I participated in a call this evening with a major investment bank, where two well-known economists and a tax policy expert discussed the economic impact on the US and the US fiscal and policy response. As with many such discussions, it was on a not-for-attribution basis.
I won’t summarise the entire call, as much of the discussion covered areas that are well covered elsewhere; instead, I’ll focus on a few interesting takeaways.
The big headlines for me are as follows:
The impact on the economy will be enormous in the near-term, with estimates increasing significantly since those from even one week ago.
The US’s fiscal and policy response is quick, large, and appropriate; it will help blunt the impact.
While the US will take on a lot more debt to pay for the response, it is not at a level that puts the US in danger.
There should be strong growth in 2021 against the weaker comparison year of 2020, and reasonably strong growth in 2022 as well.
The CARES Act
The US$2.3 trillion CARES Act phase three passed last week was the third phase of the COVID-19 crisis relief response
It includes $1.4B of direct fiscal stimulus and $850M in loans and guarantees
The CARES Act dwarfs the 2009 American Recovery and Reinvestment Act ($800B) and the 2008 TARP ($800B).
The largest components are loans and loan guarantees for large business as well as state & local governments; small business loans & grants; one-time checks and unemployment benefits; lowering business taxes; increasing health-related spending.
The expenditures are front-loaded in 2020.
The experts on the call consider the bill both well targeted, and directed to impacted businesses.
While the measures will not change the near-term growth outlook, they should reduce the medium-term impact to business and employees
Despite the large deficits that will result, “we do not believe that It will jeopardize the safe haven and reserve currency status of US assets.”
Q2 will see a sharp decline in GDP: on an annualised basis Q2, will be down more than 30%, even worse than the prior forecast of a 24% decline.
US GDP will decline 6.2% in real terms 2020, vs 2.3% growth in 2019.
Growth will resume in 2021 with a strong “catch-up” element and tailwinds from the fiscal stimulus, with growth of 5.5% in 2021 and 3.5% in 2022.
At the point of peak economic impact, the net impact of COVID-19 is more than 6% of GDP.
By far the largest impact is from Hotels, Food Service, and Car Rentals, which makes up 4.7% of GDP and will decline 75%.
Certain other parts of the economy will decline 90%, but are relatively small components of GDP: Casino Gambling, Sports & Entertainment, and Package Tours & Theme Parks. (Collectively 1.8% of GDP.)
On the other hand, expenditure on Hospitals / Outpatient Care, which is 10.6% of GDP, will increase 15%.
The sharp hit to the economy in February / March / April will begin to reduce over the rest of 2020 and 2021. The following chart shows the projected net cumulative impact, relative to a baseline assumption of 1.75% growth otherwise.
Last updated: 31 March 2020. First published: 31 March 2020
Why immunity matters
Among the most important questions about COVID-19 is whether those who recover from infection gain immunity, and for how long.
Immunity is common. It’s why most people only get chicken pox once in their lives; and why vaccinations against diseases like measles, mumps, rubella, and (back in the day) polio work. We take (or should take) an annual flu vaccine, which gives us significant protection against that year’s strain, which puts pressure on the influenza virus to mutate in order to evade those defenses next year.
There are at least four reasons why it’s critical to know more about acquired COVID-19 immunity.
First, the most plausible way that the epidemic will end is via herd immunity: when enough people in a given population have recovered and gained immunity, the virus has fewer people left to infect, and the effective reproductive number drops below one. Without acquired immunity for a reasonable amount of time, the virus could continue circulating and reinfecting us for an indefinite amount of time.
Secondly, if there is acquired immunity, recovered people can serve safely on the front lines; otherwise, we could run out of health care providers. And recovered people can also return to work safely, allowing us to restart the economy.
Finally, any prospect of a vaccine depends on acquired immunity.
What do we currently know?
The vast majority of experts I’ve read or spoken to believe that we will acquire at least temporary immunity; but we still don’t know for certain.
There are many reasons to expect that we will:
Patients recover from COVID-19 because their immune system rallies against the virus. With many viruses, this means that the immune system creates antibodies that defend against that specific virus; and in many cases, this immunity persists.
Though COVID-19 likely makes recovered patients immune, experts aren’t sure how long protection lasts
One brush with the viruses that cause chickenpox or polio, for instance, is usually enough to protect a person for life. Other microbes, however, leave less of an impression, and researchers still aren’t entirely sure why. This applies to the four coronaviruses known to cause a subset of common cold cases, says Rachel Graham, an epidemiologist and coronavirus expert at the University of North Carolina at Chapel Hill. Immunity against these viruses seems to wane in a matter of months or a couple of years, which is why people get colds so frequently.
Reports have surfaced in recent weeks of people who have tested positive for the virus after apparently recovering from COVID-19, fueling some suspicion that their first exposure wasn’t enough to protect them from a second bout of disease. Most experts don’t think these test results represent reinfections. Rather, the virus may have never left the patients’ bodies, temporarily dipping below detectable levels and allowing symptoms to abate before surging upward again. Tests are also imperfect, and can incorrectly indicate the virus’ presence or absence at different points.
But the mere presence of antibodies doesn’t guarantee protection, Wang says. Reinfections with common cold coronaviruses can still happen in patients who carry antibodies against them. And a bevy of other factors, including a person’s age and genetics, can drastically alter the course of an immune response.
Occasionally, however, mutations will alter a viral strain so substantially that the immune system can no longer recognize it, sparking an outbreak—even in populations that have seen a previous version of the virus before.
So far, SARS-CoV-2 also doesn’t appear to be undergoing any extreme mutations as it sweeps across the globe. That may be because it’s already hit on such a successful strategy, and doesn’t yet need to change its tactic. “Right now, it’s seeing a completely naive population” that’s never been exposed to the virus before, Graham says. The virus “doesn’t seem to be responding to any kind of pressure,” she adds.
Strong coronavirus measures today should only last a few weeks, there shouldn’t be a big peak of infections afterwards, and it can all be done for a reasonable cost to society, saving millions of lives along the way. If we don’t take these measures, tens of millions will be infected, many will die, along with anybody else that requires intensive care, because the healthcare system will have collapsed.
It covers some of the same ground as the early piece, but is well worth reading.
Tomas’s article led me to the impressive, interactive Epidemic Calculator. If you take only one thing away from using it, it’s that outcomes are wildly sensitive to inputs that are difficult to estimate.
This very good FT piece (paywall, but this link in theory should let you read it), The mystery of the true coronavirus death rate, covers several important topics. As the title suggests, it discusses with the Case Fatality Ratios (CFRs) have been so different in different countries, and why we should be cautious about comparing them. (See my discussion here.). It covers the thorny question of many deaths from coronavirus are indeed incremental. And it helpfully reminds us that with testing rates so wildly different by country, we really don’t have a good sense of the true rate of infection, and therefore don’t know the true Infection Fatality Ratio.
An extremely important study came out of Iceland lately, discussed here. It is perhaps the highest rate of testing (per capita) in the world of a country testing a significant portion of the population; and importantly, it screened the general population, not just symptomatic or at-risk individuals.
According to numbers published by the government on Wednesday 25 March, a total of 11 727 individuals have been tested for the virus that causes COVID-19. This translates to 32 217 on a per million basis. This is the highest proportion of tests performed by any individual country.
So they’ve tested around 3% of the population.
Importantly, two very different cohorts were tested: one representing the general population, and one focused on symptomatic or high-risk patients.
The tests have been performed on two different cohorts.
A total of 5 564 tests have been performed by the National University hospital of Iceland in Reykjavík, mostly on individuals who were symptomatic or were considered to be likely to have contracted the virus due to proximity to infected individuals or other reasons. Out of these 5 564 tests, 4 879 have been negative and 685 positive.
A total of 6 163 tests have been performed on the general population, individuals who had not been ordered to quarantine and were generally asymptomatic or showed mild symptoms. Out of these 6 163 tests, 6 111 have been negative and 52 positive.
So 12% of the symptomatic/high-risk patients were positive; but only 0.8% of the general population.
While Iceland is different in so many ways from many other countries, this is at least one data point that argues against the idea that a very high portion of the population might already have been infected, with an extremely low rate becoming ill, critically ill, or dying.
A meta-point to take away from it is that when you see reports about the distribution of confirmed cases by age group, you should first ask yourself whether the sample is random or suffers from extreme selection bias. As one Dutch professor tweeted:
It’s possible that the elderly in the Netherlands are many times more likely to get ill than those in Iceland, and that young Dutch are uniquely immune. But it’s much more likely that this reflects a combination of selection bias in the testing, and the fact that younger people are more likely to have mild cases and not go for testing.
This tool, comparing hospital resources (capacity) to projected critical cases for the US, both at a national and state level, is making the rounds quite broadly.
I haven’t looked at the methodology yet to see what caveats to attach to it, but have seen some sniping from individual states who believe it is inaccurate for their state.
I’ve focused so far on the evolution of confirmed case doubling times (i.e., the number of days it takes for the number of confirmed cases in a given locale to double) as a way of looking for early signs of progress, as do a number of epidemiologists and statisticians. There are some technical reasons to use doubling times: primarily that they are constant-slope at logarithmic scale, as you’ll have seen on the justly famous chart comparing doubling time by country.
However, there are several disadvantages of using doubling times to measure progress. They’re volatile: lumpy data (which is frequently the case in this crisis) makes them jump around in a way that can suggest more near-term progress or deterioration than is likely the case in reality. As presented, they’re coarse-grained: while it’s technically meaningful to talk about (say) a “4.2-day doubling time”, neither I nor most other sources present them in that way as it’s hardly intuitive. And finally, I’ve found that most people find doubling times hard to understand.
Today, I’m going to present a different, but mathematically equivalent, way of thinking about progress. I’ll replace doubling times with the Compound Daily Growth Rate (CDGR). I’d love to get your feedback about whether this is a more intuitive way for you to understand what’s going on.
What does CDGR mean? Anyone working in finance, and in many other businesses, will be familiar with CAGR, or Compound Annual Growth Rate. It’s simply a way of stating the annual growth rate in some figure (revenue, customers, etc.) in a mathematically useful way. (The “compound” part simple means that instead of taking the average of the growth rates over the relevant period, we calculate the growth rate that, if applied from the first period and allowed to compound, like interest, each year, gets us to the actual figure in the last period.)
Since COVID-19 is spreading so quickly, we need to talk about the daily growth rate rather than the annual rate. CDGR is a simple calculation. (Skip the rest of this paragraph if the math doesn’t interest you.) Say we’re interested in the CDGR of the cumulative confirmed case count in France over the last week, 22-29 March 2020. We divide the figure for the 29th by the figure for the 22st, take the seventh root (because it covers seven elapsed days) of that ratio, and subtract 1: (37575/14459)^(1/7)-1=15%. More generally, to calculate the CDGR for a period of d days, with c1 the count at the first period and c2 at the second, the formula is: (c2/c1)^(1/d)-1.
The relationship between doubling time and CDGR looks like this:
Obviously, high daily growth rates are bad. For many, that’s more intuitive than doubling time, where high doubling times are good and low doubling times bad.
One more technical point. For CDGR, we have to choose what period of time to use. The CDGR for a given locale will be different calculated over one day, one week, or one month. Given that the data are lumpy, we want to balance getting the benefit of some smoothing by looking over at least a few days, with wanting to see whether we’re making progress in the near-term. Somewhat arbitrarily, I’m going to use one week unless specified otherwise below.
What do daily growth rates tell us?
So what do the daily growth rates look like? First let’s look at some of the most impacted countries as well as the total for the world. I’ve highlighted US and World:
The good news is that most lines slope downwards. That means that in most countries, the rate of growth is slowing over time, exactly what control measures are meant to do. South Korea obviously cracked this early and has managed to get the growth rate almost to zero, which we all need to do. But Italy, France, Spain, and Germany are all making progress.
The news is still fairly bad. The growth rate for the world as a while is steadily increasing, not decreasing: that means that the pace of the disease’s spread is accelerating. The US’s growth rate, while decreasing recently, is still stubbornly high. And importantly, until these rates get to zero, we’re not out of the woods. A 10% daily growth rate is still very fast exponential growth! At 10% daily growth, cases double roughly every 8 days; triple every 13 days; increase 15-fold in a month; 100-fold in 50 days; and 1,000-fold in 74 days.
Is the picture different at the state level in the US?
A number of friends have also asked about what is happening in the US at the state level. Fortunately, the New York Times has aggregated data at a state level from various sources and made the data available in a GitHub repository. (As always, we should regard the data with a great deal of caution, since they likely suffer from the same limitations as country data: inconsistent standards, data lags, low / changing testing rates, etc.).
Here’s what the data say for the worst-impacted 10 states, which accounted for 79% of all reported cases in the US as of 28 March:
While it’s good news that these lines are generally sloping downwards, only Washington (the earliest state to be significantly hit) has been consistently below 20%; and per the point about compounding several paragraphs ago, sitting around 15% is not a good place to be.
More generally, most US states are still at a significantly earlier stage of dealing with the epidemic than even Italy, France, Germany, Spain, or even the UK. All of these European countries now have daily average growth rates consistently below 20% and consistently falling; by contrast, only Washington has achieved that milestone.
The news in the other 40 states isn’t much better. Yes, they have (or at least are reporting) significantly fewer confirmed cases. But only 11 are at or below 20% a 7-day CDGR, another 31 between 20% and 30%, and 9 between 30% and 40% (this includes the capitol, Washington DC).
I also looked to see if there was a relationship between the number of confirmed cases and the growth rate at a state level:
Interestingly, there isn’t an obvious relationship. That’s different than at the country level, where in general countries that have many cases are showing slowing growth rates: rapid growth leads countries to implement control measures which slows growth. I suspect that the lack of a relationship here says one or both of two things, neither that insightful. Either the confirmed case counts are so low in many states that they haven’t felt it necessary to take drastic action yet; or the underrtesting and underreporting is severe and inconsistent, so we really don’t know how many cases are in each state. Likely both are true.
Last Sunday, I shared hot-off-the-press notes from a two-hour private conference call with 15 experts and >150 major investors. The hosts of that call organized a sequel tonight following the same format, and I’m happy to share those notes now — literally minutes after leaving the call.
Again, we had 15 experts, including another Nobel laureate, a CEO of a UK private hospital, a former senior executive at the FDA, a former CEO of a regional Federal Reserve Bank, a casino owner, a restaurant entrepreneur, as well as several large investors. Again, it was a mostly US-focused conversation; again, it was under the Chatham House Rule, meaning I can share a summary of the content but not the names of the participants; and again, the notes are near-verbatim and real-time (apologies for typos). The time, we had more than 250 participants on the line.
There were a lot of fascinating topics. Some of it covered familiar ground but it was good to have key points reinforced. A few things that were particularly interesting to me:
The chief equity strategist at a major investment bank thinks the market rally last was a false one and is consistent with “bear market rallies” in the past that were false starts. He warns that we should be very cautious based on history. Incredible factoid: “1929-1932, 10 rallies of 10% or more in a long bear market. Median rally was +23% and lasted almost 20 days! But it was secular bear market”.
Emerging Markets will be badly impacted, particularly (a) given that many rely on oil and/or tourism; (b) structural deficits and debt.
What’s happening in China, and how is the economy starting to come back? Very different in different cities and sectors. Labor availability is a major factor.
It was heartbreaking to hear from the founder of a restaurant chain about the impact on his business and his employees — the outlook must be bleak for so many restaurants.
Finally, a Nobel laureate shared why deaths of despair (suicide and drugs) typically decrease in a recession!
