>However, Anthropic’s disclosure to KBRA excluded several important financial metrics, including free cash flow and earnings before interest, taxes, depreciation and amortisation. Credit rating analysts had to come up with their own estimates based on assumptions, the documents show.
>Given the limited information, KBRA produced a “risk score” to assess Anthropic’s creditworthiness rather than the traditional full financial models.
This and the fact they went to a small ratings agency does not inspire confidence in the rating.
No, but a trillion dollar company going to a ratings agency to get an assessment of its credit rating and not giving that ratings agency enough information to do that doesn’t inspire confidence in that trillion dollar company, either.
Why didn’t they give them the info? They either withheld it or don’t have it. Both make them look bad.
that's a bigger red flag. there may be reasons (bad) to not give them all the info, but going to a smaller agency is the largest of red flags when you are that big of a company.
If you look at 10-year default rates for BBB-rated paper from the big three plus EJR, AM Best, KBRA and DBRS, you find everyone tightly clustered in the 1 to 2% range except KBRA, who was up at 4% [1]. That suggests they've historically rated junk bonds as barely investment grade. I'm assuming that's what Anthropic was hoping for.
KBRA is tasked with producing a rating. Insufficient information won't stop that task. The rating has the confidence it deserves: Anthropic is hiding the bad data and as such no one can rate them highly.
No, the title is poorly phrased, I think you misunderstood.
All ratings agencies have their own scales, but broadly all ratings fall into one of two overarching categories, investment grade vs. junk.
The article explains that the rating ANT received was the highest within the junk range. So “highest junk rating” means within the junk category, it’s in the best sub-area. In other words, it’s just scratching below the investment grade rating.
This is important for institutional investors who can only buy investment grade, meaning if they can improve even one single grade up, they get access to a lot more debt.
> This is important for institutional investors who can only buy investment grade, meaning if they can improve even one single grade up, they get access to a lot more debt.
I believe some institutional investors have policies allowing some purchasing of 'highest junk' as well. Depends on the fund, etc.
You can hold both things in your mind and they aren't contradictory. It can be a bad system but it can still be suspicious if someone behaves outside the norm inside it.
Other posts made me understand what it actually measures and why is it meaningful - it tells you if you lend this entity money at the govt bond interest rate, will this entity be able to pay back more or less eventiually?
Obviously a financially rational person wouldn't lend money with worse returns it can get from the government risk-free as that would be charity, hence these are considered 'junk'.
First SpaceX, now Anthropic. I wonder what happened the last time the big rating agencies gave massive amounts of sketchy debt investment-grade ratings to avoid alienating their customers?
"The rating action, which has not previously been reported, shows Anthropic is close to achieving the investment-grade status that would allow the start-up to lower its borrowing costs and tap a wider pool of institutional investors including life insurance companies and pension funds.
Bankers for Anthropic and rival OpenAI have lobbied the largest rating agencies to grant the AI labs investment-grade ratings once they go public."
The fact that the only rating published is ‘junk’, and that it was from a small agency tells us quite a lot. Given the fact that Anthropic (or its representatives) likely ‘shopped’ for a rating means that this was the best anyone was willing to give them, and that the larger agencies probably refused to rate them without more information. I would have been surprised if the bankers hadn’t lobbied for the highest rating possible, as higher ratings increase their profits from the IPO.
The whole point is to slowly build credibility, starting with easily $wayed small ratings entities, then launder that credibility into a slightly larger one, rise and repeat until financial disaster.
The point of the headline isn’t to communicate that their debt is currently rated junk, it’s to communicate that it’s impossible for their rating to improve any further without becoming investment grade.
"The rating action, which has not previously been reported, shows Anthropic is close to achieving the investment-grade status that would allow the start-up to lower its borrowing costs and tap a wider pool of institutional investors including life insurance companies and pension funds.
Bankers for Anthropic and rival OpenAI have lobbied the largest rating agencies to grant the AI labs investment-grade ratings once they go public."
Bonds and a high risk high reward business model feel like a bad fit.
Like if there is a significant risk of bankruptcy then there must on the other side of the scale be a possibility of making bank like you have with stock or derivatives.
Subprime loans blew up because it was a bunch of opaque securities (mortgages sliced and diced together) that lenders thought were safe, but really weren't. Anthropic debt doesn't have this issue because everyone knows about anthopric and their business model.
