Everything is Relative: Florida COVID – 19 Trend Update

“It doesn’t matter what we think about a trend, it matters what the crowd thinks about it, but more importantly, how they will respond to it.”

– Mike Shell

For a quick update on the Coronavirus COVID – 19 trend, I’ll use my home state of Florida as the example.

The first cases of Coronavirus (COVID-19) were confirmed on March 1st, 2020, which occurred in Manatee and Hillsborough County. During the initial outbreak of Coronavirus in the United States, Florida’s public beaches and theme parks were under scrutiny as being areas of large crowds. Some in the news media criticized Florida for being relatively late in issuing a “Shelter-At-Home” order, finally putting it in place beginning April 3rd, 2020. Cases ramped quickly from 2 on March 4th, to over 5000 by the end of the month. Since then, however, the number of cases in Florida has leveled off, slowing the rate of change.

I focus on the direction of the trend and its rate of change.

The COVID Tracking Project has now tracked 85,826 cumulative Florida Coronavirus cases , up from 82,719 Thursday. This is a change of 3.88%. Here, I show the standard arithmetic scale on the chart.

The concern I see in the above chart is it seems to be forming a rough S-shaped curve. That is, cases trended up though April and May around the same pace, but this month the rate of change is notably stronger in the  linear price scale of an arithmetic chart. The arithmetic or linear chart doesn’t illustrate or scale movements in relation to their percent change, but instead, the linear price scale plots price level changes with each unit change according to a constant unit value. So, there is an equal distance between the data points as each unit of a change on the chart is represented by the same movement up the scale, vertical distance, regardless of what the level when the change happened. The arithmetic chart is the standard basic chart, especially over shorter time series, and it shows absolute trends.

To see how the time series unfolds with a focus on percentage of change, we changed the scale to logarithmic. The logarithmic chart is plotted so that two equal percent changes are plotted as the same vertical distance on the scale. Logarithmic scales are better than linear scales for normalizing less severe increases or decreases. Applying a logarithmic scale, the vertical distance between the data on the scale the percent change, so we can better identify changes in rates of change. Here, we see a strong uptrend in March, then the rate of change has since leveled off. The trouble, however, is it is still trending up and at its high.

Florida Coronavirus Tests Administered is at a current level of 1.5 million, which up from 1.486 million the day before, an increase of 1.72%.

COIVD – 19 Deaths have increased 1.4% since Thursday. Deaths are obviously an essential factor to track. Florida Coronavirus Deaths is at a current level of 3,154.00, up from 3,110.00 yesterday.

The steep uptrend in deaths is scary looking using the arithmetic scale showing the absolute trend in cumulative deaths. In the next chart, we observe the same trend as a log scale, which shows the rate of change is in an uptend, but has been slowing. I labeled the highest high (now) and the average over the period for reference.

Florida Coronavirus Hospitalizations is at 12,862, up from 12,673 the prior day, which is a change of 1.49%. To focus on the rate of change, here is the log scale chart.

Keep in mind, my objective here isn’t to rehash the research of others, but instead to share what I see in the trends and rates of change. As such, this isn’t a complete analysis of the virus. It’s my observations, as a quant and trend system developer and operator. The data source is The COVID Tracking Project which can only report the data as provided by the states.

ZOOMING IN TO PER DAY

The per day trends are important if we want to spot a change in trend quickly. As I warned in “In addition to the equity markets entering a higher risk level of a drawdown and volatility expansion, we now have a renewed risk of the scary COVID narrative driving more fear” a week ago, the uptrend got some attention last week. It doesn’t matter what we think about a trend, it matters what the crowd thinks about it, but more importantly, how they will respond to it.

The uptrend in Florida Coronavirus cases per day has indeed continued and with a notable new high.

I don’t like to see an uptrend like this because it’s a virus, and viruses are contagious, so they spread. In the case of Coronavirus, we can get an idea of the speed and rate of spread by the reproductive number (R0), or ‘R-naught’, represents the number of new infections estimated to stem from a single case. The reproductive number (R0) is relatively high, according to a research paper on the CDC: Assuming a serial interval of 6–9 days, we calculated a median R0 value of 5.7 (95% CI 3.8–8.9). 

