This morning began with an unusual task: Looking for a copy of Javelin analytics software on 1.4 MB floppy disks.
“WTF? Tell us about the damn virus and Depression, Ure!”
Sometimes, life is not so simple…let me get you up to speed.
A report out this morning notes that California health officials are now predicting that 56% of the state will eventually be infected. This, as California entered a “stay at home” period as of last night which officials hope will slow the rate of propagation.
It may be too early to see much impact in daily change rates – we expect the rate of change to pop back up in the afternoon (PM) report. But, as of this morning:
As you can see, until the afternoon data is out, we can’t really do much in the way of apples-to-apples. What we can do – and it should be close – is project out the next several weeks:
This implies that the Global Peak will come after mid June. However, when we start looking at the USA (population 330-million, or so) we can begin to think that IF California is typical of the whole country (it’s not) THEN we could think about 184.5 million cases, or so. And FURTHER with a 4% mortality rate (high, especially if chloroquine and other anti-malaria drugs prove efficacious) THEN we could start to ball park a finish line for this terrible adventure.
“What does this have to do with old Javelin software?”
Javelin had one feature that I’ve never found in a modern analog that made it ideal for this sort of problem: In making a forecast, you could draw an estimated curve and it would automatically back-fill data to the curve. Thus, in very short order, without having to go through the tedium of miles-deep math, you could sketch in a number of possible propagation end scenarios, and Javelin would spit out numbers. Then by simply comparing actuals vs. estimates, the “best fit” would quickly become apparent.
If you happen to know of a modern product with this kind of capability, and not on a phone, (please God, not that!) let me know what it is, where to find it, and so forth.
I decided not to buy the $300 copy on eBay because I don’t have a 5.25″ floppy drive (don’t know if anyone ever made a 5.25″ to USB external drive, either) but 1.4 MB 3.5″ we have always at the ready for old (wonderful) D-Base relatives like Jim Button’s (his real name was Jim Knopf) PC-File DB and the like. Truly a “father of shareware” as Peter Norton dubbed him.
All of which harks back to my Halt and Catch Fire days in the days back when… around the time my buddy Gaye Levy (www.strategiclivingblog.com) was writing the back-end accounting code for telcoms like pagers…oh say, there’s another lost tech…(Flash-back Friday, is it?)
Don’t mean to get off on such reminiscing, but it’s more fun than looking at the body count projections.
One further statistical caveat (while I mourn the passing of real coding and not pretending “drag and drop” is high art) has to do with the mortality rate. My method of calculation is to look at the known cases divided into the CV cemetery and presto – a number.
The more “statistically defensible” version is to add the total recovered today (86,035) and the dead (10,038) which gives a total of 96,073. This – divided into the dead count – gives a 10.44% death rate.
Not to worry, though…the death rate is nowhere near that high for a number of factors: The main one is that the “recovered” number only involves people who have “gotten in the system.” And we understand – from anecdotal reports – that half (or more) of people exposed will NOT have sufficiently bad symptoms to require getting involved with the medical community. In statistical terms, “n” (cases) is artificially low lacking the subclinical cases.
One thing that is new (and useful) to know is that almost half of all CV cases are exhibiting some kind of “digestive track disorder” (e.g. the runs).
Also useful to know is that men seem to die at a higher rate than women struck with the virus.
How Big a Rally?
Well, now, that IS the question, isn’t it? Here’s how our line-up looked this morning, comparing the 1929 collapse with what’s been going on in present-day since our Aggregate Index peaked on February 19:
What’s more, since there has been a short-term “double-bottom” there’s an admittedly weak technical case for a short-term rally lasting perhaps a week, or so.
Thing is, though, that the market will probably NOT be happy when new employment data comes out on early April. In fact, the administration is so worried about the bleak numbers ahead, according to reports like this one, they’re already leaning on state officials to slow the release of data.
Despite the technical rally – even if it lasts into next week – we still lean toward the view of former Reagan era federal budget director David Stockman who’s looking for an S&P down to 1,600.
At least quadruple witching is likely to pop higher at the open…
Crooks in the Capitol
Story to pay attention to – because it is traitorous behavior in our view – involves four U.S. Senators who sold off millions in stock holdings before the coronavirus impacts became publicly appreciated.
I have said this before – and it bears repeating: Before this is over ther world will need a virus-analog to the Nuremberg War Crimes Trials – and we will need a wall.
Points for the Paranoid:
Article in the Asia Times that reckons the U.S. and China are engaged (via the virus) in a hybrid war. We don’t like to think too much along these lines, but we have long-held that there will be a path to global nuclear conflict out of this when blame is being placed.
And Fed Repo Depot operations this morning were $53-1/4 Billion. We look for lots more in the afternoon to drive up prices into the weekend and feed some optimism to the public.
Write when you get rich,