> Subprime loans blew up because it was a bunch of opaque securities (mortgages sliced and diced together) that lenders thought were safe, but really weren't
Not quite. They blew up because the opacity meant nobody could verify that their paper was safe. That made them unsaleable. In the end, AAA-rated crisis-era paper all paid out and almost all of them paid out on schedule. But that doesn't help you if you need to fire sell them today.
Put more succinctly, a lot of people assumed that because Treasuries are highly rated and super liquid all highly-rated paper could be counted on to be liquid. Lenders thought the safe paper was liquid. In the end it was just safe.
GPU backed securities also don't have that issue because they're far more transparent. The problem with mortgage backed securities is that nobody has the time to inspect the credit worthiness of each mortgage in a MBS, each creditor has hidden correlated risk (eg. if the whole economy tanks), and on top of all that trenching obfuscates the risk. Lenders therefore had to take the credit ratings on face value. It's hard to make the mistake with datecenter/gpu/anthropic debt.
> The problem with mortgage backed securities is that nobody has the time to inspect the credit worthiness of each mortgage in a MBS
Credit worthiness of each mortgage wasn't the issue with mortgage-backed securities. It was that low value mortgage-backed securities were repackaged into a basket with a mix of securities and given a higher credit rating than what Subprime would have gotten on their own.
This repackaging is absolutely going to happen again with GPUs. If there are GPUs of different generations within the same security, it's already happened.
> low value mortgage-backed securities were repackaged into a basket with a mix of securities and given a higher credit rating than what Subprime would have gotten on their own
This is fine. Tranching is fine. It works today, it worked in 2008 and it's based on the same mathematics as portfolio theory. (A physical representation would be a dam. You can turn volatile rivers into predictable flows if you hold a bunch in reserve and then let a small amount off the top.)
Nobody knew that for certain, however, and that was a problem if you needed cash today but verifying the underlying mortgages would take months of work. (Some AAA-rated CDOs defaulted though they all paid out in the end. There has never been a AAA-rated CLO default since like the 1990s.)
> Bankers for Anthropic and rival OpenAI have lobbied the largest rating agencies to grant the AI labs investment-grade ratings once they go public. Investors expect the initial public offering this year to hit a valuation of more than $2tn.
Let’s be clear about what this means: it gives them access to the corporate bond market. What they’re lobbying for is that pension funds, life insurance, etc., can take exposure to these companies. Suddenly the everyman, your mom and dad, have significant exposure to these companies and their circular financing deals without even knowing it. This is disgusting.
What a lot of people fail to realize is that the rules around IPOs changed since the dot com bubble. Previously, employees counted as shareholders and companies above a certain threshold of shareholders were required to IPO. This is why companies like Microsoft, Amazon, etc., had the majority of their growth while being publicly traded. This allowed the everyman with a brokerage account to participate in these gains.
These rules were essentially removed by Obama in 2012 through the JOBS act. What this means is that companies can stay private for a majority of their growth phase, which is why we’re seeing IPOs being floated with $2T dollar amounts on them. They’ve privatized the gains, and they only IPO once they want to cash out.
So let’s understand clearly what’s about to happen: early VC investors in oai/ant are going to make billions selling dogshit to your parents pension funds, your parents pension funds are run by idiots who will gleefully take the AI exposure and are legally allowed to take that AI exposure because the AI companies lobbied the credit rating agencies to allow it, and when the bills come next year and these companies are completely unable to meet their liabilities, your parents have to postpone their retirements or come back out of retirement to make ends meet.
It’s truly evil. I hope the employees at these companies with dollar signs in their eyes awaiting their multimillion dollar payoff understand the system that they’re participating in and how evil it is.
Point blank: 2008 was about manipulation of the credit rating agencies. This bubble will also be about manipulation of the credit rating agencies. Absolutely deplorable.
And no, just to nip this in the bud, AI isn’t a fad. It’s the most transformative technology of our lifetimes. It’s going to change everything. But that doesn’t have anything to do with chaos in financial markets or malinvestment during the gold rush. It has to do with whether these companies have any kind of moat and whether the billions being poured into current generation chips are going to depreciate over the timeframes that investors think they will.
