LearningMachine

The following is submitted for analytical purposes and may have investment implications but political aspects have been sterilized, to the extent possible. FWIW, I think Covid-19 is likely to become endemic, within a year or two, but it will likely be an immaterial kind of endemicity. Below are the key factors considered as well as references used. -Availability of effective vaccines (% and strength of response, duration of protection) This looks promising but longer term duration is still unknown. How many in the population (at-risk and as potential spreaders) will take them is a limiting factor if vaccination rates don’t reach certain levels. Also, it’s still unclear if older and at-risk cohorts will be able to build and maintain sufficient immunity levels (available, on demand for antibodies or from “memory” cells). The typical FDA agreement includes to complete follow-up studies for safety and efficacy for at least two years. -Virus characteristics Many respiratory viruses are endemic, including the four well known common cold (beta family) coronaviruses. Longer term studies point to a very real risk of endemicity. More virulent CVs (SARS and MERS) seem to be associated with longer and more robust immunity. Covid-19 is kind of hybrid so common sense would indicate that endemicity risk is real. CVs tend to mutate less than influenza for instance (because of a very unusual and fascinating proof-reading mechanism during replication and specific to this class of RNA viruses) but this is both good and bad. The good part indicates that covid-19 will tend to be slow to genetically escape existing immunity but the bad part indicates that when a new genomic form with new characteristics is reached, this stable form may bear less favorable characteristics for the host (including for immunity) and may persist longer. Whatever residual immunity (natural or vaccine-related), reinfections are likely to be less severe which opens the eventual possibility of eradication to the same degree that it opens the door to low grade endemicity. -Population and environmental characteristics With the virus impact becoming less significant and with less public awareness or media attention, non-pharmacological measures will tend to be abandoned, not enough to cause major and widespread outbreaks over time but enough to allow endemicity in many populations. The virus has become truly global and the world has grown smaller. Seasonal and various environmental factors (for most areas of the world) may become a feature for Covid-19, following what other common cold coronaviruses have become over time. ----- So, on a weighted basis, it’s likely that Covid-19 reservoirs will persist over time but are unlikely to trigger material events on a global basis. For references, I was lucky to be able to listen to a very strong presentation 8 days ago by various ‘experts’. One of the presentations dealt specifically with immunity, reinfection risks and vaccinations. She covered, in a few minutes, most of the relevant research, some of which is mentioned in the CDC link found below (pages 12-22 for post infection immunity, pages 23-32 for reinfection risk). The study mentioned on page 28 is somewhat interesting (design, implications, despite limitations). Also found below is a link to a site monitoring progress on the vaccine front. The amount of human capital involved is mind boggling. The people involved have obvious financial incentives but there is an unusual amount of drive, especially in smaller operations. https://www.cdc.gov/vaccines/acip/meetings/downloads/slides-2020-10/COVID-Wallace.pdf https://covid19.trackvaccines.org/vaccines/ Thanks Cigarbutt, this is super-helpful. If the covid-19 reservoirs persist over time, even if it doesn't cause a material event on a global basis, would it impact behavior of some percentage of the population, e.g. not travel as much as they did pre-covid, not go to places with lots of people as much, etc.? Maybe the behavior change will depend on the perceived probability of risk, and how readily available are the alternatives to taking risk, i.e. working from home, ordering everything online, etc.? If the perceived risk is as high as it was (1) getting on 737 Max right after two crashes, maybe some would pick an alternative if available but if it as low as (2) driving a car, maybe people will accept. Wondering where on the spectrum between #1 and #2, we'll end up landing. The point on spectrum will probably change over time as well depending on how many covid-19 infections we hear about over time.

Thanks gfp. You are right. I forgot that the 10-Q did say "Approximately 70% of the aggregate fair value was concentrated in four companies (American Express Company – $15.2 billion; Apple Inc. – $111.7 billion; Bank of America Corporation – $24.9 billion and The Coca-Cola Company – $19.7 billion)." Given BAC closed at $24.09 on Sep 30, BRK owned about 1.03 billion BAC shares as of then. There is no confidentiality statement in 10-Q either.

Thanks CassiusKing1 for starting this :-). I think almost 0% probability that he is buying TSLA :-). I was puzzled when BAC was trading below his earlier purchases for many weeks, but there were no SEC filings showing up. This might explain the mystery, i.e. the statement in the latest 13F that "Confidential information has been omitted from the public Form 13F report and filed separately with the U.S. Securities and Exchange Commission." So, my guess is BAC. It is possible it is not that because banks are riskier than other companies for concentration, especially if the probability of Covid becoming endemic materializes. If not that, then maybe something in the energy sector.

