Covid-19 Spread Continues

14 min readMar 25, 2020

It’s been a wild ride over the past few weeks watching how the Covid-19 outbreak is affecting the world. In my previous post, I shared a bunch of data that informs how I’m thinking about this pandemic. Now that we have more information, I thought I’d post an update.

I’ve been rushing 24/7 to learn about this virus / prepare myself, so I feel that some of this info is coming out too late and other parts are not cleaned up enough to share. I’m going to publish and update as I can. Well cited corrections are always appreciated. UPDATE: some parts in retrospect appear overly optimistic or pessimistic, see notes in those sections for relevant info.

Fake news is everywhere

As a reminder, we continue to see fake / bad headlines (sometimes even from reputable sources), and should expect to continue to do so for quite some time. For example, Imperial College COVID-19 Response Team put out a terrifying paper (summary here). Famous statistician Nassim Nicholas Taleb (among others) were quick to post a harsh rebuke:

“While the efforts to model social response are important, leaving out critical aspects of the response yields incorrect answers. Focusing on details but using incorrect assumptions makes for bad policy advice. Where lives are at stake, it is essential for science to adhere to higher standards.”

Given the number of variables (many with powerful feedback loops), limited experience with pandemics like this, and general unknowns, the business of predicting the future is particularly challenging here. In the previous H1N1 pandemic estimates were way off, though that is by no means an argument that they will be this time as well (source):

“A major contributing factor to some of the criticisms of the management of the H1N1 pandemic was that its mortality impact was considerably less than had been predicted. Pre-pandemic global estimates of death were often in the millions while the ultimate death toll has been reported in the thousands.”

The Range of Outcomes

What’s so hard here is that the range of potential outcomes is enormous. It’s so easy to cherry-pick any factoid (positive or negative) that gives you the outcome you’re hoping for. It’s going to be obvious in retrospect, but I want accurate answers now to be able to make better informed decisions today.

The optimistic scenario is still bad: globally less than 50,000 mostly elderly people with comorbidities (cardiovascular disease, diabetes, chronic respiratory disease and cancer) die earlier than they otherwise would have, and hospitals won’t be over-run (so existing patients won’t suffer from resources being shifted away). Here’s the good fortune we would need for that less bad outcome to materialize:

Actual fatality/hospitalization rates are lower than initially reported. This would be most impactful if current infection rates are already far higher than our data shows, due to limited testing. If this were the case we could flatten the curve (and potentially even achieve herd immunity) with far fewer deaths. More on this later.
Spring weather massively reduces the infection rate. More on that later.
Creating/producing/deploying a vaccine takes < 6 months, with enough time to inoculate a meaningful percentage of the population by the Fall. I’m skeptical here because even if we get past the normal hurdles of scientific discovery / government approval / production we also need long-term testing to be sure the vaccine doesn’t cause harm.
Treatments improve. This could mean more ventilators (so we don’t overwhelm our medical system), hospital beds (likely through makeshift or “borrowed” facilities), chloroquine/azithromycin/zinc is found to be effective (and remains available), or some other expertise is developed.
Supply chains function efficiently so we don’t run out of critical supplies like ventilators, respirators, masks, gloves, hand sanitizer, etc.
People generally follow the spirit of shelter in place and other physical distancing orders.
The virus doesn’t mutate, or at least getting infected with 1 strain confers resilience against other strains. A worst-case scenario would be that patients can be re-infected weeks later, though I’m personally skeptical this is possibility.

As scary as this whole thing is, here’s what an asymptomatic Covid-19 patient (the most common case) looks like:

I suspect when this is over we’ll find a ton of people who were afraid of Covid-19 actually had it and didn’t even know.

The pessimistic scenario is real bad. The opposite of the above means that we’re physically distancing for potentially years as this thing works its way through the population and/or we have widespread deployment of a vaccine. The hospitals are over-run and don’t have enough ventilators, so death rates skyrocket. Physical distancing rules get much stricter and social isolation is now strongly enforced. As the economy crumbles, supply chains break down for even some essential goods. Social distancing leads to an increasingly depressed society.

I think the reality is going to be in the middle, but I don’t know how close it will be to either side.

What the data looks like now

This data is taken from the US CDC and seems consistent with the Chinese CDC:

Hospitalization rates and ICU admission rates (to a lesser extent) are surprisingly high for the young, but most importantly death rates remain heavily skewed toward the elderly (source, though original data is here)

Some noteworthy conclusions:

Children seem to have only mild symptoms and rarely need treatment, although they may still be asymptomatic carriers (keep them away from their grandparents!)
ICU admission — which really means breathing assistance — has a relatively small range of 2.0–10.5% (~5x) for different ages.
Fatality has a massive range of 0.1-10.4% (~100x) for different ages.

This means people of all ages will need ventilators, but their success in saving lives will be orders of magnitude better for younger people and guidelines are already starting to recommend this.

You don’t want to be a senior citizen showing up at a hospital that is out of ventilators. A common proposal for hospitals across the country is that “priority is assigned to those most likely to be saved, and most likely to live longer.”

