Weekend reading: The alchemy of turning uncertainty into risk

Comments on this entry are closed.

• 1 FI Warrior April 24, 2020, 7:32 pm

  Whatever I try to tell myself, it’s just felt like an unexpected long holiday and being an introvert, I’ve appreciated the silence and solitude. To not feel that guilty, I’ve done some routine maintenance though, fixing stuff, getting the garden ready for summer, admin in good shape, even a bit more exercise too.

  I’m keenly aware though from relatives and acquaintances, that people are losing jobs and small businesses are going bust, despite the salves pumped out on mainstream media that were supposed to prevent this. I’m not sure why the process is not working as advertised, still trying to get confirmation. Living in a small place, the post office is the backstop to a lot of functions (the banks to name just one, closed early) and just let their staff go, with the owner working 11 hour days and praying life returns before she goes bankrupt. That would hit the area hard and since they’re actually open and doing some business, where does that leave those forced to close like drycleaners, stationers, keycutting etc., not life-saving shops, but necessary for normal life.
• 2 Al Cam April 24, 2020, 7:33 pm

  One to add to your corona virus corner that adds a bit of historical perspective:
  https://www.euromomo.eu/graphs-and-maps
• 3 Al Cam April 24, 2020, 7:36 pm

  I should have said play the animation to see recent history of z numbers
• 4 The Weasel April 24, 2020, 7:52 pm

  Well, my most pressing uncertainty at the moment is that one airline may decide to force me to be their creditor as opposed to their customer. Not a great position to be in given that, in my view, FlyBe won’t be the only casualty in this.

  A bit of a first world problem, given the context.

  BTW @TI, have you come across the effectiveness of disinfectant injections against COVID-19? 😉
• 5 W April 24, 2020, 8:19 pm

  https://clausen.berkeley.edu/wp-content/uploads/2020/04/Corona_all.pdf

  Despite the economic cost of social distancing lockdown policies it would lead to an even worse outcome if these policies were abandoned before the disease replication rate is totally under control.

  The results also suggest that a massive testing and control strategy needs to be in place before there can be any easing of suppression strategies.
• 6 Snowman April 24, 2020, 8:41 pm

  This weekend reading has made me think. It is a difficult decision on whether to follow what is going on news wise or to avoid it. I’ve been a follower so far. But following is starting to make me a bit angry and I think this situation needs to be approached with a more mellow mind frame. I may switch into avoidance mode.

  Looking forward to when the pubs reopen, as someone said we are going to party like it’s 2019 when they do.
• 7 The Investor April 24, 2020, 10:34 pm

  @W — I’ve already included that link above, saw it on the other thread. 🙂
• 8 JimJim April 25, 2020, 6:57 am

  Nice titles from my favorite band @TI… Heaven: (how where I live is like without tourists) Radio head: (What I have become between Teams and Skype calls while working) Slippery People: (Getting students to respond to teams) Pulled up: (the patio and re-laid it) Wild, wild life: (came within an inch from running over a six foot grass snake west of Coniston water on my bike)
  I am passive on covid19, much more so than on investment because I don’t think that my opinion is actionable. Stay well all
  JimJim
• 9 Tony Edgecombe April 25, 2020, 7:05 am

  Just like @FIW life hasn’t changed that much for me, I already worked at home. Our regular walks are from the house instead of around the Cotswolds. No stops for cake and tea although sadly that hasn’t resulted in any weight loss.

  Probably the most surprising change has been just how much our spending has plummeted. I always thought we were pretty frugal but it has been a shock how much slack we had. I’m wondering if we can maintain that once the restrictions have been lifted.
• 10 BBlimp April 25, 2020, 9:38 am

  @weasal … if you bought direct from the airline is your credit card in play ?

  I bought from love holidays which apparently broke the chain between credit card provider and merchant ( airline/ hotel) so Amex are ‘looking into it’ but say the law doesn’t apply. You’ll be unsurprised to hear I’ve hung up my Amex.

  I’ve refused love holidays offer of a credit note and am hoping am waiting and hoping they go under so the insurer will have to step in. First World problem indeed given all that’s going on. Especially as the money will only go toward a new holiday, which could be an awfully long way away ! Hoping for Feb half term ?
• 11 BBlimp April 25, 2020, 9:41 am

  @Tony – are you sure you’d want to ? I’d swap numbers on page to get the cleaner back, get in the pub and out for meals and jet off somewhere in a heartbeat !
• 12 A Betta Investor April 25, 2020, 9:50 am

  This event surely proves two things:

  One:- All funds went down together, whether active or passive, so there is no point in trying to pick an active fund that didn’t.

  Two:- The level of the FTSE or any index is not really relevant. This collapse in equity valuations is a good reminder that the bulk of equity returns comes from dividends and reinvested dividends rather than capital growth.
• 13 xxd09 April 25, 2020, 9:52 am

  I am disappointed Investor that you are no longer posting your active plays on this board in the investment world
  I am a passive investor myself but that is not a static situation. Constant reminders from the active investing front are very valuable to reinforce my position or make me rethink it
  Constant debate is always needed- in investing as in anything else-nothing is forever even passive investing
  Some people can make active investing work -whether that is true for most investors is a good debate to have
  Maybe you will change your mind?
  xxd09
• 14 cat793 April 25, 2020, 9:52 am

  Thanks for the link above to the article by Peter Attia. I think it is the single best thing I have read so far on this pandemic. A calm, rational, articulate analysis by an expert who seems to have integrity.

  Like Snowman and TI I have been trying to follow what is going on but have become increasingly disillusioned by the mess of speculation created by the unholy combination of seemingly grandstanding, publicity seeking “experts” and sensationalist “journalists”. We know that the media are generally a bit of a joke already but it seems Michael Gove was not entirely wrong in his denigration of experts either. I guess it comes down to quality of character and honesty like so many things.
• 15 Griff April 25, 2020, 10:02 am

  I read this website and to be Frank alot of the time it goes over my head. Either it works out I am not as clever as I thought I was or, your contributors write the comments to as complicated as they can in order to elevate their own esteem. I did go heavy into vanguard equity trackers in December and am now sitting on a year’s wage in losses. Never mind. Some of your links I ignore, hence probably why I am below par in my investing nouce. But, this time I did click on the one about you and your dad. I’ll be honest. I think it’s the best thing I’ve read on your website. It really did make me think. Thank you.
• 16 MrOptimistic April 25, 2020, 10:21 am

  Perhaps the new enemy is restless boredom: all the talk of maintaining restrictive measures for years ahead is unrealistic and it is unwise to dash the hopes of early relief from this. Better to stay mum. No point having or earning money if you can’t spend it, and resentment over making sacrifices for the older generations can only rise which will encourage resistance to restrictive measures. If I was under 40 years old I’d be starting to brood on this: another couple of weeks, ok. Two years, no way !
• 17 Matthew April 25, 2020, 10:55 am

  I can help to quantify the risk of death, in my cold way

  Covid can kill, but we will all die at some point, all that covid might change is when that will be. We accept the risk of death whenever we climb into a car, I found that I couldn’t be calm on the road until I accepted that risk. We are sometimes born with defects but we accept the risk when having children as its the only way to make a genetic offspring.

  In 100 years time when people are looking at old gravestones will it really matter who or how? When its your time it’s your time.

  Enjoy the time you have, I think a lot of religion too is too much about death – as a father I didn’t brink my kid into the world to die, I brought him in to live. Life is our chance to do something while you have a physical presence in this world, be it morally satisfying or enjoyable.
• 18 Guido Maluccio April 25, 2020, 11:07 am

  @cat793 I think many of the medical experts advising our leaders are struggling not because of a lack of expertise in epidemiology but because we don’t know enough about this virus to simply apply science to policy. In a situation like this you need experts in uncertainty and risk analysis rather than virology.

  When uncertainty dominates and decisions have mortal consequences then I think you should be cautious. The outcomes from the virus are asymmetric, the risk of not acting has fat-tails. The precautionary principle should be applied when you don’t know enough about the crisis that you are dealing with. The consequences of not acting decisively to stop the virus are many more unnecessary deaths.
• 19 Neverland April 25, 2020, 12:07 pm

  @optimistic

  “all this talk of maintaining restrictive measures for years ahead is unrealistic”

  Who knows – not anyone on this board for sure

  Just as an example the uk had rationing from between 1939-1954 of one thing or another

  Unimaginable before, unimaginable now

  But reality for 15 years and nearly a decade after the fighting ended

  Just because you can’t imagine it … doesn’t mean it won’t happen
• 20 JohnG April 25, 2020, 12:36 pm

  @MrOptimistic – I actually think there could be some considerable indirect benefits from the Coronavirus lockdown; though I don’t intend to make the case that they do, or don’t, outweigh costs.

  Just a few that come to mind:
  > We’re commuting considerably less, generating less pollution and using less of our time doing something we don’t generally enjoy.
  > It may just help increase people’s perception of the value to society of some of the hard working staff in typically lower paying work; and that not everyone who needs some help from the state (furlough for example) caused their own problems and doesn’t deserve it.
  > It is challenging our preconceptions about our generally consumerist society; it’s too much to hope that our addiction to fast fashion will go away, but perhaps if people realise they can go a few weeks without buying more disposable rubbish we’ll be better off.
  > Anecdotally I’ve seen a real trend towards people who are less physically active doing some/more exercise.
  > It’s making people think about sourcing and supply chains. I’m a little worried this could lead to protectionism and encourage some nationalism, but generally think it’s a good thing that we’re considering how wise, and sustainable, it is to be entirely reliant on shipping critical things from half way around the world.
• 21 Factor April 25, 2020, 1:36 pm

  My approach to all the things that I read or hear about Covid-19 is the same as my approach to anything else that I read or hear, namely I ask myself, “Is it fact, or is it (all too often) opinion adduced as fact, or is it opinion and said to be so?”

  My acronym for opinion adduced as fact is OAF!
• 22 Sara April 25, 2020, 1:42 pm

  I work for the NHS but luckily for me far from the frontline and have mostly been able to work from home. I’ve found over the past 7-10 days that my brain has started to come out of total fear mode and is starting to accept that we are going to have to live with this virus as a risk for a long time yet. And thus to consider how to run the department on site while trying to apply physical distancing as much as possible. But it won’t be easy – I share an office and sit less than a metre from the other person. Office not actually big enough to be 2 metres apart so how are we going to manage that? Plastic screens? I’m not sure of the answers but am starting to think about them. I think we will just have to apply the rule as best we can but accept that it just won’t be possible to be physically distant in every situation. As we all experience in supermarkets now.
  And presenteeism – what about being sick with other things? Many people struggle into work with colds because you get a kicking from above if your sick rate goes over a certain level. It didn’t make you popular with your co-workers before – it’s going to make you the equivalent of Typhoid Mary now. I would hope that organisations relax the sickness rules for now – but many organisations are not renowned for their common sense….
  Sorry – this is totally not investment related – unless you make plastic screens 😉
• 23 Linda April 25, 2020, 1:51 pm

  @Griff
  There are lots of us who feel the same as you do, we don’t leave comments very often because we don’t feel qualified. I do however, read the comments & sometimes find a gem of advice that helps me.
  Don’t feel alone.
• 24 Bob April 25, 2020, 2:41 pm

  “… inform my naughty active investing.”

  Created a mental image of lady in high leather boots shouting “Bad Investor.!”
• 25 Matthew April 25, 2020, 2:47 pm

  @linda and @griff – I think the problem is that the important investing advice is unchanging and you already know, but by nature of blogs that have to keep churning out content to get views, they all have to bring up niche topics to avoid repetition. Some people I think do talk about these niches as a hobby or job or general boredom with the good advice that is passive investing
• 26 Neverland April 25, 2020, 3:33 pm

  Just been to the large local park

  People having picnics, sunbathing and on the terrace of the nearby pub which does ‘takeaways’

  So much for lockdown …

  … it only seems to apply when it’s not sunny
• 27 Dave April 25, 2020, 4:53 pm

  daily fatalities reported by the media are utter cobblers

  today TV head Iines are reporting 711 died today

  The NHS own figures (https://www.england.nhs.uk/statistics/statistical-work-areas/covid-19-daily-deaths/)

  show this number is the TOTAL who died since 11th March. NOT the daily number. Today’s actual number is 104 (yesterday24th)

  THE TOTAL IN THE PAST 40 days!, Not the daily number.

  of course for the click bait , 711 dead is a lot better than 104.
  what they are reporting is a running total.

  think about this for a minute. 60+ million people under “quarantine” for 100 people dying. Remember that these 100 people are already padded as no one now dies of regular flu, it’s all listed under Covid19

  good job no one cares about the number of people who die from cancer, malaria etc
• 28 Jonathan April 25, 2020, 5:26 pm

  But @Neverland, did they maintain approximately 2 metres distance between household groups? That is all that matters to stop virus transmission.

  Understandably the government have used proxy measures to check whether social distancing measures are being adhered to. In the daily briefings they chose car use, presumably because that is data already collected. So straight away police forces took their remit to be to reduce car journeys (looking at you Derbyshire police). That may have made the graph look better, but didn’t reduce infection risks, those dog-walkers in the drone photo were at least 2 hundred metres from anyone else.

  Similarly the presence of families are sunbathing and picnicking in parks does not necessarily mean increased risk of infection between households. Unless different families were getting together in close proximity, which is not what you said you observed.
• 29 Dawn April 25, 2020, 5:27 pm

  Emotional rollercoaster for me. Fighting to get holiday money back, refusing vouchers. 2 cruises booked later on ,no idea what’s gonna happen, do I pay balance,? do I wanna go anyway with this virus about?. Carn,t work( hairdresser) missing it, only do part time which I love. Won’t get self employed allowance ,long story as 18/19 return was half Ltd company half back self employed. Thankfully dont need it, as I have a tenant who can still pay his rent Lucky there. Also got my portfolio (cash part) to draw on. First market crash with sizable equities, but handled it fine, most of time, theres always that ,what if this time IS different that wells up every now and then. Almost grateful my parents have already passed ,the worry would be hell. Binge ate first 3 weeks, now got myself in order. Go 4 mile daily walks. See lots of folk I can chat to , Volunteer dog walking for isolaters. Weathers been fab☀️. Loads of on line chit chat to friends, clients and family. Live alone with my cute cat ,feel grateful I have a beautiful home, all paid for. No debt . No screaming kids to entertain and no partner beating me up !, not that I ever have had, no way hose, but as domestic violence is on the up, some women would prob give thier right arm to have my peace , safety and tranquility. So far so good here!
• 30 Learner April 25, 2020, 5:42 pm

  I find myself agreeing with Matthew on the overemphasis on death or mortality. I have good life insurance. Contracting a debilitating lung condition for life or suffering a stroke following c19 on the other hand, that is scary.
• 31 Neverland April 25, 2020, 5:51 pm

  @jonathan

  I find your comments a bit surreal

  There are whole bunch of new laws

  One of them says non essential businesses should be closed – a pub isn’t essential

  You are only allowed out of your house for an hour of a day for exercise, inter alia – sunbathing and having a picnic isn’t exercise and isn’t among the other allowed purposes to leave your house

  Laws are not optional just because you don’t agree with them or they don’t suit you
• 32 Matthew April 25, 2020, 6:27 pm

  @learner – indeed one of the most upsetting times at work was when someone who had a stroke was trying to say something to me and couldn’t, and got upset about it, which I really didn’t want, but I think when we do lose functionality, like with stroke, dementia, physical disability, the best thing is to not blame yourself and not expect yourself to be able to do the same things or get frustrated if you can’t – you would’t expect someone else to be fully functional so you shouldn’t expect it of yourself in that circumstance. I have learned to gracefully accept that I might need care one day, its just part of having a human body, and sometimes luck goes against you but you’d be far from alone – and in many ways we’re already lucky to live in a time with medicine, technology, relative comfort.