I’m not going into the details here, but, with a reproductive value of 5.7, an increase in new cases is material in my opinion. That is, once it trends up as we are seeing now, it seems more likely to continue.

Are new cases a function of increased testing?

Some say the increase in new cases per day is a result of more testing. That doesn’t seem to be the case. Below is a charge of cases per day with a time series of tests administered per day under it. Visually, we see no correlation. However, there are many caveats to the data. So, anyone who wants to make a cased leaning one way or another can find ways to skew it, but it is what it is. We have a material increase in cases in Florida.

QUANTIATIVE ANALYTICS

Now, we’ll take a deeper dive and apply some analytics to the trends by observing some ratios.

The Florida COVID – 19 Death Rate has been gradually trending down. Florida Coronavirus Death Rate is at 3.67%.

In the past two weeks of May, the death rate was 4.6%, so it is falling.

In our investment management, I’ve been drawing ratio charts for over two decades to determine which market or stocks has greater trend momentum than another. When the numerator (top) is trending stronger than the denominator (bottom value) we say it has stronger relative strength or momentum. In this case, I have used Florida Coronavirus Cases Per Day as the numerator (top value) and Florida Coronavirus Tests Per Day as the denominator (bottom value), which shows a clear uptrend in the cases per day relative to the tests per day. This concerns me because of the rate of spread. As you look at the ratio chart, consider that a value of 0 would mean new cases per day is the same as new tests per day. Instead, new cases is currently trending higher than testing.

Florida cumulative cases relative to tests administered is also showing some change in trend. the past few weeks. Again, not of the date collected is perfect, but it’s still representative of a statistically significant sample of the population.

My objective for trend following is to identify a trend early in its stage to capitalize on it until it changes.

Comparing per day cases to other states doesn’t mean a lot, since the data needs to be normalized. For example, what President Trump said a few weeks ago is a true statement: the number of cases are a function of testing. If we didn’t test and didn’t categorize a case as COVID, there would be no “COVID cases.” Some people, politically motivated, seem to have difficult understanding that simple statement. I’m not politically motivated, so I just say it like it is. With that said, California is winning the match of the most cases per day followed by Texas. Florida is above Arizona.

Again, this doesn’t tell us anything aside from the absolute number. A relative comparison is often necessary and this is an example. For example, we could first calculate per day cases relative to tests or population, then compare them. That’s beyond the scope of my objective today.

Here are the states that reported over 500 new cases. We are seeing some large bubbles in the southwestern United States right now.

Image

The bottom line is, we want to see these levels drifting down, not up. We want to see this trend down.

People who are at high risk should continue to operate according to the risks, but also keep it in perspective that at this point, it isn’t yet so wide spread.

In the big picture, the population in Florida is 22 million and about 86 thousand cases have been labeled COVID 19. 86,0000 out of 22 million is about 4 tenths of a percent, or 0.40%.

That’s 40 cents of $100.

Our changes of contracting COVID 19 in Florida, then, is less than half of 1% at this point.

Everything is relative.

Join 40,598 other followers

Mike Shell is the Founder and Chief Investment Officer of Shell Capital Management, LLC, and the portfolio manager of ASYMMETRY® Global TacticalMike Shell and Shell Capital Management, LLC is a registered investment advisor focused on asymmetric risk-reward and absolute return strategies and provides investment advice and portfolio management only to clients with a signed and executed investment management agreement. The observations shared on this website are for general information only and should not be construed as advice to buy or sell any security. Securities reflected are not intended to represent any client holdings or any recommendations made by the firm. Any opinions expressed may change as subsequent conditions change.  Do not make any investment decisions based on such information as it is subject to change. Investing involves risk, including the potential loss of principal an investor must be willing to bear. Past performance is no guarantee of future results. All information and data are deemed reliable but is not guaranteed and should be independently verified. The presence of this website on the Internet shall in no direct or indirect way raise an implication that Shell Capital Management, LLC is offering to sell or soliciting to sell advisory services to residents of any state in which the firm is not registered as an investment advisor. The views and opinions expressed in ASYMMETRY® Observations are those of the authors and do not necessarily reflect a position of  Shell Capital Management, LLC. The use of this website is subject to its terms and conditions.