Topic: The UK response to the COVID-19 crisis, particularly in relation to the NHS
Bio: CEO of an independent hospital in a suburban area in the UK.
NHS has 1.4M staff, 8th largest employer in the world
All non-essential work switched off
Creating extra ICU capacity
Re-recruiting people who have left NHS — 55K people who have left in last few years, 15K have come back
500K people have volunteered to help look after 1.5M vulnerable adults and children told to stay home for 12 weeks
£1.2B deal where gov’t has taken over all independent hospitals for the next 13 weeks
Expected that capacity will be overrun, especially ICU beds and ventilators.
Also don’t have enough PPE.
Opening 3 new hospitals in 3 cities. London will add 4K beds and open this week.
Topic: What should we expect from the FDA on the virus?
Bio: Former senior executive of the FDA and the NIH.
Two tests: one (antigen) lets you detect who has it now, one (antibody) lets you detect who has it and presumably has resistance. The former has been the focus but the latter will be important to know who can safely go back into the workforce etc. Also lets us get a better picture of what portion of the population has been infected without us knowing.
New, faster tests, including at-home tests, are coming.
But there will be challenges with scaling up manufacturing and with the availability of reagents
In term of therapeutics, we will likely see them before vaccines. Repurposing existing drugs plus new drugs in the pipeline
A few promising ones but no guarantees
Remdesivir: in trials, results within weeks.
Sarilumab: could help because some of the damage comes from the activation of the immune system; this modulates the immune system from cascading out of control
Antibody space is promising, these might be able to neutralise the virus. Some are taking convalescent antibodies from those who were infected to help someone who is currently infected. These will take longer but may have some results by summer.
Many in the pipeline
Many are inactivated and live, attenuated virus
Also nucleic acid, DNA and mRNA vaccine
Moderna is the one that is the farthest along, already in human trials. Need to do the testing, starting with safety testing, before you move into large-scale clinical trials to show protection against exposure.
Trying to telescope the stages to get answers more quickly. But you’ve heard others talk about 18 months and that would be rapid.
Shortages of drugs and medical products such as PPE are an issue
But also medication needed for routine medical needs. The active ingredients for most drugs, 80% come from China and much of the rest from India.
Complex supply chains for much medical equipment
Many companies reported that they had 2-3 months of reserve supply: ingredients, drugs, etc. But as this continues we may start to see significant shortages.
Topic: I got the virus, now what?
Bio: MD and Associate Professor at a major Illinois university hospital Questions:1. I am sick and feel terrible, what do I do?
Don’t assume you have it
Only 13% of people who ask to be tested in our state are positive
Call your doctor
2. What should I do to treat myself?
Early reports on chloroquine are that it’s not very effective.
If you’re getting progressively worse you might need to be hospitalised.
3. When should I get tested, if at all?
Supplies highly restricted. In IL, restricted to people who are symptomatic and at risk.
In IL we only have 1.3% death rate from known cases so the actual risk of death is likely much lower.
But some people are getting quite ill.
Risk is if we run out of ventilators etc. our death rate could go up.
4. Will I infect my family, and how do I prevent it?
Looking at China, harder to infect your family than we thought. Less than 50%.
Guidelines from WHO are straightforward, check them.
5. When do I head to the hospital?
If you’re getting worse.
Trouble breathing or coughing up blood.
In Chicago, Fire Department will take you to a hospital that has an available negative pressure room.
If you had to go you’d want to go to a teaching hospital (but I’m biased)
6. How will I get treated at the hospital?7. Should I go to the hospital, if I am an otherwise healthy 50 year old?
Yes if you are getting worse.
8. How important is hydration, and how do I do that effectively?
Hydration is very important in any illness.
Water + sodium + glucose gets water into the body. The WHO hydration formula or pedialyte work very well. Or half water + half Gatorade.
Colour of urine is an indication of hydration. You want to see pale yellow.
9. How do I deal with my respiratory issues?
If serious go to hospital.
Topic: How will the Fed continue to change its behavior in this crisis? What is the state of the economy?
Bio: Former President and CEO of a regional Federal Reserve Bank
We are already in a very deep recession; hopefully not a long one.
Not a typical recession. Caused because people stay home and businesses need to shut down.
Economy was strong before this.
Got a test of how bad it will be: unemployment insurance rates went up with 3 million filing from 350K in the prior week.
Once under control economic activity will return.
$2 trillion package and Fed actions are appropriate and will help reduce impact of the virus, but cannot fully offset the shutdown.
Q2 GDP will show significant decline. Key unknowns are Q3 and Q4, unknowable today.
Actions by the Fed
Fed funds target to zero
Targeted increases of $700B of balance sheet; but on Monday said they will buy whatever is required with no caps across the curve
Dramatically increased balance sheet by $940B over last two weeks.
Established group of seven special loan programmes. These require approval of Treasurey. Treasury provides equity investment to cover losses at the Fed and permit a 10x leverage. $30B of equity permits $300B of lending. Cover a broad set of areas.
Some are similar to programmes in the GFC. Some are new.
$354B from Treasury to backstop losses; at 10x that’s more than $3.5 trillion of ammunition
Next steps and challenges for the Fed
They need to be flexible, there will be mid-course corrections. Six areas:
1. Fed has used full range of tools forcibly.
2. They will not run out of ammunition.
3. Under a lot of pressure to loan to states and municipalities. Need to watch this as many were already under financial stresses. Need to be careful that they do not morph into programmes that deal with long-term structural deficits
4. Will they move to negative interest rates? Unlikely, they have a strong aversion, but can’t say never
5. How will they pare back the stimulus later? Has had trouble in the past, remember Taper Tantrum. Will be more difficult this time.
6. Deflation/ inflation question. We will have lower rates of inflation but I doubt we will have deflation. But inflation is below Fed 2% target so high inflation unlikely. But the extraordinary monetary stimulus the risk is non-zero.
Topic: Casinos, Real Estate, and Online Education
Bio: Casino owner, real estate entrepreneur
Market is frozen
No one knows who is going to pay rent or how much
Construction costs and land costs will come down, pricing already looking better.
Rents will probably come down.
Third crisis Ive seen in 30 years. Short memories. Expect multiples and cap rates will go back up to the prior levels if interest rates stay low.
Lots of distressed debt out there.
We own regional casinos, they are all shut, we are paying employees as are most private companies. Public companies seem not to be paying employees and seem to have serious issues.
Operating companies who lease their companies, they have debt and significant leases and are likely in trouble.
Regional casinos will fare better than Las Vegas because you don’t have to travel, people still want to have one.
Our online casino is still doing well. Russian table tennis is the only thing out there to bet on and people are doing it!
Topic: How will the COVID19 crisis impact the higher education sector in the U.S.?
Bio: Former business school dean
Near-term impacts all negative: closure of campus, universities will have to refund pro-rated room and board.
In our system that means $80M of refunds.
Tuition discounts? Open question, many are asking
Events are suspended (sports)
Campus visits are cancelled. Students are trying to choose schools, recruitment and admission will have to move online.
Uncertainty about Fall semester and whether things will be back to normal
Institutions’ big risk factors is for those that have a lot of revenue in international revenue, and those that have limited online capability — e.g., if there is a second wave.
There is some money for higher ed in the stimulus bill — $14B when we hoped for $50B. Some will go to students.
Q: Will kids learn well online? A: so far a bit clunky, but I see students putting in the time.
Will be interesting to see how this changes the use of technology longer-term for universities, which has been resisted by the existing bureaucracies / inertia.
There will be a huge bifurcation of the sector into the elite research universities, who will get every strong from the appreciation of science — 100-150 tier 1 universities will come out even stronger. They will mostly do things as they do today.
There are another 1,000 universities that confer 4-year degrees and that market is ripe for disruption. Who will provide credentials that lead to a productive career in the future?
Will be a mix of solutions in the future.
Topic: Structured Credit, Corporate Lending, and CLOs
Bio: Global head of alternative credit at a major hedge fund
Initial dislocation looks typical. Equity markets went down and Vix went up and then corporates draw down the revolvers – -this is normal.
For structured credit, two events usually trigger the selloff:
Mark to market on everything banks are lending against.
Redemptions in funds
This time we saw more redemptions in mutual funds than we would have expected
Quickly the structured credit markets break down.
What’s counterintuitive is what is sold first. Usually it’s the safest assets.
In most CLOs, much of it is AAA rated. We saw sales with quick settlements — 2 days instead of the usual 3 days.
This is big opportunity for firms with capital.
Wide range of gov’t stimulus. Usually gov’t is trying to protect employment or housing. Then we go look at the forgotten sectors —e.g., mortgage REITS have nothing to help them.. We also look at the unintended consequences, like mortgage originators. They need to have a lot of cash on balance sheet. They lock in rates for consumers, and then hold forward securities — and they are gaining in value rapidly. It’s areas like that where we find opportunities.
What’s different in this case?
Far less fraud in the system
We are focused on longer-term unemployment
Two types of investments come out of this:
AAA investments that have real cushion
Convergence trades that you can do with much less leverage, like convertible securities, SPACs trading below treasury values, etc.
Topic: What are you hearing about what is happening in industry and manufacturing in China;which industries will have U, V and L-shaped recoveries?
Bio: Senior Partner at a large consultancy.
Lessons from China where we have a large presence. The measures were like chemotherapy. Killed COVID-19 but did a lot of collateral damage.. Gov’t subsidies focused on agriculture, production and logistics, so SMEs have been badly hit. Risky businesses like gyms and spas are in danger. Cash is king, no one is paying receivables, rents are being renegotiated or going unpaid. Cargo is doing fine; amusement parks not. There will be changes in ownership. Economic restarts based entirely around labor availability. Migrant labor not trusted and not available — impacting manufacturing in bigger cities. COmpexity of supply chain hurting automotive manufacturing. Restaurant workers moved to last-mile delivery and they don’t want to go back to their old jobs. Digital is working all over and is now part of the core strategy. Tech platforms like Alibaba + authoritative gov’t gives transparency on high-risk locations, badging for citizenry. In Beijing, restaurants are open; they take your temperature and open an app that says where you have traveled recently — everyone in the restaurant has gone through that too. Saving rates in China have been high for families for a long time. And the family structure is top-heavy because of one-child policy, so cash handouts to the population has been less important. More coupons being handed out to get people to resume to spending.
U-shape vs V-shape/. Two factors: health response and effectiveness of gov’t economics response.
We think there will be regional resurgence over time.
We will avoid a banking crisis but not get a rapid recovery
GDP down 8% in 2020, not returning until Q2 2023 to pre-crisis levels. This is more significant than anything since WW2.
Public health and economic metrics are both needed.
Aerospace, gas, insurance, travel hard hit
Real estate is getting killed but likely to come back quickly. The question is the future of work — what will change permanently?
Management teams, how are they coping? Companies quickly focused on health & welfare. They are still doing furloughs and layoffs where necessary. Many companies are overconfident, holding on to too much fixed cost, not appreciating how existential this is. There will be significant corporate failures.
Topic: How has the virus crisis impacted emerging market economies?
Bio: Proprietary trader who formerly worked at several major hedge funds in Europe.
Two trajectories in the end, maybe different than in the West
Those that early suppress may see second wave in October
Those that do not have the means to suppress may have a bigger epidemic, especially those in the Southern Hemisphere that won’t benefit from the summer seasonality, But they will not have a second wave and will be done in 3-4 months.
Shocks, two key ones;
Oil shock. Significant for many countries with >20% GDP from oil — Iraq, Oman, Qatar, Saudi; >10%: Russia, Angloa. All impacted significantly. Those with large deficits have more to worry about it. And those with high debt to GDP have high probability of default. Russia will see public debt go up to 30%. Oil is a positive transfer in terms of savings to importers, good for Asia — they are getting a big subsidy, in developed Asia. This drop in oil will be more vicious than 2008 drops. Bigger capacity destruction that will not come back in 2021. Risk of much higher oil prices in 2021 when global economy sees strong recovery. Turkey and India will be very impacted, especially Turkey which has already used most of their reserves and are starting to run negative reserves.
Travel & tourism. Most impacted: Turkey, Hungary, Greece, Mexico. 5-8% of GDP, 7-10% of employment in tourism.
On policy response, EMs have cut rates, sometime aggressively. Inflation is surpassed now because of oil, but will go up in EM driven by food and depreciating currencies. Hard for them to stabilise given context of risk of default. Dollar debt is high vs 10 years ago. The outlook for EM currencies is very challenging in near-term.
Good news for EM: IMF is going to double the SDR allocations. This will give some short-term support on the most impacted poor countries. This money probably won’t prevent defaults in sub-Saharan Africa in impacted oil sovereigns.
Topic: Will explain that investors are getting drawn into a bear market rally!
Bio: Chief Equity Strategist of a major investment bank
Do not get sucked in to what happened last week.
Happened 7 times between Oct and Dec 2008 but market did not trough until March 2009.
Highest monthly vol in history, March 2020.
Annualized vol = 98 was above 1987, above Oct 1929, above Oct 2008.
Best 3-day rally since 1982. Still 25% below the high.
Context; 2008, there were 6 distinct rallies between 1 and 6 days, up 10% or more. Some almost 20% rallies.
1929-1932, 10 rallies of 10% or more in a long bear market. Median rally was +23% and lasted almost 20 days! But it was secular bear market.
There are no buybacks now for political and cash reasons. That has been the #1 source of demand for 10 years
News flow will continue to be bad
No corporate earnings for a month
Bullish for some reasons: fiscal cash injection, Fed actions are all good.
Tactically, you need three things to end a bear market:
1. Slowdown of viral spread — need to see this in the rate of new cases. Portfolio managers think it will happen 3-4 weeks, some since 6-18 months. Need to see resolution of that debate.
2. Need to see evidence that the fiscal policy is working. We are already seeing that monetary policy is working.
3. Positioning / flows: absent improvement in fundamentals, the positioning is overly bullish. People were super-long in Feb, modestly shorter over March, actually got longer last week. That’s inconsistent with prior bear markets. Need to get lower than today.
Topic: How are restaurants and hospitality going to survive the crisis?
Bio: Co-Founder and owner of medium-sized regional restaurant chain
In 3-4 weeks business entirely disrupted
Serve 600-1,000 people day, have 60-100 employees per location
Employee costs are 40%, rent is 15%
Financial reserves were almost wiped out.
We were able to reduce our food buying, but not other fixed costs.
Since the gov’t closed all restaurants on 24 hour notice, we need a lifeline as an industry.
There is little in the stimulus plan that is applicable to us.
For example, we should have all rents and mortgages suspended for six months, tax payments frozen, utility payments frozen.
And insurance companies should pay business interruption claims.
We had to let go 1,000 employees. I don’t know how many we will be able to hire back.
China has reopened restaurants but the images are not encouraging. They have to cut capacity to 1/3 to leave enough empty space.
Worried that many restaurants will not survive.
We are taking this time to try to streamline everything, target food costs at 20% after the crisis
Most challenging point will be how to reach agreement with our landlords into %-only leases, capped at 9%, with a low minimum monthly rent. We don’t know when economy will restart.
We are guessing we will have revenues at 40% of what they used to be, for at least two years.
Topic: What are the details of the Stimulus Bill that impacts small businesses and employment?
