KJP

wabuffo: Thanks for all of your contributions in this thread. I've found them compelling and persuasive and, more importantly, they seem to correctly predict what's happening. But I have a question: Under current law, the Fed is not permitted to simply directly credit the Treasury's account, correct? In other words, under current law, the United States (in the form of the Federal Reserve) cannot just print money and give it to itself (in the form of the Treasury's account at the Fed). Instead, the Treasury account must be funded in the first instance via tax revenue and debt issuance, and then the Fed can buy that debt from private parties, correct? If that's right, then what you describe as "political constraints" are also currently binding legal constraints. Of course, Congress could change the law, so perhaps it's fair to describe these legal constraints as merely political constraints. The reason I'm asking is that I have heard others (e.g., Lacy Hunt) describe the potential legal change I refer to above as a type of "crossing of the Rubicon" that could actually cause significant and generalized inflation. But is that type of inflation (as opposed to inflation in asset prices) even in such a scenario consistent with your theory? If the Fed could directly credit the Treasury's account, that would seem to lower the amount of Treasuries that would otherwise be in the system and raise the amount of bank reserves as the Treasury spends down the dollars digitally "printed," and they pile up as excess bank reserves.

Earlier in the thread I suggested your theory didn't adequately account for demand. I continue to think that's right. You've framed your analysis in terms of "effective supply," but, in the short run, the supply of city housing is quite inelastic (it's hard to quickly convert housing to an alternative use or to quickly build new housing). So, what I believe your theory depended on was a significant drop in demand for city housing, in other words, a shift left in the demand curve along a relatively inelastic (relatively vertical) supply curve. That is what I suspect happened in Detroit -- supply didn't change drastically (you need to demolish or convert housing for that); rather, the demand curve shifted left along a relatively inelastic (vertical) supply curve, causing large price declines. I doubt the supply of, for example, rental housing in New York City has changed significantly in the last 18 months. So, if market rents are back to where they were, I think that implies that demand for that housing did not change as you expected it to. Also, you believed that there would be greater demand for housing in close-in metro suburbs. Many of those areas are already heavily built up (making adding supply quite expensive or impossible) and, in any event, new construction takes time. So the supply curves in those suburbs are quite inelastic as well, particularly in the short run. Moreover, everyone needs a place to live. So, if the demand curve for close-in suburban housing shifts right against inelastic supply, we'd expect significant price appreciation, which we've had, perhaps helped along by the income effect of lower rates/higher asset prices. But those sellers themselves need a place to live, which ought to increase the demand for housing somewhere else. Where are those sellers going in your model? Also, I understod you to be talking about relative prices, i.e., you were predicting that the price of urban real estate would fall relative to the price of close-in suburban real estate. Would interest rates affect this, or do interest rates effect the nominal price levels of both types of real estate, rather than their relative prices?

The data the author used doesn't say there were no severe cases in under 12s. Rather, it contains no data at all about them. So it's possible there are hospitalizations of under 12s in Isreal. By "data" I'm referring to the "Isreali_data_August_15_2021" excel file that can be downloaded from the post.

I don't think this is correct because the entire chart appears to be limited to 12+ years old (despite the "all ages" label!).

Is that an accurate description of what the author is doing? According to the following sources, Isreal has between 8.8-9.3 million people: https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-vs-severe-disease-be-strong-when-60-of-hospitalized-are-vaccinated https://www.worldometers.info/demographics/israel-demographics/ https://en.wikipedia.org/wiki/Demographics_of_Israel https://www.timesofisrael.com/as-it-welcomes-in-2021-israels-population-numbers-9291000/ In the "all ages" numbers, the author refers only to 1,302,912 + 5,634,634 = 6,937,546 people. So, at the outset, he has excluded about 2 million people. Various sources put the 0-14 years population in Isreal at about 27%: https://www.populationpyramid.net/israel/ https://www.jewishvirtuallibrary.org/latest-population-statistics-for-israel 9 million * .27 = 2.43 million 1-14 year olds, or roughly 2 million 0-12 year olds, which is the same number of people missing from the author's "all ages" numbers. It seems to me they ought to be excluded because there is not vaccinated group of 0-12 years olds against which to measure effectiveness. Of course, you also need to exclude any 0-12 year olds from the "severe cases" data. That also appears to be what the author did, because if you download the data table labeled "Isreali_data_August_15_2021" you will see that it includes only data for people 12 and older. In addition, summing the "severe" cases produces the 214 and 301 numbers from the author's chart. So, it appears to me that the author's use of 78.7% vaccinated is correct, but his label of "all ages" is misleading, because what the chart actually shows is data for people 12 or older, i.e., the population that is 78.7% vaccinated.

