COVID-19 turned into a global pandemic because this novel coronavirus could spread weeks without detection. There are several data points that indicate that 20-50% of all coronavirus infections are asymptomatic. Almost all countries are merely reacting to the virus rather than proactively working on preventing its spread. Turkey is one of these countries. It is as if the new coronavirus came out of nowhere. Just 11 days ago Turkey had only 2 coronavirus deaths. The death toll skyrocketed 4500% to 92 in just 11 days. This means it is now too late for Turkey to prevent the spread of the virus.
I published an article on March 20 and calculated that 2 million people in the USA had coronavirus as of March 20. The model I developed estimated that the death toll in American will surpass 800 by March 26 and 20 thousand by April 15th . On the morning of March 26, the number of coronavirus deaths in America was 1042. In other words, the model I have developed produced actually very conservative estimates (i.e. the actual number of deaths and infections are higher than what we predicted).
In this article I am going to model the number of infections in Turkey and make conservative predictions about the minimum number of deaths that we should expect to see in the next 3 weeks.
In the video below (the video is in Turkish so that Turkish people could understand our predictions and take actions to protect their families), I describe how we estimated a 1% infection fatality rate for COVID-19. The actual IFR is in the range of 0.5% and 1.5%, so 1% is a relatively reasonable estimate to use in our model.
The main implication of a 1% infection fatality rate is as follows: if 100 people are infected with the new coronavirus today, only 1 of them will lose their struggle with this virus and the remaining 99 people will survive this ordeal.
Approximately 5-6 days after a person is infected with the virus, she begins to show symptoms (fever, cough, fatigue, etc.). Probably nearly half of the infected people do not show any symptoms. Therefore, it is thought that these people have an important role in the spread of this virus so rapidly. After the patients show symptoms for 5-6 days, some of them become more severe and have to be hospitalized. The hospital stay is around 14 days on average and then the case is resolved (either recovery or death).
In total, it takes around 24 days for the patient to lose her life after getting infected with the virus.
This is an important figure.
In Turkey today at least 92 people lost their lives because of the coronavirus. These people were not infected with this virus today, nor were they infected with it yesterday. These people were infected with this coronavirus at least 24 days ago.
In other words, 92 people were infected this virus on March 3rd or earlier and lost their lives after fighting with this virus for 24 days on average. We also know that only one out of every 100 people who get the coronavirus lost their lives.
This means that on March 3, in Turkey there were 100 coronavirus infections for each deaths TODAY. So, there were a total of 9200 people infected with the new coronavirus in Turkey on March 3. If you understand how we calculated this 9200 figure, it's relatively easy to understand the rest of our model.
In the USA, Italy and other countries the number of infections and deaths doubled every 3 days before they started implementing social distancing measures. The figures in Turkey also point to a similar growth rate. We know that 1% of those who are infected dies after about 24 days. Therefore, we can calculate the rate of increase in the number of infections by looking at the rate of increase in the number of those who lost their lives.
For example, in Turkey the death toll was 44 in the 24 March. This number doubled in 3 days and reached 92 (the rate of increase is a little more than 100%).
On March 21 the death toll in Turkey was 21. This number doubled again in 3 days and reached 44 (again, the rate of increase is a little more than 100% but we are going to assume that it doubles every 3 days to simplify the calculations).
Now we can start to predict the actual number of infections and project the death toll for the next 3 weeks.
We already calculated that there were around 9200 infections in Turkey on March 3rd. For the sake of simplicity let's round this number to 10000 (we can reduce the final estimate by 8% to correct for this simplification).
Since the number of infections are doubling every 3 days, we know that it will double to 20000 on March 6th, 40000 on March 9th, 80000 on March 12th, 160000 on March 15th, 320000 on March 18th and 640000 on March 21st. This means there were at least 500,000 people with infections in Turkey on March 21. Turkey started taking certain actions this week, so we will stay conservative and assume that the number of infections stayed the same since March 21.
The immediate implication of this figure is that nearly 1 out of every 150 people in Turkey is infected with the coronavirus today. Half of those infected people probably do not show any symptoms. The other half will start showing symptoms in the coming days. Probably in about a week, there will be no intensive care beds left in Turkish hospitals. By mid-April, the death toll in Turkey will exceed 5000 (one percent of 500 thousand people is 5,000 people).
Remember, these are very conservative estimates. The actual figure could be 100-200% higher than these. We expect and hope that the only reasonable reaction for the Erdogan government is to implement nationwide lockdowns to stop the spread of the virus. This will have serious economic implications and which is why we expect iShares MSCI Turkey ETF (NASDAQ:TUR) to decline at least 10% by the end of April. I personally own a small position in this ETF but it is a hedged position.
Disclosure: Long TUR. This article is originally published at Insider Monkey.