Bio: Partner with major law firm, focused on employment law litigation
There are some bills passed that will help, still uncertainty about how they will work
Essence is: paid leave benefits with tax credits for providing them, tax credits for retention, delay in payroll tax payments, forgivable small business loans based on paying employees / mortgages / rents, unemployment protection for employees, retirement fund access for individuals suffering from COVID-19
Paid leave benefits is effective April 1st, employers with < 500 employees. Just US employees, will not count related companies unless extensive overlapping control/ management. Leave for 5 primary purposes: Quarantine order, instructed to self-quarantine, symptoms of COVID-19, caring for impacted individual, caring for impacted child or if school is closed. Catch-all TBD
Amount of benefit is 100% for those under quarantine or ill; up to a cap of $511 for two weeks / 80 hours. For those taking care of children can go up to 12 weeks at 2/3 of regular pay capped at $200/day. For those taking care of an ill relative, at 2/3 but only for two weeks, same cap.
Employers will be able to claim a refundable credit on tax payments if you provide this.
Payroll tax credits for those closed due to gov’t order: refundable tax credit of up to $5000, 50% of first $10K provided to employees for the rest of the year. Also available if gross receipts decline 50% or more.
You can borrow money to pay 8 weeks of payroll and if you keep your team you don’t have to pay it back.
Topic: How has the current crisis impacted infrastructure investment around the world?
Bio: Invests in infrastructure at a major engineering company.
Substantial delays in major capital projects, and slowdowns / stops to existing projects.
1. Energy sector
Major oil & gas produces announcing delays and cancellations.
Hard to separate the effect of the pandemic from the decline in prices; the two are compounding.
Two big companies cut capex by 20% – by $10B all together.
Major multi-$B projects have been shut down for health reasons.
While many have not yet been shut down we expect more of this.
Very complex as there are local, state, and federal interactions. Everyone has health & Safety issues to deal with.
In LNG, many delays and cancellations.
I expect to see many more cuts still to come
Smaller companies have a bleak near-term outlook
Non-utilities: commercial banks are quite open to doing project finance when revenues are fully contracted. But getting those contracts right now is hard.
In US: two solar power projects closed in the last two weeks.
Gas fired power plant, project finance was pulled from the market.
A non-regulated market like Texas; spot prices down 20-30%, forward curve is down a few percent out past 1 year. Investors are looking through the trough which is positive.
Secular shift to renewables — this could accelerate.
Think there will be pension-fund type investors who will shift money out of fossil fules
Aviation has obviously cancelled many projects
Big distinction between private and public projects. Private airports in Europe have big cash flow issues.
Privately owned toll roads will have liquidity challenges
States and municipalities are seeing big cash declines in transport user fees which will impact state-level finances.
Most transportation infrastructure is owned and financed at the state level.
US non-residential construction. Declined 2009-2011, dropped 30% peak to trough.
Large infrastructure spending bill? Many are optimistic. IF there is legislation, not sure how quickly the money can be put to work; at best the impact will be delayed.
Topic: What Moral Panics Lie Before Us?
Interested in the moral aspects of the crisis.
Real danger of moral panic: when there is a phenomenon and a population starts to worry. E.g., Salem witch trials.
Feature of human psychology is the we are moralistic. When we see something that we regard as a moral wrong, we want to see people punished. E.g, McCarthyism, myths of razor blades in apples at Halloween.
In these cases, lives and families are destroyed.
Moral panics magnify these.
Worry about the intuition that people should be punished for the real or imagined offences.
Two weeks ago, going out to dinner was normal. Then it became dangerous. Now some people are saying the people who go out to dinner are akin to murderers.
Predict that there will be many ways in which this kind of thing manifests itself.
Seeing hoarding behaviour, fear. Even more unpleasant consequences will come.
First place this could come is in masks. Right now if you wear a mask: if you go out with the mask and you have a disease, that’s bad since you could spread it; and if you don’t have the disease you shouldn’t wear masks. In the next few weeks this will switch to be criticising people who don’t wear masks — social sanctions. [CCN: this is already happening in Hong Kong. See new Economist article.]
We saw this with the pillorying of the spring breaker.
Topic: Deaths of despair and deaths from the virus
Bio: Nobel Prize Winner, professor
Some are arguing that the lockdown will be a disaster and that there will be 100s of thousands of deaths of despair.
That argument is wrong.
Yes, people with less money die younger than those with more.
Deaths of despair, suicides and drug overdoses, happen to people who do not have a 4-year degree.
But it is not true the then people get poorer over the business cycle the they become more likely to die.
The reason is the deaths of despair take a long time to happen. Not like the business cycle.
Of course if it destroys the economy long-term, that’s different.
But this is likely to be more like past recessions in this sense.
We have very consistent evidence that mortality rates go UP in boom times and go DOWN in recessions.
So mortality may well fall.
Firs observed in the US in 1920. Seen over and over again in many countries and situations.
Great Depression: mortality went down, and life expectancy peaked.
Greece during GFC: unemployment tripled, 25% unemployed, but Greece and Spain saw significant increases in life expectancy that were among the best in Europe.
What happens is that there are a lot of injuries at work and driving; so this goes down.
People have more time to look after other people.
More minimum wages workers available to work in nursing homes and to look after the elderly.
The one thing that goes in the other direction is that suicide rates do go up.
But suicides, while each a suicide, were only 2% of all deaths last year. So they
Accidents, stress, more free time, more time for exercise, more time to look after other people — these are all larger contributors.
Could be that this recession, with social isolation, could make suicides worse. There is a correlation. Also many attempted suicides are saved by being taken to a hospital in time.
The counterexample is that suicides are low in wartime, if leaders can build social solidarity — e.g., Churchill in WW2. Rhetoric of common enemy can help lower suicides.
Q&A with speakers
Q: How many Americans will be infected?
A: (former senior FDA official) The number of infections will be very high, 10s of millions. But that’s not the same number of people who will have significant symptoms.
Q: Should we test a random sample of the public continuously to understand the growth? Why aren’t we doing this? Are other countries?
A: Yes we should! The CDC proposed it long ago but didn’t have the test kits. Some countries are doing it.
A: We need to get back to some kind of normal over time. One idea is to have “hot spots” and “cold spots” but that requires testing since a cold spot can be hit with new cases from outside. The antibody test I mentioned will be very helpful to determining herd immunity in a community to limit the spread. With more testing we can return to a strategy where you isolate, contact trace, and allow return to the workplace. We will have to do it in stages. Also no one knows how this virus is going to behave going forward. There are hopes about seasonality or mutations to less-aggressive patterns, but we don’t know. We’re only a few months into this. There are no clear, absolute answers.
Q: How can the Fed deal with its conflict problems?
A: (Former Fed governor) Pattern in the past for loans / guarantees, they don’t do it themselves. THey set up a special vehicle and provide funds to it. They used BlackRock in the past and have announced that they’ll do it again, to look at the collateral that large businesses would have, the viability, and Black Rock will make the decisions.
Q: States & localities always want to be bailed out by the federal gov’t. How will we deal with that conflict?
A: That’s a key question. Cities can go bankrupt, states cannot. They might come to Congress and ask for loans or assistance; or they can go to the Fed as a back door under this new authority, to get loans from the Fed. But the money is fungible; it won’t all be used for coronavirus problems, some will be used to deal with structural deficits. This is a very undesirable event. There will be enormous political pressure. Will need to be monitored closely. The Fed will be concerned about it.
I made the point a few days ago that extremely simple exponential curve-fitting has done astonishingly well in predicting the progression COVID-19.
Adam Adamou, Director of the London Mathematical Laboratory, makes precisely the same point with rather more detail and sophistication:
A few key points:
Simple exponential models work well in the early stages because they mirror the mechanics of disease spread (each infected person infects more than one person).
They are limited: they can’t predict when the curve will bend, for example.
The models that follow real underlying disease mechanics, like SIR models, more closely mirror disease mechanics including the achievement of herd immunity. But key inputs are not knowable early on, and the models are exceptionally sensitive to these inputs.
Over time, we are able to estimate these variables with more accuracy and can transition to “better” epidemiological models.
Indeed, here’s one such simple model / paper (cited by Adamou) which predicts that NHS capacity in the UK will be strained and then exceeded in “1-2 weeks” (this was on the 22nd of March):
As someone who has built such exponential models, it’s extremely important to point out what they do, and don’t predict.
These models simply answer the following question mathematically: “If the exponential growth we have seen so far continues on an exponential path, what will happen?”
Of course, there are many ways that we can change that outcome. Stringent control measures, changes to individual habits (self-imposed social distancing, washing hands), achieving herd immunity, developing a vaccine, etc. would all make the future different than the past.
But the fact that even where we have seen stringent measures put in place (Italy, Spain, France, Germany), the curve is taking time to bend to a slower-growth path, should alert us that the time to act is now (actually, it was in late January / early February) and that we need to risk doing too much rather than not doing enough.
The US, and in particularly the New York City area (my home town, where part of my family and many friends live), are showing the potential to be the worst-impacted locations in the world.
What do the data say about where the US is now and what’s likely to happen in the near future? And why does the US appear uniquely unable to respond to the challenge? (Spoiler: it’s not mainly about partisan politics.)
(Most charts below, except where stated otherwise, are from the stupendously outstanding Our World in Data. If you’d like to support their important work, you can donate here; I just did.)
Many cases, growing quickly …
The US has now surpassed 100,000 cases (28 March figure; chart below goes through 27 March):
The US is now adding more cases daily than any other country. This is the rolling three-day average as daily cases reports can be lumpy based on lags in testing and reporting:
The same article has a novel way of comparing severity in different locations that I like a lot, though it requires some thinking. This chart compares two measures: the growth rate in total confirmed cases (using a trailing 7-day measure to smooth out lumps), and the confirmed cases per thousand people, which is a measure of the degree to which COVID-19 is penetrating into the local population (“attack rate”).
There are a few things to think about when looking at this chart.
Any location–a town, a city, a country, a cruise ship, a nursing home, a family–which sees an outbreak of COVID-19 will show very rapid growth in the beginning. But we should obviously worry much more about a location growing cases from a base of many cases than a base of few cases. New York City alone had more than 26,000 cases as of 27 March, around 5% of the world’s total: a 40% growth rate in NYC is much more concerning than, say, a 100% growth rate in a small town.
What this chart shows is that New York is uniquely bad in combining three factors: a large population (20 million), a high attack rate (>2%), and a continuing high rate of growth.
And another New York Times article–actually a feature helpfully updated daily–shows US cities on the same “doubling time” chart that I’ve republished several times from Our World in Data. (This chart is of deaths, not cumulative cases; note that it’s plotted on a logarithmic scale so a straight line means consistent exponential growth.)
We see here that New York is on a particularly brutal curve, doubling deaths roughly every two days.
… and not slowing down
I’ve recently discussed the importance of doubling times and whether they are increasing or decreasing. Unfortunately, the doubling time in the US has remained stubbornly high for three weeks now. (This is an update of the doubling time chart from a prior post.)
Repeating a point from that earlier post, with updated figures:
If the US were to continue to see the number of confirmed cases double ever 3 days, the number of cases would increase 813-fold in 30 days. A 813-fold increase in the number of confirmed cases would take us from 85,991 cases on 27 March to more than 69 million cases on 25 April!
This is just a statement about how the math works, not a prediction. For many reasons, I don’t think that will happen. To name a few:
The growth rate in reported cases in the US is likely greater than the growth rate in actual cases, because the rate of testing is increasing. Over time, the two growth rates should converge.
Control measures are increasing and will slow the growth rate.
Even if nothing were done, as we get into the millions and tens of millions of cases, some degree of herd immunity kicks in. The effective reproductive number decreases as there are the proportion of infected (and therefore immune) people increases.
What’s the special challenge in the US?
Looking at different responses around the world, it’s become clear that a major pandemic exposes unique strengths and weaknesses of different cultures and different systems of government, at least in terms of their ability to respond to a pandemic.
Simplifying enormously, I think about four factors which seem particularly relevant to a society’s ability to respond to the crisis effectively:
Quality and inclusiveness of health care system.
Degree of central government control.
Individualism versus collectivism.
It goes without saying that better health care systems, which often but not always correlate with wealth, are better able to respond. But the inclusiveness of the health care system is important in a severe epidemic in at least one way: the extent to which all citizens believe and trust that they will be looked after by the system, regardless of their personal resources, may correlate with their willingness to temporarily forgo income.
Transparency has been a clear benefit in the response in many countries, with figures published daily, and extraordinary collaboration among scientists and epidemiologists to find solutions. There’s a strong argument that lack of transparency was at the root of the failure to control the epidemic earlier in China.
Considering government control as well as culture, I’ve been struck by many accounts of how China has exercised extraordinary innovation to battle COVID-19, but also used highly invasive approaches ranging from using big data to monitor and direct individual citizens, to removing infected individuals from their families. This video by a British man living in Wuhan is a particularly interesting overview of the steps the Chinese are taking on the ground today. I don’t know about you, but it’s really hard for me to imagine most Western European countries, let alone the US, tolerating this degree of invasion into the private sphere.
It’s been striking how a number of countries with roots in Confucianism, including South Korea, Singapore, and Taiwan, have been able to execute extreme lockdowns with reportedly high compliance and high effectiveness.
Looking at Europe, it’s similarly striking the degree to which in countries including France, Germany, and Switzerland have been able to implement stringent control measures. Some countries, like France, have a degree of central power that simply doesn’t exist in the US; others, like Switzerland and Germany, may have significant power within regional governments, but enjoy a high degree of trust in the competence of the government and the professionalised bureaucracy.
As relatively communitarian societies with strong memories of the World Wars, many European countries have not only accepted the sacrifice required to implement control measures, but have also managed their response on a national level. For example, in France, our local departement in the South West has no deaths and only a handful of cases. Alsace, on the other hand, has been badly hit. So the country is aggressively sending doctors and resources to Alsace, and the military is moving patients from overwhelmed hospitals in Alsace to other locations.
In comparison, the US is at the extreme of both a culture rooted in individualism and individual rights, and a highly decentralised government where many powers reside with states, cities, and other local governments. Moreover, the US must be at an all-time low in terms of faith in government.
What makes me especially pessimistic about the ability of the US to respond is the combination of extreme individualism, the relative weakness of the federal government, and the low faith in government. Watching Governor Cuomo’s daily briefings has hammed home the degree to which each city and state is choosing its own response, and to which different locales are competing for scarce resources.
A depressing outlook
At the end of the day, not every country, state, city, or region will respond effectively. Many countries in Europe are taking the difficult medicine of lockdown for long enough to get control of the epidemic, and will then be able to gradually reopen the economy while implementing aggressive track/trace/quarantine against any new outbreaks.
Imagine that one country in Europe, Covidia, either can’t or won’t implement the measures to get the epidemic under control, and the virus spreads through a significant proportion of the population. Other countries in Europe will close their borders with that country, and maintain strict quarantine rules for anyone entering the country from another. Covidia is in bad shape, but many countries are able gradually to return to normality over time.
In the US, it seems to me that we have the worst of all possible worlds. Some cities and states are moving to strict lockdown, accounting for around half of the population. But half of the population is not under lockdown.
Until every location with even a few cases implements strict control measures, COVID-19 will continue to spread. And unlike in other regions of the world, if one state or city is badly hit, it’s not obvious that there’s a way for its neighbors to close the borders.
So the US is effectively choosing intense economic pain without the benefit of the ability to come out of lockdown in 6-8 weeks. The result will be one of two scenarios, both bad.
In one, the US will need to continue to implement increasingly strict and broad control measures, extending the period of economic impact far longer than it needed to be–with catastrophic results for employment and the economy as a whole. But at least the virus will come under control over time.
In the second, the US will be politically unwilling, or unable, to do so, and COVID-19 will continue to spread rapidly. The result will be a much greater attack rate into the population; health care systems being overwhelmed around the country; more sickness and death than would otherwise have been the case; and and extended period in which economic uncertainty persists, perhaps with see-sawing between tighter and looser controls.