  When i’ve seen people go it seems to be either quite sudden or they tend to be barely/un concious – the individual usually doesnt seem to suffer too much I think – its more upsetting for family than the individual (I think)

  Lets put it this way, they dont seem to make noise or wake up in pain immediately before, I think that although what causes one to die might be painful, and treatment might be painful, death itself is like going to sleep – you lose your ability to percieve what is happening or feel pain, and conciousness goes beforehand, usually
• 33 Vanguardfan April 25, 2020, 6:28 pm

  @Dave, no, you are incorrect. The daily deaths announced give the new deaths that they are declaring today. Yes, most of these did not occur today, as there are variable delays in when deaths are declared. Most of the deaths that have occurred today will show up in future statistics, there will for sure be more than 104. The total hospital deaths announced so far is given in one of the spreadsheets in the link, and is more than 20,000.

  A more accurate figure, but which suffers from a 2 week lag, is the ‘excess deaths’ from death registrations, which includes deaths outside hospitals. This was showing something around 17,000 up to 2 weeks ago.
• 34 Vanguardfan April 25, 2020, 6:45 pm

  @neverland, there’s nothing in the law that says you can only leave the house for an hour’s exercise a day. The law specifies reasons you can leave home, and exercise is one of them.
  I agree, sunbathing and picnicking aren’t specified reasons, but I do feel for people in cities, especially those without their own outside space. It is of course harder for city dwellers to effectively socially distance. I’m just back from a glorious walk exploring a new little piece of paradise nearby, almost completely absent of people, and feel very lucky.

  What I’m feeling right now is a sense of stasis. In Feb and March, things were moving so quickly and we had to adapt to all sorts of changes on almost a daily basis. Back then, I think I assumed that the pace of learning and change would continue (I remember being surprised at an event in late June being cancelled). Now, it seems that new and useful knowledge about the virus is taking a very long time to emerge, and what is emerging, is not very encouraging.

  I think this is a result of the phenomenon described very clearly in the Peter Attia article.
  I also 100% agreed with the linked Malcolm McKendrick article, to my surprise. The policy of dumping known and suspected Covid patients into care homes is likely to have cost many lives among residents and staff. To be fair, many Western countries have had similar experience in care settings. I don’t think it’s going to be easy to get on top of, given disorganised testing approaches, PPE shortages, and the problem of asymptomatic transmission. I’d like to know if Germany has cracked this one, and also the situation in Asian countries.
• 35 Frankie Da Tory April 25, 2020, 7:06 pm

  21% in New York City doesn’t mean 21% globally. It doesn’t even mean 21% in New York State (13%), and that’s one of the top few worst affected regions on the planet.

  Massive swathes of the rest of the world have legitimately barely seen any cases. The WHO’s estimate of 2-3% globally is probably if anything too high.

  If the infection fatality rate is about 0.7%, which seems to be what the studies are saying, give or take a few tenths for regional factors such as demography, and if the number of deaths in the UK will be about 40k when all infections up to and including today have concluded, then we’re looking at about 5.7 million cases so far in the UK. Far greater than the current stats but months and months short of herd immunity.

  Say the population gets used to hearing about just under 1,000 deaths per day. In fact let’s say they can tolerate 1,500 deaths per day. Then every day there will be another 214k cases. Even at that rate it’ll take more than 200 days reach herd immunity.

  There’s a long long way to go.
• 36 Ecomiser April 25, 2020, 7:12 pm

  @Dave. Vanguardfan has beaten me to it, you are reading the figures incorrectly. The first data link in the page you linked to (COVID 19 total announced deaths 25 April 2020) shows daily deaths in April in the range 484-855, excluding 21st-24th which are marked ‘Data likely to change, please refer to data notes’. Obviously the changes will be in an upward direction.
  Are the TV headlines reporting “711 died today” or are they reporting “711 deaths reported today”? My recollection is the latter.
• 37 Ecomiser April 25, 2020, 7:28 pm

  I’m 70, long retired, and self-isolating. I’m now doing 7 days a week what I used to do 3 or 4 days a week, namely sit in front of a computer screen reading Monevator and other fine sites across the Web, followed by sitting in front of a TV screen testing and adding to my knowledge. I could keep it up for months, provided my food store keeps getting replenished.
  But I do miss those other days, meeting friends, going shopping, hopping on a bus to beauty spots then taking a long walk, whenever I felt like it.

  My big fear is that immunity to COVID19 is short-lived, meaning no herd immunity and no effective vaccine, and I could be stuck in here for the next 30 years!
• 38 Neverland April 25, 2020, 7:49 pm

  @vanguardfan

  ‘ I do feel for people in cities, especially those without their own outside space’

  I haven’t followed this much but if I understand correctly, the less any government restrictions are followed, the longer the restrictions are likely to last, the more it is likely to cost and more people die, no?
• 39 Vanguardfan April 25, 2020, 7:54 pm

  @neverland, yes, but one can have sympathy for people in a worse situation.
• 40 The Investor April 25, 2020, 8:01 pm

  @Griff — Thanks for that, very generous of you to share that, and cheers for reading the site.

  @Snowman — I hear you. I’m going to try to chill out a bit this weekend myself to recharge.

  @Frankie Da Tory — Your numbers are one way of forecasting things, fair enough. And you’re right of course about the WHO number being a global number.

  However I saw it being used several times to refute the theory that places London and New York had probably not seen massive levels of infection already. That was why it was such a striking contrast I feel when the NYC antibody numbers came out at over 21% (although as I say I’m actually disappointed they’re not higher and hope they will rise as more antibodies develop over the week. It’s a hope though, who knows).

  Regarding your scenario, a few things to think about.

  Firstly, if places outside of the big cities like London/NYC/Madrid have lower infection rates as shown by antibodies, that tells us something very interesting and potentially useful, which is the virus is probably less infectious outside of these big cities.

  Of course in some cases it could be it got to the provinces later and didn’t have time to multiply before lockdown, but I don’t find that hugely plausible given how much people move around a small country like the UK. Give or take I’m sure it had plenty of time to multiply, perhaps missing a final doubling, say.

  If it is less infectious outside of big cities (less public transport? less many-to-many mixing? less pollution?) then it could be that it can wither on the vine there over time. I.e. Perhaps there’s a good chance it will take just general physical distancing to bring R0 under 1 in these lower-antibody areas.

  A second thing to point out is there’s a lot of discussion about the size of infection required to achieve herd immunity (a term some were denying existing as valid without a vaccine a few weeks ago, interestingly, though I was not one of them).

  I’ve seen 96% quoted by a reader here, for example, which is almost certainly not going to be the case. In fact even the likes of 70-80% figures are usually drawn from much more infectious diseases such a mumps.

  In contrast I’ve seen numbers as low as 30% estimated for Covid-19. Clearly I don’t have the answer, but maybe 50% is a good ballpark?

  Anyway *if* such herd immunity was achieved at 50% then of course half of the deaths in your forecast would not happen, right, because the virus would peter out before it got to half the population?

  Obviously there’s a lot of moving parts to the speculation above, but I think one can at least see a scenario where herd immunity in the big cities and a lower R0 outside of them sees the virus become less of a threat long before the worst case scenario — especially if we can switch to a demographic/targeted policy of really protecting the old and vulnerable, so fewer of those in the death-zone get infected before immunity is reached. That’s one huge reason why I favour getting the young out sooner rather than later (and probably before even the older middle-aged).

  Of course perhaps as the WHO is cautioning today there is no immunity. I think they’re saying there’s no proof there is, rather than there’s evidence that there isn’t, and I believe there probably will be at least a temporary immunity from all I’ve read, but we’ll have to wait and see.
• 41 The Investor April 25, 2020, 8:08 pm

  p.s. Actually I’ve just realized I may not be reading your maths correctly, there’s a few workings off page. 😉
• 42 Neverland April 25, 2020, 8:19 pm

  @Vanguardfan

  A friend of mine is Peruvian from Lima. Peru is under curfew from 8 pm (from 4 pm in some parts) and only one person per household is allowed to leave the house and only to buy groceries, medicines or go to the bank on any one day. Only men can leave the house on Monday, Wednesday, and Friday. Only women are allowed outside on Tuesday, Thursday and Saturday. No one is allowed out on Sunday.

  Its just an example I know about properly. People are literally starving under lockdown in India.

  My sympathy tends to be elsewhere than people who can’t have a picnic in the UK.

  P.S. She isn’t Paddington Bear
• 43 Dave April 25, 2020, 8:42 pm

  The JustETF article “ How much risk should you take?” strongly suggests that blended portfolios of equities and bonds outperform 100% equity portfolios over the long term. This seems like a controversial point of view- is it generally true?

  I’ve always thought that a 100% equity portfolio would be expected to outperform a mixed portfolio over the long term, albeit with higher volatility. Even on a risk-adjusted basis I would be surprised if the 100% equity portfolio returns were inferior… but perhaps I’m wrong?
• 44 WhiteSheep April 25, 2020, 8:49 pm

  Regarding the threshold for herd immunity, I believe the formula typically given is 1 – 1/R0. So if R0=3, then the threshold is 67%. The basic idea is that if less than 1/R0 of the population have *not* yet been infected, and each infected individual transmits the virus to R0 others, then each remaining patient will pass on the disease to less than 1 others once the threshold is reached and the infection chain will die out. So a 50% threshold would imply R0=2, which seems lower than most estimates.

  Obviously it is a crude approximation. Even if R0 is a constant, there will be heterogeneous mixing in the population that will likely affect the simple calculation.
• 45 Jonathan April 25, 2020, 9:04 pm

  @Neverland, I wasn’t suggesting law breaking. I was pointing out the risk of proxy measures becoming the target so that effort is put in to stopping activities which are irrelevant to virus transmission. Unfortunately the guidelines (I think few of them are actually laws) appeal to some people’s sense of puritanism.

  @TI, I share your feeling that quite a significant proportion of the population may have been exposed, but what data there is still doesn’t allow any confidence. The King’s College link was interesting for suggesting symptom rates have decreased from around 10% in London to nearer 2%; for much of the country though percentages appeared about half of that. Obviously that is ambiguous, we know some individuals have had no clear symptoms and others will have had a similar symptom profile from colds or hay fever, but it might be suggestive of an overall exposure of 15-30% of the population depending on where you live. I hope down the road KCL will be able to do a random sample of their volunteer population for antibody testing so they can get a proper indication of the relationship of symptoms to actual infection.
• 46 ZXSpectrum48k April 25, 2020, 9:20 pm

  @Dave. The JustETF article is sort of appalling. Why would you compare the MSCI World Equity Index with a German Government Bond portfolio? Like why just German bonds? And why just the most recent 20 years? It’s a very arbitrary piece.

  Equities outperform pretty much everything over a long enough time frame. That’s because they are the riskiest major asset class in volatility terms. Most asset classes tend to deliver 0.3-0.6 units of return per unit of risk over the very long term.

  To be fair, the last 20 years, have been poor for many equity markets. The S&P500 has only delivered about 5.5%/annum since 2000; the FTSE All-Share just 3.6%. By comparison, the last 20 years have been very good for bonds. In many countries, long duration bonds have easily outperformed equities, delivering 7-9%. So while equities tend to outperform everything eventually, you do need to bear in mind that ‘eventually’ could be multiple decades. Hence the attraction of diversification.
• 47 The Investor April 25, 2020, 9:35 pm

  @ZXSpectrum48k @Dave — It’s a German company, I think they probably ported it over into English but kept the German data. I actually liked/included it for that reason, made a nice change! But perhaps people could have missed the Germanic origins, and thus it warranted a bracketed comment.
• 48 Dave April 25, 2020, 10:00 pm

  Hi vanguardfan

  Sorry but I have to disagree. If you click through the nhs data they clearly have daily numbers which they adjust as they get more data. They do state that on the website.

  However most the media throw out numbers to scare people eithe deliberately or because they are simply reading a figure given to them

  Ask the average person what they think is the death figure “today” I tried it with various people and they all think it’s several hundred every day.

  Ps..also a vanguard fan 🙂
• 49 Dave (not the other Dave) April 25, 2020, 10:06 pm

  Thanks spectrum, that’s what I thought. Nice to hear I’m not going mad!!

  @TI Even if it was originally written for a German audience, surely their conclusions about asset allocations are ‘bold’! Nothing to say german government bond yields will be the same in the next 20 years, etc etc
• 50 Dave April 25, 2020, 10:16 pm

  Hi vanguard fan,

  Just checked the daily mail. King of all click bait.

  Lead story
  “ NHS England confirmed a further 711 new deaths of people who tested positive for Covid-19 in the country ”

  “New deaths” they clearly imply it’s a daily total.

  The 711 is the total since March 11th.

  2 days ago the total shown on the NHS stats was 514

  They even have a column “pending verification”

  I agree with many posters here and our host that the economic destruction far exceeds “pandemic”.

  450 people a die of cancer in this country every day. 550k people admitted to hospital every year with conditions directly attributed to Smoking.
• 51 Vanguardfan April 25, 2020, 10:17 pm

  Um, yes, I did click through. What I said is correct, the deaths announced today are not all those that occurred today, because of delays in collating and announcing the deaths. We won’t know the exact number of deaths in hospital ‘today’ for some time, because numbers are added all the time. However, if you look at the ‘total deaths announced’ spreadsheet, although we can’t know the exact number instantly, its clear that we are very much in the ‘several hundreds per day’ figure. It’s certainly going to be a lot higher than 100. And of course there are all the care home deaths not included in these figures at all. I’m not really sure how else you expect them to be reported – they can’t leave out the late declarations.
• 52 Jura April 26, 2020, 7:47 am

  Dave mate you’ve lost your tin foil hat- time to pop it back on
• 53 Snowman April 26, 2020, 8:52 am

  The reported daily hospital deaths are giving a misleading impression for sure, especially in the way they are reported by the media (“number of UK deaths fall yesterday” type headlines)

  If we take the headline hospital deaths announced in England each day, which are all the covid-19 deaths registered up to 5pm the day before then:

  The 711 England hospital deaths announced yesterday on 25th April occurred (by date of death) on average on about 16th April.