I suspect young (under 40) and healthy people will be mostly fine. There will be some rare/unfortunate cases for that population, but with the risk profile of something like automobile crashes; it’s bad, but we don’t stop using cars (we do wear seatbelts and avoid drinking though).

How many cases are asymptomatic?

Unfortunately, we can’t know! It does seem that Covid-19 infects the young at a higher rate:

The most plausible explanation of this to me is that the % asymptomatic is higher in the younger population, but we just don’t know. It also underscores the value of testing scientifically at random, whereas almost every country just tests the people who show up sick (or don’t bother even then because at that point you can be pretty sure the person is infected, the wait for test results is long, and the answer may not affect treatment anyway).

I have heard people who claim that asymptomatic cases are incredibly rare (~1%), but I haven’t found good data to back that up. I’ll update if I find more info.

China claims to be recovering, doesn’t that prove this can be solved?

It’s encouraging to see that a country with over 100k infections could beat this, so we know it’s possible (although we should be skeptical of their data). That said, their actions have been very different from ours:

Lots of testing. The ratio between our confirmed and suspected cases is so large as to be flying almost completely blind.

As the saying goes “you make what you measure.” In the US, we don’t measure much. Click on the data to see just how low our numbers are. (source)

Ability to add capacity. China built a 1,000 bed hospital in just 10 days, which would be unfathomable in the US. The Army is trying to convert some existing facilities, but as our battle is nationwide (theirs was contained to a specific region) the major limiting factor is still ventilators.
Aggressive isolation policies and adherence. China was able to keep Covid-19 from spreading to every city, but it seems the US has lost this battle. The Chinese are used to authoritarianism and are far more willing to obey for the greater good. See footage of an Australian Chinese woman defying the isolation orders and being arrested (her company also fired her). Here’s intense footage of China training for those who resist (and you can find videos of them pulling people out of their homes). We can’t even get our at-risk boomers to stop risking their own death. We know that 75–80% of transmissions in China were within families, separate facilities for individuals to quarantine them are very effective but forcing their use may be unconstitutional in the US.
Smart (and scary) software to monitor people’s movements and prevent outbreaks. Check out this video of how they use tracking data to limit travel for high-risk individuals.
Cultural experience with fighting a virus, in part thanks to SARS (SARS-CoV-1, the first SARS). They already own and know how to use face-masks for example. Americans head into crowded places to buy toilet paper as if that’s their top problem.

If you want a simplified version: Spring break party-goers in Florida would be unthinkable in China. This tweet was written by Former Ms. Nevada State 2019, who is currently running to be on the Board of Trustees of her local school district:

Knowing nothing else about her candidacy, I’m going to be amazed if parents elect her.

UPDATE: A very important consideration in all of this is that China might just be lying and these numbers could be fake. Why would they be (re)closing movie theatres if there really weren’t any new cases? Maybe out of an abundance of caution against Covid-19 cases brought back to China, or perhaps they’re covering something up.

Spring is Coming

One optimistic scenario is that spring weather will save us. There’s a reason why we call the fall/winter “flu season:”

(source) What’s crazy is that the reason that flu only affects us seasonally is still actively debated!

Some data indicates we may fare the same with Covid-19, but it’s really hard to say:

Here’s another paper with a similar conclusion.

Harvard’s Center for Communicable Disease Dynamics argues that Spring will bring little reprieve.

Even at best Spring is no cure, but it could just buy us time to find a vaccine (and other ways to battle the spread of the disease) until the Fall.

Is the US going to be just like Italy next week?

Italy stats:

Be very skeptical of new cases. Once the virus has gone beyond containment, it’s more indicative of testing than of results. (source)

Deaths starting to flatten out is a big deal, watch this closesly (source)

US stats:

Again, be skeptical here as we generally test 1/10th as much in the US vs Italy (source)

If it’s not obvious why this is so bad, here’s what the lockdown in Hubei looked like (again, far more stringent and scientifically targeted vs what the US is doing):

Notice the long natural delay from lockdown to decline of cases due to Covid-19’s long incubation period, that is an example of a lockdown working well (source)

The US does have some advantages when it comes to smoking rates, fewer people living in a home (adult children in the US generally don’t live with their parents until marriage), and age. Italy has some advantages vs the US when it comes to diabetes.

Italians ICU rate was high (perhaps because of age/smoking) so maybe we won’t be hit as hard, but with the denominator (cases) so hard to measure it’s hard to predict.

US population histogram:

Italy population histogram:

Remember we saw before that age has a huge impact on surviving, but not that huge of an impact on needing an ICU.

So, what’s going to happen in the US? I don’t know, but given the lack of widespread lockdown and the confirmation that Covid-19 has already spread widely, all signs point to the US going a similar route as Italy.

With estimates of who is going to get Covid-19 at ~50% of the US population (the +/- on this is huge), who will that be?

Obviously, if you isolate perfectly you can’t be infected. But unless you’re living off the grid and not interacting with other humans, you risk exposure. I doubt you’re growing/harvesting your own food. For the rest of us, being a bubble boy is no easy way to live.

It’s great to buy in bulk at grocery stores to limit your exposure, but every time you go outside, touch something, or eat something, you add risk.