Also, here's an analysis of the Isreali data: https://www.covid-datascience.com/post/israeli-data-how-can-efficacy-vs-severe-disease-be-strong-when-60-of-hospitalized-are-vaccinated It seems every data point supports the argument that vaccines are very effective against severe disease from Delta, rather than being counterproductive as in the ADE hypothesis. I'd be interested to see any contrary interpretation the data.

Is that data different than Dallas? Slide 21 of Dallas presentation you linked to in your first post stated that, as of the date of the report, 139 fully vaccinated people in Dallas County had ever been hospitalized for COVID. I didn't see any breakdown of how that compared to the number of hospitalizations among the unvaccinated (either ever or during the last month (during the proposed Delta-ADE era)). Without that comparison, what are you inferring from the Dallas data?

I also agree that equity compensation is an expense, so the question is how to assess it. Thepupil has presented a concrete example, so I'll try to get at a more theoretical issue: The stock-versus-flow mismatch of what I've referred to as the "ex post" approach. An essential input into the "ex post" method is the actual share price performance of the firm in the years after the equity comp is given. Changes in the stock price presumably reflect changes in the expected net present value of all future cash flows flowing to the firm's equity. In this sense, the market cap of a firm is akin to a balance sheet/stock figure rather than an income statement/flow figure, and changes in equity prices reflect changes in the overall value of this "stock". So, what the ex post method does is take the overall change in value of the firm (embodied in equity price change) and assign it proportionally to the equity granted in a particular year. An income statement, on the other hand, reflects only (accrual-based) revenue and expense flows in a given year, not changes in the expected cumulative value of the firm's future cash flows. So, doesn't the ex post method distort things by inserting a stock-based measure into a flow-based statement, thereby overstating the true cost of equity comp for companies whose equity price (and presumably underlying value) is increasing rapidly? (A similar problem to putting the full cost of a 20-year asset onto the income statement in the year of purchase.) I believe the JPM method in thepupil's post is trying to get at this problem by taking stock comp out of the flow-based income statement altogether and putting into the stock-based metric of shares outstanding.

I'm also curious about this view. I had a similar discussion in the NVR thread (see March 17, 2020 posts): The ex post, hindsight method of finding the "true" cost of stock comp seems to contain a paradox: The more the stock price goes up (and thus the more equityholders have benefited) the more they appear to have been robbed by "excess" employee comp.

Today: Altice USA Over the last few months: American Outdoor Brands Leatt Corp. Turning Point Brands Nickel 28 Capital Corp.

The research that I have seen to date suggests that at least Pfizer and Moderna significantly reduce asymptomatic cases, which ought to render fewer people infectious. See, e.g., https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3790399 https://abcnews.go.com/Health/pfizer-vaccine-shows-94-effectiveness-asymptomatic-transmission-covid/story?id=76389615&cid=social_twitter_abcn If that common-sense inference is correct, then there should be fewer cases among those exposed to vaccinated as opposed to unvaccinated people. Here's one study that produces that result: https://www.medrxiv.org/content/10.1101/2021.03.11.21253275v1.full-text

Here's a podcast specifically about this thread: https://focusedcompounding.libsyn.com/website/ep-304-thoughts-on-10-different-stocks-reits-msgmsgn-and-a-bull-case-for-amazons-stock