I can understand your point of view better now. I do think working from home more (not 100%, but >0%) will be permanent now. Besides preventing the spread of diseases, there are many benefits that people have now recognized and will convince the companies to allow people to work from home more. I'm invested in a number of enterprise software/digitization companies for this reason. Regarding the transmission rate, I think that R0 for the typical flu is much lower because the larger portion of the population has already built immunity and some of them are vaccinated. If the same virus was novel like Covid-19, I'm pretty sure the R0 would be higher, although not high as Covid-19. My understanding is that Covid-19 coronavirus strains are more contagious than flu coronavirus strains because former survive outside the host longer than latter.

https://www.biorxiv.org/content/10.1101/2020.11.15.383323v1 100% certainty that Covid-19 will mutate but at a much slower rate than influenza viruses (we should be grateful this isn't "just the flu"). To simplify, let's assume the rate of mutation is related to mutations during the following: 1. Replications within a host (human or animal) 2. Jumping between hosts 3. Jumping between species This suggests that rate of mutation will be highest at the peak of the pandemic when the virus is replicating rapidly in people and jumping rapidly between hosts. We've had perhaps trillions of replications, yet vaccines are showing 90% efficiency. This shows that the virus is not mutating rapidly (which is also shown in genetic tracing). You could foresee some random mutation that allows Covid-19 to evade the vaccine, but the probability is related to the number of mutations. So the likelihood of a dangerous mutation occurring after June 2021 is much lower than one occurring in 2020. Edit to add: So that suggests you won't need a yearly vaccine solely due to mutations. You might still need one for waning immunity, but the preprint above suggests immunity could be reasonably long-lived. Thanks KCLarkin, this is helpful. Based on this, looks like probability of Covid-19 coronaviruses becoming an endemic is much lower than flu coronaviruses.

How is the transmission rate, that is, R0, going to go down for mutations of Covid-19 coronavirus with vaccine and better treatment. Flu has an R0 of about 1.3, i.e. one person can spread it to 1.3 people on average. Covid-19 coronavirus strains have an R0 as high as 5 or higher. Do we know of other ways of lowering R0 for Covid-19 strains other than behavior change? We cannot say with 100% certainty that herd immunity will continue to provide immunity for mutated strains of Covid-19. We do know that herd immunity does not provide immunity for mutated strains of flu coronaviruses. Vaccines. The transmission rate will go lower if more people are immune to it. Isn't that obvious? If the virus mutates and presents a serious threat again, I'm sure that we will develop another vaccine that works. Hence, my question at the start of the thread, i.e. what is the probability that we would need a vaccine every year? So, you agree that the probability of that is non-zero? Yes, but it won't change our behaviors to the point that the industries you pointed out will significantly change. We get flu vaccines every year. Thanks clutch, it is a very fair point that we get flu vaccines every year, and don't worry about flu that much while going on about our lives. Let's also assume that we will eventually be able to treat covid-19 without requiring hospitalization for a big percentage of people. One difference that will likely stay is that Flu coronaviruses have an R0 of about 1.3, i.e. one person can spread it to 1.3 people on average, while Covid-19 coronavirus strains have an R0 as high as 5 or higher. Where that difference might play out is in some people's behavior while they are waiting to get the vaccine for new strains. With flu coronavirus strains, because the transmitability is lower, we don't worry too much about the risk and go about our usual behaviors. With Covid-19 coronavirus strains, because of higher risk of catching it, our behaviors might be somewhat more constrained than with flu while waiting for vaccine for new strains of the year. Look at some of our behaviors currently while we are waiting for the Covid-19 vaccine. I agree with you once treatment without hospitalization is possible for a big percentage of people, we probably won't restrict our behaviors as much as now while waiting for vaccine for new Covid-19 strains, but we might not relax our behaviors all the way to how we act while waiting for flu vaccines because of increased transmitability with Covid-19 coronavirus strains. If the probability of Covid-19 coronaviruses becoming endemic like flu coronaviruses materializes, there will probably be some effect on our behaviors longer term, which will probably have some impact to whether 100% of people will go back to working from office 100% of the time.

