Since I wrote last time on COVID-19 a lot of new developments have happened towards the spread dynamics of the disease as well as measures taken by the government to prevent the contagion. Therefore I thought it would be best to revert on this in a couple of weeks time and take a status of the situation, again from numbers, stats and charts perspective. So here I am with some new numbers and plots.
Since my last blog, there was huge surge in the number of cases as well as growth rate. In the plot below we see a large surge on 1st of April with growth rate exceeding 40%, followed by around 25% growth on 2nd. 3rd April showed a very low value which could just be some statistical glitch, where the cases were probably reported next day. This trend continued for few more days, hitting 30% mark again on 6th. It was a large setback to the support systems fighting against COVID-19 but more to the morale of people who had survived two weeks of lock down and were looking at the last week in suspicion. It was argued by many that this surge was related to the Markaz incident at Delhi where a large congregation of Tabligh Jamat happened, flaunting the social distancing norms. The congregation had happened some two weeks prior, just before the lock down and it was surmised that the cases were visible after the known 7-14 days of incubation time for COVID-19. I would say that although the argument seems plausible due to the concurrence of the events, we cannot say firmly if there was any causal link.
Near the end of first lock down we saw that many chief ministers of various states imposed a further lock down in respective states. That was a signal for the stricter measures and now India has entered a second lock down until 3rd of May. This nearly one and half month of total lock down has already affected negatively the economy and there are reports that growth rates for India will be severely low this year. It is also expected to lead to large number of job cuts. However, the situation may not be as worse since the calamity is worldwide and we will just be as bad as the rest of the world (probably a little better).
So the key question is, do we see any improvements in the situation due to this two-phase lock down? This time I intend to bring simpler answers, not through some complex time series models, but via a simple chart. We saw above the daily growth rate. But to make a better sense of the numbers, we can simply take a moving average of this. Those who are familiar with technical analysis of stocks would quickly recall that this is one of the simplest indicators of the trend. One key difference must be noted here. Since growth rate is an exponential beast, simple averaging does not work here. What is more appropriate is what is called a 'Geometric Mean'. However, once we do that the rest of the logic just follows. So what do we see from these moving average charts?
We see that the longer period moving average looks smoother. This is expected since moving average is a smoothing filter. Then we see repeated cross overs between 5-day and 10-day moving averages. We can see a negative crossover around 27 March that was probably the effect of first lock down. We clearly see that there is a positive crossover around 1st of April, indicating there was significant rise in the growth rate and hence the number of cases around that. And then we see a very assuring negative crossover again around 7 April. Keen and cautious observers would also see switching and repeated cross overs every 6-7 days. But we have avoided such a positive crossover around 13 or 14 of April. In fact the 5-day moving average is now firmly below the 10-day one.
So do we take that as a good sign? Yes, but only a sign. We can see that the daily growth rate is still lingering around 10% mark. That has to go substantially below 10% to say that we are in any kind of recovery mode. And even after that, social distancing is probably going to stay till September. So we could truly meet all our relatives directly in Diwali. Till then, stay home stay safe.