COVID-19 UPDATE: Good^3. Daily tests rose 43% (vs 7D ago) yet cases basically flat = good. Positivity rate collapsed to 3.4% = good. Colleges 9% cases vs 15% in early Sept.= good

Click HERE to access the FSInsight COVID-19 Daily Chartbook.

STRATEGY: Is the labor market healthier than we realize?
The 2020 Presidential Debates start Tuesday evening (first of three) and given how close this election looks, these debates will be very important.  And of course, on the top of the mind for voters are the same topics that matter in any year — personal economic situation (jobs) and personal safety (COVID-19).  If we were to look at the stubbornly high US initial jobless claims (~1mm per week), one would get the sense the labor market recovery has stalled and correspondingly, so has the US economy.

Goldman Sachs Economists, led by Jan Hatzius, published an interesting commentary yesterday.  They note that the high level of initial claims does not necessarily reflect the actual state of the labor market. 

– Rather, they suggest that elevated filing from the PUA (Pandemic Unemployment Assistance) is accounting for the high level of claims.
– And the “loose eligibility” requirements bolstering these figures, along with backlogs, delayed filings and fraud.

This is really interesting and if correct, suggests that the underlying labor situation is actually better than implied by the economic statistics.  This, in turn, has implications for the accuracy and adjustments made to polling data.

Source: Goldman Sachs


Another sign that GS may be onto something is the latest commentary from DeepMacro, the alternative data/economics forecast firm led by Jeffrey Young.  Below is their analysis of new job offers from Feb to Sept 2020.  And he shows this based on Blue (D) vs Red (R) states.

– economic momentum looks very good in both Blue and Red states.
– and overall levels are higher in Red states.

Source: DeepMacro

Even in “purple states” (both D and R), the job offers index is quite healthy.  Anything above 100 is better than it was in February.  So, several states are above Feb 2020 levels.

– Florida is among the worst, and perhaps this also explains why FL Gov Ron DeSantis announced that executive order lifting all pandemic restrictions.

Source: DeepMacro


S&P 500 held the “line in the sand” = good.  Would be good to see VIX <25…
Last week, we talked about how the market was “pricing in the worst is yet to come” and one way we could gauge this is when we saw the S&P 500 give up 62% of the gains since June.  We used this Fibonacci level (61.8%) because this intuitively represents a 2/3 reversal of any progress since June.  And in our view, there have been considerable improvements since June, thus, a retrace to that level would be a good risk/reward.

– as the chart below highlights, it looks like this “line in the sand” held.
– the next key level is 3,363.35 and solid close above that would be a good sign (especially if VIX is <25 too)

The S&P 500 closed at 3,351.60 and got as high as 3,360.74 yesterday.  So, we are near that level of 3,363.35. 

As we noted on Sunday, it seems like the entire market suddenly turned bearish and would remain so until election day.  Because of that pronounced shift in positioning (NDX futures) plus month-end plus $4.5 trillion of cash on sidelines and near record AAII bearishness, we see good risk/reward in stocks.




POINT 1: Daily cases flat despite a 43% rise in tests (vs 7D ago)
Daily new COVID-19 cases Monday came in at 32,012, which is up 1,255 vs 7D ago. But as we discuss below, Daily cases for the past 6 days have been essentially flat. The 7D delta turned positive again after being negative for two days (see below).

Source: COVID-19 Tracking Project


Again, the daily change vs 7D ago, in our view, is the leading indicator as it is what influences the 7D moving average.  The daily cases have been flat on 7D delta for the past 6 days. 

Source: COVID-19 Tracking and Fundstrat 



Daily tests jump 43% and highest ever for a Monday…
But the big jump yesterday was in daily tests.  As you can see below, there is some daily seasonality to tests administered (makes sense, given confirmed cases is the same way and is based upon tests).  The daily tests of 938,905 are unusually high for a Monday.  Usually, we see these massive figures later in the week — Friday and Saturday.

– daily tests are 43% higher than last Monday

Source: COVID-19 Tracking and Fundstrat  


Despite 43% more tests, confirmed cases was flat.  As shown below, this implies that the incremental tests only have a confirmed rate of 0.4%.

– The extra 281,964 tests uncovered an additional 1,255 cases
– This 0.4% incremental positivity rate is why total positivity rate collapsed to 3.4%

Source: COVID-19 Tracking and Fundstrat  


3.4% is the lowest positivity rate since the pandemic started…
The positivity rate fell to 3.4%.  This is the lowest ever since the pandemic began.  In other words, the increase in testing in not necessarily uncovering massive new cases.  This is a good thing.