Earlier in the thread, a poster stated that he or she is young and healthy, and thus their chance of dying from COVID was roughly 1 in 1 million. Alas, "young" is one of those words that changes its meaning as you age, so I'm not exactly sure how old this poster is. In any event, I'm 40, give or take a few years, so I was curious if I could estimate my chances of dying from COVID, assuming I have no co-morbidities and the healthcare system is functioning normally. I went at it from top-down and bottom-up methods. Here's what I came up with: Top-Down There are about 41.65 million people in the country aged 35-44. [source: https://www.statista.com/statistics/241488/population-of-the-us-by-sex-and-age/] One plausible estimate I’ve seen is a 30% population-wide infection rate. [source: third chart here: https://covid19-projections.com/path-to-herd-immunity/] That implies that the US has had 12.5 million infections among 35-44 year olds [42 * .3] There have been about 550,000 deaths attributed to COVID in the US. The CDC has age data for 411,261 deaths, and reports 4,670 in the 30-39 age group and 11,562 in the 40-49 age group. To get the number in the 35-44 age group, I added those two numbers, divided by two, and then grossed up proportionally to 550,000: (4670 + 11562)/2 = 8,116 * (550,000/411261) = 10,854 10,854/12.5 million = .0008632 or 1 in 1158. That is unadjusted for any co-morbidities, so the “co-morbidity” death rate must be significantly higher and the “no co-morbidity” death rate must be significantly lower. Bottom-up Here's a study that attempts to control for age and co-morbidities: https://s3.amazonaws.com/media2.fairhealth.org/whitepaper/asset/Risk%20Factors%20for%20COVID-19%20Mortality%20among%20Privately%20Insured%20Patients%20-%20A%20Claims%20Data%20Analysis%20-%20A%20FAIR%20Health%20White%20Paper.pdf It appears to estimate a no co-morbidity fatality rate for 40 years olds of 1 in 1,000 [rough average of relevant age groups in figure 13]. But I think the methodology of the study means that number is too high. From the Methodology section: “For this analysis, we used a longitudinal claims subset of the FH NPIC database [a database of claims from private insurers]. This subset includes approximately 100 million covered lives.” From that dataset of 100 million people, they “identified 467,773 patients diagnosed with COVID19 from April 1, 2020 through August 31, 2020.” There are roughly 325 million people in the US, so that ratio would imply a nationwide total of 1.52 million PCR-confirmed cases (4667,773*3.25) during that four-month period. But according to the CDC, as of March 31 there were 185,867 cases and as of August 31 there were 6,026,542, or about 4x as many PCR-confirmed cases as implied by the dataset (the “Missing Cases”). [source: https://covid.cdc.gov/covid-data-tracker/#trends_totalandratecasessevendayrate] Some of the Missing Cases may arise from the fact that COVID cases are proportionately higher among the non-privately insured, e.g., uninsured, Medicaid, Medicare. But I suspect that a good portion of the Missing Cases arise from the fact that a positive test may not result in an insurance claim, particularly if you have a mild case with no need for medical attention. For example, if I had a positive result from my county testing site, I think that would show up in the CDC numbers, but I don’t think it would give rise to a private insurance claim unless I actually needed follow up care. Let’s assume that 25% percent of the Missing Cases come from population differences and 75% of Missing Cases come from positive tests that do not give rise to an insurance claim because they are very mild. That would take the “adjusted” sample size of COVID positive people in the insurance dataset up 3x without increasing deaths, because by definition these are mild cases. That alone would lower the no-co-morbidity fatality rate among 40 year olds from 1/1000 to 1/3000. Then you’d need to adjust further for (i) the number of infected people who never got tested, (ii) the apparent failure to account for obesity itself as a co-morbidity (see, e.g, https://www.acpjournals.org/doi/10.7326/M20-3742), and (iii) developments in treatment, which are implied by the downward sloping mortality curve on pg. 4 of the insurance study and the “highly statistically significant” time variable in the obesity study in the previous link. After all of those adjustments, I think this study suggests a healthy 40-year old with no co-morbidities probably has no more than a 1/5-10,000 chance of dying from COVID, assuming a functioning healthcare system. That's generally consistent with the "top down" estimate of the no co-morbidity fatality rate being substantially less than 1 in 1158. Sanity check Using the numbers from the "top down" section above, the following equation should hold: 10,854 = (41,650,000 * [Percentage of 35-45 year olds with no co-morbidities] * [infection rate] * [No co-morbidity Fatality rate]) + (41,650,000 * [Percentage of 35-45 year olds with co-morbidities] * [infection rate] * [Co-morbidity fatality rate]). I'll continue to assume the infection rate is 30%. I'll also assume the co-morbidity rate is 50% [for context, the obesity rate alone is over 40%: https://www.cdc.gov/nchs/products/databriefs/db360.htm] And consistent with my rough estimate from the "bottom up" method, I'll assume the no-comorbidity group has a 1 in 5000 fatality rate. So: 10,854 = (41,650,000 * .5 * .3 * [1/5000]) + ( 41,650,000 * .5 * .3 * (co-morbidity fatality rate)) 10,854 = 1249 + [6247500 * (co-morbidity fatality rate)] .001537 = co-morbidity fatality rate = 1 in 650 chance. To a layperson like me, that's not obviously wrong if the top down approach is roughly right that the blended average mortality rate is 1 in 1150 and co-morbidities have a significant effect on mortality. Does anyone seen anything obviously amiss here or is anyone aware of a more rigorous analysis of mortality controlling for age and co-morbidity that comes to a significantly different conclusion? As an aside, this is an estimate of the fatality rate if infected. If the disease becomes endemic, then, absent vaccination, as time goes to infinity, your chances of infection likely approach 100%.