  Whereas the hospital deaths announced on 22nd April (3 days before that) occurred on average on 17th April

  So the deaths announced yesterday actually took place on average on an earlier date than deaths announced 3 days before that!

  In the 711 deaths in England announced yesterday, 6 occurred on either 17th or 18th March for example.

  If you plot the graph by actual dates of death (rather than date reported) then you can see the daily deaths falling away in England, although you have to ignore the last week or so of figures and definitely the last few days (or try to adjust) because there are lots of deaths not yet reported to be added in.

  A seven day average of headline reported deaths gives a reasonable feel for things, but you have to watch that increasing delays in reporting deaths is distorting what you are seeing.

  In line with my approach to take a less angry approach, have to say the ONS (Official for National Statistics) are doing a great job here and their weekly statistics and analyses are very clear and open.

  (my own analysis in this post but think these figures are about right)
• 54 Snowman April 26, 2020, 9:03 am

  The peak in deaths in England was around 8th April by the way

  (looking at deaths by date of death)
• 55 Snowman April 26, 2020, 9:15 am

  And given the delay between infection and death, and a further delay as some infection is passed on to others within the same household, does that mean that the number of new infections was starting to fall even before the lockdown was announced?

  Lockdown: 23rd March
  Peak in England deaths: 8th April

  I’ll leave that for others to discuss.
• 56 Jim McG April 26, 2020, 9:31 am

  Reading the comments here, I’m totally confused. I’ve assumed that the daily deaths have been trending upward toward the current total of around 700 “per day”. So that’s how we’re now, cumulatively, at around 20,000 deaths from the virus since they started counting? Is that wrong? Never mind the simpler question over whether those 20,000 deaths are “from coronavirus” or “with coronavirus”, which you can’t seem to get a straight answer to. And what’s an “underlying health condition” that possibly contributed to that death? Age, weight, race, smoking, asthma, the flu, cancer, diabetes, hay fever, urban living, all of the above? None of the above? Some of the above? When you compare UK stats with, say, Sweden, are we comparing apples with apples? Are they recording their deaths EXACTLY the same way we’re recording ours? At which point in the data series do you start your analysis, and when do you finish? I tried to read the Peter Attia article and gave up in frustration as I just couldn’t follow it. I used to think that “Big Data” was going to be a game changer in our world but this virus is exposing it as being as vulnerable to subjective analysis as anything else is. What we need is a reiteration of simple advice that we can all follow as we exit this lockdown – be mindful of keeping your distance in social situations especially in regard to the age of people; wash your hands effectively; understand your personal vulnerabilities and act accordingly; use your common sense; don’t listen exclusively to Piers Morgan. And remember, nobody gets our of here alive.
• 57 Passive Pete April 26, 2020, 9:39 am

  @JimJim 6ft – I had to Google that, as I thought you might have been wearing your beer goggles while cycling , but no that’s as big as grass snakes can be. I’ve only ever seen one about 2 ft long. However, earlier this week I saw a dead adder on the road while cycling in the NY Moors. And I’ve been surprised by the continued amount of road kill, given the scarcity of traffic in recent weeks. Possibly the animals have got used to virtually no traffic and so have no fear when the occasional car comes along.

  @xxd09 I agree that it would be great to hear more active antics from @TI but I understand that it might confuse the Monevator brand and general site message. Maybe he can guest post on another site, or set up another ‘dark side’ brand. I still appreciate the occasional active type posts, such as the introduction to Seedr. But now I’m invested there I don’t keep up to date with the general pool I’m in.

  @Vanguardfan I’ve found Malcolm Kendrick’s posts and books very good at explaining medical issues in a simple way for the non-medically trained person to understand. His speciality is cardio vascular decease and statins. His previous post was also interesting as he described his predicament when writing out a death certificate while visiting a care home recently -another death missing from the daily Corvid 19 figures!
• 58 Matthew April 26, 2020, 10:58 am

  I read somewhere that death is the leading cause of mortality;

  https://www.theonion.com/world-death-rate-holding-steady-at-100-percent-1819564171
• 59 The Investor April 26, 2020, 11:00 am

  @Jim McG — The daily deaths peaked over two weeks ago. The data comes in with a lag. The deaths reported on the news are “with Covid”, and may not be the primary cause of death or even a factor. With that said there’s no doubt hospital wards have seen thousands of overwhelmingly elderly people die a horrible death due to Covid-19; there is plenty of Covid-19 dying going on, contrary to the more extreme view of some.

  Mostly the international stats are broadly comparable (e.g. the FT backs out care home deaths from France’s official figure, which counts them) but not entirely — Germany for example only counts deaths as due to Covid-19 if it’s on the death certificate as the cause, I believe I read.

  The true deaths figure “with Covid” is way way higher than 20,000 — probably at least 40,000. However the deaths caused “by Covid” will be lower than that, in my view likely quite a bit lower.

  The data is very confusing because the picture is very confusing; the models and forecasts out because they had few good inputs. We’ve still not great inputs. However our strategy was mostly guided by fairly catastrophic modeling, IMHO. I assume the Government is reassessing given how we’ve seen the virus actually progress.

  @Snowman:

  Does that mean that the number of new infections was starting to fall even before the lockdown was announced?

  Indeed, this is a hot topic of debate with me and a couple of my virus obsessives mates away from here (e.g. including the guy who questioned Kendrick’s early QALY maths.)

  These guys have been much less sanguine/optimistic about infection levels and quality-of-life-adjusted years lost than I’ve been, but at least one of them has joined me in starting to question the need for full lockdown when the 8 April date emerged a few days ago from his own analysis.

  My view remains that full lockdown when we did it was a completely justifiable emergency/precautionary response — but also that we shouldn’t hang about in full lockdown for any longer than we need to.

  I wouldn’t be at all surprised to discover basic physical distancing gets 80% of the gains, Pareto-style, with full lockdown maybe reducing infection by a further 20% reduction in infection at the cost of making a global recession far more nailed-on.

  The other thing I’ve believed from the start is we should have super-aggressively targeted demographics in any isolation / quarantine strategy. While I think many of these care home deaths would sadly have probably happened anyway, given the time the virus had to spread before we locked down at the start and the infectiousness of the virus, it might have helped if we’d been laser-focused on isolating the over-60s and treating them with kid gloves rather than the ‘it does not discriminate’ strategy/message. I still suspect that a lot of these elderly deaths would more or less have happened soon (doesn’t make them less sad for their families, but it does mean everyone has to look at death stats through the context of 1,500 dying a day in normal times) but perhaps we could have flattened the curve with fewer people suffering what seems at worst a truly ghastly way to die?

  As it is, it’s looking to me like the pattern is something like — virus gets into a big city, spreads rapidly, very many people get it but few suffer any terrible symptoms (e.g. that school in France with immense infection rates but almost none with symptoms, or that Boston homeless shelter where half /c.150+ were infected but at that point at least no symptoms) but a few elderly and/or vulnerable get it and suffer far worse. Also a very few who are younger/healthy who suffer from an abnormal immune response that overwhelms them.

  However in part of this spreading, the virus inevitably makes its way into places where vulnerable people are congregated — hospitals and care homes. I think it spreads rapidly there, hits many of the elderly/vulnerable, kills a lot, and massively inflates the statistics.

  This is not to denigrate the NHS or its workers, who I am genuinely full of admiration for, but it’s been well-known for many years that hospitals are statistically unhealthy places to be, where you want to spend as little time as possible. Have a Google of all the thousands who were going in for routine operations and dying of MRSA (the superbug) a few years ago, for instance.

  I suspect it was spreading rampantly in such places in New York and London in March, and that’s where we’ll find a preponderance of the fatalities in time.

  This picture is the sort of thesis that makes me very dubious it’s worth continuing to destroy 35% of GDP (with all the knock-on consequences for wider human health and happiness and life expectancy) any longer. We need to start opening up, in stages, and seeing what happens, ASAP.
• 60 Vanguardfan April 26, 2020, 11:15 am

  When interpreting the dates remember 23 March was the start of legal lockdown, not the start of all public health/distancing measures. Schools closed a few days earlier, and many workplaces were putting remote working into place the previous week. And of course the hand washing was some time before that. So it’s not a simple cut point.

  Has anyone found a good data source for covid bed occupancy at a national level? Pressure on health service capacity continues to rise until daily deaths/discharges exceed daily admissions. Focusing only on new daily events misses that point.

  Daily covid admissions are have not yet declined below half the peak number, so clearly there’s still a lot of transmission going on and I think that’s why there is great caution about lifting restrictions. There are still way too many cases to be effectively testing and tracing if that’s going to be our key means of suppression after lockdown ends.
• 61 The Investor April 26, 2020, 11:36 am

  When interpreting the dates remember 23 March was the start of legal lockdown, not the start of all public health/distancing measures. Schools closed a few days earlier, and many workplaces were putting remote working into place the previous week. And of course the hand washing was some time before that. So it’s not a simple cut point.

  Indeed, but by the same token few (outside of the tabloid loons) who are saying it’s past time to scale down “full” lockdown are suggesting we mark the event with a Bacchanalian orgy.

  As we’ve discussed before, we might start by demographics or industry sector, and we should continue with basic physical distancing indefinitely. We could open sit-down restaurants but not stand up pubs (restaurants never closed in Stockholm). Most who can work from home should mostly continue to do so, but manual workers (those whose companies haven’t gone bust already, anyway) should be out working on building sites etc. (Most of the big builders shut down operations, though some have already announced plans to restart).

  People with symptoms should continue to stay at home and do that initial quarantine routine.

  Public transport is the thorniest one for me. Maybe mandatory masks would help, but anyway I can’t see the tube not remaining an elevated R0 hotspot until herd immunity.

  This is still the makings of a recession economy, most probably. But it could help us dodge a depression.
• 62 Snowman April 26, 2020, 11:50 am

  @JimMcG

  The England daily covid-19 hospital deaths by date of death can be found here

  https://www.england.nhs.uk/statistics/statistical-work-areas/covid-19-daily-deaths/

  The top download (currently COVID 19 total announced deaths 25 April 2020) gives you the graph where you can see the clear peak around 8th April. You have to factor in that there are deaths that have happened but haven’t been recorded yet which affects the bars on the right which continue to fill in as time passes. But almost all of the deaths around the peak have been recorded and are in those bars now, so the shape is clear.

  These of course are hospital deaths only in England where the person has tested positive for covid-19. The ONS weekly statistics for England and Wales are brilliant for showing excess deaths which is pretty much the key statistic, although they are a bit more lagged time wise currently to 10th April. They also break deaths down into settings so you can see where the excess deaths are happening whether it is hospital or care home etc.
• 63 ZXSpectrum48k April 26, 2020, 12:23 pm

  @TI. When you say -35% contraction of GDP, this is just a preliminary 2Q20 q/q annualized 2Q20 forecast from the OBR. It’s currently far below other forecasts (see https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/879642/Forecomp_April_2020.pdf). Moreover, they are also forecasting a massive re-expansion by 2021 with Mar 21 GDP at +17.9%. Looking at monthly or quarterly numbers here is pretty meaningless since base effects are going to be huge over the next two years.

  It’s much better to think in terms of calender year growth numbers. The consensus for 2020 growth is around -7.5%, followed in 2021 by +4.5%. So a net loss of -3% vs. trend growth in the 2% territory. So an effective loss of say -7% vs. trend over two years. It’s a very substantial loss of economic output but not quite as dramatic as -35%.

  If we look at yearly
• 64 Jonathan April 26, 2020, 12:24 pm

  @TI, I wonder if the scientists’ reluctance to recommend easing lockdown comes back to the modelling and the shortcomings of poor quality inputs which you remark on?

  At the moment testing is simply not at a stage to give responsive information about whether any relaxation is OK or not. Once a large enough number of tests are being done, then there might at least be an indication through rising positive results among the public. But even then, it only makes sense to test people with possible symptoms which appear after about a week after infection. And the current volatility means a rise might not be noticed unless it is by about 50%. If new infections started rising that much, even immediate clamping down would lead to a greater than two times increase in hospitalisations and potential deaths.

  Once those 100,000 tests a day are being delivered, there might be a strategy which allows more responsive tracking. But they would still probably want ongoing infections to drop to less than a half of what they are now.

  For what it is worth, like you I suspect that the main driver of reduced infection rates is physical separation. While there may some other measures that can then be relaxed, the evidence is actually poor. For example, if the relatively few infections that are spreading among the public actually come from the occasions when we shop, then allowing more shops to open would cause a surge.
• 65 ZXSpectrum48k April 26, 2020, 12:25 pm

  oops pressed send before I meant to. Anyway I’ve said enough!
• 66 The Investor April 26, 2020, 12:34 pm

  @ZXSpectrum48k — I think your comment was truncated? 🙂

  Good points, and I do appreciate the extreme magnitude of the current contraction is deemed to be temporary, and set to be partially countered by a bounce back when we come out of lockdown.

  IMHO though the longer we stay in full (or even partial) lockdown the weaker the bounceback is likely to be (albeit if it happens any time soon — i.e. before structural impairment — it will certainly be very large).

  There are fewer now arguing for full lockdown until a vaccine than few weeks ago, but in that kind of world I think a very substantial retrenchment would be very possible. To some extent the economy would likely have time and incentive to adapt, but as well as the massive demand destruction in a consumer economy, I think the widespread bankruptcies and job displacements etc would result in structural drag on the economy that it would take a lot of time to improve/replace. In a very imperfect way a parallel might be the early 1980s de-industrialisation, only one we opted into. Millions would be out of work for years, and lots of skills and institutional know-how and connections would be lost as companies folded and their talent and expertise was scattered etc. And then of course there’s the debts accrued, though I do understand you’re not very concerned about that aspect even in the long-term.

  You’re probably quite right that even this extreme result wouldn’t result in an extended 35% loss of economic output. But of course we don’t need anything so dramatic to produce a meaningful impact on both people’s quality of life and their health but also, arguably, politics and beyond. (See austerity and Brexit, on the back of what, a 7% decline over two years from memory?)
• 67 xxd09 April 26, 2020, 1:08 pm

  A touch stone figure for me from UK Gov is that 600000 UK citizens die every year
  That’s 1600 a day -probably more in the winter and less in the summer
  I find it hard to relate any of the current statistics on Covid deaths to these two “hard “statistics
  Fault is mine I’m sure
  xxd09
• 68 Tony Edgecombe April 26, 2020, 2:10 pm

  “I’ve seen 96% quoted by a reader here, for example, which is almost certainly not going to be the case.”