Do you have a more optimistic story about how things will evolve in the US? Please tell me; I’ll share the best arguments in the next few days.
27 March 2020; updated 28 March to fix broken image links in some clients
Much of what I share in these notes is the result of conversations with many well-informed, intelligent, and wise people. The constant stream of suggestions, corrections, and criticism is invaluable.
One regular correspondent, who has asked to remain anonymous as he has a high-profile role in Silicon Valley, shares my interest in doubling times and second derivatives, but has taken it to another level. He’s gone so far as to assemble a small team to model this at the country level, and has given me permission to share the following.
Totally agree with your post this morning. Many countries are making progress! I am now confident that social distancing is working. Italy has seen its seventh day of ~5-6k daily new cases. That means that the doubling number is sure to keep falling. Even at our conservative Scenario B (copied below; not sure if you saw my email yesterday), I think we are within days of peak active cases for Italy (more on why that matters in a minute). Of course, with the fire now smoldering, the key question will be how it responds when social interaction resumes.
Loved your look at doubling times. When you look at Italy zoomed-in, the slope of progress becomes even more apparent. I was a bit worried about daily volatility, so I used a 2-day, 3-day, and 4-day doubling time measurement. Was remarkably smooth. Also, what do you make of the blips every seven days? I have a suspicion it’s related to slower weekend testing or a lag in reporting over the weekends.
To the point you made in you post, it’s all about the rate of change in new cases. For the graph below, I calculated the day-over-day change and then calculated the average of that change for the trailing five days. The slope of the line looks to be statistically significant … and Italy just crossed over to negative 5-day trailing growth.
So, Italy may have seen its peak reported daily new cases 14 days after lockdown. Remind you of another curve? Wuhan/Hubei! Lockdown works. It just takes 14 days before the bad news peaks.
Another idea on which I’ve been reflecting: We have been focused on Total Cases for the first few weeks, because that is roughly equal to Active Cases in the early days. However, what really matters to the stability of our healthcare systems is Active Cases, i.e “flattening the curve”. If we can keep the healthcare systems from being overrun, we can tolerate a certain number of Active Cases at a time, especially once we have effective therapeutics. Now … given just two variables, a doubling rate (in days) and an average case duration rate (also in days), isn’t there a theoretical doubling rate where Active Cases plateau? If the doubling rate (in days) is below that theoretical threshold, Active Cases will grow forever. However, once the doubling rate goes above that threshold, Active Cases will eventually peak and then start to fall. Going to see if we can model this theoretical number for COVID-19 today. That’s the “bogey” that every country should be shooting for.
To your point, on the US, it appears that there is still a long way to go. Hard to even know the trajectory until we have more days under broader testing availability. When I smooth out the doubling calculations over 4-days, the US is at a 4.6 day doubling rate. However, that was a terrifying 1.5-2.0x just a few days ago. Suspect most of this volatility is related to changes in testing availability. The next few days will be really important. FWIW, NY and CA are both at a doubling rate of ~3x, so the higher US rate is related to a bunch of other regions that are earlier on the exponential curve. Does not bode well for news coming out of the US over the next couple of weeks.
(Chris here again.). If you’re interested in going deeper into the approach this team is using, I’ll also share their description of their methodology:
Our first, rough approach (will be refined later) was to fit curves to China’s and Korea’s daily new cases over time, after they hit peak daily new cases. We then created two decay curves we could apply to Italy that were more conservative, adjusting for the exceptional containment capabilities of China and Korea. Our Scenario C, followed China on the decay, but bottomed at 20% of peak daily new cases, recognizing that many countries won’t be able to entirely eliminate transmission. Our Scenario D assumed that the decay was twice as slow as China and bottomed out at 40%.
We then applied Scenario C and Scenario D, as well as two other scenarios, to Italy’s current new case count. Scenario A took the 3/23 # of daily new cases and assumed that continued in perpetuity, just to conservatively see what it would do to Peak Active Cases. Scenario A went even more conservative, and took the average of the prior five days (including more of the peak) out in perpetuity.
The key output was then Active Cases using our four daily new case scenarios. We assumed 14 days on average from case reporting to case resolution. We thought Active Cases were far more important as a measure of burden on the health system than were Total Cases. As you can see below, our models suggest that Italy may be somewhere between 3 and 10 days from peak healthcare system utilization related to COIVD-19. More important, the current 55k Active Cases may only top out at 60-80k. Suggests that the terror of exponential growth may be behind us.
We also looked at log curves of Total Cases, which our modeling suggests are about to flatten quickly and significantly.
This is just a first draft of half a curve for one country, but it suggests to me that we may be surprised at how quickly the curves start to flatten after we implement social controls. Of course, the bigger question is what happens as we start to release those controls. I think if the public health regulators around the world can learn from China and Korea (aggressive testing, isolating, and tracing), we might be able to avoid lighting the uncontrollable community spread again.
The exponential total case curves still look scary for most countries, but if we focus on daily new case, I think we’re going to see a similar effect as Italy once we get another week into social controls for the rest of the world. My first moments of optimism in about four weeks!
A confirmed case is, or should be, “a person with laboratory confirmation of COVID-19 infection” (as defined in WHO situation reports); but different countries apply different standards, and standards have changed over time in some countries. Confirmed cases are typically reported both on a daily and on a cumulative basis. The cumulative figure therefore also includes people who went on to recover, as well as those who went on to die, and therefore is not equal to the number of current cases.
The number of confirmed cases is used to calculate the widely-discussed Case Fatality Ratio (CFR), which is simply the number of deaths attributed to COVID-19 divided by the number of confirmed cases.
For many reasons, we should care more about the number of actual cases, whether confirmed or not.
And similarly, we ultimately care about the Infection Fatality Ratio (IFR) more than the CFR, which is calculated by dividing the number of deaths attributed to COVID-19 by the number of actual cases.
Why? When you read about the estimated final attack rate (the proportion of a given population — a country, the whole world, a family, a cruise ship — who become infected over time), and want to translate that into potential fatalities, you need to apply the IFR rather than the CFR.
That’s because estimates of the attack rate are estimates of how many people will actually get COVID-19, not how many cases are detected.
For a much more detailed (and, frankly, better) discussion of this, see this section in the always excellent Our World in Data. Their section also includes a discussion on important, related topics which I haven’t had time to write about:
Why CFR is not a constant over time
Why CFR does not represent the risk of death for an infected individual
Examples of how estimates of CFR have declined over time
The complex thing about pandemics is that early mortality rate estimates tend to decline over time. Why? Here are four simple measures that matter in the context of a pandemic: (a) the population of a given geographical area (b) the total number of infected individuals, including both asymptomatic people and people that get sick (c) the total number of people that are infected, get sick and self-report (d) the total number of people that die During the haze of a pandemic, the best estimates that entities like the World Health Organization often derive are based on (a), (c) and (d), and even things like (d) are complicated by pandemics affecting older individuals with pre-existing conditions. They do not know (b) upfront, and sometimes it is never known, or only known with the passage of time.
Take the Swine Flu (H1N1/2009) as an example. Early estimates in the fall of 2009 from the WHO1 pegged the H1N1 mortality rate at 1.0%-1.3%, since they were dividing (d) by (c). Four years later, a study from the WHO and the Imperial College of London2 estimated H1N1 mortality as a function of total infections, including both the asymptomatic and the sick. Their revised H1N1 mortality rate using (b) as a denominator: just 0.02%.
So, please treat estimated infection rates and mortality rates with care, since they can mean very different things. Marc Lipsitch from Harvard has estimated that 40% – 70% of the world’s population could become infected . Lipsitch himself makes it clear that this number is an example of (b) and not (c) and that there is an enormous gap between the two, so please do not multiply population by 40%-70% and then multiply by a mortality rate assumption. The vast majority of infected people will likely not become sick, and around 80% of people who get sick develop mild infections rather than severe ones.
The Our World in Data section I linked to above illustrates just how quickly our understanding of CFR can evolve:
What are our best current estimates of CFR / IFR?
The epidemiologist Adam Kucharski, a great source, points to three papers that estimate the range in China at 1-1.5% of symptomatic cases (importantly that means CFR, not IFR):
How well have naïve, simplistic models of the epidemic done?
I’ve mentioned before that I am a great fan of Philip Tetlock’s Superforecasting.
One thing he argues is that many so-called expert forecasts are mealy-mouthed, in the sense that they are so vague as to give plenty of wiggle room. “The market is overvalued and valuations will come back to earth” is almost certainly going to be true at at least some future date, allowing the expert to sound smart today and choose the point of time in the future at which he can point to his so-called forecast. So Tetlock, and forecasting tournaments in general, require forecasters to make specific predictions linked to an objectively measurable outcome on a given date.
Kahneman discusses the challenge of decision-making in “low-validity environments.” Roughly, these are situations that are not conducive to developing high-fidelity pattern recognition/intuition, because they are noisy (outcomes are not consistently correlated with the inputs); because the true outcomes come too long after the causally-related input; or because few individuals are exposed to decisions and their outcomes with enough frequency. Kahneman argues that simple algorithms perform best in such situations, versus “expert” judgement or complex models.
Below, I’ll share what happened when I decided to build a very simple model of a very complex domain I knew little about.
My naïve model: the backstory
The origins of this blog / newsletter go back to 28 February. Several friends, who are among the smartest human beings I know, and I were e-mailing each other about this thing called COVID-19.
I had stumbled upon some of the early modeling attempts that were being discussed online, including this great interactive widget that permitted users to input a few key assumptions and get an instant forecast of the epidemic. I plugged in some plausible figures for R0 (basic reproduction number) and post-control Re (effective reproductive number) and was shocked by what I saw.
I spent the next few days looking for published epidemiologic models and was disturbed by what I found. Several teams and individuals announced that they had built models; but no one was sharing the output. That seemed … worrying. Why weren’t they publishing their predictions?
I started reading up on epidemiological modelling techniques (this book was helpful) and realised that, as someone who hadn’t done serious maths in 30 years, it was going to take me too long to get up to speed. Plus, while these models clearly were soundly based on the underlying dynamics of how epidemics spread and ultimately end, I worried that they would be very sensitive to inputs that were essentially unknowable at this early stage, and therefore not useful for prediction.
I decided to try a very, very simple approach.
We know that epidemics, at least in their early stages, follow exponential curves. I downloaded the data (first from the WHO, then from Johns Hopkins’ outstanding resources, and later from my currently prefered source, Our World in Data) and did some simple curve-fitting.
Wanting to model a curve of the form y=m*e^(b*x), I chose a least-squares fitting of ln(y)=ln(m)+b*x, using Excel (a painful mistake as the model grew larger; note to self, must learn Python).
By 4 March, I had my first model working. For a number of reasons, I chose to exclude China and to only forecast one output: the confirmed number of cases outside of Mainland China.
On the 6th, I shared the following precise forecasts with a close friend:
Date: Fri, 6 Mar 2020 14:25:18 +0000
Subject: exponential curve fit for total cases outside china
From: Christopher North
Using basic exponential growth curve on WHO data below.
I ran it with data through 3 March and it underpredicted the next two days
by 23%. I think the model is overfitting to the early days of reporting
when the cases were probably even more severely underreported than today.
Predicted doubling time is about 2.5 days.
Trebling time is almost exactly 7 days which makes for easy mental maths.
r2 = 0.9665
100K cases outside china 19 March
200K 23 March
500K 29 March
1M 3 April
5M 13 April
10M 18 April
20M 22 April
Of course this is extremely crude. There are more interesting models out
Prediction: schools and most businesses close in around 3 weeks.
I then went on to update the model daily, regularly updating the forecasts and assessing how the model was performing. Whilst I made a number of improvements to data sources, and whilst I was painfully aware of the limitations of the model, I didn’t change the underlying methodology.
I went on to published a daily update in the form of a newsletter to a few dozen friends starting on the 11th of March.
So, how did these predictions perform?
I was clearly wrong about when schools would close (although correct that they would close). Saying they would close “in around 3 weeks” was mealy-mouthed, of course. I made the prediction on 6 March (in the UK, referring to UK schools) so that implied a prediction for 27 March. In fact, UK school closures were announced on the 20th of March to take effect on the 23rd; I was a week, or one-third, off.
In terms of forecasts for specific milestones, here’s a table showing what I forecast vs what has actually happened or is highly likely to happen:
(The “Predicted” column is based on taking the most recent actual and doing simple extrapolation; it’s not an output of the model.)
It’s important to say that I don’t feel a high degree of conviction about the forecasts going through the rest of April. There are many things that stop or slow the exponential growth phase of an epidemic (control measures like we’re seeing take place around the world; behavioural change; vaccines; mutations to the virus; weather changes; etc.). And I’m fully aware of the many limitations of my model; indeed, from the beginning, I highlighted them (you can find them in the FAQs in the first newsletter here).
I’ll also add that I’m surprised that the model’s predictions have held up so well. The model isn’t aware of things like control measures, social distancing, or herd immunity. It can’t predict when the “curve will flatten” or when the epidemic will reach its peak and begin to die out. It simply assumes that exponential growth will continue along the same path.
With that in mind, it’s very depressing, as well as a clear failure of collective action, that despite having ample information at the end of February to know that we were facing the prospect of a serious global epidemic, we are still debating today in some countries and cities whether we need to take serious action.
As I mentioned in my last post, in passing:
The doubling time in the US has been around 3 for more than two weeks now. Let’s put that in mathematical context: if the US continues to double the number of cumulative confirmed cases ever 3 days for another 30 days, the number of cases will increase 500-fold. `A 500-fold increase in the number of confirmed cases would take us from 55,000 cases on 25 March to more than 27 million cases!
Again, to be clear, I don’t think this is what will happen. But the maths say it could happen if the US does not take very significant actions soon.
Today, I’m going to focus on a quantitative way to tell if we’re making progress. There’s a bit of maths, but you can skip it; everything is explained in plain English as well.
While the tide might be turning in Italy …
Over the last two weeks Italy has moved from a doubling time of 2-3 days to 7 days. France, Germany, and Spain all show early signs of a slowing of the doubling rate as well.
That leads to an interesting question I’ve been debating with a friend: how would we know if we’re making progress? It seems strange to say that we’re making progress when large numbers of new cases are announced every day, and sometimes even increasing numbers of new cases.
But in an exponential growth epidemic like COVID-19, that’s what the earliest sign of progress would look like: a slowing of the rate at which confirmed cases are increasing.
A little high school calculus is helpful here.
Many charts plot the cumulative number of confirmed cases for a given country. Those charts look scary, and they should scare us:
These curves are all increasing sharply.
What is the rate of change of cumulative confirmed cases? It’s the number of new confirmed cases daily, the first derivative (or equivalently, the slope) of the curve of cumulative confirmed cases.
Most, but not all, of these curves are also increasing. (The daily confirmed rate can be quite volatile for a number of reasons including lags in reporting, “catch-up” reporting, changing rates of testing, even changing definitions of confirmed cases.). You can see examples where the number of daily new confirmed cases ticked down and then soared upwards again.
Helpfully, the rate of change in cumulative cases is strictly inversely proportional to the doubling time. (See below for the maths.) That is, if the rate of change in cumulative cases goes up by a factor of two, the doubling time halves; and vice-versa.
But if you look at another metric, harder to illustrate, there are signs of progress; that’s the rate of change in which new cases are appearing. This is the first derivative (or slope) of the number of new daily confirmed cases, or the second derivative of the cumulative number of confirmed cases. Put in plain English, this metric measures whether the rate at which new cases are being confirmed is increasing or decreasing.
