COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes

We realize there are many people interested in updates regarding COVID-19, so please feel free to send along this commentary to anyone who has interest.  


To start, we hope everyone had a Happy Easter.  The weekend has brought incrementally better news on the COVID-19 front, internationally, NYC/NY-state and nationwide as daily case growth shows widespread deceleration and even daily reported deaths are likely apexing.  State and local governments are extending the “stay at home” orders but with diminishing intensity of the crisis, we can start to track how local economies begin to ease restrictions.  

– The White House now expects total expected deaths from COVID-19 to be around ~60,000, down from 100,000-200,000 last than 10 days ago and significantly less than original forecasts as high as 2.2 million.  The IHME projection curves are plotted below (the IHME sent us the raw data).  And tireless Ken (our data scientist) was able to render these revised death forecasts.

– As of the latest Johns-Hopkins update, the US has recorded 22,115 deaths from COVID-19, so total deaths are still expected to triple over the next month.  

– The IHME data has been quite accurate on death forecasts, particularly for NY state, but in the US overall, we believe that the widespread social distance measures is causing the realized deaths to undershoot their models.  So we believe it is possible that the cumulative death forecasts could fall even further.

COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes

 
Secondary economic ripple is to asset holders of America…

As a side note, much of the high amplitude hit to joblessness was seen early in this crisis, because of the abrupt essentially global shutdown, but there remains a secondary effect as the joblessness and social distance begin to ripple across the asset holders of corporate America–holders of assets, including the landlord, banks, suppliers, infrastructure and service providers.   So we should be aware there is a two-step hit coming to the economic data:

– first is the joblessness of the “social distance” victims including restaurants, sports/entertainment, travel-related hotels, airlines, etc.
– second will be the asset holders, as we mentioned, landlords, banks, infrastructure, etc.

And to the extent the healthcare crisis is apexing sooner, the shorter the “social distance” victim recession and therefore, the smaller the economic ripple to the second-order effect.


POINT #1: Widespread deceleration of COVID-19 across vast majority of metro areas, 28% of US sees daily case growth <5%.
We believe the best way to see COVID-19 growth is at the county level (which Johns-Hopkins provides) and a simple way to visualize the spread is to look at diffusion or distribution of daily case growth rates.  And this is showing very good progress:

–  28% of the US (based on sum of counties) is seeing case growth <5% (“green” line) and this line soared over the weekend.  Just 3 days ago, this line was at 12%.  So a full 16% of the US is seeing a massive deceleration of case growth.

– Collectively, 58% of the US is seeing daily case growth <10% per day, or cases take longer than a week to double. Good progress would mean seeing the “green” line (daily growth <5%) and “grey line” (daily case growth >5% <10%) rising while seeing all other lines fall.  And this is happening.

COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes

Daily case growth in USA now 5%, down from 8% a week ago and 13%  two weeks ago… so slowing in a sustained way
Daily case growth in the US (# daily cases/ total cases) has slowed to 5%, compared to 8% a week ago and 13% two weeks ago.  This is moving in the right direction.

– daily case growth peaked on 3/19/2020 at 33% and has since slowed sharply.  

– The law of large numbers is at play at the US has >560,000 cases in total.  Daily new cases was 28,983 and still hovering around 30,000 per day.  Thus, the US is still reporting a large number of new cases.  The US is not “out of the woods” yet.

COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes


COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes


Previously, the White House had named some cities which were seeing outbreaks which could mirror the spread of NYC.  Those were Detroit, Philadelphia, Washington D.C., Chicago and Houston.  Of these, we previously showed Houston and Texas-cities generally, were seeing much flatter growth compared to rest of US.

– The hotspots are plotted below, rebased to day 0 = date # cases >100 and the y-axis is cases per 1mm residents.
– D.C. in the past few days has broken its trajectory and is now tracking slower than NYC.  Very good news.
– Detroit also seems to be “bending” that curve and potentially seeing a flatting of cases.  Certainly, the slope of Detroit is currently flatter than NYC.
– Philadelphia’s slope is flatter than NYC currently.