I don’t like that they’re fleecing the corporate bond market after privatizing the majority of their upside. Yes, I wanted them to issue their junk bonds even earlier.
What I really want above all else is for credit rating agencies to not face pressure to rerate them favorably because they’re deemed too big to fail.
If people are getting 'fleeced' in the corporate bond market, they deserve it because they should know better. Corporate bonds are traded mainly by institutional investors and professionals, not retail like shares are.
This is the kind of wisdom that caused the great financial crisis. Funds are limited in what they can invest in. Institutional investors don’t see high returns. They’re under a lot of pressure to maximize returns within the bounds of what they’re allowed to invest in. There’s a reason they’re risk restricted: they are responsible for a lot of money and society puts a lot of trust into them to not piss away their pension funds.
We learned this lesson the hard way twenty years ago. If we let them invest in dogshit, and the credit rating agencies allow them to, they will.
When institutional investors lose, they’re not really risking their own money. They’re risking yours, or at least your cousins or your neighbors.
How are they fleecing anybody if their rating is junk?
They can lobby all they want but until we see ratings agencies bending to their lobbying in the face of extreme scrutiny from anyone who would buy these corporate bonds I wouldn't go about worrying too much.
Why would you tell people to wait to worry about it until it's happening? What more use would it be to worry about it then?
Now is when you make sure that the agencies are scrutinized before they give out inflated ratings, and make sure that this nobody agency's ratings aren't enough to get the stuff sold where it could do real damage.
You don't wait until the baby is drowning to worry, you worry when he's stumbling around next to the pool. Even then, you've gotten involved way later than you should have been.
Because for a pension fund the exposure will be much smaller if they're an early investor in a 100m Business than being a late investor in 2T business.
Basically when you invest at the peak of the bubble you're going to be hosed.
And if default risk rises, yes the value of those bonds will get crushed. But Id argue the value of equity will be more so - since they are not A) positive EBIT B) EBITDA
Those two numbers are variables that compose measures of default risk.
Yes, and there must be a reason they went with Kroll instead of someone else. It is possibly a bad reason, and we should by default be skeptical and assume such with the little information we have, until proven otherwise.
> Let’s be clear about what this means: it gives them access to the corporate bond market. What they’re lobbying for is that pension funds, life insurance, etc., can take exposure to these companies. Suddenly the everyman, your mom and dad, have significant exposure to these companies and their circular financing deals without even knowing it. This is disgusting.
What's the problem? Fund managers—who are in a position to understand the risks more than your everyday investor—can decide if they want buy these bonds or not, just like any other public corporate bond.
Most pensions and life insurers and so on are mandate constrained to investment grade only.
2008 wasn't fund managers deciding all together in unison to buy bad mortgages. It was AAA ratings on diversified MBS that should have been rated much lower giving them cover to load up.
All this, so people like us can do a job that wasn’t that hard in the first place, and in reality was actually quite comfortable, for employers who are promising to lay us off, for productivity gains that aren’t even measurable.
I just wish we could invest in the Chinese companies without the USG coming in later and claiming it's illegal to invest in them and seize the investment. It's clear when the US models financially crumble, everyone will jump on the Chinese models. (https://www.foxrothschild.com/publications/investments-in-ch...)
What's not clear is whether the west can build a separate but equal ecosystem here when VC only funds privatized models, but it was hilarious when some influencer went on and on that Chinese open weight models are unsafe because we don't have access to their training code and data. Serious WTF territory IMO.
It's because their datacenter businesses are growing so fast YoY. Why do you care about the difference in a few percentage points of interest if you're getting 37, 41, or even 82% returns YoY?
Correct hyperscalars are doing fine - as it stands.
The problem is OAI/Anthropic - but if OAI and Anthropic fail the whole thing comes apart. Goog, Meta etc will be fine as they had tremendous cash flow generating operations prior to datacenter offerings w.r.t LLMs.
If OpenAI and Anthropic somehow become insolvent to the point that they are unable to continue running inference on their existing trained models, their tens of billions of dollars of monthly revenue will need to go somewhere.
If you believe that the upside is effectively unlimited because you automate most or all jobs in existence, then it is "rational" but silly valley rationality usually ignores things that are inconvenient. But if you truly believe the upside is unlimited, then rising interest rates won't scare you.