  That was possibly me and I wasn’t quoting the herd immunity threshold. People see a figure of 80% and assume that the disease stops at that part of the population being infected. Rather this is the point at which the replication rate equals 1. There is still further to go before the disease is eliminated.
• 69 W April 26, 2020, 3:02 pm

  Quite a bit is being pinned on the hope that a large enough percentage of people have had this virus and been asymptomatic.

  However is there any evidence that this gets us closer to herd immunity? Is it not the case that those who have tested positive having had milder symptoms have generally had far less antibodies and therefore probably insufficient to confer future immunity?

  If this is found to be the case, then governments have some impossible decisions to make absent any vaccine.
• 70 Ecomiser April 26, 2020, 3:04 pm

  @xxd09 If 1600 UK citizens die each day, on average, in normal times, and 400-800 per day are dying in English hospitals of Covid19 this month, that suggests ¼ to ½ of normal deaths are being added by SARS-Cov2, ignoring the difference between UK and England, and the Covid deaths outside hospital.
• 71 Vanguardfan April 26, 2020, 3:15 pm

  Economiser/xxd09, have a look at the excess deaths statistics:

  https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/bulletins/deathsregisteredweeklyinenglandandwalesprovisional/weekending10april2020

  This gives the best estimate of the impact of the epidemic on mortality. You can’t just add the covid deaths to the average deaths per day as some deaths would have occurred anyway, and some covid deaths will not have been identified. For the week ending 10 April, there were around 8000 excess deaths. This includes the day on which the peak hospital deaths occurred (8 April), however, it’s based on date of registration rather than date of death, so we might expect the highest rate of excess deaths (in this wave) to come when the next figures, for deaths registered to 17 April, are released next week. So look for evidence of a fall in the week after that (it will be widely reported anyway).
• 72 The Investor April 26, 2020, 3:26 pm

  Edit: Please note, it seems I misunderstood the information we were getting, and not every death is being tested for Covid-19 regardless of symptoms or potential exposure. Please see the comment by @Vanguardfan below. This doesn’t mean people aren’t going into the daily death totals who didn’t die *of* Covid-19, they certainly are, but it would obviously imply fewer are. (The ONS estimated it was a contributor to death in 9/10 cases from memory).

  If 1600 UK citizens die each day, on average, in normal times, and 400-800 per day are dying in English hospitals of Covid19 this month, that suggests ¼ to ½ of normal deaths are being added by SARS-Cov2, ignoring the difference between UK and England, and the Covid deaths outside hospital.

  It’s not that simple. Remember the death stats are not “of” Covid-19, they are “with” Covid-19.

  Let’s imagine a fictitious virus, Cor-Blimey-20, which is massively infectious but doesn’t kill anyone. Let’s say it started spreading in February and by 26 April it was in half the population (50%). Let’s presume for the sake of simplicity that it was distributed entirely randomly, so it wasn’t much more common in some places than others, and thus it was distributed as randomly in hospitals where people are dying as anywhere else.

  If 1600 people died with Cor-Blimey-20 today, then 800 deaths would be chalked up as Cor-Blimey-20 related by the way we count the Covid-19 statistics, even though it doesn’t kill anyone. It would just be “there”.

  Now Covid-19 most certainly does kill people — a sizeable chunk of the old/vulnerable in infects. But still these daily numbers do not represent a clinical diagnosis from doctors saying they died of Covid-19. They are deaths that tested positive for Covid-19.

  People still seem to be struggling to grasp this, but it’s not a controversial point. (See: https://www.bbc.co.uk/news/health-51979654)

  Covid-19 is a notifiable disease, it’s checked at death, and so if it’s present it goes into the death toll.

  In many cases — maybe most — Covid-19 will have been a factor and very often the cause of death.

  But in at least some cases it won’t.

  On the other side of that we have all the deaths outside of hospitals. These again will be a mix of people dying of Covid-19 and dying with Covid-19. So again there are lots of real Covid-19 deaths not being counted by the official daily figure which excludes most notably care home deaths. But equally there are lots of simply “with” Covid-19 deaths excluded, too. Among these very aged populations.

  Roughly half of all positive-tested with-Covid-19 deaths are in patients over-80. And as has been said, 1,500-1,600 people die every day.

  Overwhelmingly these 1,500-1,600 daily deaths are people who are very old.

  It’s not heartless to see suggestive inferences in all these statistics.

  It doesn’t mean we shouldn’t care about them, it’s a price worth paying, etc etc. (My own suggestion has always been focus on supporting the isolation of them).

  It also doesn’t in itself mean full lockdown was the wrong strategy; the idea was always to reduce overall infection so few of the vulnerable got infected and also that fewer of them ended up in hospital, straining capacity.

  But it does mean the average person over the past few weeks who finished their Joe Wicks sessions, turned on the news, hears that 1,000+ people were killed by Covid-19 ‘today’, and then sees a few prominent case studies of some freakish outliers (young/healthy/similar) has got the wrong idea about what this virus has been doing.
• 73 Snowman April 26, 2020, 3:45 pm

  @Tony Edgecombe

  If 80% of the population immune is the point at which one person starts to infect less than one other, then that would imply an underlying R0 of about 5. At 80% and at an R0 of 5, 4 of the 5 people that any one person would infect would be immune, leaving one susceptible to infect.

  An R0 of 3 for covid-19 has been suggested if next to no social distancing measures are taken. Under an SIR model, if you completely let the virus spread at an R0 of 3 , even though one infects less than one at 66.7% the overshoot is considerable until the virus dies out in theory perhaps at 90%+ of the population having had the virus. That is the argument for herd immunity being at 90%+.

  But this is based on a simple SIR model which breaks down in practice for a number of reasons.

  Firstly temporary social distancing measures can be used at the time you reach around 66.7% to bring the R0 down below 3, to limit the overshoot so that only perhaps 70-80% of the population will have had the virus by the time it dies out. Once you are at 70-80% say you can relax measures, but as long as the R0 is kept at or below 3 any imported cases will not spread through the community although there will be small outbreaks that die out quickly if you don’t quarantine those coming into the country.

  Secondly the SIR model ignores the mathematics of shielding and assumes each member of the population has an an equal chance of meeting any other member of the population. At 66.67% immune and with an R0 of 3, then of the 3 people that an infected person comes in contact with, the SIR model assumes 2 are immune people and one a susceptible person. The extreme position of shielding illustrates why this is wrong. If one third of the population efficiently shield then if two thirds of the population have or have had the virus then the 3 people that an infected person comes into contact with are all people who are immune or who who currently have the virus.

  I’m no expert. And in practice it’s hard to say but 70-80% herd immunity doesn’t sound unreasonable to me but 96% does seem totally unreasonable. This assumes that having had the virus gives you immunity from getting it again.
• 74 Vanguardfan April 26, 2020, 3:54 pm

  @TI, you do realise that not ALL deaths are tested for coronavirus? And not even all hospital admissions, only those with suspected Covid? So the Covid deaths are all going to be in people who have displayed coronavirus symptoms. So whilst some of these might have a serious condition they are already dying from (eg a very ill hospitalised patient who caught it during their stay), the majority are deaths from covid (as that useful bbc article explains).
  So in your fictional example, if Cor-Blimey-20 did not cause any symptoms, no one would be tested for it and precisely no deaths would be recorded as due to it.
• 75 The Investor April 26, 2020, 4:33 pm

  @Vanguardfan — My understanding was indeed that all deaths (or very nearly all) are now routinely tested for Covid-19. So perhaps I have misunderstood this, if you do know different. I would love to see a source as I have struggled to find it very clearly set out anywhere, e.g. even in this ONS deep dive:

  https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/bulletins/deathsinvolvingcovid19englandandwales/deathsoccurringinmarch2020

  I agree a majority of the hospital Covid-19 counted deaths seem to be Covid-19 related. On reflection “maybe most” as I put it was too ambiguous.

  I am always leaning “maybe” partly because everything is so uncertain and partly because I am still not sure what we should think of (say) an 85-year old who has been in hospital for six weeks with 3-4 underlying health conditions who ultimately succumbs to Covid-19. It’s killed them, in once sense — certainly in the daily death toll sense — but did it really kill them?

  This isn’t pedantic, it’s at the heart of one part of the debate I feel. But I’m not at all well-placed to address/understand it.
• 76 Dragon April 26, 2020, 5:10 pm

  “… virtual drinks over video …. Joe Wicks’ workouts … Unlimited Internet porn ….”

  I think TI has accidentially let slip his daily routine! 🙂

  The question is, sequentially, or simultaneously?
• 77 Vanguardfan April 26, 2020, 6:27 pm

  @TI, I don’t have a reference, I am sure such a thing doesn’t exist, my source is experience of certifying deaths and of notifiable disease surveillance. I also did a quick check around those I know who are currently dealing with Covid patients in hospital. Approaches will vary of course, but this is the gist.
  Patients admitted to hospital are tested for covid if they have clinically suspected disease, symptoms that could be covid (even if not the working diagnosis) or known exposure. If they are being admitted for something completely different then they wouldn’t routinely be swabbed (for virus testing). If someone dies, and a post mortem is indicated (a minority of deaths), pathologists are routinely swabbing before doing a PM, for their own safety. If positive they do a limited PM to check for signs of covid pneumonia, if present covid will be recorded as cause of death. Someone coming into hospital and dying who is being treated clinically for something else, who has no signs or known covid exposure, will not be tested and will not be part of any covid death data.

  I am told that some hospitals are expanding testing to include screening of those with no indications of covid, one colleague whose hospital has done some testing of this type found only one positive.
  So no, it is absolutely not the case that hospital inpatients and deaths are routinely tested for covid. Your idea that there may be loads of hospital deaths in whom coronavirus is an incidental but unimportant finding doesn’t really relate to the reality. There will be some frail people admitted to hospital who catch covid there and die swiftly because of that, in whom covid won’t be the underlying cause of death, but it’s not going to be a big proportion.
• 78 The Investor April 26, 2020, 6:41 pm

  @Vanguardfan — Okay, I guess will have to see about the proportions in the fullness of time (we do agree it’s not the majority!) but that is useful information pending a written source so thank you very much for spelling it out.

  I’ll curb my ‘of/with’ distinction enthusiasm for now accordingly.

  (It’s not a massive part of my thesis as to what’s going on, but it was a small part of it).
• 79 TCA April 26, 2020, 8:22 pm

  Thanks for posting the JustETF link. Made for very interesting reading.
• 80 Naeclue April 26, 2020, 8:40 pm

  @Dave, it is good to see someone talking about finance! I actually thought the justETF article was ok and I don’t think they really suggested that balanced equity/bond portfolios outperform 100% equity portfolios in the long term. The article was really about how diversification across uncorrelated assets improves risk adjusted returns – the main free lunch in investing. It just so happened that over the last 20 years, equity and bond returns were very similar and in such circumstances mixed equity/bond portfolios can outperform 100% equity.

  In general, if you have 2 uncorrelated asset classes with returns R1 and R2, then the expected return on a regularly rebalanced portfolio will be greater than weighted average returns of the asset classes. So for example, for a 50:50 portfolio, the expected return will be greater than R1/2 + R2/2. If you have a 60:40 portfolio then the expected return will be greater than 0.6*R1 + 0.4*R2, etc. Provided the 2 asset classes provide similar returns, this effect can mean you get better returns from a regularly rebalanced portfolio of both assets than a portfolio of the highest returning asset. The closer the 2 returns and lower the correlation, the bigger the rebalancing bonus. In practice, the returns need to be very close to each other, as they are in the justETF article, for a equity/bond portfolio to beat 100% equities over 20 years and the rebalancing bonus would not normally be as big as 0.5%.
• 81 Indecisive April 26, 2020, 10:19 pm

  @Snowman #55, @TI #61

  Indeed, but by the same token few (outside of the tabloid loons) who are saying it’s past time to scale down “full” lockdown are suggesting we mark the event with a Bacchanalian orgy.

  The “Did infections start to decline before lockdown?” has become the new Brexit in my family (and split on identical lines). Like Brexit, a lot comes down to semantics over what is meant by “abandoning lockdown”.

  Given the panic and many people distancing 2+ weeks before ‘official’ lockdown, I’m not surprised by an earlier peak, but don’t see that as meaning lockdown wasn’t needed (to keep people doing what is a tiresome nuisance). I want lockdown to end and people to be trusted and encouraged to distance (where’s the nudge theory when you want it?). My family want lockdown to be abandoned so they can go back to normal, doing whatever they damn well please without socially distancing because “it’s inconvenient”.
• 82 Naeclue April 27, 2020, 8:25 am

  For a bit of entertainment I picked out the annual returns of UK equities and gilts from the Barclays Equity Gilts Study. For the 20 year period ending 31 Dec 1999, equities returned 19% per year, gilts 13.2. Here are the annualised returns from portfolios rebalanced at the end of each year

  %equity return weighted average return
  0 13.2 13.2
  10 13.9 13.8
  20 14.5 14.4
  30 15.1 15
  40 15.7 15.5
  50 16.3 16.1
  60 16.9 16.7
  70 17.4 17.3
  80 18 17.9
  90 18.5 18.4
  100 19 19

  There was a rebalancing bonus of 0.1% to 0.2% for every portfolio, but the 100% equity portfolio still did the best as the return from equities was so much greater than gilts.
• 83 Naeclue April 27, 2020, 8:46 am

  Another interesting thing I just spotted from the Equity Gilts Study is that in the 25 year period from the end of 1945 to end 1969, treasury bills (essentially cash) did much better than gilts, with cash just keeping up with inflation, gilts underperforming by about 50%. With gilt yields so low now, maybe the next 20 years will show a similar result.
• 84 Naeclue April 27, 2020, 8:57 am

  Just to add to my last comment, in the 1970s, cash performed about the same as gilts. It is really only from the 1980s that gilts have outperformed cash. The large drop in yields boosting gilt returns.
• 85 Boltt April 27, 2020, 9:01 am

  Hi TA/TI – I don’t recall reading any articles about preference shares (Here or anywhere) , is there one, or could there be one? Their frequent cumulative nature could be very attractive currently (inflation risk less so!)

  Whilst I’m making requests have you tried adding survey tools (Financial Samurai does occasionally) – it would be nice to see the opinion profile of the readership.

  Thanks for the good job.

  B
• 86 Vanguardfan April 27, 2020, 9:35 am

  @naeclue, is there any UK equivalent to treasury bills?
• 87 The Investor April 27, 2020, 9:41 am

  @Boltt — There’s a few articles from the ‘old days’ when I discussed my active ideas.

