The narrative around COVID-19 and the US economy has become a bit messy in the past few weeks. Since late June, daily cases were surging throughout the US and prompting many states to roll back some of easing measures. And this in turn causes high-frequency economic activity to slow. And in turn, this causes investors to rotate heavily into NASDAQ/tech/work from home vs epicenter/cyclicals.
There were signs early last week that daily COVID-19 case trends were peaking, but as we commented last week, this weekend was critical as 7/24/2020 (Friday) was 7D after the “official new high.” And if US cases were peaking, it was critical for new COVID-19 cases not to make a new high this weekend.
That is good news.
As shown below, daily cases over the past few days continued to edge lower. On Sunday evening, daily cases fell below 60,000 and are 4,000 lower than 7D ago. So even adjusting for potential seasonality, the progress in daily cases is indeed trending lower.
Source: COVID-19 Tracking Project
As further evidence daily cases have peaked, this is now the 7th consecutive day that the daily cases delta (vs 7D ago) is flat/down. As shown on the chart below, this is a break in trend compared to much of June and July when daily cases delta (vs 7D ago) was surging.
Source: COVID-19 Tracking Project
Even Dr. Scott Gottlieb is also noting that new epicenter states, FL, CA, AZ, TX, or F-CAT, likely peaked. His comment below suggests that we are seeing a peak in these states. The rolling over in these states has been in place for >7 days now. But as we wrote in several recent pieces, only if there is a sustained decline will any healthcare experts be willing to deem a peak is in place. Thanks @apegator for the shared tweet.
https://twitter.com/ScottGottliebMD/status/1287344075662856197
Even hospitalizations now look to be contracting. See both new gross admissions (newly admitted) (left chart, 33 states) and “net admissions” (net of discharge) are now down noticeably. So, from many facets, it looks the surge in COVID-19 is increasingly looking behind us.
STRATEGY: Non-NASDAQ 100 outperformed +420bp last week as cases peaked…
Last week was a rough week for markets. Perhaps lost in the turmoil is a potential rotation underway. Take a look at the chart below. We divided the S&P 500 market cap into two buckets:
– S&P 500 components which are in the NASDAQ 100
– S&P 500 components not in NASDAQ 100
Source: Fundstrat
As daily cases peaked, the non-NASDAQ constituents are starting to outperform ala late March…
Daily cases peaked last week (and did not make a new high this past weekend) and a notable rotation took place:
– NASDAQ 100 stocks (also in the S&P 500) fell 4.1% last week (since 7/20)
– S&P 500 stocks NOT in NASDAQ 100 rose 0.1% in the same timeframe
In other words, the NASDAQ surge from June 8th to July 20th reflects markets becoming uneasy with the surge in new cases. But if daily cases have peaked (which we believe to be the case), then investors will again see reduced risks of a re-closing and as such, cyclical stocks, aka epicenter groups, should outperform.
High yield bonds rallied strongly in the past week, thus, while equity markets wobbled, HY was rock solid…
Last week, HY bonds posted a strong performance, with spreads rally 48bp (pretty significant). And this improvement is certainly notable considering the extended rally in equities and the rising doubts about the ability of equities to stage further gains.
– however, if HY bonds remain strong = supportive of epicenter stocks/Cyclical stocks
– we view HY as a virtual leading indicator, and its relative resilience bodes well for further equity gains
Take a look at the commentary from JPMorgan’s Credit Research team. Notably, HY bonds staged a strong rally in the past week and the improvement was across all credit tiers (BBB, higher quality and CCC, lower end).
Source: JPMorgan Credit Strategy
POINT 1: Good weekend –> no “new highs” and Sunday daily cases fall to 59,447
As we mentioned last week, Friday’s update 7/24/2020 would be critical as 7/17/2020 was the prior high for new cases. And as shown below, daily cases came in at 59,447 and are meaningfully lower than 7D ago.
– data on weekends tend to be slow/lagged, so to better judge the progress, it is helpful to look at 7D change
– this smooths out the potential weekend seasonality
Source: COVID-19 Tracking Project
6 states reported meaningful 3D increase in cases
Louisiana 3,840 vs 2,084 (Fri) +1,756
Tennessee 3,140 vs 2,091 +1,049
Michigan 1,041 vs 594 +447
Washington 1,025 vs 762 +263
Alaska 186 vs 57 +129
Minnesota 862 vs 767 +95
Total 6 states +3,739
6 states reported a meaningful 3D decrease in cases
Florida 9,344 vs 12,444 (Fri) -3,100
Texas 5,810 vs 8,701 -2,891
Georgia 2,765 vs 4,813 -2,048
California 8,259 vs 9,718 -1,459
Arizona 1,973 vs 3,357 -1,384
South Carolina 1,191 vs 2,001 -810
Total 6 states -11,692
Source: COVID-19 Tracking Project
POINT 2: F-CAT passing peak…AZ + FL set to surpass NY/NJ for case prevalence…
The good news continues with F-CAT, with these 4 states looking to be past their peak (see Gottlieb’s statement above). In fact, the daily new case trends for the F-CAT states below show this remains pretty promising.