How is the transmission rate, that is, R0, going to go down for mutations of Covid-19 coronavirus with vaccine and better treatment. Flu has an R0 of about 1.3, i.e. one person can spread it to 1.3 people on average. Covid-19 coronavirus strains have an R0 as high as 5 or higher. Do we know of other ways of lowering R0 for Covid-19 strains other than behavior change? We cannot say with 100% certainty that herd immunity will continue to provide immunity for mutated strains of Covid-19. We do know that herd immunity does not provide immunity for mutated strains of flu coronaviruses. Vaccines. The transmission rate will go lower if more people are immune to it. Isn't that obvious? If the virus mutates and presents a serious threat again, I'm sure that we will develop another vaccine that works. Hence, my question at the start of the thread, i.e. what is the probability that we would need a vaccine every year? So, you agree that the probability of that is non-zero?

How is the transmission rate, that is, R0, going to go down for mutations of Covid-19 coronavirus with vaccine and better treatment? Flu has an R0 of about 1.3, i.e. one person can spread it to 1.3 people on average. Spanish Flu had an R0 about 2.2 to 2.9, and effectively lowered further due to less travel sometime after World War I was over. Covid-19 coronavirus strains have an R0 as high as 5 or higher, exacerbated further by dense populations and travel. Do we know of other ways of lowering R0 for Covid-19 strains other than behavior change? We cannot say with 100% certainty that herd immunity will continue to provide immunity for mutated strains of Covid-19. We do know that herd immunity does not provide immunity for mutated strains of flu coronaviruses.

If he's buying a basket of pharma stocks, then I wonder has he bought more than the ones listed in this 13F? I would not be at all surprised if he's bought Glaxo (London listed) and Sanofi (Paris listed). Both are good companies and would fit into his basket approach. He has also owned both before, and both are arguably cheaper than when he owned previously. He could have bypassed the 13F listing requirement by buying both stocks on the local exchanges, something he has done before (Tesco). Didn't he own Sanofi before? Yes, he sold out after some time because Sanofi didn't go anywhere in terms of making money for shareholders.

Thanks KCLarkin, would it be possible to share the link to the source? Which one of the below are you saying? (a) 100% certainty that Covid-19 coronavirus will not mutate, unlike the flu coronaviruses, or (b) 100% certainty that the same vaccine we get once for Covid-19 coronavirus will be able to provide immunity against all mutations of Covid-19 coronavirus, unlike the vaccine for flu coronaviruses.

From Barron's article at https://www.barrons.com/articles/why-warren-buffetts-berkshire-hathaway-fell-out-of-love-with-jpmorgan-51605629567: Not sure if they have take into account NEAM figures in that $9 billion total purchase amount.

Hi petec, would it be possible to share your reasoning? If I count up their debt, and do an EBITDA/EV earnings yield, I'm not getting anything remotely close to 15% earnings yield, that is, before corporate and personal income taxes. Are you thinking their EBITDA is going to go up a lot with their pricing power during inflation, or that their expenses include a lot of R&D expenses that need to be amortized not expensed directly, even though they have patent expirations coming that will need R&D almost as sustaining capital expenses?

Day trading is the new value investing? Here is what Buffett and Munger said in 1999 about buying a group of pharma stocks at below market multiple: Yes, they might be trading at below market multiple, especially if you compare to some tech stocks, but these pharma stocks are not cheap.

Wondering if he is still thinking there is a non-zero probability that Covid will become an endemic, requiring a vaccine very year. Another interesting thing to note is that all the pharma stocks he bought (ABBV, MRK, BMY, PFE) seem to be looking into Covid vaccines and treatments. Wonder if it is going to be another pump and dump scheme like BRK seems to be doing for GOLD and might do for Snowflake, i.e. get some attention to these stocks, and ride on a high probability that one of them could have a vaccine or treatment take off, and then dump them.

Some folks in the medical and academic communities believe that there is a good chance Covid will end up becoming an endemic like the common cold or the flu, which needs a new vaccine every year. Wondering if anyone in the group has looked into credible sources to help figure out the probability range for this happening. Once we know the probability range, if the high end of the probability range is high enough, we can talk about potential implications to airlines, remote work, density in cities, etc. Here are some articles I've run into: * https://www.huffingtonpost.co.uk/entry/covid-19-endemic-pandemic-epidemic_uk_5f99a2abc5b6aab57a0eb788?guccounter=1&guce_referrer=aHR0cHM6Ly93d3cuZ29vZ2xlLmNvbS8&guce_referrer_sig=AQAAAFY2D4Q9zh3-O_LifL8Km8l7eait0vUsQ5u9Sxy6bWucWEPwkcijmsuJlk64j2OaBsW8pgUSJLPoqSvXmyaze-412bD9YFU55UvwlZJnlWgodlmVyuMuVErL-iV3MShTAg32rSybeVNwo5lX7mQs8x9perwMjJ9083wvT0KzdJGZ * https://science.sciencemag.org/content/370/6516/527 * https://www.publichealth.columbia.edu/public-health-now/news/will-covid-19-virus-become-endemic * https://www.sciencedaily.com/releases/2020/10/201015101820.htm#:~:text=Evidence%20suggests%20COVID%2D19%20could,temperate%20regions%20similar%20to%20influenza.

