Coronavirus Tracker Update 4/11/20
|
|
Firstly, I hope everyone has a joyous and happy Easter and is able to celebrate with family this weekend.
- GOV ABBOTT TO ANNOUNCE REOPENING PLAN TUESDAY. Next week I'll have a full analysis of whatever is released. My understanding is they are targeting the first week of May.
- CDC RELEASED NEW RESEARCH ON PANDEMIC RESPONSE. The CDC's view of social distancing? "..our systematic reviews suggested that social distancing measures could be effective interventions to reduce transmission and mitigate the impact of an influenza pandemic. However, the evidence base for these measures was derived largely from observational studies and simulation studies; thus, the overall quality of evidence is relatively low.” Simulation studies means more models, so the justification of the current models is primarily just more models. It's models all the way down. How's the record of epidemiological modeling? It's not very good, as I describe below.
- TEXAS CASES STAY LOW. One of the greatest challenges of this crisis has been the lack of available data. It is astonishing to me that there is not some kind of daily dashboard of forecast vs. actual hospitalizations, since that's the resource we are supposed to be managing. Below are the best numbers I've been able to assemble. Texas has 1,491 coronavirus-related admissions as of April 9 and the revised IHME forecast projects 2,259. The state has 6,199 ICU beds and 58,000 hospital beds. The IHME forecast also no longer projects Texas to have a shortfall of ICU beds or ventilators.
|
|
|
- HOUSTON CASES ALSO LOW. The data availability for the Houston area is similarly poor. Harris County does have a dashboard, but it lacks any information about hospitalizations, which again is the resource we are supposed to be measuring. The Houston CBSA has about 3,100 cases - which should yield 90-150 hospitalizations. The metro area has about 1,200 ICU beds and 13,000 hospital beds. I am unable to compare it against the forecast used to justify SAH order because neither the UT School of Public Health nor Judge Hidalgo's office will release it.
|
|
|
- EMPLOYMENT. In February, there were about 135,000 people unemployed in the Houston MSA. That gave us an unemployment rate of almost 4%. In March, another 116,000 filed - moving the rate to 7%, with 250,000 unemployed. Patrick Jankowski at GHP said Thursday he thinks 250,000 new filings as a result of the shutdowns is not out of the question, and I think he's right. Given the difficulties people report with filing and the antiquated computer system Texas uses, it's entirely possible that we are already there. That would be about 12% unemployment. That's getting close to 1980s-level unemployment. But the recovery should be relatively fast, because not that many businesses have had close. I compared all 290 industries in the Houston CBSA with the SAH order, and it turns out 82% of business are essential, and that covers almost 75% of employees. Of the non-essentials, about 1/3 work in education, so they are likely to return to work once the illness has passed. The other important thing this suggests is that most of the layoffs are from reductions in activity, rather than because businesses have had to close, making it much more likely for employees to be rehired quickly. Our data analysis also showed that essential employees averaged about $67,000 per year compared to non-essential employees at $41,000 - making the impact more likely to hit the rental markets harder than the ownership markets.
|
|
|
- MOST TEXAS JOB LOSSES RESTAURANTS/MEDICAL/RETAIL. The Texas Workforce Commission released unemployment claims for the last four weeks by industry. Looking specifically at the last two weeks, 280,000 Texans have filed for unemployment; 30% were in restaurants/hotels, 12% in health care and almost 10% in retail. Although the medical losses are troubling, most of those workers should return to work once regular procedures are taking place. That leaves about 40% in hospitality and retail - again reflecting the impact of Stay at Home orders and suggesting that many of those will probably be able to go back to work quickly once the orders have passed.
|
|
|
- CANCELLATIONS SKYROCKET. It wasn't all going to be good news. A week ago, we didn't have this news - brokers can take up to 5 days to report cancellations in the MLS. Beginning the week after the SAH order, cancellations almost tripled. The decline in sales for April and May will likely be very significant; but even after the cancellations in late March and early April, there are still 2,968 new homes are pending, and over half of them went under contract in March. I cover my forecast numbers for April and May in the webinar. Source: Houston Association of Realtors MLS.
|
|
|
- Sales appear to have remained steady. A quick review of new home sales YTD for 2020 vs 2019 show 2020 still ahead of last year, and ahead in every week of the year except week 13 - which includes March 24 when the SAH order was issued. But it would appear that sales in this week were merely pushed forward to the next week. (The graph trails off at the end because the week is not complete, with both years having almost identical sales for the incomplete week). Source: Houston Association of Realtors MLS.
|
|
|
- Good news/bad news on the national fatalities forecast. The bad news is there were over 10,000 additional fatalities in the US in the last week. Over half of all US fatalities are the NY/NJ area. The good news is the forecast has fallen from about 150,000 expected to 65,000 expected. This is not, contrary to media reports, because social distancing is working, but because the model received an influx of new data more accurately reflecting infection rates, hospitalizations and fatality rates: "As we obtain more data and more precise data, the forecasts we at IHME created have become more accurate," Murray said. The model assumed current social distancing before and after the revision. What's more US probably overstates deaths relative to other countries because in the U.S. you are counted as a victim of the pandemic if you die while testing positive for the virus, even if something else causes your death.
|
|
|
Although it would appear the forecasts for New York were not very accurate, either, according to the State of New York. The chart below shows the various model projections for New York vs actuals, which are shown in the purple dashed line at the bottom.
|
|
|
- Epidemiological models are not very good. Apparently the poor predictive value of these models is quite common. The London Sunday Times notes that inquiries by the British Government into models used epidemics and 2001 and 2009 found they were invalid, led to expensive and ineffective policy solutions, and government leaders treated modelers like "court astrologers." Why are they so bad? This 2018 study found only a third of models were verified against actual data and this study in the American Journal of Epidemiology from 2009 showed none of the top 6 models predicted hospital usage due to respiratory viruses "provided accurate or consistent estimates for both viruses and outcomes." It has a more prescient warning for us now: "Sophistication of statistical methods may have been interpreted as assurance that results are more accurate. Without validation against epidemiologic data, with viral etiology confirmed in individual patients, the accuracy of statistical methods in ecologic studies is simply not known. Until these methods are validated, their methodological limitations should be made explicit and proxy estimates used cautiously in guiding public policy."
|
|
|
Again, still doing webinars. The webinar covers the national and local economic situation and Location Strategy's current forecast for Houston, and I find I've been running about two weeks ahead of most media reports and government actions.
Stay well.
|
|
|
HAVE A QUESTION OR TOPIC YOU'D LIKE CONSIDERED? CLICK THE BUTTON BELOW TO SEND IT TO ME.
|
|
Sincerely,
Scott Davis
Location Strategy, LLC
832.304.3478
|
|
|
|
|
|
|