  Scroll down here:

  https://monevator.com/tag/preference-shares/

  I picked up some NWBD and LLPC near-par at the height of the sell off but have largely reduced those positions away now after they partially recovered (I retain small holdings, not sure if I’ll keep them medium/long-term.) Read up on the Aviva preference shares farrago of a few years ago if you haven’t already; there is now deemed a risk of preference shares being redeemed at par. Of the two mentioned here I think LLPC is a bigger risk than NWBD for this, but please do your own research and this is definitely not a recommendation. 🙂
• 88 Naeclue April 27, 2020, 9:46 am

  @Vanguardfan, yes, UK treasury bills. These were the ones I was talking about above, but not much point retail investors holding these as you can easily get better returns by shopping around for FSCS protected deposits and NS&I deposits/bonds. Institutional investors don’t have this luxury. They can put money in banks and become unsecured creditors, buy short dated corporate bonds, again becoming unsecured creditors of the issuers, or buy treasury bills/short dated gilts and have the government as the creditor.

  The return on treasury bills in the Equity Gilts Study is a good proxy for the return on cash, but retail investors can almost always find better returns on cash than are available on treasury bills.
• 89 Snowman April 27, 2020, 10:16 am

  Indecisive, interesting your family is discussing the did infections decline before lockdown issue. I don’t see the need to have a polarised view here, but I think it does help the understanding of when the R0 may have reduced below 1. To build on the dates I gave earlier

  Peak of deaths: 8th April (very clear from death data by date of death, why the daily briefings still use that messy reporting date graph I don’t know)
  Full lockdown: 23rd March
  Peak of infections (estimated as peak of deaths less 23-26 days): 13th-16th March

  As you say it is important to remember that a number of stringent measures were sequentially taken prior to lockdown, and public behaviour was changing also. So I think it is very difficult to support an argument that no measures at all were needed to stop the virus spreading and the NHS being overrun (although I’m not sure many were advocating that, more there was a debate on the scale of the restrictions e.g. something like the pre 16th March position vs the 16th March position vs full lockdown). My own position at the time was that I wanted lockdown to happen before it did as I thought there was a real danger the NHS was going to be overrun.

  But IF infections started reducing around 16th March it is interesting to consider what measures were in place then. On 16th March new measures were added in and after these measures the position was as follows:-

  – Quarantine for 14 days introduced (if family member has symptoms). Isolating for 7 days with symptoms was already in place
  – Advice to avoid pubs, clubs, theatres, and other such public venues (introduced)
  – Advice to work from home if at all possible (introduced)
  – Advice to avoid non-essential travel (introduced)
  – No government support for public gatherings (introduced). Cheltenham festival went ahead some days before on 10th-13th March but some other events had already being cancelled.
  – Vulnerable groups, over 70s or health conditions, advised to self-isolate (introduced, but had been pre-warned earlier)
  – schools still open (closed 4 days later)

  So you can make an argument that we could return to a regime similar to that after the 16th March charges, without infections getting out of control and the NHS being overrun. But even if you accept that you could stop the NHS being overrun with such a lifting of full lockdown, you might still choose to suppress infections even further by maintaining the lockdown.

  It’s understanding what the data says is possible and how that would impact infections and NHS capacity, that I’m interested in. If we can broadly understand that, then we can have the debate of what approach to take.

  Incidentally if anybody has data to improve, challenge or support the 23-26 days infection to death assumption I would be interested to see it.
• 90 ZXSpectrum48k April 27, 2020, 10:53 am

  @Vanguardman. The UK T-bill market is very small. Currently only around £85bn exists. All are 6-month T-bills, except for one which is 3 month. Unlike the US, where they are a key part of asset base for money market funds, in the UK they are only really issued for Government Exchequer cash management. The best short term proxy for “risk-free cash” is typically a short-dated Gilt, currently say 22Jul20, which yields an eye-popping 0.07%. Don’t spend it all at once.

  Be careful using returns from the Barclays Equity-Gilt study for Gilts. It’s not a market-cap weighted index. The construction of the Barclays Gilt index has also been altered so many times, it’s not even comparable with itself over time, never mind the market.
• 91 Naeclue April 27, 2020, 11:17 am

  Construction of Barclays Gilt Index.

  “The Gilt Index measures the performance of long-dated gilts. From 1899 to 1962 the index is based on the prices of undated British funds. During this period the undated stocks were a major part of the gilt market, but over the years the effect of high interest rates on their prices, together with the growing number of conventional long-dated issues, meant that undated stocks became less and less representative of the market as a whole.

  Since 1962, the Barclays Gilt Index has been based on a portfolio of long-dated stocks, selected on 1 January each year. The portfolio was chosen to represent as closely as possible a 20-year security on a par yield, and contains a weighted combination of four long-dated stocks with a mean life of 20½ years (so that the average life of the stocks for the year in which they are in the portfolio was 20 years). The combination and weightings of the four stocks are chosen to have the minimum possible deviation from a par yield. Small issues (less than £1bn) are excluded and in any year none of the four stocks has been allocated a weight of more than 40%, or less than 5% of the index.

  During the late 1980s there was a steady contraction in the number of issues that satisfied the criteria for inclusion in the Gilt Index. As a result of the lack of issues of new long-dated stocks and the fall in the remaining life of existing stocks, the universe of eligible stocks narrowed sharply. By the end of 1989 there were four stocks with a life of more than 20 years, and only two of these were over £1bn nominal.

  Thus from the beginning of 1990 the index has been constructed to represent a portfolio of 15-year par yielding gilts.”

  Not perfect, but not unreasonable. Had they stuck with 20 year gilts from 1990 returns would probably have been better.
• 92 ZXSpectrum48k April 27, 2020, 11:45 am

  Naeclue. This is just one example of why it’s bad “The combination and weightings of the four stocks are chosen to have the minimum possible deviation from a par yield.”

  This is a totally stupid assumption. First, why 4 stocks? Second, there is just no reason to have the combination close to par. These days that assumption would exclude the bulk of the index, since most issues are trading way above par. It’s a lazy approach to index building that could introduce substantial errors over time.

  I would just use the Gilt subindices of the two big government bond indices, Citi WGBI or JPM GBI. I don’t know how easily available they are to retail though, tbh.
• 93 The Investor April 27, 2020, 1:12 pm

  Meanwhile, back on Covid-19, a reader shared the following interesting Reuters story which once again highlights the potential that we’ve seen super-high infection rates, expecially in institutions — and very interestingly asymptomatic:

  As the virus spread, they began mass testing.

  They started with the Marion Correctional Institution, which houses 2,500 prisoners in north central Ohio, many of them older with pre-existing health conditions. After testing 2,300 inmates for the coronavirus, they were shocked. Of the 2,028 who tested positive, close to 95% had no symptoms.

  “It was very surprising,” said Chambers-Smith, who oversees the state’s 28 correctional facilities.

  As mass coronavirus testing expands in prisons, large numbers of inmates are showing no symptoms.

  In four state prison systems — Arkansas, North Carolina, Ohio and Virginia — 96% of 3,277 inmates who tested positive for the coronavirus were asymptomatic, according to interviews with officials and records reviewed by Reuters. That’s out of 4,693 tests that included results on symptoms.

  https://www.reuters.com/article/us-health-coronavirus-prisons-testing-in/in-four-u-s-state-prisons-nearly-3300-inmates-test-positive-for-coronavirus-96-without-symptoms-idUSKCN2270RX
• 94 Naeclue April 27, 2020, 1:17 pm

  @ZXSpectrum48k, “The combination and weightings of the four stocks are chosen to have the minimum possible deviation from a par yield.”. What they are doing here is constructing a hypothetical gilt, priced at par to enable them to identify a coupon. They then report the change in price of this hypothetical bond over the course of the year in order to separate out capital performance from interest. In the construction they are doing their best to avoid picking any gilts that might have artificially distorted that construction.

  For the purpose of my calculation on mixed equity/gilt portfolios in the 80s and 90s, I doubt it would have made the slightest difference to the conclusion that 100% equities beat all other portfolios if I used a different index, or constructed my own from individual gilts. I might have a look at this later, although neither of the index series you mention go back far enough. War loan price history might be an interesting comparison. The embedded call option would have been way out of the money to have had much of an influence on prices.
• 95 Naeclue April 27, 2020, 1:54 pm

  @TI, really interesting prisons article as such high sample sets. Shame there was no demographic breakdown though. There is this remark as well “Some people diagnosed as asymptomatic when tested for the coronavirus, however, may go on to develop symptoms later, according to researchers.”
• 96 ZXSpectrum48k April 27, 2020, 2:18 pm

  @Naeclue. The whole point of an index is to represent the market. You build a bond index by having clear and consistent entry and exit rules for the bonds (liqudity, maturity, market-cap), you construct a total return index for each bond on a daily basis, you weight by market cap, allowing for auctions and buybacks on a monthly basis etc etc.

  Picking specific Gilts to avoid “artificial distortions” is not what an index does. What you don’t do is try to construct some hypothetical par bond. That’s not an index at all. That’s might be useful if you want to define some par yield curve. Now does it make a huge difference? Depends on your need for accuracy, yield levels and the dispersion of yields along the curve. It won’t change whether Gilts beat Equities or vice versa but you could easily be talking about return errors of 1%/annum which in my view is huge.

  I probably come over as being pedantic but good index construction matters. If someone took just 4 stocks out of the FTSE or S&P or starting talking about constructing a hypothetical stock that represented the whole index, with the ‘right’ dividend, it wouldn’tbe taken seriously. It doesn’t for bonds either. I do get frustrated that people things govt bonds are somehow totally homogeneous, that there is no yield curve dynamic, no bonds that trade cheap or dear.
• 97 Grumpy Old Paul April 27, 2020, 2:25 pm

  @The Investor, Naeclue
  Re Reuters article. Age. See https://www.bop.gov/about/statistics/statistics_inmate_age.jsp . Only 2.8% over 65.

  Obviously we don’t know about the individual prisons but have no reason to believe that they differ markedly from the US average. The prison population is in excess of 2 million but not all are included in the FBOP stats.
• 98 The Investor April 27, 2020, 2:46 pm

  @Grumpy Old Paul — Interesting. To be clear these mass infectious / few-or-no symptoms stories are interesting to me because they (re-spins broken record) may suggest the virus is much more widespread than has been commonly believed, not that the virus isn’t serious/killing old people (it’s pretty much always been clear it takes a heavy toll there.)
• 99 Hare April 27, 2020, 3:07 pm

  @TI What about a monthly post called ‘The Naughty Corner’ for active investing.
  ‘Naughty Corner: [that post’s title]’

  I do mostly passive investing (meaning I buy and don’t sell unless crystallising a gain/annual moving from taxable account to tax wrappers). But we’re in deaccumulation and on HL which means often the best passive option are active (e.g. low charge ITs which provide consistent income and have .45% lower platform charge than funds).

  Your active posts, links and comments have provided valuable learning and insight and so do the comments on those posts.

  Labelling active posts ‘the naughty corner’ in the title with a standard first paaragraph with the usual disclaimers, keeps the passive message while providing learning for those of us who need it.
• 100 Vanguardfan April 27, 2020, 3:53 pm

  @TI. What this shows is that the virus spreads quickly in crowded enclosed communities.

  I think it’s important to keep thinking back to the principles of disease transmission. It requires connection between people. The closer and more prolonged the contact, the more likely the transmission. So outbreaks occur in connected time-space clusters, they don’t progress in an orderly ripple throughout the whole population. You can’t generalise the experience of outbreaks in prisons, nursing homes, cruise ships, migrant dorms and other confined populations to what may be happening in the free living community (which is clearly not one homogeneous population).

  This has implications for the design and interpretation of antibody studies (I realise this was an antigen study). If you want to answer the question ‘what proportion of the population has been infected’, then the first thing you need to do is define the population of interest. This might be a confined population such as a prison, or it might be a geographically defined area, or perhaps a workplace. Then, if the group is too big to test everyone, you need an unbiased sampling strategy. That means essentially that everyone in your sampling frame should have an equal chance of being picked. Ideally you have a complete list of your sampling frame, and make random picks. In practice, there are usually biases that creep in somewhere (people missing from the list, people who don’t respond, people who are unreachable – nearly always these people are not typical population members. Studies asking for volunteers, rather than choosing and approaching participants, are especially prone to bias). Finally, interpretation and generalisability. Strictly speaking, your sample can only be generalised back to your sampling frame.

  Clearly, if an infection becomes so widespread in a geographical area that most people have had it, then antibody prevalence studies are easier to interpret and less prone to bias. Earlier on, when maybe the overall prevalence might be in single figures, you might expect the distribution within a population to be very lumpy and therefore much harder to get an accurate ‘average’ from a small sample.

  All these things need to be considered when interpreting prevalence studies. Ask yourself, what is the population they are trying to measure? Are the participants likely to be representative of that population, or is the sampling very biased? To what extent can these results be generalised to the population I’m interested in?
• 101 Naeclue April 27, 2020, 4:14 pm

  @ZXSpectrum48k, surely you must be able to see the difference between constructing an index with 4 stocks and constructing a hypothetical gilt with 20.5 years to maturity from 4 long gilts? Yes it would make a difference to include more gilts, but nothing like the difference it would make to include more stocks. They are not trying to track the cap weighted whole gilt market, just an idealised gilt that they replace each year.

  There are often kludges in index construction. Just look at the DJIA or Nikkei. Even the S&P is committee selected.
• 102 Naeclue April 27, 2020, 4:21 pm

  It would be good to see more stats, such as the comprehensively sampled prison stats done for care homes. That should easily be possible to do.
• 103 The Investor April 27, 2020, 4:52 pm

  @Vanguardfan — All that’s fair enough, and I agree it tells us that the virus spreads easily in confined populations.

  However I infer more than that, because the article says these 95% of these very large populations were asymptomatic. (i.e. Not even ‘mildly suffering’. Asymptomatic. Notwithstanding the caveat that some may become ill later).

  You may recall back in early April you asked where the likes of me and @Snowman were getting our concept of large numbers of infections from? One notably rancorous board contributor predictably replied “wishful thinking”.

  The reality was I was inferring the strong likelihood of lots of more infections from data snippets such as this. Now it’s surely undeniable.

  Even if the virus spreads much less rapidly in a crowded city like London versus a prison, it suggests the cases/symptoms we were seeing were almost a drop in the bucket, assuming the 95% asymptomatic held true in wider population. (A big assumption I know).

  If we go back far enough the early estimates were that *maybe* 20% of virus carriers were asymptomatic. Later it because c.50%. Now we have 95% asymptomatic in this particular prison, which is another rare instance of a whole infected population being tested at once.

  I don’t claim it’s a model for understanding how the virus spread through Putney or Wolverhampton. I say it tells us that vast numbers of people can demonstrably be asymptomatic, and we can infer things from that.