I think the issues are:
1) that timeline has to line up with the broader macroeconomic factors (unlikely imo)
2) anthropic/openai/etc have to actually capture the value (unlikely given deepseek/glm/etc)
3) for all jobs to be automated, there needs to be a huge shift in the social contract and welfare state, which requires competent and functioning government (lol) and taxes on destruction of labor rather than labor itself (lmao) otherwise you get millions of unemployed people and a deeply unstable society, and a lot of blown up datacenters.
I think the value of AI is unlike anything that has come before, I just don't think the big players will necessarily be the ones capturing it. There will be a lot of spectacular implosions over the next couple years, and AI will continue to march along.
Every now and then I read articles from the Financial Times just to confirm that they are still a terrible source of information.
BB+ is technically junk. But the it is one notch from investment grade. Also Junk ratings are normal for young. So the framing is sensationalized.
"Growth has slowed" is technically true but misleading. Q2 to Q3 still went from $11.5bn to $16.8bn. And at those levels, it’s not wildly unexpected that it has to slow down.
And the bias is obvious when they write "Despite limited disclosures" which makes it sounds suspicious, when it’s normal thatrivate companies don't publish financials.
It’s also technical errors. Broadcom agreed to support the residual value of the chips, they did not guarantee the loan.
When journalists and news organizations are this terrible at basic reporting, Within their own specialized domain, it's hard to find sympathy for their complaints about AI-generated content taking over.
Edit: after another look I found that the agency that gave the rating also is a new small one with not very good record eighter (Kroll Bond Rating Agency).
Alternative title: Small Agency, which is previously has been fined for Ratings Without Method, Guesses that Anthropic's Numbers are close to normal for startups; FT Reports the Guess as News.
‘Growth has slowed" is technically true but misleading. Q2 to Q3 still went from $11.5bn to $16.8bn. And at those levels, it’s not wildly unexpected that it has to slow down.’
The growth slowdown is decidedly unexpected. The market priced in a $2T valuation that assumed it wouldn’t slow down.
>However, Anthropic’s disclosure to KBRA excluded several important financial metrics, including free cash flow and earnings before interest, taxes, depreciation and amortisation. Credit rating analysts had to come up with their own estimates based on assumptions, the documents show.
>Given the limited information, KBRA produced a “risk score” to assess Anthropic’s creditworthiness rather than the traditional full financial models.
This and the fact they went to a small ratings agency does not inspire confidence in the rating.
Why didn’t they give them the info? They either withheld it or don’t have it. Both make them look bad.
that's a bigger red flag. there may be reasons (bad) to not give them all the info, but going to a smaller agency is the largest of red flags when you are that big of a company.
If you look at 10-year default rates for BBB-rated paper from the big three plus EJR, AM Best, KBRA and DBRS, you find everyone tightly clustered in the 1 to 2% range except KBRA, who was up at 4% [1]. That suggests they've historically rated junk bonds as barely investment grade. I'm assuming that's what Anthropic was hoping for.
[1] https://www.ft.com/content/d97e1e86-1e90-4c5c-b972-052d43a36...
Rating agencies will absolutely refuse to rate if provided insufficient information.
...
"If we don't work with them they will go to our competitors. Not our fault. Simply the way the world works."[1]
[1] https://www.youtube.com/watch?v=mwdo17GT6sg @ 3:31
...and score still showed "high probability of being junk"?
All ratings agencies have their own scales, but broadly all ratings fall into one of two overarching categories, investment grade vs. junk.
The article explains that the rating ANT received was the highest within the junk range. So “highest junk rating” means within the junk category, it’s in the best sub-area. In other words, it’s just scratching below the investment grade rating.
This is important for institutional investors who can only buy investment grade, meaning if they can improve even one single grade up, they get access to a lot more debt.
I believe some institutional investors have policies allowing some purchasing of 'highest junk' as well. Depends on the fund, etc.
Then you have been misinformed. Rated bonds by and large fail at the rating predicted by the big three [1].
[1] https://www.ft.com/content/d97e1e86-1e90-4c5c-b972-052d43a36...
Obviously a financially rational person wouldn't lend money with worse returns it can get from the government risk-free as that would be charity, hence these are considered 'junk'.
