Newsletter · 2026-10-06 · Ashutosh Agarwal

Oura Pulls Its IPO as Anthropic's Prospectus Draws Fire and Yields Test Every Deal - Capital Markets: IPOs, M&A & Exchanges - Week of October 6, 2026

Capital Markets: IPOs, M&A & Exchanges for the week of October 6, 2026 (podcasts published September 29 to October 6). Oura pulls a $2 billion IPO on the eve of pricing despite a 4x oversubscribed book, podcasts dissect Anthropic's leaked S-1 ($4.6 billion of 2025 revenue, $518 billion of compute commitments, a $2 trillion target), the 10-year hits 5.25%, ICE and OKX file a tokenized-stock venue, the CFTC probes Kalshi's volumes, AMD buys World Labs for $8.2 billion, and private-credit funds brace for mergers.

Capital Markets: IPOs, M&A & Exchanges

Week of October 6, 2026: Oura Pulls Its IPO as Anthropic's Prospectus Draws Fire and Yields Test Every Deal

Oura walks away, Anthropic's prospectus gets picked apart, and 5% bond yields test every deal on the calendar.

On Tuesday, September 29, Oura was hours away from pricing. The smart-ring maker had a book that was reportedly four times oversubscribed, revenue up 74%, and a Nasdaq ticker waiting.

Then it pulled the deal.

That same weekend, Reuters published details from Anthropic's confidential IPO filing. That filing is the S-1, the prospectus a company files with the SEC before selling shares to the public. It showed a company growing at a staggering pace, losing a lot of money, and asking for a valuation north of $2 trillion.

Put those two stories next to a 10-year Treasury yield sitting at roughly 5.25%, its highest in about 19 years, and you have the theme of the week. The IPO window has not closed. But buyers have stopped paying whatever sellers ask.

Bloomberg's equity capital markets reporter Anthony Hughes summed it up on Bloomberg Intelligence:

"The sellers of these assets are wanting more for them than the buyers are prepared to pay. And that's basically that disconnect is widened."

The lead: Oura blinks, and the IPO market shows its nerves

What happened

• Oura postponed its US IPO on the eve of pricing. The Rundown's Zaid Admani (Public.com) laid out the numbers: Oura planned to raise about $2 billion at roughly a $13.5 billion valuation. It has 5.7 million paying subscribers, and revenue for the first nine months of the year rose 74% to $1.21 billion. The company says it recently turned profitable. (The Rundown, "Oura Pulls it's IPO Last Minute, FICO's Monopoly Comes Under Fire", Sep 29, 2026)
• The price range valued it at up to $15.5 billion. On Lifers with Christina Farr, Stephanie Davis cited a target range of $40 to $44 a share, valuing Oura at $14 billion to $15.5 billion fully diluted. (Lifers with Christina Farr, "Rounds: Take the Money and R1", Oct 1, 2026)
• The demand was there. The right price was not. Morning Brew Daily reported the book was about four times oversubscribed, but Oura couldn't get the price it wanted. (Morning Brew Daily, "Oura Unexpectedly Delays IPO", Sep 30, 2026)

Why buyers balked

Three separate podcasts pointed at three separate problems.

1. Most of the money was going to existing shareholders, not the company. On The Rundown, Admani flagged that roughly three-quarters of the shares on offer were being sold by existing investors. They were cashing out, not raising new money for the business. Oura itself was expected to net only about $500 million:

"Almost all that money would have gone towards taxes related to employee stock options. So that means very little of the money being raised from this IPO was actually going towards the business."

He also noted that at $13.5 billion Oura would have been valued at more than eight times sales, roughly the multiple Apple trades at today. (The Rundown, Sep 29)

2. The valuation looked rich for what is still mostly a hardware business. Davis ran the numbers on Lifers. She was generous: same seasonality as last year, plus 75% growth again next year. That gets you to about $2.8 billion of 2027 revenue. Even then, the deal priced Oura at five to six times revenue.

• For comparison, Fitbit went public at about two times revenue. It peaked at five times before being bought for less than one times.
• Subscriptions are only about 20% of Oura's revenue. The other 80% is ring sales, which Davis called "a huge hole to fill every year from hardware sales."
• Oura's gross margin is about 50%. Health-tech peers trading at similar multiples, like Hinge, Doximity and Veeva, run gross margins in the mid-70s to mid-80s. (Gross margin is the share of each sales dollar left after the direct cost of making the product.)

(Lifers with Christina Farr, Oct 1)

3. The investors who matter most stayed home. Hughes told Bloomberg Intelligence the roadshow was "pretty slow going." Roadshows are the week or so of investor meetings before a deal prices. Long-only funds, the slow-moving mutual funds and pension managers that banks want as anchor buyers, "were hesitating," and "many of them passed." Hughes said the company was "probably looking at pricing below range if they really wanted to get this done," and chose to walk rather than cut the price. He also flagged concerns about the lockup, the period after an IPO during which insiders are barred from selling. (Bloomberg Intelligence, "Smart Ring Maker Oura Becomes Latest Company to Delay US IPO", Sep 29, 2026)

It's not just Oura

This is the part that matters for the rest of the calendar.