I haven’t seen anyone plotting this curve. It’s easy enough to do, but I don’t have time right now to download the data and plot it. But helpfully, because of the relationship between the rate of change of daily new confirmed cases and the doubling time, there’s a simply way to observe this second derivative, and that’s through the change in the doubling time.
If the doubling time for a given country is increasing, the second derivative (again, the rate of change in which new cases are appearing) is decreasing; and vice-versa.
That’s why the first sign that we’re making progress is likely to be a sustained increase in the doubling time. Your best daily source for this data is (no surprise!) the great table in Our World in Data (OWiD).
I downloaded the OWiD data, which mostly comes from the ECDC, and took a quick look at how doubling time is evolving:
China, the top line, has seen doubling time steadily increase, meaning the rate at which infections are increasing is decreasing — a lot! South Korea and Japan show the same trend. Unfortunately, the World (all cases in the world) doubling time is decreasing significantly, meaning the rate at which cases are increasing worldwide is increasing rapidly.
Let’s look at that same chart with China, South Korea, Japan, and World removed to make it easier to see what’s happening in Europe and the US. (I made the lines curvy, which they’re not–these are calculated daily–as it’s easier to see the individual countries.)
It might be easier to see this in a table:
Good news: Italy looks like they’re making progress! And there are early signs of progress in Spain and France too.
Of course, it will take many more days of ratcheting up the doubling time to begin to feel more confident. And with a mean lag time of 3-4 weeks in fatal cases between infection and death, the number of deaths will continue increasing rapidly even when the rate at which of confirmed cases is growing is slowing. And then there’s the questions of what happens when we begin to relax control measures–could there be a second wave? Those are all questions that merit further reflection.
... the US is really, really scary
The doubling time in the US has been around 3 for more than two weeks now. Let’s put that in mathematical context: if the US continues to double the number of cumulative confirmed cases ever 3 days for another 30 days, the number of cases will increase 500-fold.
`A 500-fold increase in the number of confirmed cases would take us from 55,000 cases on 25 March to more than 27 million cases!
Now, for a lot of reasons, that’s not going to happen. First, some of the rapid increase in the US must be the result of more testing identifying a higher proportion of cases. Second, a number of cities and regions have rolled out, and are increasing the intensity of, control measures. Finally, no epidemic can increase exponentially forever: at some point, the virus runs runs out of new people to infect (even if we don’t develop a vaccine), and the evidence is starting to confirm that there should be at least partial / temporary immunity (and likely better than that) after recovery.
Still, I am really, really worried about the US for reasons that go beyond the numbers. The response so far has been piecemeal: state-by-state and city-by-city. Unfortunately, SARS-CoV-2 doesn’t care if you live in New York or Florida. And unlikely countries, it’s not really clear that states or cities have the option of closing their borders to their neighbours, or what the consequences would be of trying.
To date, my updates have been dipping into the flow of information to try to capture high-quality signals. I don’t know about you, but I can’t keep up anymore; there’s just too much information coming at me at once. Plus, all too often I struggle to know how to integrate even very high-quality information with what I already thought I knew and believed, and therefore how to update my expectations.
For my own benefit, at least, I’m going to try a slightly different approach for the next few days to see if it’s more satisfying. I want to try to document what our best / most reliable information is on key topics (and how we know it, and how reliable we think it is — it’s so-called “epistemic status — as well as what critical questions remain.
The idea is to create these as a resource that builds cumulatively — and, it goes without saying, with a lot of help from friends — and then have future updates be with reference to that baseline understanding.
I reserve the right to modify the format as we learn more about what works, but the initial thought is a set of individual posts/pages on key topics / questions, updated as appropriate.
This document apparently represents the current views of the “International pulmonologists’ consensus group on COVID-19.” Much of it is beyond me, but it’s fascinating to see how doctors are coming together to share knowledge and best practices.
I found this summary of what is known about why the disease affects children less significant than adults very interesting (and consistent with what I’ve read elsewhere):
Don’t try this at home
There’s been a lot of discussion about the potential of chloroquine, starting with President Trump’s tweet. Here’s why you shouldn’t try to treat yourself.
How best to support COVID-19 efforts?
Recommendations from my friend William MacAskill, Professor of Philosophy at founder of the Effective Altruism movement. Will has an extremely high bar about both the potential impact of projects, and about the extent to which projects represent “value for money” from a donor’s point of view, in terms of impact per dollar deployed.
I wasn’t surprised to see the Johns Hopkins Centre for Health Security on the list, but all of the others were new ones to me.
Last updated: 31 March; first posted: 24 March 2020
One acute contributor to uncertainty in this crisis is that we really don’t know how many people currently have, or have previously had, COVID-19.
First, some definitions.
A confirmed case is, or should be, “a person with laboratory confirmation of COVID-19 infection” (as defined in WHO situation reports); but different countries apply different standards, and standards have changed over time in some countries. Confirmed cases are typically reported both on a daily and on a cumulative basis. The cumulative figure therefore also includes people who went on to recover, as well as those who went on to die, and therefore is not equal to the number of current cases.
The number of confirmed cases is used to calculate the widely-discussed Case Fatality Ratio (CFR), which is simply the number of deaths attributed to COVID-19 divided by the number of confirmed cases.
For many reasons, we should care more about the number of actual cases, whether confirmed or not.
And similarly, we ultimately care about the Infection Fatality Ratio (IFR) more than the CFR, which is calculated by dividing the number of deaths attributed to COVID-19 by the number of actual cases.
Why? When you read about the estimated final attack rate (the proportion of a given population — a country, the whole world, a family, a cruise ship — who become infected over time), and want to translate that into potential fatalities, you need to apply the IFR rather than the CFR.
Why do we think the number of confirmed cases underreports the number of actual cases?
A1: Only a small proportion of each country has been tested. I highly recommend Our World in Data’s section on testing. Here are the two key charts showing total tests and tests per million people:
[Note: since I posted this, Our World in Data withdrew their charts on testing as they no longer regard the data as reliable; see hear for an explanation. I have left these charts as the best data I’ve found so far, but they should be regarded with suspicion.]
As you can see, there’s an enormous degree of variance in testing in both absolute numbers and relative to population.
In general, the lower the rate of testing per capita, the more sceptical we should be about the number of reported cases. Even South Korea has tested less than 1% of its population. But countries that test more of the population, and who aggressively test patients at risk (those presenting symptoms, self-reporting or otherwise; those who may have had contact with an infected patient; etc.) are likely reporting as confirmed cases a higher percentage of the actual cases than other countries.
What other factors impact the reported data on confirmed cases?
We’ve frequently seen both lags between testing and reporting, and inconsistent frequency of reporting by some countries. This can lead to an apparent large jump in reported cases, when several days’ worth of confirmed cases are reported all at once, which can appear to overstate the rate at which actual cases might be growing. Similarly, lags in reporting can make it appear at other periods that cases are not growing. My sense is that these issues are improving over time, with many countries falling into a regular and predictable rhythm of reporting; but I don’t know that for sure..
If a country is increasing its rate of testing, could that make it appear that cases are growing faster than they really are?
Yes! I’m so glad you asked that, because this is very likely happening and it may be hiding some of the progress that’s being made.
In simple terms, if a country is increasing the degree of testing over time, they are likely therefore capturing an increasing proportion of the actual cases over time. During the period that the degree of testing is increasing (which I hope and believe is happening now in many countries to a significant degree), even a country that is seeing a slowing rate of growth of actual cases may appear to show a steady or increasing rate of growth of reported cases.
A simple example will illustrate this. Imagine that country Covland has 100 actual cases on day 1, and actual cases are growing at 10% per day. Now imagine that on day 1 it is only detecting 10% of cases, but then rapidly ramps the degree of testing to capture 10 percentage points more cases each day, up to 100%. Here’s what happens:
So Covland appears to have a much higher rate of growth than it actually does.
I think it’s highly likely that this is happening now in European countries.
Is there any way to assess the degree of underreporting to try to estimate the number of actual cases in each country?
Yes, there are several, but they all have limitations.
A recent paper on this, published 23 March, uses Adam’s preferred methodology. While we don’t know the number of actual cases, we do have a strong sense of the number of deaths from COVID-19. We also have ranged estimates of the average time between infection and death, and (with very wide error bars) ranged estimates of the percentage of cases that result in death. Putting those together, we can look at the number of deaths that happen in a given country on a given day, and estimate how many people must have been infected 3-4 weeks ago to result in that many deaths.
[30 March update: the authors updated their estimates; see here. Chart below updated to match.]
Here is the output of that analysis:
Of course, this is at best a rough estimate, and probably only useful for making order-of-magnitude estimates.
Back when I was maintaining my own COVID-19 models, I used a simplified version of this methodology and concluded that the US and many European countries had 15-30x underreporting (i.e., were detecting and reporting only 3-6% of total cases). Interestingly, even then, I estimated that Germany and the Scandinavian countries were reporting a high proportion of the actual cases.
However, while other estimates I’ve seen or heard about come to similar conclusions, this is not a universally held view.
As an example of a much more extreme estimate, using a different methodology, this article in the FT on 24 March discusses a paper with the claim that SARS-CoV-2 may have already infected half of the UK population. I am sceptical of this claim, but if it were true it would completely change our understanding of the disease; and in particular to revise down our estimate of the Infection Fatality Rate.
On 30 March 2020, researchers at Imperial College published a paper (discussed in this article) which, while primarily oriented to assessing the effectiveness of control measures put in place in 11 European countries (conclusion: consistently effective in reducing the effective reproductive number), also tries to estimate the percentage of the population that is infected in these countries. Here is the key table, presenting ranged estimates at a 95% confidence interval:
I would be cautious with these figures as the study had to make a number of important assumptions (including that the impact of control measures was roughly consistent across these countries). Still, it gives us a sense of what the possible ranges might be.
If there is significant underreporting, doesn’t that mean that the Case Fatality Rates we read are potentially overstated, and that the final Infection Fatality Rates will be much lower?
This is such an important, and complicated topic, that it deserves its own page. I’ll try to get to that in the next few days, but in the meantime go read the section on this in Our World in Data (the anchor on their page isn’t working; scroll down). In the meantime, I’ll say very simplistically: yes and no.
Yes, the CFRs we were hearing early in the epidemic from Wuhan of 3-5% are almost certainly too high as estimates of the final IFR. They likely suffer from underreporting, but they also reflect the fact that Wuhan’s medical system was overwhelmed. CFRs will vary significantly depending on context (age and health of population, availability of care, etc.).
Yes, if underreporting is high, reported CFRs will be greater than the actual IFRs.
However, some people with COVID-19 today will die, and so CFRs can also understate the ultimate IFR. To give an obvious example, if 100 people catch COVID-19 today, and 5 ultimate die of it, the CFR appears to be 0% until the first death happens.
Why is this post labeled “draft”?
It’s a first effort. I need to get feedback and input for a period of time before being confident enough in the accuracy of the information I present here to remove “draft.” Of course, I’ll continue updating it after that.
Better and worse outcomes seem more likely; expected case appears more painful
I’ve spent much of the last three days thinking more about expected, better, and worse cases, and how I should update my priors in light of a constant flood of new information.
On the one hand, I’m increasingly optimistic that there is a path toward the better case, and I’ve increased from p=20% to p=25%. We now have multiple existence proofs (China, South Korea, Taiwan, Hong Kong, maybe Japan) of a playbook that works. The New York Times describes it very well: Let scientists lead; ramp medical capacity rapidly; implement extreme social distancing (total lockdown); ramp up testing; isolate the infected, including removing people from their families (the secondary attack rate seems has been estimated at 10-15% with a lot of questions); aggressive identify, track, trace, and quarantine every suspected case and every contact of that suspected case; and prevent new infections from coming across the border since the potential for reintroduction is very high even after successful controls. At the same time, we need to continue the rapid push for therapeutics (promising) and a vaccine (likely far away); and to help poorer countries both for humanitarian reasons, as well as on the selfish grounds that otherwise the virus could find human reservoirs.
On the other hand, even in the expected case, the current and potential economic impact is almost unimaginable (see yesterday’s post to get a sense of it), with nothing to compare it to in any living human’s memory. With aggressive economic support from governments and actions by central banks, it’s possible to imagine how we bridge six weeks or even three months that lets the world return to its current productive capacity within a few years, avoiding mass long-term unemployment, bankruptcies, and worse. But if a total lockdown needs to go on for many months; or if governments implement half-measures or are unable/unwilling to maintain strict control measures and then have to re-implement them again several months later; the economic damage could be severe. And that’s still leaving aside many known unknown risks like a severe second wave; the virus mutating in a more dangerous direction; failing to build health care capacity; zero-sum behavior by countries; rising geopolitical tensions; and so on. I fear that there’s a real risk of an own goal here, and have raised my risk of a worse outcome to 25%; and at the same time have understood that the economic impact under the expected case is worse than I previous thought.
What should our base rate be on the time to economic recovery?
I love this white paper from JP Morgan, which looks back in history to ask how quickly countries rebound from crisis and economic hardship. I highly recommend reading the full paper. Some key points:
Looking across 12 major dislocations since 1870, “With the exception of the Great Depression and certainly in the post-war era, it generally took less than four years to regain the peak, and sometimes less than three.”
The market currently (17 March; we’re further down since then) implies more than a decade to regain the prior earnings peak:
German and Japan after WWII, and China after Mao’s Great Leap Forward, did take a decade or more to recover; but COVID-19 should be much less destructive of the world’s economic capacity under almost any scenario.
“Sometimes, even when wars do result in damage and ruin, the speed with which economies rebound is faster than the speed with which they decline. The US Civil War was the deadliest conflict in US history, with 6x the number of war dead as a percentage of the US population than World War II. The Civil War left the southern US in complete shambles […]” but farm income recovered within a few years of the end of the war:
Finally, asset prices often reach their bottom well before things start to improve, and even when the news is relentlessly bad (see many charts in the paper).
Reasons for optimism from an immunologist
I’ve started actively follow Dr James R. Baker’s blog (yes, we seem to use the same WordPress template) and recommend it for balanced, expert discussion in plain English. I particularly liked his recent post, Seven reasons to be optimistic. Read the whole thing, but the headlines are:
Social distancing does work
Underreporting is probably very high, so the actual Infection Fatality Rate (IFR) is likely to be much lower than the Case Fatality Rates we see reported. Plus we might be building herd immunity faster than we think. (Reminder, IFR is based on the actual number of infections, whether we know about them or not; CFR on the confirmed cases. We only know CFR, but we should care about IFR.)
We’re building medical capacity quickly
Coronavirus could mutate to become less pathogenetic. (CCN: I’m sceptical.)
Summer weather could help. (CCN: Still sceptical, but more evidence is pointing this way; I need to do a post just on this topic.)
Existing anti-virals could help as therapeutics.
While we’re working on vaccines, we’re not counting on them. (CCN: I also need to do a post on why an effective, safe, widely distributed vaccine is likely very far away.)
What’s happening in Iran?
A close friend whose family left during the revolution, and who now lives in the West, reports:
“Iranian doctors are very liberally using the hydroxycloroquine combined with citromycin combo and reporting very good results on patients that have respiratory distress. I have been speaking to a few doctors in the family every couple of days to get the pulse. [Apparently] the country is practicing zero social distancing (in the middle of a big national holiday when everyone is visiting everyone) and has a weak medical system.”
A useful resource comparing the pressure on health care systems across Europe
“European countries have been facing enormous pressure on their healthcare systems […] We calculate the relative pressure on healthcare systems by analyzing the proportion of COVID-19 related deaths (intensity approach) and COVID-19 active cases (magnitude approach) in real-time relative to the number of hospital beds, number of medical doctors, and healthcare expenditure, in European countries.”
