Posted on

Why Don’t You Have Your Credits Go To D Instead?

Weird times makes for weird finance stories, and today was no different. 

2020 04 09 15 31 13

First up is SoftBank, which is the procylical, risk-increasing gift that keeps on giving. Apparently CEO Son pledged 80% of his SoftBank shares to banks as collateral during the recent sell-off. Collateral against … what, you wonder? Against personal loans, of course. This company is a profligate producer of risk. Remarkable.

2020 04 09 15 37 54

And then there’s LendingClub, which, as it turns out, is struggling with what used to be lousy credits becoming … lousier credits. This is no surprise, but why not simply re-reate credits, rather than dropping D credits and making Cs the new Ds, as it were? You can still invest in Ds, they’re now just called C, more or less. This is delightfully Spinal Tap-is, of course, so kudos them.

Posted on

Multicollinearity, Diseases of Modernity, and How Everything Causes Everything

I’m regularly asked to explain how it is that the severity and mortality of the current outbreak are tied to so many factors, all of which seemingly have high explanatory power on their own. Whether it’s diabetes, COPD, obesity, air pollution, age, or anything else, they all, according to various studies, explain the differential risk faced by individuals. 

Here is an answer I gave recently, arguing that there are (at least) two ways to look at this, one more epidemiological, and one more holistic and almost philosophical:

  1. There is obviously multicollinearity here, with many of these putative explanations captured, to a greater or lesser degree, by other explanations. An obesogenic environment, for example, is closely tied to weight, air pollution, gender, age, and diabetes. Epidemiological explanations are, in all but the most trivial cases, over-determined: there are more causes than consequences.
  2. If everything explains differential Covid-19 severity, then it might arguably be characterized as a disease of modernity, like depression. Its spread and mortality then become merely the clinical manifestations of the environment in which we live. Trying to parse out a single risk factor or causal chain for increased severity or mortality then becomes somewhat futile, like parsing a central causal factor for depression, beyond modernity, of course.
Posted on

Ads , Pharma, and Target Markets

SXzpGWsT

I love it when a financial services provider has this sort of madness on its front page. It’s like watching a golf tournament, seeing endless pharma ads for incontinence or erectile dysfunction products, and realizing “I’m not their target market and I should stop doing this”.

Posted on

Virus Transmissibility, Retest Reliability, and Reinfection Rates

Some studies and related data this morning are driving me crazier than usual, so … rejoinders:

  • A CDC study trumpeting an R0 of 5.7, without saying that it was in China, in the outbreak’s early days, during Chinese New Year travel with as many as 415-million people on the road, staying in crowded multigenerational housing, and with nosocomial panic spikes. It might have briefly spiked to that level, but it almost certainly isn’t best characterized as an R0 5.7 virus
  • People shouting about low antibodies in “some” recovered patients, without noting that there is no test/retest reliability in the indicator. This is very early data. 
  • People shouting about reinfection rates, without noting the high false negative rate in nasopharyngeal swab for virus clearance that allows some people to be released prematurely. While coronaviruses can cause reinfection in recovered patients, it has usually been at low levels. 

Humans exhaust me when they find narratives they can’t abandon.

Posted on

Data on the Changing World of Data

Some remarkable datapoints from an Akamai virtual conference with analysts today (via CIBC):

  • Since the week of Feb 10th, Akamai has seen a 30% increase in traffic in four countries with early lock-downs (China, South Korea, Japan, Italy) vs. RoW.
  • Since the week of March 9th as more countries implemented lock-downs, global traffic saw 30% Y/Y growth vs. 3% on average
  • Peak traffic has more than doubled Y/Y to 167 Tbps.
  • AKAM execs calling it the “greatest spike in Internet traffic the company has ever experienced“
Posted on

Scaling Mt Stupid

I like to remind me as often as possible that while I may feel like I’m climbing the Slope of Enlightenment, that it’s more likely I’m still lost somewhere on the sheer walls of Mt Stupid.

C62780ef 1a03 445c 9e4f 280471d5325b