• Three postponements in three weeks. Hughes counted Oura, an insurance company the week before, and another deal before that: "So it's becoming a trend." (Bloomberg Intelligence, Sep 29)
• The list keeps growing. On The Information's TITV, co-executive editor Martin Pierce named the metals company Amira, Holtec Nuclear Energy, and the SoftBank-backed SB Energy as other recent delays. He was blunt: "I really don't think it's the companies. I think it is the overall market." Holtec's own delay notice cited the war in the Middle East and rising rates. Pierce noted the long bond "is hitting levels that hasn't hit since 2002." (The Information's TITV, "Inside Oura's IPO Postpone, Anthropic's AI Discount Cuts", Sep 29, 2026)
• The Fed is watching too. Minneapolis Fed President Neel Kashkari told Bloomberg Talks that the Fed reads deal activity as a signal of how tight money really is. If deals are getting done, oversubscribed, with markets wide open, that suggests policy may not be biting. He pointed to a shelved IPO like Oura's as a sign of tentativeness creeping into financial markets. In other words, the rate hike is starting to work. (Bloomberg Talks, "Minneapolis Fed President Neel Kashkari Talks Economy", Oct 1, 2026)

The counter-view: this is a pricing problem, not a broken market

• Oura doesn't need the money. On Squawk Pod, Andrew Ross Sorkin argued the Oura deal "was really a cash out" for early investors: "It does not appear that this is a company that's going to need that money. And they could, I believe that they could do an IPO in six months, a year, year and a half." (Squawk Pod, "Tech in DC: SEC Chair Atkins, Speaker Johnson & Senator Warren", Sep 29, 2026)
• Biotech bankers don't see a read-across. On Biotech Hangout, one host said he had spoken to a banker on the pulled tech deal, who thought "there was a deal to be had just at a slightly lower price." The banker's view was that "in the tech world, it's either, you know, go big or go home," so he's "not seeing a lot of read through" into biotech. The biotech IPO market is on pace for 30-plus US listings in 2026, with 23 already done. T-Rex Bio, Iambic Therapeutics and City Therapeutics are coming next. (Biotech Hangout, "Episode 198", Oct 2, 2026)
• The regulator is playing the long game. SEC Chair Paul Atkins told Squawk Pod that 583 companies have gone public in the last 18 months, up 75% from the comparable stretch at the end of the Biden administration, raising about $208 billion. "We're trying to make IPOs great again," he said, adding that "there will be ups and downs as market conditions change. That's the way it should be." (Squawk Pod, Sep 29)

Why it matters: A pulled deal at 4x oversubscribed tells you that headline demand is a vanity number. What counts is whether the long-term buyers will pay up, and right now, with 10-year yields above 5%, they won't. Expect more deals priced at or below the bottom of the range, more cut sizes, and more issuers choosing to wait.

Anthropic's prospectus: the week the bulls and bears read the same document

Reuters' leak of Anthropic's draft S-1 dominated the business podcasts after September 29. Almost everyone agreed on the numbers. They disagreed sharply on what they mean.

The numbers most shows agreed on

• 2025 revenue of about $4.6 billion, up roughly 12x, or more than 1,000%, from 2024. (Prof G Markets, "Steve Eisman: One Company Could Break The AI Boom", Oct 2, 2026)
• An operating loss of about $8 billion. The headline $42 billion net loss includes about $34 billion of non-cash accounting charges tied to revaluing financing instruments. Scott Galloway on Pivot and Rory O'Driscoll on 20VC both stressed this point. (Pivot, "AI's Rocky Road to Wall Street", Oct 2, 2026; 20VC, "Instinct Raises $1B at $10B Valuation | AMD Buys Fei-Fei Li's World Labs for $8.2B", Oct 1, 2026)
• Q2 2026 revenue of $11.5 billion, a sign of how much the business has grown since last year. (Generative AI 101, "Anthropic's IPO Leak & the $518 Billion Rent Bill", Oct 5, 2026)
• $518 billion of future compute commitments to cloud and chip suppliers. Generative AI 101 broke out $111 billion to Google and $110 billion to Amazon. (Generative AI 101, Oct 5)
• A target valuation above $2 trillion, with the listing expected after the November 3 midterm elections. (Times Tech, "Why is Anthropic asking people to fund AI doom?", Oct 1, 2026)
• About 80 pages of risk factors against roughly 48 pages describing the business. (Pivot, Oct 2)

The bear case, in their own words

Concentration is the red flag that surprised Scott Galloway. On Pivot, he said the thing that jumped out was not the losses:

"A company trying to obtain a $2 trillion market capitalization with this level of concentration of client base was the thing that jumped out to me."