  Of course as you say it suggests that when the virus gets into crowded situations (certainly a prison, possibly a tube train) it spreads rampantly. And since we are testing live virus infections here, you would have to ‘catch’ it at the right time to see the high infection rate. Because mass testing is still rarely being done on entire populations, for reasons I don’t understand (we had 20K-a-day spare testing capacity as of last week) we don’t have much data on that. However similarly strongly positive samples have been seen before (e.g. In that Boston homeless shelter).

  Presumably testing a couple of weeks later (perhaps after someone died?) would show just the tails of those infections, and far fewer infected, and we’d be back to thinking a lower number were asymptomatic.

  It also suggests to me (with less confidence) that where the infections have been contracted in care homes and hospitals, these probably weren’t a few unlucky people who got the virus as it snuck through a cordon of steel. I suspect that many will have had it. This would be important if true because it could mean many of these places are over the worst.

  All this matters because we’re a country in lockdown blowing up our economy on the grounds we’re stopping transmission of a virus based on the prevailing and uncertain view of the virus of mid-March. At least as I read it there’s a different still-uncertain view now; I don’t think anyone seriously believes there hasn’t been *massive* amounts more infection than was officially thought likely on the understanding of the virus back then, the question is how much?

  In 6-8 weeks time hopefully we’ll have better antibody data, gathered in the way you suggest.

  If I was the US government I would want to test all those prisoners who tested positive for Covid-19 for antibodies, for example. We can then see how far away from 100% antibody creation we’re seeing, for instance.

  Anyway, at that point I suspect we’ll all know and agree that e.g. London saw a very high level of infection, a couple to a few million, mostly asymptomatic. But we’ll be debating some other aspect of the virus as being uncertain and people will again be asking for strong proof and evidence.

  My attempt here (not least as an active investor) is to figure out what’s going on ahead of time.

  And of course my view here could certainly be wrong — I’m no expert, and I’m not seeing enough antibodies in NYC for my liking, as I’ve said.

  However I continue to think the understanding of the virus is moving in my direction — far FAR more infections than was realised, lower fatality rate, less need for a continuing *full* lockdown (no argument with doing it initially as a precaution), more of a case for shielding, continued physical distancing (the status pre-full lockdown) until we know better, and perhaps easing out of lockdown in stages and seeing what happens.

  Also, we have giant new hospitals lying empty in London and Birmingham based on a different set of uncertain assumptions (and a sincere hooray for that) so let’s not pretend clueless observers like me have a monopoly on fuzzy estimations. 🙂
• 104 Simon T April 27, 2020, 4:53 pm

  Can anybody do a article on the FCA Investment Pathways and the implication. (from what I can see – it kills the choice in SIPPs when you go into drawdown). Luckily it’s being moved til next year.
• 105 Linda April 27, 2020, 4:54 pm

  I am loving this thread, coronovirus with a smattering of finance. Helping to pass my solitary day without too much boredom. Thanks everyone.
• 106 The Investor April 27, 2020, 5:05 pm

  p.s. @Vanguardfan — In my enthusiasm I forgot to say that I found that post of yours a very helpful addition to my understanding of biases etc in sampling and testing, so thanks for that.
• 107 Vanguardfan April 27, 2020, 5:12 pm

  @TI just in brief, on one point. The timing of testing for antigen is very important. All infections start asymptomatic, for a median of about 5-6 days. So if you want to estimate the overall proportion of infections which are never symptomatic, you have to do longitudinal (ie repeated) testing within the population. I believe this is how the Vo study was done in Italy. I haven’t read this prison study so can’t comment in any detail on the possible limitations.
  I’m not arguing any particular stance here, I’m just trying to explain a little of the methodologies which make data interpretation not straightforward. In any study, you have to look at the design and understand what it can tell you, and what it cannot (and what the range of possible explanations for the observations might be). It is important to keep an open mind where the range of possible explanations consistent with the data is very wide.

  @naeclue. In general, widespread testing is not done in nursing home outbreaks, because once you have established that there is an outbreak, further testing doesn’t change the necessary control measures. So as a matter of policy only a few cases will be tested, then other symptomatic cases will be presumed on a clinical basis. In this current outbreak, I think the priority for any increased testing in nursing homes should be the staff. Even now the practical barriers for care home staff to access testing are considerable, and this is one of our most serious failings in my view. That and the lack of proper PPE and training and support in care homes.
  So yes, you could test widely in homes, but I think you’d have to have a clear purpose for what you’d hope to learn, particularly in a context where testing resource is scarce and needs to be prioritised.
• 108 Vanguardfan April 27, 2020, 5:19 pm

  Btw yes I agree that a good follow up study in this prison would be an antibody study. That would tell you what proportion of infections (in a population like this one) generate antibodies. Then of course what you want is a re-exposure to infection…..
• 109 Vanguardfan April 27, 2020, 5:42 pm

  @ti, there was never a cordon of steel around care homes, not even so much as a flimsy ribbon. On the contrary, there was a policy decision to discharge patients into homes to free up hospital beds, knowing that some of these would likely be carrying the virus, and without any testing. Care home managers were just left to fight the fire with inadequate training, support or equipment – these are not hospitals but they’ve been expected to provide care way beyond their capabilities. Our care homes death stats are looking pretty ugly.
• 110 The Investor April 27, 2020, 5:53 pm

  there was never a cordon of steel around care homes, not even so much as a flimsy ribbon

  @Vanguardfan — I agree, I was being facetious. From the start I’ve argued the focus should have been on protecting the elderly/vulnerable. You may remember my post from mid-March where I bewailed that in that ‘voluntary’ state of lockdown in mid-March I was seeing 60-80-year olds milling all around my area of West London, chatting and foraging. By that point my mum was already in self-isolation on my instruction. That should probably have been the case for all of them, as much as possible.

  I realize these weren’t care home residents milling around. I do think though that a reluctance to officially and boldly state this is overwhelmingly a disease that kills old people — if for an understandable reason, to try to enforce full lockdown and its aims — hasn’t helped their cause. (Last I looked over 50% of deaths were over 80s, and the vast majority over 70/80, and there weren’t that many over-80s, relatively speaking, to begin with.)
• 111 ZXSpectrum48k April 27, 2020, 6:12 pm

  @Naeclue. Of course I see the difference but you don’t seem to see that constructing a hypothetical 20.5 year (and any maturity) bond from a small selection of bonds just isn’t an index. How exactly are you going to take account of the cheapness/dearness of specific bonds, repo effects, supply-demand etc. Why 20.5 years and not 15 years or 11.11years? Of course it makes less difference on bonds but why not just do it correctly? It isn’t even a good method by which to construct hypothetical par bond of a certain maturity. If you wanted to do that you’d use a two-stage curve spline: construct a best-fit zero forward-forward curve from all the traded Gilts and smooth it with a parabolic or exponential spline.

  I suppose I’m sick of all these so-called ‘investment experts’ who do their asset allocation analysis using full indices for equities but somehow then decide that govt bonds can be replicated using the 10-year Treasury with some back of the envelope approximation for the return.

  I know at the end of the day, no one on this site cares about bonds or anything other than equities. Active investing can’t possibly ever work. It’s all a zero net sum game or other nonsense that isn’t actually true. I’ll wander off …
• 112 Penelope April 27, 2020, 7:50 pm

  Elvis has left the building.
• 113 Snowman April 27, 2020, 8:28 pm

  Don’t know if anyone’s still watching the daily briefing. I’d resolved not to watch it again because it was irritating me so much. But anyway I watched it today.

  They are now answering questions from members of the public. First question from a member of the public was from Lynne in Skipton and what a stonkingly brilliant and powerful question it was

  “I am missing my grandchildren so much, please can you let me know if after the 5 criteria are met, is being able to hug our family one of the first steps out of lockdown?”

  I have to say that blew me away.

  We do need to think a bit about how we are going to allow people to met their families who don’t live with them. If a relative of someone who is shielding has not been out in public for a week or so and let’s keep it simple and say they live on their own, it must be reasonable for them to visit that shielding relative? And it is for them to decide if the risk is worth it?
• 114 Boltt April 27, 2020, 10:09 pm

  Sensible looking economist article, talking about the different between actual v expected deaths, hopefully the link works (otherwise on LinkedIn).

  https://www.economist.com/graphic-detail/2020/04/16/tracking-covid-19-excess-deaths-across-countries?utm_campaign=coronavirus-special-edition&utm_medium=newsletter&utm_source=salesforce-marketing-cloud&utm_term=2020-04-25&utm_content=article-link-6

  B
• 115 Ruby April 27, 2020, 10:10 pm

  @ Snowman – I live alone and that’s what I’m going to do rools or no rools.

  @ ZXSpectrum48k – I have an equity background but (maybe because of that) find you easily the most interesting poster on the board so hope you’ll persevere.
• 116 Vanguardfan April 27, 2020, 10:12 pm

  @snowman, the principle behind the legal lockdown is to enforce social distancing at a population level – its not about individual protection. So I would expect any continuing ‘shielding’ recommendations for at risk individuals to be advisory rather than mandated. The law is not going to mandate whether hugs are permitted. In making personal family decisions, it would help to have some guidance. Hugs from grandchildren who live with doctors or care workers might not be terribly sensible, for example.
  There will be much more difficult issues, such as how to protect vulnerable individuals who can’t safely return to work, but still need income.
• 117 The Investor April 27, 2020, 11:23 pm

  New York City antibody rate is now up to 24.7%. I suppose the likeliest explanation for creeping higher from last week’s 21%-ish is statistical noise, but will be very cool if it reflects antibodies having a few more days to develop — and if it continues!

  Still very early days of course.
• 118 Vanguardfan April 27, 2020, 11:47 pm

  @TI do you have a link to the NYC study? Eg detail of where were they sampling – there is a wide variation in cases in different districts: https://gothamist.com/news/coronavirus-statistics-tracking-epidemic-new-york
  A couple of figures to add in to that – NYC has a population of 8.4m or so, and apparently 21,000 excess deaths (NYT) – official covid deaths are about 16,000 but excess deaths is likely more accurate.
  So if a quarter of people have been infected, that’s about 2.1m and would imply an infection fatality rate of about 1%.
  All very simplified back of an envelope of course, but all quite plausible without assuming anything that contradicts the ‘conventional’ understanding of the disease.
• 119 The Investor April 28, 2020, 12:09 am

  @Vanguardfan — I’m literally off to bed and just checked comments as usual passing the PC, but just to say these are from Governor Andrew Cuomo’s daily briefing. It’s worth watching at least one of them, they’re better and more grown-up than our briefings IMHO.
• 120 The Weasel April 28, 2020, 7:58 am

  @Snowman & @Ruby.

  I saw a headline on one of the red tops about how one the exit strategies could be for every household to pick ten close friends/relatives to socialise with and stick to those, hence widening the fence, so to speak.

  That would certainly work with family circles, in my view, as long as everyone keeps taking the same precautions they are taking today.
• 121 MrOptimistic April 28, 2020, 8:30 am

  @ZXSpectrum. I find your posts interesting, they make a change from the equity fanboy stuff I read on other blogs. Not least they caused me to remind myself about ‘ splines’ and to spend a rewarding 35 seconds looking up ‘ convexity’ .
  I still have a big chunk of cash sitting uninvested. My previous intent was to go 40:60 equities to bonds or something like that, with a slice of cash and gold to pander to my endofdays gloom.
  Bonds were supposed to be the maiden aunt in this ménage but they haven’t been acting in character of late and now I am not so sure about allocations.
  Equity markets seem to be unnecessary at some fundamental level. Would it really matter if we all just stopped swopping shares between ourselves with cash adjustments, and on what are values founded? All well and good to say a P/E of 13 is ‘cheap’ but there is no reason other than history to assert such a thing, and the problem with history is it is always behind us.
  So, what does the current position if the bond market imply for an asset allocation decision which seeks a return of rpi plus 1% ?
• 122 FlyByNight April 28, 2020, 8:49 am

  @Vanguardfan: Then of course what you want is a re-exposure to infection…..
  Just wondering – is there any data which suggests immunity can be built up? There doesn’t appear to be any at the moment. But this could knock the idea of a vaccine on the head and may mean we have to live with Covid long term.

  Which would leave the dilemma over whether we change society to protect people as they become vulnerable, or accept lower life expectancy.

  Was also wondering if the virus could cause long term health issues? I’m not an expert but couldn’t help wondering this after reading the David Lilienfeld link above. But could someone who is asymptomatic today end up with health issues caused by Covid later? In which case not getting infected (if possible) would be an advantage.
• 123 Simon T April 28, 2020, 8:55 am

  @flybynight – one of the stories up there in the links deals with long term health I think. From memory 10 years of your life, or did I skim it and not understand it..?
• 124 Simon T April 28, 2020, 8:58 am

  EDIT – I skimmed it without taking it in. https://www.telegraph.co.uk/news/2020/04/23/coronavirus-cutting-decade-victims-lives-scientists-say/. (must have read another version today because its not in the links), its historical and not future, ie it cut the lives of those who died
• 125 The Investor April 28, 2020, 9:12 am

  @Vanguarfan — I agree with your napkin maths. While this antibody number is creeping up, it’s definitely much lower than I’d hoped.

  So the ‘many more infected’ thesis as of *now* is shaken (I’d still maintain it was a worthwhile point to make 4+ weeks ago when people were talking far lower numbers, but only in understanding the virus then, not in prognosis for the long-term).

  I suppose the best hope for the ‘much more infectious / less deadly’ theories now currently looks like: R0 falls with summer (likely IMHO but only gets us to autumn) / R0 drops with lower infection numbers/immunity than supposed (no evidence yet I’d say) / virus is so easily defeated by young people that in many cases their macrophages etc are doing the job and they don’t even get antibodies (potentially we could infer this perhaps by the overwhelming lack of notable symptoms in the young, but I wouldn’t call that evidence, and also as far as I know we don’t know what it’d mean for them as future carriers).

  Bottom line, I’m leaning back towards a modified version of consensus of the virus spread (though not quite yet tactics to defeat it). 🙁

  I need(ed) to see NYC antibodies at 40-50% I guess.
• 126 Ruby April 28, 2020, 9:19 am

  @ MrOptimistic – I would be really interested to see a few thoughts from @ZXSpectrum on fixed interest allocation for passive investors. Currently, all I do, and I suspect many others do the same, is plonk it in Vanguard Global Bond Index Fund or similar and hope for the best. The characteristics of that fund mean one misses out on a lot of other potential sources of fixed interest return – longer duration bonds, index linked being just two. I know the Gospel according to St Lars advises against!
• 127 MrOptimistic April 28, 2020, 10:52 am

  By coincidence this was in today’s Times ( may be behind a paywall). Also discusses the scars over 50’s carry from the formative years in the 1970s.

  ‘Anyone who claims to know assets are fairly priced now is delusional’

  https://www.thetimes.co.uk/article/past-economic-recoveries-are-a-poor-guide-to-the-challenges-facing-us-now-3bm0lkqfb
• 128 Grumpy Old Paul April 28, 2020, 11:05 am

  @Ruby,
  I too really appreciate ZXSpectrum’s comments in so far as I understand them. Like others, I do have to look up some of the terms he uses though!
  I’d also appreciate his views on bond allocations for passive investors. I use Vanguard Global Bond, Vanguard LS20 and Vanguard Inflation-Linked Gilt index. I adjust my equity holdings to compensate for the 20% equity within LS20 but that is probably a slightly obsessive attention to detail. Yes, I’m aware that LS20 has a slug of Global Bond and inflation linked gilts. I just wanted to get more diversity within my bond holdings with relatively low volatility. I’m aware that inflation-linked gilts are volatile and are very expensive at the moment but I like the fact that they seem to be uncorrelated with anything else and offer me some protection in the event of inflation taking off in the future. I rebalance very infrequently and so have not attempted to take advantage of the wild movements of gilts in March; I also would have had to be very quick off the mark and use ETFs rather than funds for speed of transactions. All way above my pay grade!
• 129 Snowman April 28, 2020, 11:09 am

  From the ONS statistics released today up to 17th April

  11,854 more excess deaths (relative to 5 year average) registered for week 16 in England and Wales (E&W)

  This adds to the 7,998 excess deaths for week 15, and 6,082 excess deaths for week 14

  So excess deaths in E&W based on the past 3 weeks of data of registrations = 25,934
  (3 weeks is the period over which excess deaths are detectable, you could add in about 1,000 deaths for the previous week before that)

  19,112 deaths are shown officially as covid-19 based on registrations up to 17th April, and
  22,351 deaths are shown officially as covid-19 based on dates of death up to 17th April

  So around 30,300 excess deaths occurred up to 17th April in England & Wales, due to direct or indirect affects of covid-19 (25,934 x 22,351/19112)

  Hospital daily death count was 14,451 in E&W at 17th April, so quite a bit lower than the 30,300 figure.
• 130 Grumpy Old Paul April 28, 2020, 11:41 am

  I too agree with @Vanguardfan’s napkin calculations. A more widely quoted figure for IFR is around 0.65% (E.g. 0.66% in ps://www.medrxiv.org/content/10.1101/2020.03.09.20033357v1 – not peer reviewed).

  This would suggest that the number of people who’ve had COVID-19 in NYC might be 50% higher than the 24.7% figure cited.

  More generally, I’ve seen no discussion of the proportion of 65+ people who are either asymptomatic or suffer mild symptoms. Anyone have any clues to the answer? Surely, given the proportion of deaths within care homes now evident, there should be some focused epidemiological work on care home residents.
• 131 MrOptimistic April 28, 2020, 11:56 am

  @Grumpy. Based on what I have read the IFR might be in the 1 – 1.5% range. Wouldn’t that square your calculations and bring the implied infection rates back in line with the actual NYC test results?
  Generally the excess death data seems the best basis we have. The UK flu reports show a very abrupt increase at weeks 14/15. I think napkin calculations would best be based on that sort of data.
  As an aside, with that IFR I would note that the probability of death before next birthday for a 66 year old is just above 1%. So for me Covid implies a risk commensurate with what I faced already ( and didnt spend too much time fretting about, let alone turning the population into house prisoners and juicing the economy). It’s an odd logic. I thought the NHS was there to protect me, not vice versa!
• 132 The Investor April 28, 2020, 12:04 pm

  Surely, given the proportion of deaths within care homes now evident, there should be some focused epidemiological work on care home residents.

  I’ve just had something of a heated argument with a friend of mine — and Chris Whitty fan — about exactly this.

  We’re blowing up the economy at £2.4bn a day on the powers-that-be’s say-so, and so while I’ve personally been prepared to cut them a lot of slack given the nature of the thing, I still think their actions (or lack of…) should be held to a high standard.

  And this care home situation — in a crisis where the *one thing* we always knew for sure were the elderly were overwhelmingly the most vulnerable — is becoming somewhat frightful.

  Last week we had 20K out of 40K spare testing capacity. Even if it was hard to coordinate e.g. NHS workers to the testing stations to use that capacity, it should surely have been possible to get a few e.g. coordinated army medics out and back in to make use of this capacity and give us more information.

  But Whitty et al seem so reactive.

  Why aren’t we doing mass population tests using the spare capacity in say a single prison or a care home or hospital where there’s an outbreak to get a sense of how widespread a known infection is/becomes?

  Why are we not planning to retest exactly the same population in 2-4 week for antibodies to see how many known cases develop them etc?

  Indeed, why is Whitty/the government kicking most of his antibody tests out into the far future? (300K over the next 12 months is the current plan as I understand it. Feels necessary but not sufficient.)

  If Whitty admits we have population-level-good antibody testing now (as he did yesterday) then why isn’t he doing a random antibody sample of say Brent in London?

  Why are we sitting around trying to infer things from the likes of NYC and Sweden when our own gov has the capacity and resources to get us better data?

  Whitty comes across to me as an affable chap giving one of those after work lectures I like to go to at the LSE — but in the middle of a crisis that is still ongoing…

  Perhaps it doesn’t make any difference — Sweden, after all — but I think I’d prefer a wartime consigliere.

  The argument I’ve had back is that the CMO doesn’t have this kind of power, and it’s not his fault, so it should be the government doing it (e.g. Hancock).

  Then this quickly gets political; the suggestion is because Johnson and Cummings purged the cabinet of those with backbone (to get through a No Deal in December) there’s nobody with the sense of responsibility to properly step-up now / when Johnson was recovering.

  I also appreciate some might say getting e.g. population level data doesn’t help us because the diagnosis would be the same — full lockdown ideal, partial lockdown if not.

  But given the retort is so often “we don’t have that data” and we’re losing perhaps £1bn a day in GDP permanently (generously assuming 50%+ comes back in a post-lockdown bounce) you’d think the money and resources could be found?
• 133 Grumpy Old Paul April 28, 2020, 12:41 pm

  @Investor,
  To me the kick-ass arguments for epidemiological studies on the care home population , immediate representative sampling of the wider population and associated longitudinal studies are:

  1) If you don’t know what’s happening in care homes how on earth are all you going to manage the emerging crisis?

  2) If you don’t know what’s going on in the general population, how are you going to find out, in a timely manner, if an incremental step out of lock down threatens to overwhelm the NHS?

  3) If you don’t know what’s going on in the general population, if there is a seasonal fluctuation in COVID-19 infections, how are you going to detect an autumnal upsurge in cases swiftly enough to prevent the NHS being overwhelmed?

  4) Once the longitudinal studies and the associate data capture and analysis are in place, they can easily be repurposed to give early warning of future flu outbreaks and any other future epidemic.
• 134 Snowman April 28, 2020, 12:58 pm

  The ONS data (w/e 17th April) showing that care homes are taking the brunt (about 40% of the excess deaths).

  Compared to average death figures in these settings for week 16, hospital deaths and home deaths have roughly doubled, but care homes deaths have more than tripled

  Of the roughly 12K excess deaths, roughly

  5K are in care Homes (+ 239.6% vs 5 year average)
  2K at home (+ 95% vs 5 year average)
  5K in hospital (+97.1% vs 5 year average)
• 135 Vanguardfan April 28, 2020, 1:20 pm

  @GOP, there’s already quite extensive flu surveillance:
  https://www.gov.uk/guidance/sources-of-uk-flu-data-influenza-surveillance-in-the-uk
  https://publichealthmatters.blog.gov.uk/2020/01/09/flu-detectors/

  I would imagine the Covid early warning system for the autumn will be tacked on to these systems, especially the sentinel GP practices.

  Incidentally here’s a link to a digest of covid related surveillance data. Some won’t be that helpful right now as normal access via GPs isn’t being used for Covid.
  https://www.gov.uk/government/publications/national-covid-19-surveillance-reports
• 136 MrOptimistic April 28, 2020, 2:15 pm

  @Vanguardfan. I think the flu reports already effectively capture Covid via the reporting of acute respiratory incidents and excess death data. The next UK report is out on Friday but the previous reports tagged the increase in excess deaths and the abrupt rise in care home outbreaks. Of course come the flu season they will need to find a means to separate and discriminate.
  We might be in for an interesting autumn/winter with the threat of a combined flu/covid attack.
  Then, of course, there is H5N1.
  I reckon there will be a shortage of flu vaccine this year not least because everyone has woken up to the risk.
• 137 Vanguardfan April 28, 2020, 3:45 pm

  @MrO, yes indeed
  This is a better link to the flu reports, they don’t publish as often ‘out of season’.
  https://www.gov.uk/government/statistics/weekly-national-flu-reports-2019-to-2020-season

  @TI, NYC will certainly get to 45-50% immunity, but it may well take another 16,000 deaths…

  On infection fatality rates, worth noting that it’s not a fixed number, even if we could estimate it precisely. It will vary by population, key factors being the age distribution and the availability of health care. Frankly I think ours might not look too great. I’m more and more of the view that the NHS has only ‘coped’ by locking out anyone old and frail. Not to mention all the non covid work. Still, you can’t conjure up doctors and nurses out of nowhere, even if you can build new warehousing facilities for them to work in.

  My feeling is that we will now probably manage ok through the summer, with a gradual lifting of some restrictions (which I think will start quite soon, my original prediction was within 8-12 weeks), but there is likely be another nasty wave coming at us when the weather turns in the autumn and we all huddle indoors again. There is still seems to be plenty of susceptibility in the population, as far as we can tell, to provide the fuel. It won’t look the same as the first wave because we start from a different place, but it could be just as damaging. We might even be back to lockdown again…..however much no one wants it (looking at news from Germany).

  It would be nice to think the government will use the few months relative breathing space to prepare furiously, and competently, but I’m not holding my breath. I expect more of the same logistical difficulties, equipment shortages and rigid over centralised one-size-fits-all policies (part of the issue with matching testing supply with demand is the overly centralised delivery, and lack of integration with local primary care and public health services).

  Lots of uncertainties of course, not least the extent of immunity, but also factors like the weather, the type of flu season, and the huge pressures to have something like a Christmas season in retail and leisure sectors….and to top it off there’s the end of year Brexit crash out to look forward to.
  I do hope I am being too pessimistic. Eventually of course we will live past this, and life will go on, and we will adapt if needed to any continuing health threat. Who knows, we might even get a vaccine.
• 138 Vanguardfan April 28, 2020, 3:55 pm

  Btw on lockdown easing. In my neck of the woods there’s a distinct feeling of people starting to get back to work: contractors working outside, garage services, a few more food outlets. More traffic. It’s worth reflecting that lots of businesses and activities that stopped in March are not actually prohibited, and I think we’ll see more activity happening within the current rules, whatever the messaging.
  I’d really like to see schools starting back in June, at least the primaries, but I can sense there will be a huge fear factor here. This is actually the area where I most consider the public health benefits vs costs are not justified by the evidence.
• 139 MrOptimistic April 28, 2020, 6:02 pm

  Yes definite increase in activity today.
  In medical science, as in investing, in the absence of facts every opinion is valid .
  Since these folks have lowered the paywall it would be impolite not to read it
  https://www.google.com/url?sa=t&source=web&rct=j&url=https://nymag.com/intelligencer/2020/04/we-still-dont-know-how-the-coronavirus-is-killing-us.html&ved=2ahUKEwi_nbTry4vpAhVUeMAKHce1AlkQFjAAegQIBBAB&usg=AOvVaw1vAItDAPnVf7TVKLegCJ-Q
• 140 The Investor April 28, 2020, 11:36 pm

  @MrOptimistic — That NY Mag article on how Covid-19 acts/kills is pretty daunting. Will include it in the links on Friday.
• 141 Vanguardfan April 29, 2020, 7:27 am

  An interesting and informed read:
  https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(20)30985-5/fulltext
• 142 The Investor April 29, 2020, 7:55 am

  @Vanguardfan — Thanks for that. Good digest.
• 143 The Investor April 29, 2020, 1:39 pm

  Market popping on news that Gilead’s anti-viral drug Remdisivir is seeing positive news emerging from a double-blind controlled trial with NIAID in the treatment of Covid-19.

  Worth recalling we went through this with the same drug a couple of weeks ago. But the NIAID is described as the ‘gold standard’. Note the study hasn’t been released yet AIUI.
• 144 MrOptimistic April 29, 2020, 3:54 pm

  @Vanguardfan. ‘Delivery of efficacious vaccines is not a competitive race to the finish’.
  You think ?
  One of the puzzles to me at least is why they doubt post- infection immunity. If the survivor’s immune system didn’t see the infection off with antibodies, how did the patient overcome the disease? If the antibodies were agents in this doesn’t immunity ( to some degree and for some time) come with the turf?
  I always wondered with the flu vaccine why immunity wasn’t for life ( Spanish research paper showed it waning after 100 days). The answer seemed to be that the virus mutated. But getting flu twice in a season isn’t a well documented occurrence.
• 145 Vanguardfan April 29, 2020, 7:01 pm

  @TI #143 – I find it interesting that the ‘results’ of the US trial were leaked (and unavailable for scientific scrutiny) on the day the China trial (which should no effect) is published.
• 146 Vanguardfan April 29, 2020, 7:18 pm

  @MrO. Not sure what you meant by that quote? (It’s not my work…)
  I’m not a virologist or immunologist, but here’s what I can tell you.
  Influenza virus mutates frequently, most of the mutations are minor, occasionally there is a dramatic shift which can then spread rapidly, potentially causing a pandemic (as population immunity is low). This is why pandemic flu is a constant threat/expectation. There are several strains of flu virus in circulation at any time. Flu vaccine has to be reformulated every year, to match the predominant circulating strains in the months before flu season. That’s why the flu vaccine is a annual jab. Some years it’s more effective than others, it’s a bit of luck/guess work involved.

