Quantian Profile picture
Aug 7 14 tweets 5 min read Twitter logo Read on Twitter
Since it's Sunday and I'm bored, I figured I would finally, definitively answer the question of "why do people think the economic is bad (and is that the case?)". This is because it is easy to answer, and the answer is funny and will make many people mad.
First of all, we will ignore everyone on Twitter and look at objective data, in this case the University of Michigan's consumer survey asking what people think economic conditions are. Spoiler: People think the economy is very bad, worse than any time other than 1980 and 2008! Image
Ok, now let's think of all the variables that might cause them to think that and build a model. I picked the following:
Inflation rate
Inflation rate change
Unemployment
Unemployment change
Housing prices
Real wages
Dollar strength
Interest rates
Stock prices
and went from there
Now, let's build a simple linear model to predict confidence from those 9 variables, but only fit it from 1978-2019, excluding COVID. The R^2 is *very* good- 87%- and the p-value is essentially 0, so consumer vibes can be almost perfectly be explained by objective data. Great! Image
(The results line up with what you'd expect, with a few surprises: people really want low levels of inflation, low unemployment, rising asset prices, high interest rates, etc. Two surprises: there's a slight preference for rising inflation, and real wages don't matter at all.)
Now, let's extend that chart above to include COVID. Same model as above. Oops! Now it's wildly wrong. Consumers should think the economy is really good (105-ish), *if they cared about the same things they did from 1979-2019*, which they don't. Economists HATE inconsistency! Image
So let's re-estimate the model, then, and figure out what changed. The answer is that three major things changed: consumers care a lot less about unemployment than they did, prefer housing prices to fall rather than rise now, and, more than anything else, HATE high interest rates Image
This third effect completely dominates everything else, and explains 104% of the 29-point gap between the model and current levels. The other effects largely offset each other, but that one is THE huge outlier.
So, the clear-cut, data-backed, but-for reason why Biden is unpopular is because Jerome Powell is making people's mortgage rates go up, and they're not happy about it. This may surprise you, but it really shouldn't: Americans LOVE borrowing money, and HATE being told they can't.
You could probably tell a of story about how America is now a deeply levered shitco that doesn't care about income as long as change in net cash stays positive, but like 500000 people have done that already and I need to get back to to work to pay off my credit cards now. THE END
(One quick robustness check here, before people complain about break points: if you fit your model *only on data from the 1980's*, it's still able to predict the next three decades almost perfectly. There really was a structural break, and it really did happen during COVID) Image
Two postscripts:
1. Some people say that data wildly diverging after the fitting period means it's overfit. It's not, you can train on much smaller portions (~12%) of the pre-2020 data and get the same fit and divergence. There is a structural break in 2020. Image
2. Other people make the (better) complaint about multicollinearity in the data. If you LASSO the variables instead of using OLS, you still see very similar results: housing flips negative, inflation becomes more salient, unemployment less salient, rate hikes hugely salient. Image
@profplum99 did a similar analysis with his own preferred set of variables and reached a similar conclusion that housing price and mortgage spreads are major drivers, with the CPI <> Unemployment salience flipped as well

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Quantian

Quantian Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @quantian1

Jun 26
Junior data scientist interview question: Assume you generate points X = N(0,1); Y = N(0,0.1). Rotate the (x,y) dataset 45 degrees, so they look something like pic below (line is y = x). If you were to calculate the OLS regression y = b1*x + b0, what is E[b1] as n->infinity?
(this question and the answer were previously buried in my replies, don't cheat or i will bully you for it)
The responses to this were bifurcated between “this is a hard question no one would get” and “this is a trivial question everyone should get ”. Surely if “data scientist” means anything, it means you should be able to answer basic linear algebra and mathematical stats questions!
Read 4 tweets
Mar 6
As I strive to provide #content for my followers, I typed up my ratings from this weekend. Of the 200+ wines I tasted, ~150 were producer samples of guaranteed quality, and ~100 of those were reds (I don't drink enough white burg to have an opinion). Editorializing to follow:
The scoring method was just A+ to F, since it needed to be simple enough to be robust to me getting smashed over the course of the tastings, since I didn't spit. The older wines were Friday, all the 2020s were Saturday.
It is probably not surprising that Dujac, Hudelot-Noellat, Salon, etc. had the best wines. However, I was very impressed with the Robert Chevillon NSG wines, and the Didier Fornerol regional appellations punched way above their weight. Denis/Arnaud Mortet was also great.
Read 5 tweets
Feb 26
Doesn’t take a genius to read behind the lines here and conclude that the agencies most likely to have actual relevant data from China think it’s natural, whereas the one who would have the least (the FBI, a *domestic* agency) has the highest confidence it is a leak (“moderate”).
The current agency distribution is 4-2-2 natural-leak-undecided, and curiously despite the breathless reporting on the FBI+DOE the identities and arguments of the 4 is are unremarked upon, despite their evidence being persuasive enough that an interagency panel sided with them.
Also, look, not to be tweeting about COVID in the year of our Lord 2023, but doesn’t lab leak prove that all the things lab leakers were complaining about were justified? If it’s a dangerous engineered Chinese pathogen then of course vax mandates, shutdowns, etc. are a good idea!
Read 4 tweets
Oct 27, 2022
Absolutely fascinating to read the memoirs of the founder of Huawei's trip to America in 1994 (translation by @ZichenWanghere) today pekingnology.com/p/ren-zhengfei…
Ren is impressed with the natural environment and beauty of the US, and is amazed that in New York ("the largest and most prosperous city in the United States, but also the dirtiest one with the worst social order", New Yorkers btfo) has a space like Central Park ImageImage
He complains that Chinese in America are too frugal, while American spend lavishly and make many friends, which allows them to easily succeed in business and commerce Image
Read 7 tweets
Oct 10, 2022
This is the greatest article I have ever read. Every part of it is so perfect it is beyond any parody wsj.com/articles/glori…
They attempted to make anti-woke credit cards out of bullet casings, but it was foiled by the fact that brass would interfere with the security chips!
Read 4 tweets
Aug 9, 2022
Militaries are built to accomplish specific goals, which dictates their cost- for instance, France's is "fuck around in Africa", China's is "take over Taiwan", Pakistan's is "fight India", etc. The US's is "win a war against anyone anytime anywhere in the world in under a month".
The US can deploy an 800-man Ranger battalion anywhere in the world within 12 hours, a 4,000 man rapid response brigade within 24, an armored Stryker brigade within 96, and an entire armored division within 5 days.
Many people look at US aircraft carriers and assume it must be slow and defenseless because that’s how you would balance a video game. In fact, they are the fastest *because of their size*. A Chinese 055 stealth destroyer has a hull speed of ~32kn, a nuclear powered carrier 44+!
Read 4 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us on Twitter!

:(