🦔A Wall Street Journal analysis published Wednesday found that U.S. life insurers hold $40 billion in bonds graded by Egan-Jones Ratings, a small firm the WSJ found grades bonds one to three notches higher than peer agencies. Egan-Jones is facing an SEC probe and a lawsuit from former employees who say they were pressured to inflate grades. Bermuda's regulator dropped the firm from its approved list in January. The largest holders include TIAA, Mass Mutual, and insurance companies controlled by Dodgers owner Mark Walter, whose debt is tied to a $16 billion federal fraud investigation. My Take Anyone who saw The Big Short will recognize what is happening here. Rating agencies get paid by the bond issuers whose debt they grade, insurance companies use those grades to determine how much money they set aside in reserves, and inflated grades mean insurance companies are undercapitalized against the bonds they hold. Moody's and S&P ran this playbook with mortgage bonds and everyone still remembers how that ended. Egan-Jones is doing the same thing with private credit right now, and most Americans have annuity money or pension money that touches this debt without their knowledge. Federal investigators are already tracing $16 billion in loans that moved through third parties before reaching the books of insurance companies Mark Walter controls, and Egan-Jones graded most of that debt. If investigators confirm the loans were mislabeled to hide his affiliation, the reserves shrink, and the policyholders who paid premiums for decades learn that Mark Walter's TV broadcasting company was sitting behind their retirement all along. The SEC has known about this pattern since 2023 and Egan-Jones keeps rating deals. Regulators typically move only after an insurance company collapses. Hedgie🤗

🦔SoftBank just borrowed $10 billion using its OpenAI stake as collateral to buy more OpenAI stock. A consortium led by Goldman Sachs and JPMorgan arranged the loan at 7.88%, nearly triple what a margin loan against public stock would cost. The banks couldn't independently value the collateral because OpenAI is private and its shares don't trade, so SoftBank had to add corporate guarantees to close the deal. SoftBank has now committed $65 billion to OpenAI. The company holds $135 billion in total debt. My Take The banks that arranged this loan know what they're taking on. A normal margin loan against public stock runs closer to 2.5%. This one costs 7.88% because if OpenAI's valuation drops, there is no exchange where the banks can force-sell the collateral. Secondary market data already shows five sellers for every buyer of OpenAI stock, and that stress shows up in the loan pricing. Half of everything Masayoshi Son borrowed now sits on one private company that has never turned a profit. This is the same investor who put $18 billion into WeWork before that valuation dropped from $47 billion to $8 billion. Banks like Goldman and JPMorgan don't hold this debt on their own books for long. They package it into structured products that pension funds and insurance companies buy without visibility into the collateral terms. Hedgie🤗

🦔AI companies are bulk-buying rare books, scanning them through high-speed machines that cut the spines off, and shredding the originals. A service called ISBNdb facilitates orders of up to a million books and keeps buyers anonymous. Pre-2022 books are premium because they're free of AI-generated text. A federal judge ruled the practice is fair use because eliminating the original means only one copy exists at a time. Anthropic hired the former head of Google Books partnerships to obtain "all the books in the world." My Take This got to me. A bookseller told 404 Media that rare books with almost no surviving copies are being fed into this pipeline. Books that survived wars, fires, and centuries of handling are being shredded so an AI can learn to write a better marketing email. ISBNdb's website literally says "'AI company destroys two million books' is not a headline that generates sympathy," and they still built an entire business around making it happen quietly. They offer NDAs as a feature. They coach clients to call it "digital preservation." I've covered AI companies scraping the internet, torrenting libraries, and stealing music. This is worse because it's irreversible. You can re-upload a website. You can reprint a bestseller. You can't replace the last three copies of an 18th-century botanical text once someone shreds them for training data. And the judge said it's legal. So it's going to accelerate. "We shred rare books and offer NDAs so nobody finds out" is a legitimate business model in 2026. What a timeline. Hedgie🤗

🦔Meta announced it will sell excess AI compute capacity to outside customers. The company spent $145 billion on AI infrastructure this year, put chips in fabric tents to speed up deployment, drafted 6,500 engineers into AI training they call soul-crushing, and cut thousands of jobs to fund the buildout. CoreWeave stock dropped on the news. My Take Six months ago these companies said there wasn't enough compute in the world. Now Meta is selling its leftovers. A company doesn't sell surplus of something it needs. CoreWeave went public on the promise that demand for AI compute would outstrip supply for years. SpaceX started selling compute too. Now Meta joins them. Every new seller of excess capacity makes it harder for the next company to justify a premium price, and every one of them undercuts the scarcity story that the IPOs were priced on. Three companies are now competing to sell compute they built on the assumption someone else would buy it. The demand that was supposed to absorb all of this hasn't shown up, and nobody wants to be the last one holding unsold capacity when the market figures that out. Hedgie🤗