– FL as well
– CA has a tenuous peak…
– AZ looks well past the peak
– TX is improving
Source: COVID-19 Tracking Project
Source: COVID-19 Tracking Project
AZ and FL have surpassed NY/NJ in case prevalence…
What is surprising is that AZ/FL have surpassed NY/NJ in total case prevalence (cases per 1mm residents).
– Similarly, TX is now converging on CT levels
– if one believes in the potential for a breakpoint in the infection/herd immunity, these prevalence levels could be important
Source: COVID-19 Tracking Project
The 4 nucleus cities of F-CAT continue to see a decline in daily new cases…
We have written in past commentaries about how 4 cities in each of the F-CAT states is the origin of the largest outbreak. We call these the nucleus cities. As shown below, these nucleus cities are seeing continued decline in new daily cases.
– we view this as affirming that F-CAT is past its peak.
Source: Johns Hopkins
POINT 3: VIMMP: Relapse of VA, IL, MI, MD, PA, or VIMMP, argues against “herd immunity”
As we mentioned at the top of this commentary, the 12 states that were last to re-open are seen as the most cautious and waited until COVID-19 new cases had dropped to levels that were seen as low risk. Below are the new daily cases per 1mm residents these states had at the date of open:
In the intervening months, the COVID-19 path has diverged in these states:
– 5 states have seen cases remain in decline/flat: NY, NJ, MA, RI and CT (outbreak in Fairfield)
– 5 states have seen cases relapse: VA, IL, MI, MD, PA, or VIMMP
– 2 states/regions are in the middle: Delaware and DC
Source: COVID-19 Tracking Project
5 of the states that opened last has seen an absolute crushing decline in cases…
These 5 states – NY, NJ, CT, MA and RI have seen daily cases absolutely collapse since their respective peaks. This is highlighted below.
– in 5 of 5 cases, daily cases are down >90% since the peak.
– CT, however, has seen a slight spike –> teen parties?
Source: COVID-19 Tracking Project
Daily cases in VIMMP, after falling 88% have since quadrupled…
The daily cases, in aggregate, for VIMMP, is shown below. The daily cases peaked in late April and by early June, or less than 5 weeks, fell ~90% to a low in early June. But like nearly every state in the US, somehow, cases had begun to surge. As we have written about in past commentaries, the timing and sequence fit well with the nationwide BLM protests. But regardless, since then, cases have quadrupled.
Source: COVID-19 Tracking Project
Why are there divergent paths? Even as these states waited so long to open?
The obvious question is why some states are faring quite well (NY tristate + MA + RI) and others are seeing a relapse (VIMMP). We are just not sure. But the following issues come to mind regarding states that were late to re-open:
– argues against herd immunity as these states saw a rise in cases, despite high case prevalence and waiting to re-open
– what caused the second surge? it looks like “teen parties” are actually coming up a lot in the media
– daily deaths have not risen, so despite a rise in daily cases, there is no corresponding rise in mortality
VIMMP has lower COVID-19 prevalence compared to NY Tristate…
Looking at cumulative cases per 1mm Pops, we can see that the VIMMP states have a lower overall COVID-19 prevalence. So this raises a natural question:
– are VIMMP states still needed to see further spread?
– if so, are the NY tristate and MA + RI suggestive that case prevalence > 15,000 is when infection rates slow?
Source: COVID-19 Tracking Project
Anecdotal reports of teen parties being a source of new outbreaks…
I don’t know if this is purely anecdotal or if this is a primary source of new spread. But I am hearing from a lot of parents and friends about teen parties and subsequent COVID-19 outbreaks. In fact, here are a few headlines from the NY tri-state area:
– Greenwich, CT party in a Belle Haven basement led to dozens of new cases
– New Canaan, CT had several teen parties that led to new outbreaks
– Teen party in Middleton, NJ leads to a new cluster of cases
This may be purely anecdotal and good media fodder. But this is interesting. We wonder if this is behind the surge in cases in VIMMP.
https://www.ncadvertiser.com/news/article/Coronavirus-parties-in-New-Canaan-cause-15415374.php
https://www.cnn.com/2020/07/24/us/coronavirus-cluster-nj-teen-party/index.html
Despite the 4X rise in daily cases at VIMMP, daily deaths are continuing to decline…
But the daily deaths are diverging from daily cases. Despite a surge in cases at VIMMP, daily deaths have not risen. In fact, as the chart below shows, daily deaths have continued to drift lower.
– If this is “teens” getting COVID-19, the risk of severe outcomes is lower and hence, could explain this lower mortality
– but if younger eventually infect older Americans, deaths could rise.
Source: COVID-19 Tracking Project
