1 April 2020 (promise: no April Fools’ jokes today)
Epidemiological modeling is hard
A friend sent me this excellent FiveThirtyEight post, Why It’s So Freaking Hard To Make A Good COVID-19 Model. It’s an accessible read that gives real insight into why different models give such different results, and why even the best epidemiologists are constantly changing their forecasts. I wrote related posts, here and here, about why simple approaches to modelling have been the best approach so far (but have limited future use).
How healthy people die from COVID-19
From the excellent blog Pandemic Pondering: Why totally healthy people die from coronavirus:
[Y]oung people who die from COVID-19 infections have a very different illness. They seem to get acutely ill and develop a systemic attack on many of their organs. While these patients can also die from respiratory failure, there is no predisposing illness. They often have failure of the other organs. This is been ascribed to an acute inflammatory response, now being called “cytokine storm.”
[…]
People who die from acute coronavirus infection seem to have an exaggerated innate response and a delayed specific response, so that they fail to clear the virus. The body continues to flail at COVID-19 with inflammation in an attempt to wake up the immune system. However, this causes organ damage and can lead to death.
Should we be worried about oil prices?
A few days ago, a friend shared Goldman Sachs’ report on oil prices, which made my head spin. I had no idea that oil was so complicated, or the predicted near-to-medium-term future so counterintuitive.
I can’t share the report, but here is an article summarising the key points; and here is blog making similar points. Here are my highlights:
- Demand for oil has cratered, which obviously drives prices down.
- What is driving prices down even further is the lack of storage. Many oil wells are expensive to shut down and restart, so they keep pumping despite the lack of demand. But the world will quickly run out of storage given continued (even if reduced) supply and much lower demand.
- We are already hitting these storage constraints.
- The impact is different on different kind of oil producers. “Waterborn” oil producers (e.g., oil platforms in the North Sea) having direct access to tankers, which are an ample source of storage; while landlocked producers (e.g., the Permian basin in the US, Canada) face extremely limited storage.
- As a result, in some locations the price of oil could go negative as producers without access to storage pay others to remove it rather than shutting down their wells.
- So in the near term oil prices could be even more impacted.
- As if that wasn’t complicated enough, the report further predicts that oil prices will go up very significantly post-crisis–and that we will even face potential oil shortages.
- The reason is that the economic shock will take some of the world’s production capacity permanently off-line; and some of that capacity, once offline, will be uneconomical to bring back online. We believe the upstream sector could lose as much as 5.0 million b/d of oil supply capacity.”
- Goldman’s conclusion is that we could in fact face an oil shortage as a result, in 2021 or beyond, which could impact the economic recovery:
“This will likely be a game changer for the industry,” the bank said.
“Big Oils will consolidate the best assets in the industry and will shed the worst … when the industry emerges from this downturn, there will be fewer companies of higher asset quality, but the capital constraints will remain.”
“Paradoxically, this will ultimately create an inflationary oil supply shock of historic proportions because so much oil production will be forced to be shut in,” it added.
Potential good news on serological tests?
A friend shared a Citi research report, published today (1 April 2020), with a positive outlook for the speed with which widespread serological (antibody) testing will be available in the US. I can’t share the report itself, but this article summarises the key points. Here are a few takeaways:
From the article:
“Citi’s global health-care, strategy and economics teams say governments and health-care providers will be able to supply 60% of U.S. individuals of working age with antibody tests by the end of April, and 95% by the end of May.”
“Individuals with elevated antibody levels will then be able to return to the workforce with minimal risk of reinfection or transmission, they say. How many? Such tests could enable between 20,000 and 400,000 of sidelined U.S. workers with previous exposure to COVID-19 to cease lockdown and immediately and safely return to work. Soon after, 90 million workers, representing 60% of the U.S. workforce, could return.”
Additional key points from the report (my summary of what Citi says, not my point of view):
- This would be a significant positive for equity markets.
- There could be 160K and 3.2M undiagnosed cases of COVID-19 in the US.
- “Antibody based testing for COVID-19 is inexpensive, rapidly scalable: [A]ntibody based SARS-CoV2 testing requires small volume blood sampling either by a pin prick or phlebotomy. It has zero operator error (such as poor sampling through incompetent nasopharyngeal swabbing required for PCR). The currently available point of care tests indicate after 15 minutes whether a patient has detectable antibodies to SARS-CoV2.”
- If so, the healthcare sector could be significantly undervalued at the moment.
Citi envisions four main benefits:
- Lower COVID-19 and non-COVID-19 related fatalities by allowing sidelined
healthcare professionals and key personnel to return to work. - Restart the economy.
- Shift back to contact tracing using PCR testing.
- Longitudinal sampling to determine durability of resistance.
Subjective Probability 
One thought on “How healthy people die, negative oil prices, serological tests, and more”