Hi Jacob! Glad you enjoyed. I laid out my workflow in a different response to Teal elsewhere in the comments here, but the tech stack is all R and mostly pulls from FRED, the BLS, NASDAQ data link, or Yahoo Finance. If you really want to look here's all the code I've used for my charts! https://github.com/Miles-byte/Apricitas
Epic! I was thinking it might have been Python or R, I had done my fair share of Python code back in college but never really got super great at it, looking at your charts makes me want to get back into it.
hi, great article! One point I am getting hung up on is why wealth inequality, aging population and low productivity would be deflationary. I would think low productivity and less workers would create more inflation as the widgets produced would be lower for the same amount of money in circulation, so cost per widget is higher.
In light of todays Inflation Numbers would you consider writing short break downs? I think theres room for short pieces on interpreting economic releases that may not merit a full article.
OK, I know, I am stupid. But, a few days ago we did our food etc. shopping and prices were considerably higher than a few weeks earlier. How will raising the interest rate make my groceries cheaper ?
I remember the Volker-years and how raising interest to a high level brought us in Europe a long and deep and brutal recession. I bought a small 2-room walk-up apartment, interest rate on the mortgage was 12,5 %. Because of the recession this dwelling lost half its "value" in three years time.
I fail to see how higher prices, unemployment, unaffordable housing, failing businesses, many people on welfare, etc. make for a good economy.
Powell will (and has on many occasions) admit that they don't believe rate hikes will lower gas or food prices, they are specifically targeting more cyclical components like housing and other core services.
In still do not get it. Housing prices go up too far (over and over) and the only option to force them down is bring misery to ordinary people ? There must be a better way.
The rate hikes will also not bring down my bills from the doctor and the dentist. Au contraire, as they are faced with just about everything being more expensive too.
I'd love to get a sense of your process and your 'tech stack'. You're quite prolific, so clearly you have a system that works well for you.
* Data acquisition: do you use a paid data aggregator (Haver/Macrobond, etc...), or do you download data directly (using FRED, website downloads etc..)?
* Do you just use Excel, or do you use R/Python/Tableau for data manipulation + visualization?
At some point it would be cool if you did a post about your process. I'd find it super interesting, and I suspect others would as well.
Thanks Teal! I am glad to hear you're enjoying the newsletter.
I don't use a paid data aggregator—the data for almost all of the charts comes either directly from APIs (mostly FRED, the BLS, Yahoo Finance, Nasdaq data link, and the EIA API) or is manually downloaded.
All the charts themselves are made in R, and that's also where I do the vast majority of the data cleaning/manipulation. It makes it very convenient to just plop in my design theme for new charts and re-run the code every time there's a new data release.
I might write more about my process sometime during the holidays as a fun "behind the scenes post", but most of the time articles are developed like this:
1) I see a big question or interesting data come up (usually online) and get inspired. Sometimes it's a reader comment, sometimes it's a wrong hot take I find interesting, sometimes it's a chart I think is cool. Whatever it is, it goes into a draft document.
2) Over time I keep adding to the draft document—collecting cool data, interesting papers, short notes, and a broad outline of what I want to write. The time in this stage is highly variable—my recent post on exchange-rate pass through spent about two months here, my post on inflation expectations spent more than a year here.
3) I select an article from the draft documents to publish this week. Usually this will just be an idea that has had enough time to marinate in stage 2, but sometimes I have to rush something out because there's breaking news or something I feel needs highlighting immediately.
4) I build out the charts for the final document. Oftentimes I can partially recycle or just straight-up reuse old charts (there are more than 300 charts I have preexisting code for at this point, so there will usually be something I've discussed previously that's relevant to highlight and I can draw on existing stock without becoming repetitive) but most pieces will require 4-5 new unique charts, and sometimes drafts start with different charts than the final posts.
5) I start writing around the charts. This is an important distinction—I write after the charts are made. Sometimes I have been forced to write before adding charts because of time or data constraints and it always leads to a worse product. The writing works best for me when it is a connective tissue that builds a narrative around the data I am showcasing.
Thanks! That very much answers my question. I'm fine-tuning my own workflow, and this is very helpful. I'm an R user too, and just built a custom ggplot2 theme. I work mostly with international economic & financial data, and I've been slowly building up my data acquisition and processing framework.
If you do feel up to writing a "behind the scenes" post, I would love it, and I suspect others would as well.
I have a lot of admiration for your work. Keep up the great work.
Could it be a path dependency, that a sudden jump in rates after an extended period of lower rates shocks borrowers and the economy into recession well before the demand effect (such as it may be) starts to work on inflation? I.e., stagflation? That would imply that a longer slower tightening path could protect the economy by allowing restructuring to occur while the inflation cure gradually takes effect? Would that be a better path?
Certainly, the Fed's ham-handed messaging combined with their cluelessness/silence about what it's going to take in total this week isn't boosting anyone's confidence.
Another technical but accessible article.
A suggestion for the decline in productivity, in the UK, is early retirement.
Beautiful article! I love these charts so much I'm curious to know what your tech stack looks like to make them as customizable as I assume they are?