Unsurprisingly, the charts show that Italy and Spain are under unique pressure today; do they give a hint of the next countries to feel the crunch?
The one chart I look at every day
Is the Our World in Data chart showing cases by country, on a log scale, normalized to t=0 on the day they first reported 100 cases. This tries to answer the question of how the different trajectories compare by country, adjusting for the fact that the disease took hold / took off on different dates. Here’s the chart as it currently stands (this is embedded so should be up to date):
The key question, of course, is whether countries like Italy, Spain, the US, the UK, etc are beginning to bend the curve since implementing increasingly strict control measures, as China, South Korea, and other Asian countries did.
As a reminder, this chart plots confirmed cases. That means we should expect a lag, perhaps even a significant lag, between the date a country imposes strict control measures, and the date we see the curve bend, for two reasons. First, there’s probably a 6-8 day lag (maybe more) between a decline in the actual rate at which infections are increasing, and a decline in the rate at which confirmed cases are increasing, given the lag from infection to symptoms, from symptoms to a test, from testing to confirmation, and from confirmation to reporting. Second, most or all countries implementing strict control measures are also increasing the rate at which they test, so the gap between actual cases and confirmed cases should be closing over that period (since we’re detecting a higher proportion of the actual cases through more testing), which could have the effect of making the rate at which confirmed cases are increasing appear to accelerate even if the rate at which actual cases are increasing is declining.
I had the opportunity this evening to participate in a private conference call where 15 speakers, each (except for one — me!) an expert in his or her field, shared brief thoughts on the COVID-19 epidemic. The audience was primarily large investors and business leaders, and so as you’ll see much of the commentary was around the economic, financial, and market impact. Most participants were in the US so it was also a US-centric conversation.
The conversation was under the Chatham House Rule, meaning I can share a summary of the content but not the names of the participants. I’ve shared near-verbatim notes taken live (as I type this, the call has just ended) but removed identifying information about the speakers. To give you a sense of the speakers, we had a Nobel laureate, a federal judge, a former members of the NEC, an IMF senior executive, and senior leaders / former leaders at major investment banks, hedge funds, and private equity firms.
Because my brief comments echoed (with some small updates) points I’ve made here previously, I’ve moved them to the end, though I was the first speaker.
As you’ll see, the tone around markets and the economy were very negative in the short term. I think the verbatim notes mask the fact that there was a strong consensus that this is something we will get through in a period of 1-2 years and that there will then be a strong recovery in the economy and the markets; that governments and central banks will do everything within their power to mitigate the impact; and that the health issues are going to be addressed successfully. Still, for some sectors and businesses, the impacts are clearly going to be enormous.
It was particularly interesting to hear from experts on Mexico, Africa, and on the real estate sector.
How did I update my priors after the conversation? On the negative side, I am only beginning now to appreciate the magnitude of the economic impact, with reportedly a 24% rate of contraction on an annualised basis in GDP in the US — something not seen in living memory. The impact on individuals and companies in the near-term could be enormous. On the positive side, the Federal Reserve’s readiness to do whatever it takes, and to ramp up firepower never before deployed, has clearly impressed these sophisticated market participants. And there was enormous optimism about the speed at which we will ramp medical capacity and deploy therapeutics.
Speaker: former National Economic Council member Topic: What will the Federal and State Response be to the ongoing Crisis?
Every gov’t focused on the health care surge with no constraints
Big question of how you tide people over during period of much lower economic activity. Enough to eat, shelter, get health care.
How to you maintain institutional integrity: Market integration; avoid mass bankruptcy and spikes in unemployment that make it harder for the economy to recover when controls are released.
No sense of fiscal constraints. $1.8 trillion currently, so around 9% of GDP. That’s meant to last for around 3 months and potentially more infusion after that.
A cheque to every American adult tapering from $75K/year income, tapering to $99K/year, after that nothing, plus $500/child. That’s around $250B. Another $50B of assistance to small business, not clear what form.
Then TARP money to protect [sarcasm] the strategically important golf course and hotel sector with maximal discretion to the administration to decide who to save. A lot of the negotiations in Congress today are around imposing constraints on that. Limitations like not laying off workers, exec compensation, stock buybacks are being negotiated.
Speaker: chief equity strategist for a major investment bank Topic: Solvency & liquidity
Solvency & liquidity:
usually 200K unemployment claims/week — >2 million this week. 24% Q2 annualised rate of contraction in GDP. In 1958 down 10%, in 1980 down 8%, 4Q08 8.6%
Stimulus approaching 2 trillion, 10%
Lots of concerns about liquidity to survive the next 90 days. Many small businesses will not survive. Larger companies have a better position.
Most companies will have a loss in Q2
No earnings reported until 20 April through May. Lots of pre-announcements and negative commentary to be expected.
Used to think 2450 would be the low, now I think 2000 could be the low.
Very powerful rally after that up to 3000 at the end of the year.
Why am I so optimistic? Money flow from hedge funds, retail, foreign investors. Market was 2 standard deviations above average, now 1.5 stddev below average. All the low points in 4Q2018 were 2.5 stddev below average. We’re not fully cleaned out in the positions, market goes lower before it goes higher.
U-shaped not V-shaped as it’s Q2 and part of Q3.
Speaker: Well-known investor from major bank and hedge funds. Topic: What are the implications for the March Quarter End Rebalancing by Real Money on the Equity Market?
Fresh capital moves slowly, it’s existing capital that moves markets.
Change in value between equity & bonds causes investors to rebalance; it calls for a lot of selling bonds and buying equities.
If you had 50/50 as target, equities have gone down 20%, you have $40 in equities and you want $45, so you need to buy $5 which increases equity holdings by 10%. Similar numbers for allocations up to 70/30 or 30/70
There could be global 40 trillion of assets managed this way to target allocations.
Approaching from another direction, of this 40 trillion, 50% target in equities would mean that a high % of total public market equities are managed this way. 10 trillion are with Vanguard, Black Rock, State Street so that’s ballpark correct. So the rebalancing buy of equities is $2 trillion — that 4% of total global market cap.
Q: Did we see this in 2008 after the sharp equity decline?
A: We ultimately did see some of it. But in the last decade the share of assets managed this way have grown dramatically.
Impossible to say how long this will take. The sell-off has been much more compressed than 2008. Hard to compare.
Q: Your utility function is different when you make money than when you lose money. How does this impact investors psychologically? And if volatility is much higher than we thought, does that mean I have to reduce my equity weighting?
A: Long-term volatility isn’t that much higher based on this; and this significantly increases expected future returns.
Our high-net worth investors have mostly held or increased exposure after a period where they reduced.
Speaker: Hedge-fund investor specialised in fixed income Topic: How are fixed income markets working? Anything crazy in spreads?
Last few markets is unprecedented level of illiquidity. Result of changes of market structure that made dealers less important as market-makers, and others more important. All the other market-makers have disappeared and we’re left with banks who are providing very little liquidity.
Treasury, interest-rate swaps: bid/offer compared to normal, they are 8x wider than usual on average.
Volatility also extraordinary, swinging in both directions, gigantic rate moves.
Policy makers, Fed: the Fed has been all over it. Lots of QE so far but they are going to pump in much more liquidity. They’re spending >$50B/day in the treasuries market and will keep going at that level until it’s fixed.
They’re going to spend $100B in the mortgage market. Maybe yield-curve control. Maybe $100s of billions more. They clearly are signalling that they’ll do anything it takes.
A lot of fear out that about markets being closed. T-Bills are a proxy for cash but if you think the market is closed for even a week then a T-Bill isn’t as good as cash. They need to make it very clear that markets will not close.
Think the Fed is going to succeed in getting treasuries and mortgage markets to trade in an orderly way.
Less clear what will happen in corporate credit. That’s the next rabbit they need to pull out of the hat — we didn’t have a programme there in 2008. How can they backstop that market to prevent companies from going bankrupt?
Speaker: Real estate investor / developer at one of the largest real estate investors in the world. Topic: Senior Living and Other Real Estate Disasters
Seniors are the most impacted by COVID-19, and they are an important tenant.
Our buildings are used to occasionally going onto lockdown in seasonal flu from time to time; we have been on lockdown for some time and have had no cases.
Senior living move-outs is 50% since average length of stay is 2 years. Also there could be high deaths in this population.
Some of the largest operators in this space have high leverage. One company has stock down 80% and had only 1x lease coverage prior to this — it’s severe. Another has 62K units with a corporate guarantee and their stock is down 70%. Sum-of-the-parts is that they are trading at 50% of replacement cost.
Hotels — we own large hotel chains. We shut down a 30+ chain of hotels, we don’t know when it can reopen, and revenue is approaching zero in most hotels. Most companies are announcing layoffs and furloughs. Other hotels depend entirely on large events like conferences and trade shows — when will they again be full? Same thing with hotels that depend on tourism.
Marriott and Hyatt are down 40-50% and they are asset-light.
Las Vegas: all hotels are closed. When will they reopen? Will gambling resume? Many stocks down 60%+
Retail — all sales have dried up other than grocery/pharmacy. Those tenants are begging for rent relief from landlords.
Apartments are holding up the best; people need to live somewhere. People are sheltering in place. Many places banning evictions. Home sales and rentals are at a standstill. NYC: no tours permitted.
Student housing? Don’t know when schools will reopen, and can’t charge students when not there.
Office buildings? Technically not closed but office workers aren’t there; only maintenance etc.
Speaker: Mexican PE investor Topic: What is happening in Mexico?
History — in 2009 Mexico had dress rehearsal with H1N1 in early March that had an early curve that looked like China’s. CDC says 12K people were infected / suspected; other studies say 55K-370K. Around 100 deaths. Mexican governments suffered many of the same issues in the US, like lack of testing. They were very transparent; within six weeks implemented social distancing, closed schools and offices. By beginning of May economic activity was returning to normal on an accelerated basis.
This is much worse of course.
GDP lost around 1% in 2009 with a bad Q1 and Q2.
This time Mexico is 2-3 weeks behind the US. Two weeks ago no economic impact, not even in airlines or cargo/logistics. Last 10 days that has changed, starting to see some sell-off.
Mexico gets hit three ways: oil shock which is fiscally important; tourism is a large part of the economy; and the supply chain that feeds into US industrial production.
Currently economists expect -4% to -2% GDP decline vs +1% a month ago.
Government is copying the early US script. Said two weeks ago, “don’t worry, keep your activities;” last 72 hours they have changed the script and now gov’t employees will not work from Monday. Private sector, especially multinationals, banks and law firms implemented WFH a week ago.
One doctor says they’re treating 45 suspected cases in a single hospital
Speaker: Major PE investor in Africa. Topic: What happens when the virus hits Africa?
Outlook is grim. High urban concentration, weak health care system, 16% of the population but only 1% of health care spending.
Kenya has perhaps 50 ventilators in the whole country and they’re probably the largest spender on health care.
Lots of the playbook won’t work there.
You can’t have everyone shelter in case; they don’t have savings, refrigerators.
50%< 25, very young population, small older population
Fewer people living with chronic conditions because it would have killed them already
Few nursing homes
Some experience with Ebola and other pandemics
Big question for the world is how the climate impacts this. Currently concentrated between 30-50% attitude. Don’t know.
Not much testing in Africa. We were infected early; so far not seeing many deaths. Doesn’t seem to be spreading as quickly as elsewhere.
Big issues because of decline in oil, commodities from China. Tourism is important in many countries.
Many countries were already in recession prior to this.
At minimum pressure on currencies.
Impact on GDP won’t be as big as in the developed market. Less formal supply chains, more informal economy that is more flexible.
Speaker: biotech investor. Topic: What are you hearing from the NIH? What are you hearing from your private company investments?
Lockdown is there to buy time, keep hospitals from overcrowding, build supplies. It absolutely will decrease the number of cases.
Everyone is saying most important thing is: we need testing. Who has it now? Who has had it? Tests will cost $2.50-4 so very cheap.
Once you relax after lockdown, need to use tests to identify and stop breakouts.
The treatments will get better quickly. Will reuse existing treatments and develop new ones.
Vaccine best case: 12-18 months.
Could you test temperature at high-volume areas like subways, offices, grocery stores so that people feel confident?
Think we can get the economy back on track in 2-3 months.
Speaker: Former senior IMF executive Topic: How will the IMF and the other international organizations respond to the crisis?
Worried that there will not be the geopolitical cooperation we need among major parties.
We don’t know how long this will last, lots of uncertainty.
If we start handing out cheques, how do we stop?
We can do demand stimulus, not sure how well this will work.
Q: Intra-EU transfer payments — e.g., Germany to the south? Risk of Italian default? Will the ECB refinance Italy?
I don’t know. Italy: the key thing is the European Stability Mechanism, which was created to deal with this kind of crisis. Can borrow as much as member countries allow. Debt constructed to be as close as possible as a joint/severable liability. If there was a political consensus, they could do it with massive firepower. The question is the political consensus.
Speaker: Senior corporate banker on the credit side Topic: How are the banks doing in these crazy times and why is it different this time?
This week, unprecedented draws — one big company drew > $15B!
Covers the whole credit spectrum.
Lots of asks for new credit. There are the impacted industries like airlines and cruise lines; many clients who have lost faith in the commercial market including many A1 issuers; and people who have realised that even if they don’t have an immediate liquidity need, they want more insurance.
This breaks the mechanism of the old revolver market, which was priced so low because no one ever drew it down.
What’s different? No flight to quality. In the past, the best companies didn’t feel the pain; today, they all are, including mega-caps.
The large-caps enjoyed low cost of commercial paper; now they have to revisit entire capital structure.
Investment-grade market usually recovers quite quickly. We’ve done deals > $63B in total supply; but even when you over 60-90bps better pricing, the bonds are not selling as we’d expect.
Huge new issue calendar coming up.
The old credit desk culture doesn’t exist any more.
The energy collapse is a double whammy.
Banks are incredibly healthy.
Expecting big bifurcation between the best and the rest. Big, healthy will be at the front of the queue. Smaller, market cap collapsed, will not.
European banks: some of the market caps for big wealth managers vs some of the commitments they’re still writing worry me.
CARES act — what does that mean for airlines and cruises? Where do the banks sit vs government?
Big tech company asked: what do we do about payables? No one knows what the credit quality is. Can the Fed provide a mechanism to allow companies to extend AP days?
Q: Bernanke said that in the Great Depression, banks would not lend to small enterprises. Who is going to lend to the small players in this situation?
We’re providing support up and down the size spectrum on the credit side.
Lots of fintech people since we can do this better than we can — but in a credit we’re seeing that tested.
Speaker: Senior investment banker. Topic: What are you hearing from your corporate clients and access to markets?
Many of our clients are PE/VC owned.
Corporate portfolio managers, public credit/equity funds, PE funds: if they are already invested, they are focused on damage assessment.
Lots of corporates, the stronger ones, are thinking about how we take advantage of the position.
No one we have heard is having trouble drawing down revolver. Everyone is doing it except some high-yield borrowers who have maintenance covenants, and would have new targets they’d have to it. They are the only exception.
Public equities: 10 days ago people were doing converts; that’s stopped last week.
Lots of capital out there to be put to work. Question is how much goes to new names, how much has to go to existing portfolio.
We’ve heard of 20 structured credit funds who are preparing to help be a solution but they will charge a big price.
Speaker: Fintech founder specialised in Big Data Topic: What do you see from big data?
What data are we looking at?
Bloomberg survey: 10 days ago forecast was +1.4% for the quarter and the year. Those clearly no longer apply. Very rapid change.
