Forecasting the Wuhan Coronavirus

Date of article: 6 February 2020

This article is quite a long read, but it's important and very relevant, not only for investors and traders.

First, I'll explain why stocks in Europa and US are going up, almost ignoring 2019-nCoV, and second I'll explain what investors are missing.

All market research firms are following the coronavirus very closely and many of them are expecting the coronavirus to slow down. They do this by studying the actual reported numbers (just like I'm doing here), rather than going by R0-models.

As discussed before: this is good in my opinion, because using Basic Reproduction Numbers (or the R0) has turned out to be unreliable (the actual transmission numbers are due to many other factors: R0 isn't a good indicator) for predicting the future numbers of nCoV-infections.

The mathematicians doing these calculations also look to the speed of daily increases (this is just like what I have been doing), which is the derivative of the acceleration: also known as the 'jolt'. Below I've made a new chart to show my version of the 'jolt' of Chinese reporting nCoV data.

Based on these mathematical (in my opinion accurate) approaches, market research firms like Gavekal conclude (as quoted by Bloomberg) that the virus will peak early March.

However, these market research firms and institutional investors are forgetting one very important point..

Their estimates are very useful if the 2019-nCoV would be an outbreak with homogeneous infectivity conditions. That is however certainly not the case. Thailand or India will not be able to handle the virus like China is doing (lockdowns, mass surveillance, or ordering a pizza and getting a note with the temperature of the cook, the delivery guy and the waiter).

If you don't take into account the heterogeneity of the infectivity conditions, exponential growth in infections in future epicenters (e.g. Thailand or India) will not be flagged, if China is able to contain it within their borders (which is currently 99% of infections).

That's why I have been looking to growth in infections and infection rate in subsets of China and I will do so in the future (once there's enough data) for countries outside China. E.g. looking to growth in China, excluding the epicenter, or looking to three megacities (see below)

Summarized: yes the containment of the coronavirus is definitely looking good in most places in China, outside the epicenter. However, new exponential growth in China-neighbouring countries is likely, in which case the virus peak will turn out to be far beyond March 2020.

Added on March 8th, 2020:

To visualize what I had been saying since late January, I have added the following two charts. It illustrates how future epicenters will not be flagged when only studying the totals, as mentioned above.

article by Newsnodes



Date Country page
3 May,23:26 Hungary
25 Apr,21:35 Moldova
24 Apr,14:07 Iran
6 Apr,00:20 Spain
1 Apr,23:43 Portugal
31 Mar,08:36 Paraguay
30 Mar,11:57 Belgium
27 Mar,10:50 Paraguay
26 Mar,10:10 Paraguay
26 Mar,10:09 India


Date Country page
18 May,18:03 Finland
9 May,22:18 Sri Lanka
8 May,00:28 United States
7 May,22:43 Brazil
7 May,22:00 Ethiopia
7 May,21:59 Moldova
7 May,21:58 Turkey
7 May,21:56 Algeria
7 May,21:54 South Africa
7 May,19:55 France

© 2024
| About us |