Hi Jacob! Glad you enjoyed. I laid out my workflow in a different response to Teal elsewhere in the comments here, but the tech stack is all R and mostly pulls from FRED, the BLS, NASDAQ data link, or Yahoo Finance. If you really want to look here's all the code I've used for my charts! https://github.com/Miles-byte/Apricitas
Epic! I was thinking it might have been Python or R, I had done my fair share of Python code back in college but never really got super great at it, looking at your charts makes me want to get back into it.
Definitely have to read this twice. Does anyone on the FOMC read your masterpieces? Thanks for your hard work and research.
Thank you Robert! I don't know, but I hope so haha
Great article! Very helpful overview.
Thank you! I'm glad you enjoyed
Thank you for this article
But... how high will they go?
Perhaps nobody seems to know, but surely we can make a guess as to a ceiling which they are unlikely to reach.
For example, I feel 99% certain that they will NOT go to 20%, as during the Volcker era
Haha fair! Glad you enjoyed the piece.
hi, great article! One point I am getting hung up on is why wealth inequality, aging population and low productivity would be deflationary. I would think low productivity and less workers would create more inflation as the widgets produced would be lower for the same amount of money in circulation, so cost per widget is higher.
In light of todays Inflation Numbers would you consider writing short break downs? I think theres room for short pieces on interpreting economic releases that may not merit a full article.
Planning on writing on the inflation numbers this weekend!
OK, I know, I am stupid. But, a few days ago we did our food etc. shopping and prices were considerably higher than a few weeks earlier. How will raising the interest rate make my groceries cheaper ?
I remember the Volker-years and how raising interest to a high level brought us in Europe a long and deep and brutal recession. I bought a small 2-room walk-up apartment, interest rate on the mortgage was 12,5 %. Because of the recession this dwelling lost half its "value" in three years time.
I fail to see how higher prices, unemployment, unaffordable housing, failing businesses, many people on welfare, etc. make for a good economy.
Powell will (and has on many occasions) admit that they don't believe rate hikes will lower gas or food prices, they are specifically targeting more cyclical components like housing and other core services.
In still do not get it. Housing prices go up too far (over and over) and the only option to force them down is bring misery to ordinary people ? There must be a better way.
The rate hikes will also not bring down my bills from the doctor and the dentist. Au contraire, as they are faced with just about everything being more expensive too.
As always, very much enjoy your analysis.
I'd love to get a sense of your process and your 'tech stack'. You're quite prolific, so clearly you have a system that works well for you.
* Data acquisition: do you use a paid data aggregator (Haver/Macrobond, etc...), or do you download data directly (using FRED, website downloads etc..)?
* Do you just use Excel, or do you use R/Python/Tableau for data manipulation + visualization?
At some point it would be cool if you did a post about your process. I'd find it super interesting, and I suspect others would as well.
Thanks Teal! I am glad to hear you're enjoying the newsletter.
I don't use a paid data aggregator—the data for almost all of the charts comes either directly from APIs (mostly FRED, the BLS, Yahoo Finance, Nasdaq data link, and the EIA API) or is manually downloaded.
All the charts themselves are made in R, and that's also where I do the vast majority of the data cleaning/manipulation. It makes it very convenient to just plop in my design theme for new charts and re-run the code every time there's a new data release.
I might write more about my process sometime during the holidays as a fun "behind the scenes post", but most of the time articles are developed like this:
1) I see a big question or interesting data come up (usually online) and get inspired. Sometimes it's a reader comment, sometimes it's a wrong hot take I find interesting, sometimes it's a chart I think is cool. Whatever it is, it goes into a draft document.
2) Over time I keep adding to the draft document—collecting cool data, interesting papers, short notes, and a broad outline of what I want to write. The time in this stage is highly variable—my recent post on exchange-rate pass through spent about two months here, my post on inflation expectations spent more than a year here.
3) I select an article from the draft documents to publish this week. Usually this will just be an idea that has had enough time to marinate in stage 2, but sometimes I have to rush something out because there's breaking news or something I feel needs highlighting immediately.
4) I build out the charts for the final document. Oftentimes I can partially recycle or just straight-up reuse old charts (there are more than 300 charts I have preexisting code for at this point, so there will usually be something I've discussed previously that's relevant to highlight and I can draw on existing stock without becoming repetitive) but most pieces will require 4-5 new unique charts, and sometimes drafts start with different charts than the final posts.
5) I start writing around the charts. This is an important distinction—I write after the charts are made. Sometimes I have been forced to write before adding charts because of time or data constraints and it always leads to a worse product. The writing works best for me when it is a connective tissue that builds a narrative around the data I am showcasing.
Hope that helps answer your questions!
Thanks! That very much answers my question. I'm fine-tuning my own workflow, and this is very helpful. I'm an R user too, and just built a custom ggplot2 theme. I work mostly with international economic & financial data, and I've been slowly building up my data acquisition and processing framework.
If you do feel up to writing a "behind the scenes" post, I would love it, and I suspect others would as well.
I have a lot of admiration for your work. Keep up the great work.
Could it be a path dependency, that a sudden jump in rates after an extended period of lower rates shocks borrowers and the economy into recession well before the demand effect (such as it may be) starts to work on inflation? I.e., stagflation? That would imply that a longer slower tightening path could protect the economy by allowing restructuring to occur while the inflation cure gradually takes effect? Would that be a better path?
Certainly, the Fed's ham-handed messaging combined with their cluelessness/silence about what it's going to take in total this week isn't boosting anyone's confidence.