As of March 14th: how often do people visit a retail establishment based on mobile phones? Down 13-15% nation-wide, seasonally adjusted. Concentrated in the northeast and the west coast which tracks the outbreaks.
Even around New Rochelle, when containment started and before NYC controls, we saw aggregate activity in NYC down. Friday-Saturday was quite active, but retail visits overall were down even including more grocery shopping.
Putting all data sources together, including job offers, even some data that we use for China as well as the US (e.g., NASA air pollution) — all are very negative. We can see the early effects, very quickly.
Think we still have more downside surprises to come.
Looking at China:
January 23rd was t=0 on lockdown.
Sharp decline in January and February.
Looking at things like mobility, activity started to come back somewhat exactly 1 month after the quarantine started.
Now, 2 months from t=0, in the first 3 weeks of March, it’s about flat vs February. This is consistent with “U shape” — straight line down, across, and then going back up somewhat but not the full amount.
Hard to know what Chinese GDP will be. Maybe 10% annualised decline.
We aren’t halfway back to where we were. Some parts of the economy are recovering more quickly than others — e.g., factory activity ramped sooner than people-to-people. That part of the economy is still fairly week.
You can go out again; e.g., Shanghai is like 60-70% of normal in terms of street traffic.
Second-round infections from returnees may cause some back-and-forth.
Speaker: Professor of Economics Topic: Economic impact
The VIX / volatility, and the S&P 500
The VIX is around 80% right now. This goes up a lot as correlations increase, and we saw a huge spike in correlation — everything went down together. No diversification, everything is based on average volatility.
Actual composition of the index changes dramatically as well as the actual asset betas take over.
The level of the VIX isn’t something I pay attention to. I look at the distribution of strike prices to understand the tail of losses and gains might be, what the market is saying right now.
Even though the Sharpe ratio goes up as the expected gain over the expected loss goes up, under most asset allocation models, volatility / risk dominates in the short-run.
With a big increase in volatility, that should mute the amount of risky assets that individuals/asset allocators will tend to hold.
So the expected tail losses in the market — the worst 10%, integrating across that — is around 51% right now; normally that’s 10%. That suggests a 51% drawdown possibility in the market, similar to 2008. The ratio of upside to downside, in log space, is less than 1 at the current moment.
5-7 standard deviations to the downside.
This looks like November 2008 in terms of high volatility and high downside.
The change in these measures is interesting. If we use market prices as a guesstimate, we have to ask about the change in the Sharpe ratio and what is the change in vol. Lately, the change is all negative. Upside/downside has fallen and expected tail loses have increased. You might get a V-shape but the options market is not saying that.
Q: When vol goes up, small stocks often outperform large stocks. Is that priced in? A: right now it doesn’t matter because the correlations are so high; you get no diversification. Small stocks have even higher beta so more downside. The options data suggests smaller companies are more exposed.
Speaker: Former treasurer of a major investment bank Topic: What learning about liquidity can we apply from previous financial crises?
Real interest rates of zero were penalising cash and pushed capital up the risk curve which meant it wasn’t there to cushion an event like this.
Market participants are realising that what they thought was cash isn’t cash, and they need more than they thought, so they are selling less liquid assets.
What is cash? Currency, demand deposits if banks remain open, Fed.
Not: ETFs, mutual funds backed by corporate bonds or equities.
Removal of historical shock absorbers makes this worse.
Banks and dealer balance sheets were taken off-line by Dodd -Frank.
Challenge for the market: despite intervention by the Fed, the financial system is not equipped to intervene. Banks do not have the staff and skills to prudently assets risk etc.
Fed is the only option. Must step in, expand remit, deploy balance sheet. They will have to go further, maybe even all the way to equities.
Amount needed? $22 trillion global GDP, $15B/day needed based on declines so far. At $2 trillion that’s 133 days.
Speaker: PE investor with connections to defense Topic: What are you hearing from your portfolio companies, especially defense?
I’m a middle-market LBO investor focused on defence, government services, energy.
Little impact on my portfolio so far. Government is being accommodative, letting people work from home where they have never done so. Even secured employees at the Air Force. Unpredecented.
State & local is almost entirely remotely. NYC: strangely net positive. Each company spending what they need to, worry about it in the future, no fiscal limits.
We had a health-care company (GP offices) — we closed a big investment on Tuesday, even in the middle of the crisis.
1/3 of my friends’ portfolio companies are strongly exposed to impacted sectors. People are in shock, none of us expected this. These are small caps, we don’t have covenant light deals. They don’t have the balance sheet to absorb this. In normal cases the lenders would move in and take the companies over when the default; will the accommodate in this case?
Speaker: A Federal District Judge Topic: What will happen with regard to the Federal Judiciary? Will the Judiciary give incredible latitude to Federal, State and Local Officials?
Much of what we do is incompatible with social distancing!
Most urban districts have taken drastic measures, closed their doors for 3 weeks. No trials, no sentencing.
The grand jury met last week, probably won’t this week.
Have extended all deadlines by 21 days to start with.
Seeing similar actions across the country.
Federal government have limited powers as enumerated in the Constitution. Local/State governments have more general police powers to take measures necessary to accommodate public safety.
Round-table discussion among speakers:
Q: What will Fed do, buy corporates?
A1: They will want maximum flexibility to put out fires.
A2: I asked a Fed governor. This tool is not in their toolbox and requires legislative approval.
A3: Bloomberg just posted: Congress likely to pass legislation Monday clearing the way for more action from the Fed. Might allow them to purchase corporate bonds.
Q: Will Fed buy individual companies?
A: Fed would just be a counterpart to a bank.
Q: The Judge was suggesting that there were limitations on the ability to use the military in civil disturbances. If we did have riots or other disturbances, as unlikely as that seems right now, what would the political limitations be on using federal troops?
A: Depends on circumstances. A spike in gun sales suggests some people worried about stability; but you’d have to get to extreme scenarios to consider anything like that. A: Can’t imagine the election being cancelled.
Q: What are the next triggers for politicians? Will it be when hospitals get overrun and we see it on TV? What will the government do when they are overrun?
A: The real question is the extent to which our system can redirect resources from one part of the country to another. Will doctors come from (eg) Texas to NY to help out? Difficult. Everyone understands that we’re in this together, a medical problem not a political problem.
Q: Can options tell us about what the expectations are for the duration of impact?
A: No, they are most liquid only 1-2 months out, can’t predict further out with any confidence.
Q: What happens with e-commerce? Are there capacity constraints, can they scale?
A: Big acceleration of mix changes that were already under way — from physical to online — in many markets. Ecommerce, new releases move from movie theatres to Netflix, restaurant meals to DoorDash and Deliveroo. On capacity, limitations are on deliver and on supply chain, but on the former no problem because can scale to peak capacity as long as they can get drivers; supply chain is the ultimate limitation — can they get the goods?.
Q: How will retail investors behave vs institutional?
A: Investors who think over the very long-term (e.g., retirement, estate planning) don’t change plans much because of high volatility. The fact that equities can go down 50% is not a big surprise to long-term equity investors.
A: Do investors really know that? Does the actuality of it happening shock them? In theory with a high VIX we could touch 1500 S&P.
A: Average daily move in March has been 6% per day! 10% moves some days. Most clients think it goes lower. (But why don’t they sell if they think it’s going lower?). They have been selling.
Q: What are you optimistic about?
A: G20 meeting this week might agree special drawing rights for the IMF to provide liquidity for low-income countries. They might agree to accelerate funding to IMF. In EU, might agree to allow the European Stability Mechanism to increase its war chest.
A: We will get fast innovation on testing.
A: Asian countries — they were not that early, and some were not that aggressive. Each country that did strong controls brought it under control, both infections and death rate.
A: Germans, French, Washington — a lot of the shackles and constraints on policy are being thrown away.
A: We’re all adapting really quickly; e.g., working from home is working.
A: Expected returns in the equity markets are much higher than they were relative to the risk-free rate of return.
A: We can reallocate resources to solve problems.
Q: To the investors, are you putting risk on or risk off? A: This is temporary, economy will come back together. But we’ve never shut 1/3 of the economy down before for 3+ months. Hard to guess what the interruptions will be and how they will be solved. Logistics are much more advanced today, science will move forward. It’s a liquidity problem short term.
Q: What about food distribution? Will we have food shortages? A: No, it’ll be fine.
Speaker: Christopher North
Topic: How should we think about COVID-19?
I’m going to give you my sense of what our Bayesian priors should be in this situation. Let’s consider three outcomes: an expected case, a better case, and a worse case.
1. Expected case.
In many countries (including most of Europe, UK, US) we face 6-12 weeks of severe and potentially escalating control measures. After an initial peak that stresses emergency services, this works to bring the rate of spread down and allows health services time to build capacity.
Countries like China, South Korea, Taiwan, Singapore, and Hong Kong who got it under control early may be successful at keeping most new cases out with aggressive track/trace/quarantine and extremely tight border controls.
But many countries won’t be able to implement or sustain sufficient control measures; in particular, poorer countries risk runaway rates of infection and becoming reservoirs for the virus.
For those reasons, the virus persists beyond the first wave, and even countries that successfully managed the first wave face a further 6-15 month period with an acute tension between relaxing control measures to permit resumption of economic activity, vs maintaining tight controls to keep the spread at a slow rate.
In my expected case we may well get helpful therapeutics soon, and will certainly ramp testing and health care capacity quickly. But it’s unlikely we get an effective vaccine anytime soon — certainly not this year and quite likely not next year either — since an implausible number of things would have to go right.
.Quite plausible that we get to 25-40% global attack rates with infection fatality ratios (IFRs, which unlike CFRs are based on the actual rate of infection, not the confirmed rate) of 0.5-1.0%.
So that’s the expected case which since the publication of Neil Ferguson’s Imperial College paper seem to have become the received wisdom that many governments are now acting against.
2. What are the main ways it could turn out to be better than the expected case? Here are the ones that seem most salient to me.
Several Asian countries’ measures have worked really well. The effective reproductive number dropped from around 3 to around 1 within a week of control measures in Wuhan, and then down to 0.3 within 5-6 weeks. If enough countries are willing to take harsh medicine quickly and follow it up with strict track, trace & quarantine, and strict border controls, we might see more substantial easing of controls and resumption of economic activity possible earlier than three months from now. This is the most promising path in my mind to a better outcome.
We could get really lucky with the virus through mutation or weather. I wouldn’t count on that.
There’s a fringe theory that there may be several orders of magnitude more people infected with very mild cases than we think, which could mean very low IFRs and faster herd immunity. .While we think that underreporting of 10-30x is common, and IFRs may well be lower as a result, no epidemiologist I know of thinks it could be 1000x underreporting. . We won’t know until we get a serological test.
A different reason for optimism is that some really smart people think that we will get through this better and faster than the expected case. Bill Gates and Elon Musk are two prominent examples, but there are a fair number of epidemiologists who disagree with Ferguson.
Finally, as we’ve learned from domains ranging from Malthus to shale oil , it’s never a good idea to bet against human ingenuity.
3.How could it turn out to be worse than the expected case?
The big risk is that many countries’ measures are insufficient. European countries are not yet meaningfully bending the curve despite severe control measures, and US, UK and other countries are not implementing strict controls fast enough.. Yes, we should expect something like 6-8 days latency between control measures and a slowing of confirmed cases given time lag from infections to symptoms to testing to reporting. Yes, countries are ramping testing so confirmed cases may be growing faster than actual new cases. But we should be scared until we see many countries bending that curve.
It could turn out that recovered patients have limited or temporary immunity; while that is unlikely for a number of reasons, until we have a serological test and more time, we don’t know.
Many health care systems are not ramping fast enough, so the death toll could be higher than it would have been.
We could relax control measures too quickly and, like the Spanish flu, have later waves that are much deadlier than the first.
We should expect an RNA virus to mutate easily, and it could mutate to become deadlier.
My probabilities are evolving quickly. Even four days ago I had the expected case at 60%, better case at 20%, and worse case at 20%. Today, I’m probably closer to expected 50%, better case 30%, and worse case 20%.
I’ve been overwhelmed (in the positive sense) by the number of thoughtful e-mails I’ve received with ideas, resources, good articles/reports, suggestions and questions. I can’t do justice to all of this in a single update.
Yesterday I posed the question: “What are the arguments against the current ‘expected case;'” and in particular, what are the best arguments that the human/health and economic impact may be less severe than markets and the media are currently suggesting?
I’m still digesting the input I’ve received in the last two days and need another day or two to organise my thoughts on this.
In the meantime, for today, I’ll share a hodgepodge of interesting materials from a variety of sources. In the spirit of engaging with counterarguments, they’ll be lopsided towards those who argue that things aren’t as bad as they look.
A strongly worded (would you expect anything else from Taleb?), but largely polemical (rather than introducing new evidence), piece arguing that lockdown followed by track & trace followed by tight border controls can be effective in controlling COVID-19; basically, why Ferguson is wrong.
Good Judgement’s questions all have an evaluation date of 31 March 2021, one year from now. The algorithmically aggregated probabilities [corrected from my early misstatement that these %s reprented the share of forecasters] are as follows:
Reported cases – 50%: Between 53 and 530 million cases reported by then – 14%: worse than that — implying less than 14% probability of global attack rate will be above 6%! – 36%: fewer than that
Deaths: 50%: 800K-8M WW – Only 13% predict 8M+.
So in summary, forecasting tournaments don’t think it’s as bad as some are predicting from the perspective of confirmed cases and total deaths. (Still, 800K-8M deaths worldwide is a lot, compared to around 500K for flu.)
In fact, if you listen closely, Ackman is arguing something close to or even less severe than the current “expected case” / received wisdom: that intense intervention, early, can reduce both the health/human and economic impact.
This is a great example of how looking a mix/proportions among cohorts, rather than ratios within cohorts, is misleading. By the chart above, 39% of total severe (hospitalized) cases were patients 54 and under. So that means it doesn’t disproportionately cause severe cases in the elderly, right? Wrong.
If you look at the table below instead, it’s clear that on an in-cohort percentage basis, the disease disproportionately causes serious cases in older cohorts. No cases under 20 required ICU admission, and only 2-4.2% of cases 20-44.
(Why does such a test matter? Current PCR testing based on saliva swabs can only detect the current presence of the virus; it can’t tell us who has had COVID-19 and recovered, nor whether an individual has antibodies that could confer resistance.)
SEROLOGICAL TESTING As diagnostic testing for SARS-CoV-2 infection ramps up in the United States, many questions remain regarding the number of cases and asymptomatic infections that are going undetected, both in the United States and around the world. The PCR tests currently used to diagnose COVID-19 patients are effective at identifying active infections by detecting virus currently present in the specimens, but they are not able to determine whether an individual was previously infected after the patient has recovered. For this, serological tests are needed. Serological tests identify the presence of antibodies, which were generated as a result of prior infection. A study published on March 18 (pre-print) describes the development and initial testing of an ELISA serological test by researchers at the Icahn School of Medicine at Mount Sinai, in collaboration with colleagues from multiple international institutions. Based on tests using human samples from both uninfected individuals and recovered COVID-19 patients, their preliminary findings indicate that the new serological test can effectively detect the target antibodies. Additionally, the researchers note that the test “do[es] not require handling of infectious virus” and that production is “amenable to scaling,” which could allow for rapid production in order to conduct larger population surveys. Our World in Data switches from WHO data to EDC dataI’ve mentioned the outstanding Our World in Data website many times before. I was fascinated to see that they have stopped using WHO data and have moved to the EDC data set instead. I’ve also been very frustrated with the WHO data for reasons ranging from high latency (lag in reporting), to errors in reporting, to inconsistency with presumably authoritative sources. https://ourworldindata.org/coronavirus
Sobering interview with a doctor who helped defeat smallpox.