What do we think is the buying power of each of these folks working for a company currently located in a city that is not very expensive compared to the buying power of the analyst or software engineer working for a company that had already picked to be located in a big city for talent? Probably lower, right? While WFH a lot more of the time will be standard, WFH 100% of the time will probably take some time to pick up. Assuming it picks up for some percentage of these folks and that these folks are able to move, what is the radius of the area they will be targeting within their new region? Will it be 1-mile radius (p* 1 square mile) like it was for some of those rich analysts and software engineers, or will it by 20-mile radius (p* 400 square miles)? Probably latter, right?

Totally agree that high tech employers want to be able to attract top talent wherever they are. This is why big tech companies are leasing new NYC office sqft even in the middle of the pandemic. They want to be able to attract the top technical talent in the biggest metropolitan region of almost 20 million people. My point is that the game has changed a little. In order to attract and retain top talent, tech companies are now also competing over offering WFH and hybrid work benefits. This means for that newly hired Facebook engineer, he no longer is required to get to office daily, and his/her desirability of being able to walk to work has gone down, and the living options for that engineer have gone up quadratically within the region. The desirability of being close to cities is still there, but his residence doesn't have to be walkable distance to work. Now, that engineer might still want to be in the wider city region of slightly bigger radius for dating life and other reasons, e.g. expanding their desired region from within 1 mile of work (pi * 1 square mile) to within 10 miles from work (pi * 100 square miles). However, for folks with families, dating infrastructure might not be as important a reason, and they can expand their radius for the area farther from within 1 mile of work (pi * 1 square mile) to within 20 miles of work (pi * 400 square miles). The desirability of being close to cities is still there, but work doesn't have to be in immediate vicinity within the region anymore. I am already starting to see that folks who used to want to live within 1 mile radius of work are now willing to look much farther out within the region, while still having access to all the city has to offer.

The demand to relocate out of expensive cities seems to be high according a survey by Blind, which verifies employees by asking them to provide their company email address. Roughly third are willing to relocate even with a pay cut. In addition, 40-45% will relocate without a paycut, depending on city/company. Here are results broken down by company and by city: https://www.teamblind.com/blog/index.php/2020/09/14/44-of-professionals-are-happy-to-take-a-pay-cut/ https://usblog.teamblind.com/wp-content/uploads/2020/05/PayCut.pdf https://docs.google.com/spreadsheets/d/1zF_jxowBZYkiJIeatZAm3soelVBpoFf1TbjEZVwxBpA/edit#gid=171959972

I meant during and after GFC, Detroit residential vacancy hit above 20%, causing banks to give away houses for $1 in some cases to avoid liability for property taxes and vandalism. I am not saying it will get that bad in every city. What I am saying is 20% vacancy can cause more than 50% drop in prices. I'm also not saying prices will fall 50% in cities with certainty. What I am saying is the effective supply of options is going up way more than 20% for that well-paid analyst or software engineer, and that in history, we have examples of even just a 20% effective supply increase causing a big drop in prices. So, I think probability is high that there will be some drop in prices for residential real estate in expensive cities. Many factors will determine the amount of that drop, e.g. how fast the trend materializes, how much existing housing is already available in desirable locations [desirable defined in the new era where they don't have to be within 5 miles from work], how fast can developers build new houses on the outskirts in good school districts that these folks would want to move to, the magnitude of new supply from developers vs. the demand, how much regulations and zoning slow down these developers, etc.