In other words, it looks like the other cities are beginning to see a real flattening of case growth (cases per 1mm residents) and that is positive.  If other cities are outperforming NYC, we could see a much faster resolution to apex in COVID-19.

COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes

Even New Orleans has apexed cases way ahead of NY state…
And New Orleans has moved far past peak new cases, reaching this apex considerably faster than NY state.  Recall, our previous charts that showed New Orleans parishes had much higher cases per 1mm residents than NYC.  But in aggregate, the growth of COVID in LA is way down.

COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes
COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes

Most of the US is now seeing case growth slow to 2X takes 2 weeks or longer… 
The extent of the slowing of case growth is evident by looking at the state level. We calculated, based on a state’s daily case growth, the number of days for the cases to double.  We need to point out this is a “snapshot” in time, and each day, as more states see a slowing, the number of days to 2X will also increase.

– The chart illustrates that COVID-19 is past the exponential stage — the only states seeing 2X in 4 days-ish is South Dakota.

COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes

POINT #2: Population density vs Weather– is weather an unappreciated factor? warming weather = good
There remain a lot of unknown and unanswered questions about the COVID-19 outbreak,  NYC was the worst outbreak, and it seems most metro areas are doing better than NYC (except Detroit, Philadelphia, at this point).  

Why was NYC so bad?  
Was it population density?  
Ethnic mix (more Italians and Chinese?) 
Hygiene? 
Deliberate spread (terrorism?) 
or Weather?

You might be tempted to think this doesn’t matter — if things are getting better, then that supersedes.  But answering this does matter, to the extent we address a few unknowns:

– what additional factors slow COVID-19 spread, besides the known factors of –> masks, social distance, hand washing, etc.
– if weather is a factor, this could be a positive surprise as warming weather could lead to a significant slowing of spread.

Surprisingly, In NYC, COVID-19 case spread is INVERSE to population density…
Several of our clients commented they felt NYC did considerably worse because of population density.  Our team took zip-code level data and plotted COVID-19 cases per 1mm residents vs population density.  The results might surprise you:

– the most densely populated zip codes in NYC had among the lowest COVID-19 cases.

– even looking at boroughs specifically (the 5 boroughs are color coded), this was true across all 5 boroughs.  The least dense areas had the highest COVID-19 spread.

COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes

Zip code data argues against the idea that high population density caused greater spread…
How can we explain this?  We cannot. But this sure seems to be an argument against urban density causing COVID-19 spread to be worse.


Weather… is it an unappreciated factor?
You might recall this image from a study, titled “Temperature, Humidity and Latitude Analysis to Predict Potential Spread and Seasonality for COVID-19” conducted by UMCP + several other universities (SSRN paper here) and this image below shows the cities with breakouts fall into a narrow band of temperatures.

COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes

And last week, we cited a recent study by Tamma Carleton (University of Chicago) and Kyle Meng (UC Santa Barbara) asserts that COVID-19 transmission is very seasonal  (Preview of study here and Article summary here).  This is not yet peer reviewed and was posted on Medrvix.org and it has intriguing conclusions.  The key abstract below:

“We find robust statistical evidence that a 1°C increase in local temperature reduces transmission by 13% [-21%, -4%, 95%CI]….we project that changing temperatures between March 2020 and July 2020 will cause COVID-19 transmission to fall by 43% on average for Northern Hemisphere countries and to rise by 71% on average for Southern Hemisphere countries.”

Basically, the authors suggest that each 1 degree Celsius rise in temperatures reduce transmission by 13%. Their work isolated weather from public health responses, densities of cities and other factors.  And if they are right, the spread of COVID-19 should slow sharply as the weather improves across the USA.