An outstanding interview with Larry Brilliant. What a bio! “Brilliant, a technology patent holder, has been the CEO of public companies and venture backed start-ups. He was the inaugural Executive Director of Google.org, […] the first CEO of Skoll Global Threats Fund […] Brilliant currently serves as the Chairman of the Board of Ending Pandemics, and is also on the boards of the Skoll Foundation, Salesforce.org, The Seva Foundation, and Dharma Platform.”
I’ve excerpted a few of the most interesting sections below.
Since it’s novel, we’re still learning about it. Do you believe that if someone gets it and recovers, that person thereafter has immunity?
So I don’t see anything in this virus, even though it’s novel, [that contradicts that]. There are cases where people think that they’ve gotten it again, [but] that’s more likely to be a test failure than it is an actual reinfection. But there’s going to be tens of millions of us or hundreds of millions of us or more who will get this virus before it’s all over, and with large numbers like that, almost anything where you ask “Does this happen?” can happen.
Is this the worst outbreak you’ve ever seen?
It’s the most dangerous pandemic in our lifetime.
By slowing it down or flattening it, we’re not going to decrease the total number of cases, we’re going to postpone many cases, until we get a vaccine—which we will, because there’s nothing in the virology of this vaccine that makes me frightened that we won’t get a vaccine in 12 to 18 months. Eventually, we will get to the epidemiologist gold ring. […] That means, A, a large enough quantity of us have caught the disease and become immune. And B, we have a vaccine. The combination of A plus B is enough to create herd immunity, which is around 70 or 80 percent.
Now that we’ve missed the opportunity for early testing, is it too late for testing to make a difference?
Absolutely not. Tests would make a measurable difference. We should be doing a stochastic process random probability sample of the country to find out where the hell the virus really is. Because we don’t know. Maybe Mississippi is reporting no cases because it’s not looking. How would they know? Zimbabwe reports zero cases because they don’t have testing capability, not because they don’t have the virus. We need something that looks like a home pregnancy test, that you can do at home.
Are you scared?
I’m in the age group that has a one in seven mortality rate if I get it. If you’re not worried, you’re not paying attention. But I’m not scared. I firmly believe that the steps that we’re taking will extend the time that it takes for the virus to make the rounds. I think that, in turn, will increase the likelihood that we will have a vaccine or we will have a prophylactic antiviral in time to cut off, reduce, or truncate the spread. Everybody needs to remember: This is not a zombie apocalypse. It’s not a mass extinction event.
Should we be wearing masks?
The N95 mask itself is extremely wonderful. The pores in the mask are three microns wide. The virus is one micron wide. So you get people who say, well, it’s not going to work. But you try having three big, huge football players who are rushing for lunch through a door at lunchtime—they’re not going to get through. In the latest data I saw, the mask provided 5x protection. That’s really good. But we have to keep the hospitals going and we have to keep the health professionals able to come to work and be safe. So masks should go where they’re needed the most: in taking care of patients.
How will we know when we’re through this?
The world is not going to begin to look normal until three things have happened. One, we figure out whether the distribution of this virus looks like an iceberg, which is one-seventh above the water, or a pyramid, where we see everything. If we’re only seeing right now one-seventh of the actual disease because we’re not testing enough, and we’re just blind to it, then we’re in a world of hurt. Two, we have a treatment that works, a vaccine or antiviral. And three, maybe most important, we begin to see large numbers of people—in particular nurses, home health care providers, doctors, policemen, firemen, and teachers who have had the disease—are immune, and we have tested them to know that they are not infectious any longer. And we have a system that identifies them, either a concert wristband or a card with their photograph and some kind of a stamp on it. Then we can be comfortable sending our children back to school, because we know the teacher is not infectious.
From a major biotech investor A professional investor shared this on the condition that the source not be divulged (the summary is mine); as you’ll see, it’s firmly in the camp of “this is manageable, not the zombie apocolypse”
One specialist investor in biotech has shared with their partners that they think drug companies are at extremely attractive valuations. They make several interesting points. First, they think the impacts to drug companies of the crisis (both economics and health) are small — mostly delays in clinical trials and some temporary disruption to sales . Secondly, large biotech companies have experienced one of the largest drops in valuation in history. Third, many biotech companies have a strong cash position and therefore the ability to not only weather, but to invest through the crisis. Fourth, they think that after a period of strong control measures, following China and South Korea’s playbook, the US and other countries will be able to gradually relax controls, implementing “track & trace” with quarantine.
Apologies for no update yesterday. We’re discovering that home schooling three children and keeping the house running is a 12-hour/day job, and are having to manage work and a bit of exercise in the early and late hours. Hopefully we’ll get more efficient over time but updates might not be daily.
Today’s update is relatively brief.
Many friends have sent fascinating papers, articles, resources, questions, and observations. Thank you and keep them coming!
Question: What’s are the arguments against the current “expected case”?
If you’ve been following these updates for a while, you will likely have concluded (mostly correctly) that I subscribe to what is emerging as a kind of received wisdom “expected case.” (I’ll go on to describe this in a moment.)
But given the speed with which many policymakers and governments have swung from relative inaction to adopting a set of increasingly stringent control measures–no doubt from a mixture of expert advice and political considerations–and given the extreme economic and human impact of these measures, it’s worth asking (as several friends have been asking) the following question: What are the best arguments in the other direction — that argue that the impact will be less severe than we think; that less stringent control measures are required or that control measures are required for a shorter period of time; etc.?
In the meantime, here is a high-level, hand-waving summary of how I’ve tried parse out my own beliefs and the relatively likelihood of each.
Even as I write this, my thoughts are evolving quickly on this; what I’ve written here represents what I thought as recently as 48 hours ago; but as I engage further with the best counterarguments, I’ll likely end up reducing p(expected case) and increasing p(less bad case).
I should also add a few methodological points as a prefix.
I’m strongly influenced by Philip Tetlock’s work on forecasting, best summarized in his outstanding book Superforecasting. Most relevant to my thinking here is the puzzle of what base rates to use in such a rare situation as COVID-19, where we don’t know the extent to which past examples are a good guide as to what to expect.
Finally, you’ll note that I group together a number of very different topics within each case: disease progression, health impact, government reaction, economic impact, etc. There’s a different approach one can take to forecasting, which is to string together conditional probabilities; but I think the interaction of these elements is so complex and non-linear, with lots of unknowns and lots of feedback loops, that this just gives false precision.So with that, I describe three cases with my subjective probabilities.
1. p=60% Expected case. In many countries (including most of Europe, UK, US) we face 4-12 weeks of severe and (in some countries) escalating control measures. After an initial peak that stresses emergency services, this begins to bring the rate of infection down and allows health services time to build capacity. The virus persists after that and we face a 6-12 month further tension between relaxing control measures and seeing higher infection rates. Some countries can’t or won’t sustain long-term controls and see second and perhaps further waves; so we may see-saw back and forth between significant and more relaxed controled measures. The economic impact is severe for several months and there may be further waves of panic that drive markets lower. We reach 25-40% attack rates with infection fatality ratios (IFRs, which unlike CFRs are based on the actual rate of infection, not the confirmed rate) of 0.5-1.0%. Then it mitigates somewhat, and 3-9 months from now we begin to return to a “new normal” where most economic activity returns, albeit with some sectors permanently impacted. In the meantime, many businesses may have gone bankrupt, many individuals are out of work, some businesses are nationalized and/or bailed out, etc. While the world’s productive capability is not significantly diminished in the medium-term, the massive fiscal stimulus and expansion of government balance sheets may store up inflation (possibly asset price instead of monetary), and will cause huge distortions of its own over time.
2.p=20% Less bad case.
Whether because control measures prove effective and allow countries to move from mitigation back to containment (tracking & tracing all contacts, quarantine); warmer weather mitigates the virus’s propagation; a vaccine and/or therapeutics are announced that calm people down; etc; we begin recovering within 3 months. Health care systems rapidly build capacity to keep fatality rates low. While the virus persists, people accept that the risk for many individuals is not that much higher for most people than seasonal flu (for which many don’t choose to get vaccinated) or other normal but risky activities, and we live with some degree of behavioral changes (hand washing) and a degree of social distancing as a new normal. Final attack rates (% of population infected) come in well below the 25-80% ranges widely discussed, and/or IFRs are below 0.5%. In this scenario, we may have already seen the market bottom and despite a bad recession over two quarters, economies rapidly recover. No long-term damage is done to the world’s productive capacity.
3. p=20% Worse case. Like the expected case but worse. It could be that control measures are unsustainable or ineffective and we see multiple waves of the epidemic; attack rates are more in the 50-80% range than the 25-40% range; IFRs are greater than 1%. Perhaps there is mutation in the virus to become deadlier; or we see annual recurrence; or poor countries have extreme outbreaks and become long-term reservoirs of the virus. Global tensions linked to COVID-19 drive geopolitical conflict, trade wars, currency wars; etc.
Bill Gate’s AMA on Reddit
Speaking of arguments against the emerging “expected case”, Bill Gates participated in an “Ask Me Anything” session on Reddit yesterday, 19 March 2020. I thought it was fascinating (thank you, William, for sharing it!) and excerpt several of the most interesting sections below.
Overall, you’ll see that Bill is more optimistic than Neil Ferguson (in fact, he directly attacks the Imperial College study) and thinks extreme control measures are only required temporarily. That would be great news if true.
Besides the summary below, I also recommend checking out the Gate’s Foundation-funded Institute for Disease Modelling website, which has some highly curated resources and sources. I was particularly interested in their summary of our best understanding of key parameters like R0, Serial Interval, CFR, etc.; and the sources of that understanding.
The format below is that all bullet points are direct quotes from BIll’s responses; they are preceded either by the exact question put to him, or grouped under a thematic heading.
Comments on Imperial College / Ferguson Paper:
Fortunately it appears the parameters used in that model were too negative. The experience in China is the most critical data we have. They did their “shut down” and were able to reduce the number of cases. They are testing widely so they see rebounds immediately and so far there have not been a lot. They avoided widespread infection. The Imperial model does not match this experience. Models are only as good as the assumptions put into them. People are working on models that match what we are seeing more closely and they will become a key tool. A group called Institute for Disease Modeling that I fund is one of the groups working with others on this.
(What about the NYTimes report that just came leaking a government document saying this will be 18 months with “multiple waves”?). There are many models to look at what will happen. That article is based on a set of assumptions derived from Influenza and it doesn’t match what has happened in China or even South Korea. So we need to be humble about what we know but it does appear that social distancing with testing can get the cases down to low levels.
The goal is to keep the number infected to a small percentage. In China less than .01% of the population was infected because of the measures they took. Most rich countries should be able to achieve a low level of infections. Some developing countries will not be able to do that.
Is there any chance that the 18 month timeline for development of a vaccine can be shortened, and by how much?
This is a great question. There are over 6 different efforts going on to make a vaccine. Some use a new approach called RNA which is unproven. We will have to build lots of manufacturing for the different approaches knowing that some of them will not work. We will need literally billions of vaccines to protect the world. Vaccines require testing to make sure they are safe and effective. Some vaccines like the flu don’t for the elderly.
The first vaccines we get will go to health care workers and critical workers. This could happen before 18 months if everything goes well but we and Fauci and others are being careful not to promise this when we are not sure. The work is going at full speed.
A therapeutic could be available well before a vaccine. Ideally this would reduce the number of people who need intensive care including respirators. The Foundation has organized a Therapeutics Accelerator to look at all the most promising ideas and bring all the capabilities of industry into play. So I am hopeful something will come out of this. It could be an anti-viral or antibodies or something else.
(Thoughts on chloroquine/hydroxychloroquine?). There are a lot of therapeutic drugs being examined. This is one of many but it is not proven. If it works we will need to make sure the finite supplies are held for the patients who need it most. We have a study going on to figure this out. We also have a screening effort to look at all the ideas for Therapeutics because the number being proposed is very large and only the most promising should be tried in patients. China was testing some things but now they have so few cases that that testing needs to move to other locations.
What do you think of the current approach the Netherlands is currently taking to combat this virus? They are not going to a full lockdown but rather try to spread it controllably in order to work towards ‘herd immunity’.
The only model that is known to work is a serious social distancing effort (“shut down”). If you don’t do this then the disease will spread to a high percentage of the population and your hospitals will be overloaded with cases. So this should be avoided despite the problems caused by the “shut down”. If a country doesn’t control its cases then other countries will prevent anyone going into or coming out of that country. I think the Netherlands will end up doing what other countries are doing.
The current phase has a lot of the cases in rich countries. With the right actions including the testing and social distancing (which I call “shut down”) within 2-3 months the rich countries should have avoided high levels of infection. I worry about all the economic damage but even worse will be how this will affect the developing countries who cannot do the social distancing the same way as rich countries and whose hospital capacity is much lower.
(What do you think about China’s approach?). After January 23 when they realized how serious it was they did strong social isolation which made a huge difference. Of course that isolation created a lot of difficulties for the people involved but they were able to stop the case spread. Other countries will do it somewhat differently but a combination of testing and social isolation clearly works and that is all we have until we get a vaccine.
How long will this go on?
This will vary a lot by country. China is seeing very few cases now because their testing and “shut down” was very effective. If a country does a good job with testing and “shut down” then within 6-10 weeks they should see very few cases and be able to open back up.
I think people in the US will be able to largely isolate for 2-3 months. If they can access testing including a home test kit then they will understand who is infected. I keep saying how important the testing piece is.
Won’t a rebound happen after the shutdown ends?
It depends on how you deal with people coming in from other countries and how strong the testing effort was. So far in China the amount of rebound being seen is very low. They are controlling people coming into the country very tightly. Hong Kong, Taiwan and Singapore have all done a good job on this. If we do it right the rebounds should be fairly small in numbers.
When will this all end?
To bring it to small numbers globally we need a vaccine. Many rich countries will be able to keep the number of cases small (including the US) if they do the right things but developing countries will find it very difficult to stop the spread so a vaccine is critical. A group called GAVI helps buy vaccines for developing countries and they will play a key role once we have a vaccine being made in volume.
Do you believe the news coming out of China though? It’s hard to at this point
China is doing a lot of testing. South Korea is also doing a good job of testing. Once China got serious in January they have been quite open about their cases so yes the good news is they are seeing very few infections at this point. The US needs to get its testing system organized so we see what is going on.
How is the economy likely to recover after all if this in your opinion
Yes eventually. The economic impact of the “shut down” will be large but if it is done well (including the testing piece which I keep mentioning) eventually we can open back up.
[I]t feels like our testing has not increased. Our number of confirmed cases are starting to lag behind other states. What do you think gives? Effective social distancing or lack of testing?
The testing in the US is not organized yet. In the next few weeks I hope the Government fixes this by having a website you can go to to find out about home testing and kiosks. Things are a bit confused on this right now. In Seattle the U of W is providing thousands of tests per day but no one is connected to a national tracking system.Whenever there is a positive test it should be seen to understand where the disease is and whether we need to strengthen the social distancing. South Korea did a great job on this including digital contact tracing.
Two other articulate points of view
Both of these are consistent with “expected case,” though neither is alarmist. Michael Lin: “This is not the zombie apocalypse.”
I created a WordPress blog in late March 2020. Prior to that, I sent out a newsletter via MailChimp but did not have a site. The links below are to the MailChimp newsletters I haven’t yet imported to this blog.