Agreed, you wouldn't want to lose these. That said, assuming you were close to your work before, would you now willing to be 30 minutes away from some of these needs if you could get an estate now for the same price, e.g. say with a farm, view, horses, heliport, airport, or something else you like? Maybe you have to go too far in your area, but do you think some small percentage of people might be willing to consider that option for their next house? For each extra mile you're willing to add now, the options grow quadratically with new options being available in area of size pi * (new_distance ^2 - old_distance^2). So, if you were willing to be only 5 miles away from these options before, you could cover only pi^25 square miles. Now, if you can go 20 miles, your options for living go up to pi^400 square miles. If you are willing to go 30 miles, your options for living have expanded to pi^900 square miles. I feel the oligopoly power of the living options within pi^25 square miles has certainly diminished by the availability of living options in p^900 square miles, while keeping access to things that are important. All it takes is a small percentage of people to start exercising this option of leveraging increased effective supply as 20% incremental supply can lead to moves in prices as 20% vacancy in Detroit showed us. Here, the effective supply has increased quadratically by many multiples not just 20%.

Thanks Pupil for sharing your words of wisdom. Shelter vs. farm products: A rich analyst doesn't pay a big percentage of annual income for products from farms either because there is no oligopoly-type ownership concentration so far with farms. If they had the option to buy farm products from only a few non-competing farm owners, those few options could have easily extracted a big percentage of their annual income, but fortunately that is not the case so far. Both shelter and farm products come from real estate but former has been taking much more income because of need for close proximity to work so far, which has created oligopoly type structure by reducing options. I had done a survey of a group I am part of on what percentage of people would consider moving if WFH became permanent. I got about half the people. Redfin did the same survey and found about 50% people ready to move in some cities: https://www.redfin.com/news/wfh-leaving-new-york-san-francisco/ This is what I am starting to see in practice as well. The folks I'm seeing do this are not into FIRE. Rich working single folks are moving for multiple reasons, e.g. to be close to their family and buying a house in another city, etc. Rich working folks with families are buying bigger properties farther away from work. I hear you on rich folks like to be together. Here, I'm seeing working rich folks in the tech who are getting the flexibility to move and starting to leverage that flexibility together. This is similar to how we saw white flight in the 60s, where those who could afford a car, started moving to suburbs. This time also, the reason for many is not FIRE but a bigger property, newer homes, etc. Savings are just icing on the cake. All it would take is about 10-20% move to make a big impact on pricing as 20% vacancy did in Detroit.

Pupil, would it be possible to elaborate what you meant here so that we can get the benefit of your wisdom here?

This is counterintuitive, but if you actually allow more real estate developers, fewer zoning restrictions and let them build lot more, that would increase the supply so that price can come down to construction cost of buildings. Anyway, Covid should help now by effectively increasing supply by decreasing demand for needing to take the train to downtown everyday, so that folks can move further out or to other towns.

Twitter, Microsoft, Facebook, and Google have all announced some form of hybrid work post-covid, where they will let their employees work from home 100% of the time or at least more of the time than pre-covid, permanently. Some folks at some of these and other tech companies had already started to move, e.g. to be closer to their family in another city, or buy a bigger estate farther away from work. Now, they are thinking of making these moves permanent. Overall the # of options folks can pick from for their living situation has gone up more than 10 times in some cases. Beyond the cost of the building itself, I think of real estate as deriving its value from having small local oligopolies. For example, if someone needed to be able to get to work within 10-30 minutes for five days of week, they had a limited number of available multifamily landlords to rent from. Now that they don't have to go to office at all or go only 2-3 out of 7 days, for folks to keep the same total commute time per week, they could have more than 10X the number of multifamily landlords they could rent from. Same goes for the number of options of houses they would look at before and what they can look at now. I understand things are not that simple as some people will still want to be in a certain area because of family or school district, but prices are determined by marginal supply/vacancy. In Detroit, 20% vacancy had a huge impact on house prices. For prices to be impacted, all it would take is a small percentage of folks exercising their options to pick from suddenly increased # of options. In the public markets, if we suddenly had a company's oligopoly position destroyed such that we had 10X the number of companies available to buy the same product, that company's profitability and value would fall drastically. What do folks think would be the % degradation of real estate values where the oligopoly position has gone down to the level of suddenly 10X the number of options being available to buyers. Could it approach to becoming a free market where cost of housing comes down to cost of actual building itself in some cases as is already the case with rural homes, where some folks are moving?

Actually, it is more than that. In the same article, it says: Employees can work from home full time with manager approval. They can also move to a new location for remote work, with salaries adjusted based on geography if their manager approves. Imagine the repercussions on currently expensive real estate prices close to Microsoft. All it would take is a small percentage of people moving to exurbs for big estates or closer to their families in other cities to impact the incremental supply/vacancy rate. 20% vacancy/incremental supply had a huge impact on Detroit house prices.