I'm curious what you mean: (i) the alleged symptoms don't exist or include arguably non-objective symptoms (e.g., anxiety), (ii) the alleged symptoms exist but there is no causal link between them and past COVID exposure, or (iii) something else?

Constructive criticism. Thank you for the info*. The conclusion relays an impression of an incomplete picture. It's like if a company would describe the effect of currency movements on its balance sheet by focusing on the differential exposure between the components of the liabilities. The valuation 'narrative' of the last few years is based on a low interest rate environment. Rising rates (i'm not saying this will happen; in fact i think (on a weighted basis) this is unlikely to happen, at least for the 'risk-free' part) would trigger a reappreciation of the asset side also. But individual net exposure may vary and the idea that debt can be inflated away is an attractive one. *The info doesn't seem to include nonfinancial corporate loans (kept on banks' balance sheets) which are still quite a significant amount and which (the last time i checked) were about 85% variable. The increasing rate exposure that scorpioncapital describes also needs to take into account a dynamic aspect with rolling refinancing risk (cost of capital may be higher and more 'floating') and a very unusual bunch of potential fallen angels. I think wabuffo has been making a similar point, assuming I understand his references to aggregate assets. But the dispersion of assets and liabilities is not uniform throughout the population; a few are very very rich and many others are running an "asset light" business model. Is there a way to decompose the aggregate statistics to understand where, say, the middle quintile is with respect to assets and liabilities? I think policy in a democracy/republic would tend to gravitate to what favors that median group, rather than what might make sense if you looked only at the aggregate numbers.

The following is based on nothing but my armchair pontification and is likely worth what you paid for it. Real wealth comes from the amount and quality of good and services in the economy, which in turn is driven over time by investments in and discoveries of new and more efficient technologies/ways of doing things. Inflation -- meaning a general rise in nominal price levels throughout the economy -- provides an impetus for investment, because the alternative of putting it under your mattress will lose value over time. Deflation -- meaning a general decline in nominal price levels throughout the economy -- retards investment, because you can gain relative wealth simply by putting your coin under a mattress and because earning a return on $100 invested today is more difficult when nominal price levels are declining (your customers will have fewer nominal dollars to pay you). So mild inflation creates a gentle push in favor of investment that over time that leads to more real wealth.

