Wow, amazing piece. I’ve thoroughly enjoyed your recent private credit pieces and can say this meets their calibre. You’ve done some great analysing of the effects of JP Morgan’s choices on the software industry. I look forward to reading the next piece.
Wow, so much information. Can notice you’ve spent some time on this topic.
I still believe the BDC Sector downfall will continue because of the software - sector further declining, but rarely there will be big bankcrupcies.
Private credit has its place and demand for Private credit will continue for years to come, as Bank are getting more and more strict regulation.
Arcc, Trinity Capital, Capital Southwest & Hercules Capital are among strong players. Golub as well, but not sure about, as you referred to Dividend cut + refinancing going from 2,5% -> 7%.
The quality of your work never ceases to suprise me. As a long-time reader I am pleased to see you reffering to and updating your old work, with your specific reframing of “Private Credit Is Lying To You”. I can see you put emphasis on continious learning which has become exceedingly rare amoung todays authors. I hope in a few months time you will revisit this topic and revise what you have written here. This marks the difference between those who simply write, and those who treat their work as a living argument.
Uhhhh, so which magic exchange is is CRWD trading 368.99 on?
On which exchange did it drop 5.1% after earnings?
It's one thing to have AI write a lot of words, it's another thing to feed it bad data, but how can you present yourself a serious analyst using AI as a helper when the "SaaS analyst" isn't even bothering to proofread the piece? Or maybe they did, but didn't realize that that CRWD was trading $50 over the price that created all this "analysis"?
Just for funsies, I plugged this into an AI-checker and it came back 9% human and 91% AI.
I had noticed some things in earlier pieces that were probably AI hallucination and hand-waived them away -- but this is too egregious to ignore.
I thought you had some interesting takes on things and had covered some angles I hadn't thought of, but I'd rather not be paying to read AI slop where I have to double check every "fact".
You're right on the CRWD price. The scorecard pulled $368.99, which was the after-hours print from the March 3 earnings release, not the March 13 close. The correct figure is $441.78. The return flips from -5.1% to +13.6%. It has been corrected.
The error made CRWD look like a victim of the reflexive loop at the time of writing. It was not. It had already broken through.
We use AI as a research and drafting tool, the same way a Bloomberg terminal is a data tool. I do not outsource the analysis, the framework, or the calls. The CRWD price error was a proofreading failure, mine, not a hallucination. I should have caught it. I did not. It will not happen again.
But here we are again -- in the spirit of accountability, please go look at the after hours trading activity of CRWD on the earnings release date and tell me if you think that describing $368.99 as "the print" is in any way an accurate reflection of how it traded after hours as the earnings report was digested. Is the VWAP price < 370? <380? <390?
And a proofreading error is one thing -- typos, we all make them, whatever. But what was in the "article" wasn't a typo. An entire block of analysis about one of your strongest conviction plays was written based on a WILDLY inaccurate price.
So, if you wrote the piece (and not AI), and did so based on an incorrectly pulled price level, I guess that means that you have been unaware of the stock performance of one of your "strongest conviction" plays for a solid 10 days after record earnings?
Fair catch on the CRWD price — that's a data error and it undermines the specific example. The structural argument about bank-driven markdowns forcing deleveraging independent of company performance doesn't depend on any single stock's price, but you shouldn't have to do the fact-checking yourself. Fixing it.
Did you pull data around actual exposure to declining rate environment across the BDC sector? It’s common practice for BDCs to build in floor rates, that ensure loans through their maturity hurdle cost of capital - specifically put in place to mitigate against declining interest rate environment risk.
Curious your thoughts on the insurance channel. Apollo, KKR, Blackstone all own insurers now and are deploying policyholder float into the same loan book you’re describing, while offloading part of that liability to their own reinsurers in the Cayman Islands.
Seems like a second loop running in parallel, slower but potentially larger. Does that add a transmission channel?
Discovering this Substack is like discovering Bitcoin in 2011. Tremendous, deep value work here.
Thanks! Very kind of you.
Wow, amazing piece. I’ve thoroughly enjoyed your recent private credit pieces and can say this meets their calibre. You’ve done some great analysing of the effects of JP Morgan’s choices on the software industry. I look forward to reading the next piece.