"The rating action, which has not previously been reported, shows Anthropic is close to achieving the investment-grade status that would allow the start-up to lower its borrowing costs and tap a wider pool of institutional investors including life insurance companies and pension funds.
Bankers for Anthropic and rival OpenAI have lobbied the largest rating agencies to grant the AI labs investment-grade ratings once they go public."
It is not like Anthropic was once this stable investment grade company that is now shopping around because the debt is being downgraded.
The headline is just click bait.
"The rating action, which has not previously been reported, shows Anthropic is close to achieving the investment-grade status that would allow the start-up to lower its borrowing costs and tap a wider pool of institutional investors including life insurance companies and pension funds.
Bankers for Anthropic and rival OpenAI have lobbied the largest rating agencies to grant the AI labs investment-grade ratings once they go public."
Okay thanks
As in "main ratings agencies" see no reason to bother
I’d reserve judgement until i see them report more detailed financials and the big ratings agencies step in with assessment.
Like if there is a significant risk of bankruptcy then there must on the other side of the scale be a possibility of making bank like you have with stock or derivatives.
Not quite. They blew up because the opacity meant nobody could verify that their paper was safe. That made them unsaleable. In the end, AAA-rated crisis-era paper all paid out and almost all of them paid out on schedule. But that doesn't help you if you need to fire sell them today.
Put more succinctly, a lot of people assumed that because Treasuries are highly rated and super liquid all highly-rated paper could be counted on to be liquid. Lenders thought the safe paper was liquid. In the end it was just safe.
Credit worthiness of each mortgage wasn't the issue with mortgage-backed securities. It was that low value mortgage-backed securities were repackaged into a basket with a mix of securities and given a higher credit rating than what Subprime would have gotten on their own.
This repackaging is absolutely going to happen again with GPUs. If there are GPUs of different generations within the same security, it's already happened.
This is fine. Tranching is fine. It works today, it worked in 2008 and it's based on the same mathematics as portfolio theory. (A physical representation would be a dam. You can turn volatile rivers into predictable flows if you hold a bunch in reserve and then let a small amount off the top.)
Nobody knew that for certain, however, and that was a problem if you needed cash today but verifying the underlying mortgages would take months of work. (Some AAA-rated CDOs defaulted though they all paid out in the end. There has never been a AAA-rated CLO default since like the 1990s.)
A link to the criminal mastermind that makes free speech possible here.
Let’s be clear about what this means: it gives them access to the corporate bond market. What they’re lobbying for is that pension funds, life insurance, etc., can take exposure to these companies. Suddenly the everyman, your mom and dad, have significant exposure to these companies and their circular financing deals without even knowing it. This is disgusting.
What a lot of people fail to realize is that the rules around IPOs changed since the dot com bubble. Previously, employees counted as shareholders and companies above a certain threshold of shareholders were required to IPO. This is why companies like Microsoft, Amazon, etc., had the majority of their growth while being publicly traded. This allowed the everyman with a brokerage account to participate in these gains.
These rules were essentially removed by Obama in 2012 through the JOBS act. What this means is that companies can stay private for a majority of their growth phase, which is why we’re seeing IPOs being floated with $2T dollar amounts on them. They’ve privatized the gains, and they only IPO once they want to cash out.
So let’s understand clearly what’s about to happen: early VC investors in oai/ant are going to make billions selling dogshit to your parents pension funds, your parents pension funds are run by idiots who will gleefully take the AI exposure and are legally allowed to take that AI exposure because the AI companies lobbied the credit rating agencies to allow it, and when the bills come next year and these companies are completely unable to meet their liabilities, your parents have to postpone their retirements or come back out of retirement to make ends meet.
It’s truly evil. I hope the employees at these companies with dollar signs in their eyes awaiting their multimillion dollar payoff understand the system that they’re participating in and how evil it is.
Point blank: 2008 was about manipulation of the credit rating agencies. This bubble will also be about manipulation of the credit rating agencies. Absolutely deplorable.
And no, just to nip this in the bud, AI isn’t a fad. It’s the most transformative technology of our lifetimes. It’s going to change everything. But that doesn’t have anything to do with chaos in financial markets or malinvestment during the gold rush. It has to do with whether these companies have any kind of moat and whether the billions being poured into current generation chips are going to depreciate over the timeframes that investors think they will.