Technip Energies

Did you mean to link to the interview of Richard Aboulafia of Teal Group rather than the interview of John Rogers? I couldn't find the Aboulafia interview online. I enjoyed this Q&A from the Aboulafia interview: Answer: "The International Air Transport Association says we get back to the 2019 traffic peak in early 2024. I say late 2022." Question: "That's fast." Answer: "The traveling public now has a half-century-record level of savings waiting to be thrown back at the vacations that they haven't been taking. This is a recipe for the fastest recovery every." Question: "Which companies benefit the most from people flying again." Answer: "I could never pretend to be an investor, but I'm happy to offer five companies best positioned for the recovery comeback: Raytheon, Northrup Grumman, Safran, Lockheed Martin, and Howmet Aerospace." Question: "What do you like about those five." Answer: "Lockheed Martin is the world's most important maker of combat-aircraft products, and that's a strong market. They are also strong in missiles. . . . Northrup Grumman: lots to like, strong emphasis on investing in the future of defense. That focus won them the two legs of the nuclear triad, how the U.S. delivers nuclear weapons, that are being replaced -- ground-based ICBMs ad the strategic stealth bomber." Those two investment theses have nothing to do with Aboulafia's purportedly variant view on the timeline of the comeback in commercial travel, nor anything to do with a "recovery comeback." The questioner never acknowledges this or asks why Aboulafia's top "recovery" picks have nothing to do with his bold call on commercial aviation's comeback. More broadly, what percentage of commercial travel is business travel? Will businesses adapations to COVID have a significant lasting effect on business travel? A good and prepared reporter ought to be asking these types of follow up questions.

+1 +2 I've also been adding here and there to Altria and Lockheed Martin.

I think we were having different conversations. I was still trying to respond to JRM's question about why policymakers don't seek to create deflation, which I understood to refer to a general fall in nominal price levels across the economy. So, on the theory that policy in a "one person, one vote" system is driven by the desires of the median household/voter, I was exploring the situation of that household/voter vis a vis deflation.

If I understand your source correctly, it's an aggregate number across all households. How do you decompose it to look at the exposure of, for example, the median household/voter? Unsurprisingly, the Survey of Consumer Finances shows that the median household is well behind the mean in terms of financial assets and net worth. My original comment was responding to JRM's question about why policymakers don't actively seek to create deflation. Putting aside what would actually have to be done to create deflation, my initial take was that the median household/voter would be opposed to deflation because they are wage earners who would be greatly opposed psychologically to any cuts in their nominal wages and who perhaps are short the dollar due to long-term, fixed rate debts (home mortgage and education loans) exceeding their exposure to financial assets that might benefit from deflation (e.g., bonds), though perhaps this is mistaken to the extent these liabilities can be refinanced into negative rate loans. I had not considered the big pension entitlement asset listed in your source which may benefit some people in this situation. thepupil seems to be asking about the opposite scenario -- inflation/rising rates causing housing prices to decline. This may temporarily affect the homeowners' balance sheet, but if their wages are rising with inflation, don't they get to pay off debts with cheaper and cheaper dollars? So would their income statement look better even if their balance sheet might be underwater for a bit?

Which Fed survey are you referring to? This one? https://www.federalreserve.gov/econres/scfindex.htm

Isn't the individual US consumer/voter also often massively short the dollar via a highly levered 30-year fixed-rate mortgage and, perhaps, other debt such as school loans? And how easy is it to reduce nominal wages alongside deflation?

I would try to find something your child is interested in, e.g., sports --> Nike, cosmetics --> L'Oreal, TV --> Netflix. If the first company "analysis" goes well, then try another more generic company in the same or a related industry. Someone learn a lot just by trying to understand the differences in gross margins between, say, Nike and Jerash Holdings. (Of course, this is stolen from Buffett's lessons about See's Candies, etc.)

It does. Thanks for pointing it out. Maybe I should think less and buy good companies like PGR, but the overanalytical bear in me says at some point people start pricing in a secular decline in premiums due to autonomous driving making driving much safer, and I don't want to own this when that happens. Maybe that's too far off to care about now. If you normalize margins for typical (non-COVID) driving/accident volume (11.5% EBIT margin on premium written was my two-minute back of the envelope) and assume zero investment gains (probably have unrealized losses YTD give rising rates), is Progressive trading at 14x 2020 "normalized" earnings?