He laid out the math: a valuation of "roughly 435 times last year['s] revenue," compute commitments "more than 110 years of its 2025... revenue," one quarter of revenue from two customers, and 47% of sales routed through Amazon and Google. Those big tech partners are "simultaneously suppliers, investors, and potential competitors." His summary of the risk section: "This may be the first IPO where the risk factor is not our business might fail, but our business model might succeed beyond our ability to control it." (Pivot, Oct 2)

Ed Zitron went further on Monetary Matters with Jack Farley. He focused on the fixed obligations:

"They're $518 billion. There's $413 billion of them that are non-cancellable. $252 billion in non-cancellable obligations across Microsoft, Google, and Amazon. $161.2 billion in Broadcom TPU lease obligations."

His concern is how those bills get paid. Even if Anthropic becomes profitable, he argued, the cash to cover $100 billion to $200 billion a year in spending "is going to come through debt at the single worst time in history to raise high yield." He also flagged that those $252 billion of commitments are already baked into analysts' forecasts for Google, Amazon and Microsoft. That creates what he called "counterparty risk here at scale": if Anthropic stumbles, the hit spreads to its suppliers. (Monetary Matters with Jack Farley, "Ed Zitron on Anthropic's IPO (S-1), AI Debt, and Counterparty Risk", Oct 1, 2026)

New Constructs put it in its "Danger Zone." David Trainor told Money Life with Chuck Jaffe that his firm titled its report "Anthropic is the most ridiculous IPO of 2026." The last time they used that title was for WeWork in 2019. His argument:

• A $2 trillion value implies future profits bigger than "five of the most profitable companies in the market today combined." Meanwhile competitors sell "virtually identical products at a 20th, a 50th of the price."
• Anthropic has already raised $120 billion privately. Compare that to what other companies raised before going public: Amazon $8 million, Google $25 million, Meta $2.3 billion, Uber $24 billion.
• His conclusion: "This is an exit strategy. Nothing more." He thinks Wall Street is pushing the deal partly because so much of the AI trade depends on Anthropic continuing to buy chips and compute.
• He also argued the leak was "selectively disclosed," which he believes would violate the SEC's fair disclosure rules.

(Money Life with Chuck Jaffe, "New Constructs' Trainer calls Anthropic is 'the most ridiculous IPO of 2026'", Oct 5, 2026)

Steve Eisman thinks the timing tells you something. On Prof G Markets, the investor famous for calling the 2008 crash offered what he called a "conspiracy theory":

"I think Anthropic might be going public now because the first half of the year looks really good because of token maxing and the lack of open weight models."

He expects the third quarter to show "something of a slowdown," and the fourth quarter "definitely." His bigger worry is the whole chain: "If you look at the hyperscalers, 70% of their AI revenue is from OpenAI and Anthropic... If there's a problem with those two companies, then I think the whole chain is in trouble." On the AI business more broadly: "None of these companies have any moats." A moat is a lasting competitive advantage that keeps rivals out. (Prof G Markets, Oct 2)

The more measured view

The 2025 numbers don't matter. Q3 does. On 20VC, Rory O'Driscoll dismissed the leak as containing almost nothing new: "4.5 billion in revenues. 8 in compute expense. 5 in other expense." The one interesting fact was that two customers made up 25% of revenue. His view: "The Q3 revenue number is 90% of the data required to make a decision on pricing on Anthropic." (20VC, Oct 1)

The deal might work. The stock afterwards might not. On the same episode, Jason Lemkin had a different worry: "I worry the IPO will be successful. They'll hit their number, whatever they ask for." But a month or two later, with the negative headlines piling up, "we may see a drift below the IPO price." O'Driscoll pushed back. The people who buy at the IPO are "a small number of highly compensated managers," not the public: "70% of Americans might think data centers suck, but they're not running Fidelity Growth." (20VC, Oct 1)

The SEC chair called the risk disclosures unusual. Atkins wouldn't discuss the filing directly, but told Squawk Pod: "This is an unusual type of disclosure. I haven't seen it yet. This is being handled by the staff." (Squawk Pod, Sep 29)

OpenAI stays private, and Galloway says that's the real story

Sam Altman told Bloomberg Talks at DevDay that OpenAI is in "a very steep growth period" but declined to confirm a reported $70 billion annual revenue run rate. OpenAI is targeting $30 billion of new funding at a $1.4 trillion valuation while pushing its IPO back, citing safety. (Bloomberg Talks, "OpenAI CEO Sam Altman Talks IPO Planning, DevDay", Sep 29, 2026)

Galloway's take on Pivot is worth reading in full:

"The private markets can now finance companies at a scale that lets them postpone the discipline of the public markets... OpenAI can get public company money without public company scrutiny."