Wow, so much information. Can notice you’ve spent some time on this topic.
I still believe the BDC Sector downfall will continue because of the software - sector further declining, but rarely there will be big bankcrupcies.
Private credit has its place and demand for Private credit will continue for years to come, as Bank are getting more and more strict regulation.
Arcc, Trinity Capital, Capital Southwest & Hercules Capital are among strong players. Golub as well, but not sure about, as you referred to Dividend cut + refinancing going from 2,5% -> 7%.
Let’s see. Interesting times indeed.
Excellent work. And thanks for the update especially on ADBE which I had asked about in my comment on your previous post.
The quality of your work never ceases to suprise me. As a long-time reader I am pleased to see you reffering to and updating your old work, with your specific reframing of “Private Credit Is Lying To You”. I can see you put emphasis on continious learning which has become exceedingly rare amoung todays authors. I hope in a few months time you will revisit this topic and revise what you have written here. This marks the difference between those who simply write, and those who treat their work as a living argument.
We agree. We highlighted this piece for its novel insight in Issue 002 today.
Fantastic analysis.
Another great piece
Another fantastic piece!
Uhhhh, so which magic exchange is is CRWD trading 368.99 on?
On which exchange did it drop 5.1% after earnings?
It's one thing to have AI write a lot of words, it's another thing to feed it bad data, but how can you present yourself a serious analyst using AI as a helper when the "SaaS analyst" isn't even bothering to proofread the piece? Or maybe they did, but didn't realize that that CRWD was trading $50 over the price that created all this "analysis"?
Just for funsies, I plugged this into an AI-checker and it came back 9% human and 91% AI.
I had noticed some things in earlier pieces that were probably AI hallucination and hand-waived them away -- but this is too egregious to ignore.
I thought you had some interesting takes on things and had covered some angles I hadn't thought of, but I'd rather not be paying to read AI slop where I have to double check every "fact".
Is there a refund button?
You're right on the CRWD price. The scorecard pulled $368.99, which was the after-hours print from the March 3 earnings release, not the March 13 close. The correct figure is $441.78. The return flips from -5.1% to +13.6%. It has been corrected.
The error made CRWD look like a victim of the reflexive loop at the time of writing. It was not. It had already broken through.
We use AI as a research and drafting tool, the same way a Bloomberg terminal is a data tool. I do not outsource the analysis, the framework, or the calls. The CRWD price error was a proofreading failure, mine, not a hallucination. I should have caught it. I did not. It will not happen again.
Accountability makes the work better.
I agree that accountability makes the work better
But here we are again -- in the spirit of accountability, please go look at the after hours trading activity of CRWD on the earnings release date and tell me if you think that describing $368.99 as "the print" is in any way an accurate reflection of how it traded after hours as the earnings report was digested. Is the VWAP price < 370? <380? <390?
And a proofreading error is one thing -- typos, we all make them, whatever. But what was in the "article" wasn't a typo. An entire block of analysis about one of your strongest conviction plays was written based on a WILDLY inaccurate price.
So, if you wrote the piece (and not AI), and did so based on an incorrectly pulled price level, I guess that means that you have been unaware of the stock performance of one of your "strongest conviction" plays for a solid 10 days after record earnings?
I am a paying subscriber.
I would like a refund.
Fair catch on the CRWD price — that's a data error and it undermines the specific example. The structural argument about bank-driven markdowns forcing deleveraging independent of company performance doesn't depend on any single stock's price, but you shouldn't have to do the fact-checking yourself. Fixing it.
Did you pull data around actual exposure to declining rate environment across the BDC sector? It’s common practice for BDCs to build in floor rates, that ensure loans through their maturity hurdle cost of capital - specifically put in place to mitigate against declining interest rate environment risk.
Curious your thoughts on the insurance channel. Apollo, KKR, Blackstone all own insurers now and are deploying policyholder float into the same loan book you’re describing, while offloading part of that liability to their own reinsurers in the Cayman Islands.
Seems like a second loop running in parallel, slower but potentially larger. Does that add a transmission channel?
This is excellent. I came down to write this after reading a few paragraphs- now I am going back to finish what I was reading. Nice work.