The biggest outbreaks in the US do fall into a longitude between 38° to 45°…  suggesting weather could be a factor
This map of the US, based on county data, shows gradients of cases per 1mm residents.  That red zone in Idaho (Mountain region on western part of map) is Blaine County, ID which has recorded 19,721 cases per 1mm residents (2% of the population has COVID-19).   

COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes

If we aggregate the data (see below), there are more cases in the US between the longitudes of 38° to 45°.  And the chart on the right is the spread between % of cases and % of population.  So this highlights that there are more cases within these longitudes.

– NY state is the worst outbreak.  So even excluding NY state, this is true.

– Even ignoring NY state (from all the data), the longitudes 38° to 45° have a higher number of cases compared to their proportionate population.


COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes


COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes


We are not making the assertion that weather is a differentiator.  But it is hard to ignore the relationship noted above. And we also showed NYC density was inverse to COVID-19 spread. But the data does suggest that warmer and colder regions of the US have seen fewer cases.  Granted, these other regions could have different ethnic mix, hygiene, etc.

– But if it is weather, then warming temperatures could accelerate the US efforts to slow COVID-19.



POINT #3:  NYC/NYS deaths peaked, tracking closely to IHME forecasts 
The data over the weekend for NYC and NY was positive overall.  And perhaps the most encouraging from Cuomo’s press briefing is the change in hospitalizations.   This figure is down to 53 on 4/11, was down from 1,421 on 4/2 and as the chart shows, not only is the trend way down, but it may be that NY state will see “net discharges” in coming days.

COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes

NYC is showing steady leveling of new cases reported by the 5 boroughs shown below.  The lumpiness of NYC daily new cases is a contrast to NY state which has shown much steadier gains.  

– but what is encouraging, is NYC daily deaths has been making a choppy but steady decline.  The latest daily reported deaths was 457, which is below the 807 reported a week ago.  And this number is important because mortality declining would affirm NYC has the worst behind it.

COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes
COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes



NY state overall is showing steady declines in the number of newly reported cases.  This remains encouraging, as NY/NYC were the worst of the outbreaks. 

COVID-19 update. 58% of US sees case growth <10%. NYC case spread inverse to pop density based on zip codes

STRATEGY: The combination of more imminent COVID-19 Apex, plus potential for faster “re-start” and continued market skepticism, keeps us in the “half-full” camp and see a potential S&P 500 close above 2,800 as further confirmation…

Earlier in this crisis, the bad news cascaded upon bad news, and markets went into a free fall as the sellers dominated in the “get me out of everything” — since then, stocks have staged a V-shaped rally.  

This V-shaped move is consistent with precedent declines of >30% as the 10 prior declines saw markets retrace 50% of their losses within 0.5X the time taken to see the initial decline.

This raises the natural question — did stocks already record their “low low”?  

Logically, one should expect a re-test.  After all, a re-test is evident in other prior markets (most of the time) and it is also logical given the second wave of bad news coming.  

A re-test is so widely held a view, that it is more than consensus.  It is a central view with a narrow right tail that sees “bottom already in.”  And uncertainty is widespread–how deep is the contraction, how long, when will healthcare crisis turn. In a risk asset world, uncertainty is bearish.  Thus, the consensus is unusually bearish (uncertain = bearish) and expects a re-test.

This does show the “half-full” view is arguably a better risk/reward given those factors.

But if markets have bottomed, there are some reasons to explain this:

– Healthcare crisis peaks fast enough to avoid massive destruction of the global economy 
– Consumers will be way more resilient, so the recovery is V-shaped vs consensus L- or something worse 
– Fiscal policy was fast enough to stave off a deeper contraction 
– Fed has moved quickly enough to stave off a deeper financial market crisis 
– Stocks over-reacted to the pandemic

We realize Consensus certainly does not expect this.  

More from the author

Disclosures (show)

Sign in to read the report!

We have detected you are an active member!

Ray: 092064-dd9f98-b8bba0-d44388-032b69

Want to receive Regular Market Updates to your Inbox?

I am your default error :)

Events

Trending tickers in our research