Source: COVID-19 Tracking and Fundstrat




POINT 2: 7 states account for almost the entire rise of US cases in the past week…
Over the past few days, it is becoming clearer that the overall “flat” trend in US cases is masking a mix shift.  This mix shift shows states which contributed to many cases earlier this year are not the ones driving the rise in recent cases:

– NY tristate + MA + RI was responsible for much of the March/April surge
– FL, CA, AZ, TX + 25 states accounted for the June/July surge

This chart below shows the 7D delta of the weekly number of new cases (smooths it out).

Source: Fundstrat


And these states have, for the most part, not seen a resurgence of cases.  Instead, the remaining 19 states (22 w/territories) are the states where cases have been stubborn.  

– But within those remaining states, 7 account for most of the recent increase.  Below is a map showing 7D delta below and we highlight those states.

Source: COVID-19 Tracking Project


These charts below show the actual numbers.  The first chart is the daily cases (3D avg), sorted highest to lowest.  And the chart on the right is the 7D delta of this figure.  A negative value on the second chart is good, as it means that a particular state is seeing cases fall vs 7D ago.

– I highlighted 7 states (see red markings) and these 7 states are seeing sizable increases in 7D delta. 
– These are on an absolute basis

The 7 states with the highest 7D delta in daily cases (3D avg):

– Wisconsin         +354
– Washington       +225
– Illinois                +224
– Michigan            +201
– North Carolina   +190
– Utah                  +167
– New Jersey       +148

Source: COVID-19 Tracking and Fundstrat


There is no real unifying reason to explain these states.  But these are the states which are contributing to the disappointing improvements in COVID-19.  And as the chart below highlights, the middle part of the US is where cases are still hovering near new highs.  

– These charts are mostly available in the chartbook, which we send daily.

Source: COVID-19 Tracking and Fundstrat




POINT 3: Past 2 weeks, US colleges 8.6% of new cases vs 14.9% in early Sept
Our data science team, led by tireless Ken, compiled additional data on colleges, thanks in large part to the statistics produced by the NY Times.  As you know, we have written several comments about colleges and universities, and for the most part, we believe colleges/universities have managed COVID-19 surprisingly well. 

In other words, those colleges which have students returning to campus, have not necessarily seen cases surge at a higher rate among our findings so far.  Foremost, few, if any colleges, are seeing super-spreader like growth in cases. Rather, cases appear, and colleges seem to manage them.  And positivity rates at many colleges are below the corresponding level of their local state, arguing it is safer to be on campus. In other words, it has actually been a decent policy to welcome students back to campus.

Ken and his team took the data from ~1,700 colleges and calculated what share of overall COVID-19 cases are driven by these colleges, looking at several time intervals.   And with these snapshots, calculated how many cases reported by colleges/universities account for the total share of US COVID-19 cases.

– During the first week of Sept, colleges accounted for ~15% of US cases
– But since then, this figure has fallen to ~9%

Source: NY Times and Fundstrat


If colleges were a problem, one would expect college share of cases to be flat or rising, not FALLING
This is what surprises us.  We know many of the new cases are younger people and thus, we expect the prevalence and therefore cases to be higher at colleges.  And if the rate of spread was consistent with the rest of the US, college share of cases should be flat (~15%) or rising.

– instead, across the US, college share of cases is actually falling.

We highlight the US tier of states below to show this share of cases.  And as you can see, it is the same story throughout the US. We think this speaks to how colleges are actually managing the COVID-19.  Instead of seeing an exponential rise in cases, we are seeing a surge, likely due to the moves onto campus.  And then, once past that surge, cases are falling vs the rest of the US.

Source: NY Times and Fundstrat



In case you are wondering, which colleges are contributing to the rise in overall college cases.  The top 20 are shown below.  In total, about 138,000 COVID-19 cases have been seen at colleges since mid-August.  And these 20 are seeing the greatest rise.

– eyeballing this, it certainly looks to be a good overlap between these schools (location) and states with higher cases.

This is actually kind of good news.  If these schools manage to contain cases, we should see the associated states report fewer cases.  The key, obviously, is for colleges to manage cases well.

Source: NY Times and Fundstrat