COVID cases from winter and testing

Based on COVID SIR model, fitting data to (a) best information about unknown/known case ratio, (b) portion of declining mortality attributed to including less critical cases in reported data, (c) projections which fit incoming new data.

I have a background in safety and modeling the combination of human behavior and complex systems (spacecraft, transportation, etc.) to determine accident or crash rate. Since late March I’ve been modeling COVID, gradually developing models as sophisticated as any available.

In July I became aware that the relative number of unknown cases was diminishing fast enough to undermine projection accuracy. I went on vacation for a month instead of looking into it, as I was totally burned out.

Recently the US and European media have been reporting “spikes” in cases and talking about severe lockdowns. That scared me, because I think lockdowns are more damaging that people realize. According to one paper, half as many people died from the lockdown in the US in March and April as from COVID for reasons such as having heart attacks and not going to the ER.

A seasonal rise in cases – similar to the flu – was baked into COVID from the start. This was WELL KNOWN by April. Many studies were done based on temperature and humidity and several papers were published. The average seasonal effect for the US is 20%, but colder regions like NY have a 30% effect.

Spikes over the summer were from “easing.” Spikes now are likely to be from cold weather.

But the spike now, should it be so large? Where I live (Houston) I do not see any “easing” going on. It gets tighter every week for the last two months.

So reluctantly I began two weeks ago to look into the changing case ratio. If you do a lot of testing, the number of reported cases goes up, even if the “actual” number of cases goes down or remains the same or only goes up slightly.

It is easy to check this. If less critical cases are included in the totals, mortality should look like it is going down. It has, in the US from over 8% of reported cases to only about 1%.

Of this factor of 8, by fitting projected model data to actual new data coming in, it appears a factor of 4 is due to increased testing, and a factor of 2 is due to improved medical understanding of and handling of critical cases. Not everyone needs intubation, and steroids keep the body from killing itself. COVID blocks a friend/foe identifier on lung cells, and the body attacks itself and dies even while the virus load is declining.

The results are startling. Our lying, hateful president is finally right about something, like a stopped clock is right twice a day. Who’d have thought.

We should not go into a severe lockdown. Instead we should skip phase 3 trials and immediately start vaccinating people. Approximately 20,000 people in the US will die for every month we delay.

We’ve never had a situation like this is vaccination history. We’ve had situations where the number of people who might die from a bad vaccine was a few hundred or thousand, and the number of lives saved by the vaccine was lower, maybe a few dozen. But in this case the numbers are reversed. Dramatically.

Forks in the Road 2014 & 2019

I was investigating whether other market segments were keeping up with the S&P 500, which seems to be ignoring the plight of ordinary people and charging ahead in spite of COVID-19. To my surprise, the biggest forks were in mid-2014 and early-2019.

In 2020, there was a sharp divergence with the S&P 500 getting ahead in April by about 8%, but after that the various segments mostly came along together:

Comparison of large and small and micro caps and US vs. world markets, 2020, from Yahoo Finance

But the forks in 2014 and 2019 both have lasted up through the present:

Same comparison from 2015

In 2014, mid-year, it is easy to identify suspect events that led to a continuing 9.5% per year differential, with emerging markets and world-other-than-US essentially flat while the US grew rapidly:

  1. Russian invasion of Ukraine and subsequent sanctions on Russia
  2. Emergence of ISIS and counter attacks
  3. Unrest in Yemen
  4. Fracking boom and downturn in oil prices

These problems led to mass migrations, affecting Europe. So Russia, North Africa, and the Middle East have all been in turmoil since. The crash in oil prices caused economic decline in parts of the Middle East that avoided war, and in Venezuela, while the US benefitted from fracking and becoming an energy exporter.

If you have other ideas for the fork in 2014, please leave a comment below.

The causes of the 2019 fork inside the US, which happened early in the year, January or February, are harder to identify. Democrats took over the House of Representatives, causing some stalemate in government. This gradually led to focusing on the border wall and impeachment instead of governance. But did this really affect the fortunes of smaller companies more than larger ones? The mechanism is unclear to me.

It is interesting to plot the price of oil over these periods.  There is a correlation with both 2014 and 2019, which would seem to indicate that the world economy does better with a higher oil price.  But of course this may be a consequence, not a cause.

Oil prices last ten years, from https://www.macrotrends.net/2516/wti-crude-oil-prices-10-year-daily-chart

Perhaps you can think of another correlate. Please leave a comment below, and a chart if possible.