COVID-19 and the Proliferation of Junk Science

OVERVIEW

Over the past year a plethora of strategies to combat COVID-19 were put into action. However very few of these strategies were backed by publicly available credible data. In fact, even as COVID-19 cases per 100 tests dropped sharply through spring 2020, alarmists published misleading studies and governments took policy action basis these studies. These issues were exacerbated by the human mind’s susceptibility to recency bias, fondness for doom scrolling (i.e., alarmist news) and inability to appreciate probabilities. Investors who were able to address all or some of these issues benefited from the panic created around COVID-19.

A year into COVID-19, we are none the wiser

We have now crossed the first anniversary of the emergence of COVID-19 and we are none the wiser about its origins. China is largely blamed by the western media for the initial outbreak and the alleged cover-up by them, which supposedly allowed the disease to spread globally. This version is obviously disputed by Beijing, which has repeatedly claimed that the virus originated outside China (China raises doubts about COVID-19 origin, 2020: Source – CBS news).

The ability of the disease to spread rapidly alongside overly alarmist academic studies (see The shocking Coronavirus study that rocked the US and the UK, 2020: Source – Financial Times) put governments across the world into overdrive as they deployed diverse strategies to combat this crisis. And yet, looking at the global data (see Exhibit 1) it is not apparent that: a) there is any discernible pattern to the incidence of COVID-19 across four large countries; b) lockdowns (or a lack thereof) have a significant impact on the incidence of COVID-19.

In fact, it is the rapid drop in cases per 100 tests in March/April 2020 that led the Marcellus team to conclude in early April 2020 that the COVID-19 threat was being radically overplayed – see Marcellus’ April 2020 blog. At a time when western ‘experts’ were saying that India will see in excess of 100 million COVID-19 cases (see alarmist study from CDDEP), the Marcellus team estimated that India’s case count was likely to be around 10 million. The method used by Marcellus team to arrive at this estimate is highlighted further on in this note.

Is the current euphoria around COVID-19 justified?

Given the chaos which we saw in 2020, it is remarkable therefore that now there is a sense of euphoria that we are past the worst of the damage inflicted by COVID-19. Part of this elation is due to the imminence of the vaccination drives which have been initiated globally. Yet if we look at the data shown in Exhibit 1, the most relevant metric – cases per 100 tests – has been rising for the advanced economies since October 2020 even as countries like India are showing a steady drop since July 2020. (Note: It is pointless to focus on total COVID-19 cases because as testing rises, so do total cases. A more useful metric is new cases per 100 new tests.)

So, after everything we have been through over the past year, how should investors really think about COVID-19? And how should we conceptualize the disease as we go about our day-to-day business?

Rational thinking helps

Firstly, we need to treat the grand claims (both pessimistic and optimistic) about COVID-19 (and other issues) with a degree of scepticism. In March 2020, three of the most alarmist studies around COVID-19 were published:

The University of Oxford published a study about COVID-19 infections. It stated that the UK already had considerable instances of the infection by March itself that had made most of the country’s people immune to the disease (The Impact of Host Resistance on Cumulative Mortality and the Threshold of Herd Immunity for SARS-CoV-2, 2020: Source – Medrxiv.org). As shown in Exhibit 1, this does not seem to be the case at all.

A celebrated study by Imperial College likened the COVID-19 pandemic with the 1918 Influenza pandemic in terms of the severity (Imperial College study, 2020: Source – Imperial College website). This, it turns out, was a gross overestimation of the risk posed by COVID-19. The 1918 flu epidemic claimed 228,000 British lives (Deaths during 1918 pandemic, 2021: Source – Metro news). COVID-19 has so far claimed 82,00 British lives (British COVID-19 fatalities, 2021: Source – worldometers.info).

The Centre for Disease Dynamics, Economics and Policy (CDDEP) along with John Hopkins University (who later pulled their name out – read more here) published a study about COVID-19 hitting India particularly hard and affecting more than 125 million people (Predicted infections from COVID-19, 2020: Source – CDDEP website). So far India has seen 10.5 million COVID-19 cases and given the trend seen in Exhibit 1, it seems unlikely that the CDDEP’s 125 million estimate will be the actual outcome.

Secondly, it is crucial to not get influenced by the recency bias – which makes a person give more value (fear) to information that he has come upon recently relative to what he learned a few years ago. For example, in 2018 in India, road accidents had a Case Fatality Rate (CFR = total deaths from recognized cases/total recognized cases) of ~32% (Road accidents statistics, 2019: Source – Autocar pro website) vis a vis the 1.44% CFR of COVID-19.

To understand the power of recency bias, let us frame the situation in a different way by taking numbers for the state of Maharashtra. 39 people were killed for every 100 road accidents in Maharashtra in 2019 (Road accidents in Maharashtra, 2019: Source – Highway police, Maharashtra government) vis a vis 3 people per 100 COVID-19 cases as of 31st December 2020 (COVID data widget, 2021: Source – GitHub widget for live tracking). And yet, if you look on the streets of Maharashtra, you will many people are riding their 2-wheelers with masks but without helmets!

Similarly, tuberculosis, also an infectious disease, had a CFR in India of 17% in 2019 (TB statistics, 2019: Source – Worldhealthorg website) significantly higher than that of COVID-19 (1.44%) in India. In fact, when compared with other infectious diseases prevalent in the country, COVID-19’s death count (151,160) as well as death rate (0.01%) is the least compared to the top 5 communicable diseases in India, as evidenced in Exhibit 2.

And lastly, most people find it hard to correctly recognize probabilities of events happening and tend to overstate them when uncertainty is put into the mix. In a March 2020 blog, Marcellus used Bayes’ theorem, which is a formula to calculate the conditional probability, or the probability that an event will occur given some information, of an event (read more here), to predict various probabilities relating to COVID-19. (Read: Reverend Thomas Bayes and a Big Buy Call on India, 2020: Source – Marcellus Investment Managers)

According to the data available in March 2020, as shown in Exhibit 3 and the calculations done in Exhibit 4, even if one had symptoms like cough & cold, the chances of one having contracted COVID-19 was only 5%. Furthermore, in the absence of any symptoms, this percentage fell to a mere 0.42%. [Interestingly, Marcellus’ March 2020 estimate that around 10mn Indians could contract COVID-19 was more accurate than the scientific studies published by Western ‘experts’. In fact, since our 10-million estimate is very close to the actual number, the probability of an Indian contracting COVID-19 is 0.77%, identical to our March 2020 estimate! All credit for such accurate predictions should go to Reverend Thomas Bayes for creating a remarkable way of estimating conditional probabilities.]

Events like COVID-19 are a test of investors’ ability to stay rational in the face of mass hysteria fuelled by junk science which gains rapid circulation thanks to social media. To counter junk science (and the waves of alarmism and optimism created by it), we need scepticism regarding outlandish claims alongside the ability to work out rational probabilities ourselves.

Mark Mobius is the Founder of Mobius Capital Partners LLP. Saurabh Mukherjea and Nandita Rajhansa are part of the investments team at Marcellus Investment Managers.