First you say you don't like that they're accessing the corporate bond market, which is something all public companies can do.
Then you say they should've IPO'd earlier when they were even earlier stage and riskier .. which would allow them to issue bonds even earlier?
What is your position? Or are you complaining for the sake of complaining?
What I really want above all else is for credit rating agencies to not face pressure to rerate them favorably because they’re deemed too big to fail.
We learned this lesson the hard way twenty years ago. If we let them invest in dogshit, and the credit rating agencies allow them to, they will.
When institutional investors lose, they’re not really risking their own money. They’re risking yours, or at least your cousins or your neighbors.
They can lobby all they want but until we see ratings agencies bending to their lobbying in the face of extreme scrutiny from anyone who would buy these corporate bonds I wouldn't go about worrying too much.
Now is when you make sure that the agencies are scrutinized before they give out inflated ratings, and make sure that this nobody agency's ratings aren't enough to get the stuff sold where it could do real damage.
You don't wait until the baby is drowning to worry, you worry when he's stumbling around next to the pool. Even then, you've gotten involved way later than you should have been.
Basically when you invest at the peak of the bubble you're going to be hosed.
And if default risk rises, yes the value of those bonds will get crushed. But Id argue the value of equity will be more so - since they are not A) positive EBIT B) EBITDA
Those two numbers are variables that compose measures of default risk.
Calm your panties down. Kroll != S&P.
What's the problem? Fund managers—who are in a position to understand the risks more than your everyday investor—can decide if they want buy these bonds or not, just like any other public corporate bond.
2008 wasn't fund managers deciding all together in unison to buy bad mortgages. It was AAA ratings on diversified MBS that should have been rated much lower giving them cover to load up.
[1] https://www.wsj.com/livecoverage/stock-market-today-dow-sp-5...
[2] https://www.tradingview.com/news/seekingalpha:173ffc32a094b:...
[3] https://www.cnbc.com/2026/10/15/coreweave-credit-default-swa...
You can't really rip off bond-holders when you are not positive EBIT and EBITDA...
All this, so people like us can do a job that wasn’t that hard in the first place, and in reality was actually quite comfortable, for employers who are promising to lay us off, for productivity gains that aren’t even measurable.
The problem is OAI/Anthropic - but if OAI and Anthropic fail the whole thing comes apart. Goog, Meta etc will be fine as they had tremendous cash flow generating operations prior to datacenter offerings w.r.t LLMs.
I think the issues are:
1) that timeline has to line up with the broader macroeconomic factors (unlikely imo)
2) anthropic/openai/etc have to actually capture the value (unlikely given deepseek/glm/etc)
3) for all jobs to be automated, there needs to be a huge shift in the social contract and welfare state, which requires competent and functioning government (lol) and taxes on destruction of labor rather than labor itself (lmao) otherwise you get millions of unemployed people and a deeply unstable society, and a lot of blown up datacenters.
I think the value of AI is unlike anything that has come before, I just don't think the big players will necessarily be the ones capturing it. There will be a lot of spectacular implosions over the next couple years, and AI will continue to march along.
BB+ is technically junk. But the it is one notch from investment grade. Also Junk ratings are normal for young. So the framing is sensationalized.
"Growth has slowed" is technically true but misleading. Q2 to Q3 still went from $11.5bn to $16.8bn. And at those levels, it’s not wildly unexpected that it has to slow down.
And the bias is obvious when they write "Despite limited disclosures" which makes it sounds suspicious, when it’s normal thatrivate companies don't publish financials.
It’s also technical errors. Broadcom agreed to support the residual value of the chips, they did not guarantee the loan.
When journalists and news organizations are this terrible at basic reporting, Within their own specialized domain, it's hard to find sympathy for their complaints about AI-generated content taking over.
Edit: after another look I found that the agency that gave the rating also is a new small one with not very good record eighter (Kroll Bond Rating Agency).
Alternative title: Small Agency, which is previously has been fined for Ratings Without Method, Guesses that Anthropic's Numbers are close to normal for startups; FT Reports the Guess as News.
The growth slowdown is decidedly unexpected. The market priced in a $2T valuation that assumed it wouldn’t slow down.
https://www.reuters.com/business/finance/investors-temper-an...