I still take tons of precautions, but I’m under no illusion that I’m in total control. It’s a frustrating place to be in because you’re constantly assessing the cost-benefit of each decision. Should I eat vegetables that aren’t cooked? How often do I need to disinfect my apartment? I’m going to be so sick of cleaning when this whole thing is over.

My estimation is that the places hardest hit will be the densest, as they naturally have the most social interaction and even a few cases can spread like wildfire. They also have higher real estate costs, so people live in smaller apartments and their lives are designed around going out regularly to meet their needs (fitness at gyms, food at restaurants, social lives at bars, etc). They are starting social distancing from a much worse position.

What effect will this have on the market?

The data from above tells me it’s going to get much worse before it gets better. I believe that when investors realize how bad things are, they will panic sell nearly every asset for cash.

The good news is that once we lock down the whole country (which feels inevitable) we’ve seen from other countries that we can recover — although we’ll have to move up the learning curve fast on testing, containment, social practices, etc. We’re starting from behind. We don’t even have hand sanitizer.

Gold has done far better than the S&P500 over the past month, but worse than holding cash. When the government is printing trillions of dollars it would seem like a good time to hold the yellow rock, but right now we’re seeing interest rates go down. My personal bet is that gold is great when you think inflation will increase in the future, but bad in times where you think th eopposite. I’m holding off on buying gold in the short term, but I think it will be an excellent buy once it seems interest rates seem to be bottoming out.

Meanwhile, the real economy is in a tailspin:

(source). The NY Times has some similarly jarring data as well.

If you look around, you won’t see people at restaurants, hotels, on flights/cruises, at concerts, in ubers, etc. This virus is almost surgically designed to cripple an economy, and it takes a toll on the human psyche as well.

I don’t know if this AirBnB video is real or not, but anecdotally I hear stories of AirBnB hosts that bought a bunch of property and are now underwater. This palpable anger/frustration is only going to get worse

While politicians might be taking aggressive action to limit the damage, the reality is that productivity is going to decline substantially in Q1, and if the situation doesn’t improve quickly productivity is going to be decimated in Q2. The economy is a like a giant container ship, it’s not so easy to nimbly steer. With interest rates near zero the fed doesn’t have as many tools left in their toolbox. We could still see negative interest rates, the fed buying US stocks, and dumping their gold reserves.

My prediction is that because the spread of this virus is so underestimated, the effect it will have on the market is equally misunderstood. People are acting as if this drop means that everything is priced in and things will soon return to business as usual. I don’t buy it for a second. While guns may be selling like hot-cakes, actual panic hasn’t set in yet. That happens when people on social media post videos of them fighting for scarce ventilators to save their grandparents. When that happens, everyone will realize that it’s too late and wonder how we got here.

As one article put it:

In terms of the stock market, we don’t think stock prices are currently reflective of the possibility of a daily death toll of over 1,000 and a national lockdown for a period of a couple of months.

Is this going to hurt Trump?

It all depends on how the recovery goes. In the bad scenario, if your sibling loses their job, your parent/grandparent dies, and your portfolio is down ~50%, there isn’t going to be anything Trump can say to save himself.

Here’s a preview of what Biden’s future attack ad on Trump will look like:

However, if this thing blows over in the next 2 months and the stock market recovers, he can claim that he took strong action. Cash payments to Americans (and trillions of dollars injected into the economy) will buy a lot of happy voters. It’s good politics for Trump to move hard to the left for re-election; very few Republicans are going to vote for a Democrat because their candidate is too liberal, while independents might not bother showing up for Trump’s opponent if Trump is already implementing liberal policies.

It’s worth noting that for an elected official these pandemics represent a lose-lose scenario. Act early enough to prevent the spread of the disease and you’ll be seen as crippling the economy for no reason (even more-so if it works), act too late and you’ll be criticized for allowing it to spread. Assuming a politician is a completely heartless rational actor, then the optimal election game theory is to wait too long until the problem is apparent, then take very aggressive action and claim that you were “tough” on the issue and handled something “unprecedented” (data from other countries not withstanding). That sounds roughly like what Trump is doing, though he seems to have done a terrible job on execution when it comes to things like testing, ventilators, adding ICU beds and masks/sanitizer.

So far his approval numbers are good.

That said, I still don’t believe panic has set in yet, and that’s when the actual pain begins.

How does this compare to other horrific events?

Covid-19 has taken the whole world’s attention, is it too much? The short answer is that we don’t know how bad it’s going to be yet, but the uncertainty and newness certainly make it scarier.

For reference, here’s how many people die in other ways:

The flu kills 25–50k people in the US (290k-650k people globally) each year. Like Covid-19, these deaths are heavily weighted towards the sick and elderly, but unlike Covid-19 it also kills infants and children.
Car crashes kill ~35k people in the US (~1.25MM people globally) per year.
The attacks on 9/11 killed ~3k people, and as a society we obviously reacted very strongly to that.
Heart disease kills ~650k people in the US per yr, and cancer kills ~600k people in the US per year

Here’s a morbid infographic to end on if you prefer: