Keeping an eye on things
The screen was not a pretty sight this morning.
There was an interesting article last week in the Actuarial Post about the impact of the crisis on life insurers. Unfortunately they got the main point wrong, arguing that the asset side would be hit by “declines in bond yields”.
Fujian looks like the first province to have completely eradicated the corona virus, if the figures are to be believed. See chart above. 296 cases diagnosed, 295 recoveries and 1 death. Thus the two methods of computing the case fatality converge.
CFR = D(R+D) = D/C = 1/(295+1) = 0.34%.
or 1 in 296. With a population of 38.6m, the ratio of fatalities to population is thus 1 in 38.6m, proving again as I argued yesterday, the best chance of surviving the virus is not getting it.
The chart above shows cases (C), deaths (D) and recoveries (R) in the Chinese province of Henan, 22 January – 7 March 2020 (source John Hopkins).
As you can see, the number of cases rises rapidly until early February, then the rise slows, and from mid February tails off completely. With a delay of about two weeks the number of recoveries rises, then falls, following the same pattern. Thus there are now 1,272 diagnosed cases in Henan, with 1,243 recoveries, and only 7 unresolved cases, and 7 deaths.
With a population of 95m in Henan, that means that the fatality ratio of cases resolved is 1 in 58, but as a ratio of the general population it is only 1 in 4,333,000.
Does that mean there is no cause for worry?
Chair: Mel Stride
Members: Rushanara Ali, Steve Baker, Harriett Baldwin, Anthony Browne, Felicity Buchan, Angela Eagle, Liz Kendall, Julie Marson, Alison McGovern, Alison Thewliss
You can see them all on Parliament TV at the first hearing of Wednesday 4 March 2020.
Subject: Appointment of Andrew Bailey as Governor of the Bank of England
Witnesses: Andrew Bailey, Chief Executive, Financial Conduct Authority
Steve Baker gives a particularly firm questioning to the governor-elect.
The Institute has confirmed Louise Pryor as next President-Elect. Congratulations to Louise, who I worked with many years ago on data issues.
Never thought Eumaeus would say that, but if the data is right (sourced from John Hopkins) then the chart above shows how, for three Chinese provinces (Guangdong, Henan, Hunan) the rise in new cases has been almost completely stopped.
Bruce Aylward (Assistant Director-General of the World Health Organization, and the leader of the WHO team that visited Wuhan in China) explains here how they did it.
His main target, and the message bears repeating endlessly, is the fallacy that only an authoritarian regime can accomplish what was done in China.
We know it’s a bad dangerous disease. How do we bring the mortality down? You concentrate your resources, you shorten the time frame to make sure you can identify very quickly, you look at the high risk ones and get them into the specialised centres in case they crash, these are the things that we should focus on.
[EDIT] For balance, here is Gordon Change explaining how China has “bought the WHO”, and questioning the statistics (see chart above) apparently showing that infections are “on the downslope”. “Whenever you have statistics supporting a leader’s policies, you’ve got to be very concerned”.
As we have reiterated, the conclusion depends on the accuracy of the data, and accuracy of data is hard to determine when all we have is the data. As Wittgenstein said, you can’t check what is said in the newspaper by buying a second copy.
Kevin questioned yesterday whether the deferment rate is affected at all by any illiquidity premium. Here is a mathematical approach to the same question.
The IFoA working party on equity release mortgages chaired by our friend Craig Turnbull has just issued an interesting ‘discussion note’ about equity release valuation. You might have thought that the WP might have had something to say about some of the rude things we have said about the subject, or about the IFoA or even about the WP itself, but no.
At one level, the WP’s non response is admirable. From a scientific perspective, however, it isn’t helpful to ignore research that gives conclusions they might not like. Better to confront and rebut, otherwise people might be tempted to draw their own conclusions.
There is also nothing about the ostrich elephant in the room, which is whether the industry are getting it wrong, like getting NNEG valuation an order of magnitude wrong.
Data so far suggests that of the 3,711 people on the Diamond Princess cruise ship, 705 contracted the virus, i.e. 19%. Much higher than in any place in China, as we have stated passim, but then a cruise ship is a confined space with a lot of people in it. So far, 6 people appear to have died, which is 0.85% of those diagnosed. The mortality relative to the whole population is therefore 6 divided by 3,711, i.e. 0.16%.
These figures conflict somewhat with the crude distribution we published here, but we did say to treat all figures with caution. There is a conflict because, as is well known, mostly old people like us go on cruises, but we are seeing nothing like an 18% case fatality rate.
[EDIT] A useful paper on the Diamond Princess age distribution is here. Table copied below.
| Age group | Symptomatic cases | Asymptomatic cases 1 | Total | Crude asymptomatic ratio 2 | Persons aboard3 |
| 0-9 | 0 | 1 | 1 | 100% (95%CI: 2.5%, 100%) | 16 |
| 10- | 1 | 1 | 2 | 50.0% (95%CI: 1.3%, 98.7%) | 23 |
| 20- | 18 | 2 | 20 | 10.0% (95%CI: 1.2, 31.7%) | 347 |
| 30- | 18 | 5 | 23 | 21.7% (95%CI: 7.5%, 43.7%) | 429 |
| 40- | 18 | 7 | 25 | 28.0% (95%CI: 12%, 49.4%) | 333 |
| 50- | 27 | 22 | 49 | 44.9% (95%CI: 30.1%, 59.8%) | 398 |
| 60- | 73 | 56 | 129 | 43.4% (95%CI: 2.5, 100%) | 924 |
| 70- | 92 | 136 | 228 | 59.6% (95%CI: 53.0%, 66.1%) | 1015 |
| 80- | 27 | 25 | 52 | 48.1% (95%CI: 34.0%, 62.3%) | 215 |
| 90- | 2 | 0 | 2 | 0% (95%CI: 2.5%, 84.2%) | 11 |
See table below, sourced from Financial Times market data. The FTSE is down 11% since the start of the current market crisis (around the middle of February 2020), but it looks as though insurers have been disproportionately affected.