The vaccine prevents infant deaths from a variety of causes, and sharply reduces the incidence of respiratory infections.
The vaccine seems to “train” the immune system to recognize and respond to a variety of infections, including viruses, bacteria and parasites, experts say. There is little evidence yet that the vaccine will blunt infection with the coronavirus, but a series of clinical trials may answer the question in just months.
As noted, the vaccine is used widely in emerging markets.
Kolanovic looks at COVID-19 cases and mortalities per million as a function of the percentage of the population older than 60 using a sample of the 100 most-affected countries. Not surprisingly, age matters, but there’s a lot of dispersion in the data.
The outliers above the trend line (Italy, the Netherlands, Belgium, Spain, Sweden, UK, Switzerland, France and the US) either never had a BCG vaccination program or else discontinued it. In fact, as Marko puts it, “virtually all countries disproportionally impacted by COVID-19 do not have current BCG programs”.
Note that Japan is the mirror image of Ecuador. Ecuador has a high mortality rate for an EM and is among the only developing countries without a universal BCG vaccination program. Japan, on the other hand, has a very low mortality rate despite being the oldest country in the developed world, and it just so happens that Japan has a BCG program.
This being Kolanovic, he doesn’t simply speculate – rather, he does the actual math.
Here’s how Marko tests the BCG link hypothesis statistically (and I’m just going to use a block quote to make sure there’s no chance of leaving out something important):
We have correlated mortality and COVID-19 prevalence adjusted for age (i.e., residuals above or below the trend line in the charts above) to BCG vaccination status. For the BCG status score we assigned 0 if the country never had a program, and a fractional score if the country has or had a program in the past. The fractional score was calculated as the number of years the program was in place expressed as a percentage of the average population age. Multiple/Booster BCG vaccines were added to the score as an additional number of years based on the age when administered.
I’ll save you the suspense: The results of the regressions are statistically significant. BCG vaccination status matters, or if that’s too definitive a statement, we can simply say that statistically speaking, there’s a correlation between the presence (or not) of vaccination programs and COVID-19 prevalence and mortality.
Kolanovic writes that in his study, “the reduction that can be linked to BCG status… is estimated to be a factor of ~2”. More simply, that translates to a roughly ~50% reduction of COVID-19 prevalence and mortality.
Researchers at the University of Michigan came to similar conclusions. “The presence of national policies for universal BCG vaccination is associated with flattened growth curves for confirmed cases of COVID-19 infection and resulting deaths in the first 30-day period of country-wise outbreaks”, they wrote earlier this month. (The link is to the actual study, but the University released a summary you can read here.)
In addition to the BCG hypothesis, Kolanovic reminds you that EMs are typically younger which, as most readers are likely aware, and as baked into all the analysis mentioned above, matters quite a lot for COVID-19 mortality.
“The average age in Africa is 19.7 years, MENA and India 26.8y, and Latin America 31y, which can be compared with, for example, an average age in Europe of 43.1y”, Marko writes, emphasizing that “for an exponentially increasing mortality curve, this makes a huge difference”.
Finally, he cites multiple studies which suggest higher temperatures and humidity may help dampen (no weather pun intended) COVID-19 case rates. Obviously, many emerging markets have higher temperatures and humidity versus developed economies. On this point, Kolanovic writes only that it “appears plausible”, and in any case, could help at the margins when taken in conjunction with everything noted above.
Putting it all together, the relative youth of EMs almost surely reduces COVID-19 mortality versus, for example, Europe, and while correlation doesn’t equal causation, there is a statistical case to be made for the contention that BCG vaccination status matters, and could further inoculate (figuratively and perhaps literally) developing economies from the virus.
Of course, no statistical model can fully account for the myriad vagaries of emerging market politics. There’s always the chance (and this is me talking, not Kolanovic) that questionable decision making by leaders in developing economies makes a bad situation worse. Additionally, EMs are more vulnerable through weaker healthcare systems, external funding needs and fragile currencies, among other potential stumbling blocks.
That’s not to suggest mistakes haven’t been made by officials in developed markets – they clearly have. It’s just to state the obvious, which is that it is impossible to predict how things will play out when you pit a highly contagious pathogen against a personality like, for example, Brazil’s Jair Bolsonaro, who fired his health minister this week for refusing to endorse Bolsonaro’s characterization of COVID-19 as “a little flu”.