  Re doubting immunity. The production of specific antibodies to a new antigen is a relatively late immunological response, I believe it takes about two weeks post infection. It’s not by any means the only way a body fights infection – there are much quicker non specific responses. (I don’t know loads about this, but I’m quite sure the authors of the paper do). Also, antibody levels decline over time. Because this coronavirus is new, we have no direct evidence about whether the antibodies detected by the various tests are effective in actually preventing reinfection, if so for how long, and what levels might be practically effective. Scientists have hunches based on what is known about other corona viruses (the ones that cause some colds) – iirc these tend to provoke some immunity, but not particularly long lasting. (Perhaps an actual virologist will give a more expert explanation…). So the assumption is that there is probably some immunity, but most likely not permanent, and it is likely that less severe infections provoke lower, or perhaps no effective, antibody response. But it’s not yet known, hence caution.
• 147 Snowman April 29, 2020, 8:01 pm

  Interesting interview with a vaccine expert, Dr Jerome Kim, giving a clear insight into what is involved with attempting to develop a vaccine.

  https://www.youtube.com/watch?v=5cYWd0N8nO4
• 148 Linda April 29, 2020, 8:39 pm

  @Vanguardfan

  I worry about schools going back, my son & wife are both shielded, if their children go back to school, they could bring the virus home & kill their parents.
• 149 The Investor April 29, 2020, 8:56 pm

  @vanguardfan — It is depressing how this is getting politicized. Various talking heads on CNBC were doing victory laps today saying you should trust American innovation and science over the authoritarian system of China. If I had to choose I certainly would, but the last 3-5 years of politics in the US and here has fouled up so much.
• 150 MrOptimistic April 30, 2020, 9:10 am

  @Vanguardfan. The quote was from the paper. Struck me as odd. Perhaps just academics trying to raise themselves above ‘ trade’ 🙂
• 151 Tony Edgecombe April 30, 2020, 10:55 am

  @snowman

  “I’m no expert. And in practice it’s hard to say but 70-80% herd immunity doesn’t sound unreasonable to me but 96% does seem totally unreasonable.”

  I never quoted 96% herd immunity, rather that once you get to herd immunity the transmission still soldiers on. There will be an overshoot: https://twitter.com/CT_Bergstrom/status/1252077930286444545
• 152 Vanguardfan April 30, 2020, 11:27 am

  @tony Edgecombe, that’s a super-informative thread and and author. We haven’t heard enough about what ‘herd immunity’ means.
• 153 The Investor April 30, 2020, 11:42 am

  @Tony — That’s a very clear graph and useful thread, thanks for sharing. Will include in the links this weekend even though it’s a little old.

  (Not sure why the author is talking 2% IFR mind you, which is way higher than other estimates but we all have our pet numbers I guess.)
• 154 Vanguardfan April 30, 2020, 12:21 pm

  Not sure where you are seeing 2%, but the IFR isn’t relevant to the explanation of herd immunity.
  However, from what else I’ve read, he thinks the current data suggest IFR of something like 0.5% (but of course it will vary by population demographics, amongst other factors).
  Apparently 0.15% of total NYC population has already died, so the IFR in that population cannot be lower than that.
• 155 Snowman April 30, 2020, 12:35 pm

  Sorry Tony Edgecombe.

  I think you are defining herd immunity (threshold) as the point at which the replication rate starts to fall below one, and I’m defining herd immunity as the % of the population that get the virus after the overshoot above that point, the overshoot providing some protection against community spread. And the confusion has been whether we are talking about your herd immunity definition or mine. So when I say you suggested herd immunity 96% I meant my definition not yours.

  So I think you are saying that very roughly perhaps the R0=5, that at 80% having had the the virus, the replication rate falls below 1, and perhaps 96% get the virus? Is that right?

  And I’m saying perhaps R0 = 3 at least at the beginning of the outbreak, that at 66.7% having had the virus, the replication rate has fallen below 1, and perhaps 70-80% end up getting the virus.

  Here is the previous conversation as it was a while back

  @ Tony Edgecombe

  “I’ve seen 96% quoted by a reader here, for example, which is almost certainly not going to be the case.”

  That was possibly me and I wasn’t quoting the herd immunity threshold. People see a figure of 80% and assume that the disease stops at that part of the population being infected. Rather this is the point at which the replication rate equals 1. There is still further to go before the disease is eliminated.

  @Snowman

  If 80% of the population immune is the point at which one person starts to infect less than one other, then that would imply an underlying R0 of about 5. At 80% and at an R0 of 5, 4 of the 5 people that any one person would infect would be immune, leaving one susceptible to infect.

  An R0 of 3 for covid-19 has been suggested if next to no social distancing measures are taken. Under an SIR model, if you completely let the virus spread at an R0 of 3 , even though one infects less than one at 66.7% the overshoot is considerable until the virus dies out in theory perhaps at 90%+ of the population having had the virus. That is the argument for herd immunity being at 90%+.

  But this is based on a simple SIR model which breaks down in practice for a number of reasons.

  Firstly temporary social distancing measures can be used at the time you reach around 66.7% to bring the R0 down below 3, to limit the overshoot so that only perhaps 70-80% of the population will have had the virus by the time it dies out. Once you are at 70-80% say you can relax measures, but as long as the R0 is kept at or below 3 any imported cases will not spread through the community although there will be small outbreaks that die out quickly if you don’t quarantine those coming into the country.

  Secondly the SIR model ignores the mathematics of shielding and assumes each member of the population has an an equal chance of meeting any other member of the population. At 66.67% immune and with an R0 of 3, then of the 3 people that an infected person comes in contact with, the SIR model assumes 2 are immune people and one a susceptible person. The extreme position of shielding illustrates why this is wrong. If one third of the population efficiently shield then if two thirds of the population have or have had the virus then the 3 people that an infected person comes into contact with are all people who are immune or who who currently have the virus.

  I’m no expert. And in practice it’s hard to say but 70-80% herd immunity doesn’t sound unreasonable to me but 96% does seem totally unreasonable. This assumes that having had the virus gives you immunity from getting it again.
• 156 The Investor April 30, 2020, 2:23 pm

  Not sure where you are seeing 2%, but the IFR isn’t relevant to the explanation of herd immunity.

  I am seeing it in the thread you also enjoyed on Twitter and thanked @Tony for, where the author states:

  “In India, for example, this would cost about 19 million lives with a 2% infection fatality rate.”

  I didn’t say it was relevant to the explanation of herd immunity. 🙂 I said the author of that thread is for some reason city an IFR twice as high as I’ve seen as consensus elsewhere.
• 157 The Investor April 30, 2020, 2:25 pm

  p.s. I guess as he mentions India he could be of the view that IFR vary wildly from country to country, which would be fair enough and fits what we’re seeing so far.
• 158 The Investor April 30, 2020, 6:02 pm

  Study from Nature, suggests mean time until antibodies seen in blood is just 13 days:

  https://twitter.com/MarcGozlan/status/1255893936469610497
• 159 Snowman April 30, 2020, 7:29 pm

  The problem with antibody tests at the moment is not in detecting the antibody response in those who have had severe symptoms. They seem to be talking about patients in that article so they do seem to come into that severe symptom category.

  The difficulties seem to be in the asymptomatic and those who are able to fight off the virus easily through their innate immune system. It’s harder to study the asymptomatic because the asymptomatic are harder to identify. And it’s hard to detect the low level of antibodies, and if you set the bar too low in terms of the test you can end up with false positives.

  So it seems to me we have to be a bit careful with some of the population serology studies we are seeing because they may be under-estimating those who have had the virus because they may not be picking up those who had a mild encounter with the virus.

  The immune response in these mild cases and detecting that response through an antibody test, are areas where understanding and tests will hopefully develop.
• 160 Snowman April 30, 2020, 7:47 pm

  Chris Whitty was rightly very cautious about remdisivir as an effective treatment at the daily briefing. He and Patrick Vallance pointed out that the Chinese study showed no statistically significant results.

  https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(20)31022-9/fulltext

  “Our trial found that intravenous remdesivir did not significantly improve the time to clinical improvement, mortality, or time to clearance of virus in patients with serious COVID-19 compared with placebo”

  And they mentioned that the American Gilead claims of its effectiveness was not backed up by any published data, and that no statistically significant reduction in mortality was claimed, only that the treatment time was claimed to be reduced.
• 161 Snowman April 30, 2020, 7:52 pm

  Pleased to hear that Boris pretty much said we will be advised to wear face-covering when the lockdown measures are loosened.

  And the World Health Organisation are now saying Sweden are a model for countries going forward!

  How things change
• 162 MrOptimistic April 30, 2020, 8:39 pm

  The only logical definition of herd immunity is when R0 becomes less than 1 surely. Since R0 isn’t zero then clearly infections continue. But just like modelling neutron reactions the reaction fizzles out, but it doesn’t generate a mushroom cloud. Once RO is less than 1 the epidemic is in decline and will eventually disappear.
• 163 Indecisive April 30, 2020, 8:55 pm

  There’s a lot of confusion about herd immunity in these comments. I found a new pre-print on the topic, which is relatively accessible: https://www.cell.com/pb-assets/products/coronavirus/immuni4358.pdf

  Found via a co-author’s tweet: https://twitter.com/LB_Barreiro/status/1255168499342872579
• 164 Indecisive April 30, 2020, 9:16 pm

  @Snowman, #161

  And the World Health Organisation are now saying Sweden are a model for countries going forward!

  Well I had to check that out 🙂

  I watched an excerpt from today’s WHO press conference, but will refer to their transcript, as it’s much easier to reference. You can find it here https://www.who.int/docs/default-source/coronaviruse/transcripts/who-audio-emergencies-coronavirus-press-conference-29apr2020.pdf?sfvrsn=aaa81d24_2 and it’s linked to from the “Media resources” page on their website (not linking as Monevator puts comments with 2+ links into the moderation queue)

  Sweden first gets a mention at 00:41:12, and MR first addresses how different Sweden’s approach is. My extracts:

  I think there’s a perception out there that Sweden has not put in place control measures and has just allowed the disease to spread. Nothing could be further from the truth. […] it’s very much relied on its relationship with its citizenry
  and the ability and willingness of citizens to implement physical distancing and to selfregulate

  When they talk about Sweden being a model it’s rather more nuanced. This is between 00:42:56 and ~00:45:00 and for ease of reading I’ll reproduce it in full:

  I think if we are to reach a new normal in many ways Sweden represents a future model of, if we wish to get back to a society in which we don’t have lock-downs then society may need to adapt for a medium or potentially a longer period of time in which our physical and social relationships with each other will have to be modulated by the presence of the virus. We will have to be aware the virus is present and we will have to, as individuals and families and communities, do everything possible on a day-to-day basis to reduce the transmission of that virus.

  That may mean adjusting the way we live our lives and I think maybe in Sweden they’re looking at how that is done in real time. So I think there may be lessons to be learned from our colleagues in Sweden but again I wanted just to emphasise, Sweden has not avoided controlling COVID-19. It’s taken a very strong strategic approach to controlling COVID-19 across all of the elements of society.

  What it has done differently is that it really, really has trusted its own communities to implement that physical distancing and that is something that remains to be seen, whether that will be fully successful or not.
• 165 The Investor April 30, 2020, 9:54 pm

  @Snowman:

  So it seems to me we have to be a bit careful with some of the population serology studies we are seeing because they may be under-estimating those who have had the virus because they may not be picking up those who had a mild encounter with the virus.

  The immune response in these mild cases and detecting that response through an antibody test, are areas where understanding and tests will hopefully develop.

  This is why I was suggesting the other day that we should be doing a population sample for the live virus, and then following it up a month later and see who got antibodies. After a couple of these I presume we’d have a rough heuristic for how much to gross up the +ve antibody number seen in a population, even though it would obviously be useless (as most of the tests still are AIUI) on an individual level.

  Perhaps Whitty is going to do that with his live Covid-19 infection population sampling he’s finally beginning, but as far as he’s said it seems to be more focused on surveillance and calculating R0. But no doubt there’s tons of this stuff going on in various countries by now.
• 166 The Investor April 30, 2020, 9:58 pm

  Also, a promising update on the situation re: children (which may be relevant about reopening schools, too):

  No child has been found to have passed coronavirus to an adult, a review of the evidence in partnership with the Royal College of Paediatrics has found.

  Major studies into the impact of COVID-19 on young children suggest they “do not play a significant role” in spreading the virus and are less likely to become infected than adults.

  https://www.theage.com.au/world/europe/experts-fail-to-find-a-single-case-of-children-passing-virus-to-adults-20200430-p54ohi.html
• 167 Vanguardfan April 30, 2020, 10:24 pm

  @indecisive, nice paper.

  For assessing possible immunity, we only need antibody tests not virus tests. Infection that doesn’t produce antibodies won’t contribute to herd immunity.

  Re children, I think there is absence of evidence rather than evidence of an absence of transmission. Ie as yet very unclear. Here’s a counter view:
  https://zoonosen.charite.de/fileadmin/user_upload/microsites/m_cc05/virologie-ccm/dateien_upload/Weitere_Dateien/analysis-of-SARS-CoV-2-viral-load-by-patient-age.pdf
• 168 Snowman April 30, 2020, 10:38 pm

  @TI

  Re Chris Whitty’s population sampling, not good news, apparently the first data is starting to come in and he told the Select Committee about 6 days ago (at 1hr 25min)

  https://www.bbc.co.uk/iplayer/episode/m000hpkj/select-committees-science-and-technology-committee

  “Unlikely that any part of the UK maybe outside London will have a sero-prevalence much above 10% “
• 169 MrOptimistic April 30, 2020, 11:28 pm

  Yes, that’s what the WHO said a few days ago. New York etc were below 25,%. It’s all a bit too frenetic at the moment, need a calmer reflection than we are getting.
• 170 Snowman May 1, 2020, 7:28 am

  Like this article on antibody testing

  https://www.livescience.com/coronavirus-antibody-tests.html

  “Some people who have beaten COVID-19 may not generate antibodies at all, but that may not mean they’re not immune”

  I think it is actually suggesting that the body may produce antibodies but for the asymptomatic or those with very mild symptoms perhaps they may be at a very low level that aren’t detectable by the test or that the body produces different antibodies to the ones the test picks up.

  I’ve said before it would seem surprising if the body would fight off a virus with ease but not leave guards at the door for the next attack, but then perhaps the body assumes it can fight it off again so doesn’t need the guards at the door.
• 171 Indecisive May 1, 2020, 12:46 pm

  Today’s report from the ONS titled “Deaths involving COVID-19 by local area and socioeconomic deprivation: deaths occurring between 1 March and 17 April 2020” is fascinating reading.

  The media outlets are picking up on the deprivation index correlation; I went straight to the map where you can enter your postcode and see how many have died in your local area. V interesting to see the lack of spread across the country.

  https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/bulletins/deathsinvolvingcovid19bylocalareasanddeprivation/deathsoccurringbetween1marchand17april (can they come up with a more catchy URL?)
• 172 Snowman May 1, 2020, 3:02 pm

  Good find Indecisive.

  I like the following chart also which comes from ONS local authority data because it is based on actual date of death rather than date registered.

  https://twitter.com/DarkishJungle/status/1255798365067120641

  It shows a peak in excess deaths occurring (i.e. by date of death not date registered) in week 15 (the week ending 10th April).

  This ties in with the England and Wales official covid-19 hospital deaths which shows a clear peak at 8th April.

  And you can see there have been around 28,300 excess deaths that have occurred in the past 4 weeks in England up to 17th April.