(Pivot, Oct 2)

Why it matters: The Anthropic listing is the deal the whole IPO market is waiting on. Hughes said as much on Bloomberg Intelligence: "A lot of people are focused on Anthropic later in the year and a lot of the other stuff doesn't really measure up." If it prices well after the midterms, the window probably reopens for everyone. If it slips or prices far below $2 trillion, the Oura-style standoffs get worse.

Rates: the backdrop every banker is staring at

The bond market is now the single biggest variable for deals.

• The quarter in one line. Bob Doll of Crossmark said the 10-year yield rose from 4.38% to 5.26% in the third quarter, close to a full percentage point in three months. The Fed hiked by a quarter point in September, its first increase since July 2023. Long-dated Treasuries fell 10%, and the forward P/E on stocks, the price investors pay for each dollar of expected earnings, dropped by three points. His warning: "Decisive breakouts in the U.S. 10-year Treasury yield have been the catalyst for equity setbacks this decade. Yields are currently in the process of breaking out, thus our bias would be to de-risk if yields do not soon pause." (Doll's Deliberations, "Quarter-end Review: Bond Market Riot", Oct 5, 2026)
• The highs. The Rundown said the 10-year hit 5.25%, a 19-year high, and the 30-year hit 5.57%, its highest since 2004, with oil near $100 a barrel. (The Rundown, Sep 29) The TreppWire Podcast tracked the 10-year from 5% to 5.29% in the two weeks after September 16, with no new Fed action behind it. (The TreppWire Podcast, "425. The Payoff Problem", Oct 2, 2026)
• A softer jobs report gave a little relief. Squawk on the Street covered September payrolls of 29,000 against an 84,000 estimate, with unemployment rising to 4.2%. The odds of an October hike dropped to about 20% from 70%. (Squawk on the Street, "11AM Hour: September Jobs Below Expectations", Oct 2, 2026)
• Kashkari sees a big gap between the market and the Fed. He said the actual two-year yield is "around 488 or 490," while the yield implied by the Fed's own forecasts is "a little above four." He called this "a very, very large gap." He added that markets "will send a signal regardless of what the Fed is telling them," and noted that after the financial crisis, the markets turned out "more right than the Fed was." (Bloomberg Talks, Oct 1)
• Wellington thinks long rates have overshot. Fixed-income portfolio manager Brij Khurana told Alpha Exchange that the 10-year rate expected ten years from now is "close to around 6.2%," which "looks almost 100 basis points too high to me." A basis point is one-hundredth of a percentage point. (Alpha Exchange, "Brij Khurana, Fixed Income Portfolio Manager, Wellington", Oct 1, 2026)

Big Tech's borrowing binge is reshaping the bond market

On Unhedged, the FT's Emily Herbert shared Goldman Sachs figures. Investors have lent AI-related companies about $500 billion so far this year, with hyperscalers accounting for about $200 billion. Hyperscalers are the giant cloud builders like Amazon, Alphabet, Meta and Oracle. Goldman expects hyperscaler borrowing to reach "about a trillion dollars of debt" by around 2030.

Treasury Secretary Scott Bessent and Fed Chair Kevin Warsh have both suggested this corporate borrowing is pushing up Treasury yields. Rob Armstrong disagreed: the core buyers of Treasuries, such as central bank reserve managers and insurers, hold them no matter how many corporate bonds are on offer. (Unhedged, "Hyperscale my debt!", Oct 1, 2026)

Why it matters: Higher yields make future earnings less valuable today, and that hits high-growth IPO candidates hardest. Every new prospectus is now being judged against a 5% risk-free alternative.

Exchanges: the race to trade around the clock

NYSE's owner teams up with a crypto exchange

• ICE and OKX are launching a tokenized stock venue. Thinking Crypto reported that a joint venture between OKX and NYSE parent Intercontinental Exchange (ICE) has notified the SEC it plans to launch a venue for trading tokenized stocks. These are blockchain-based digital versions of regular shares. It will start with more than 60 NYSE-listed companies. Former New York Governor Andrew Cuomo, the venture's co-chair, said the goal is to "show how 24-7 on-chain markets can make trading and settlement more efficient, accessible, and global." (Thinking Crypto News & Interviews, "HUGE SHIFT! CFTC'S NEW CRYPTO REGULATION & OKX EXCHANGE NYSE TOKENIZED STOCKS!", Oct 6, 2026)
• The cash has to keep up with the trading. The same episode covered Kraken operator Payward partnering with Singapore Gulf Bank for round-the-clock dollar settlement. Payward's Mark Greenberg described the problem: "Settlement stops when the business day does while their markets do not." (Thinking Crypto, Oct 6)

Nasdaq and NYSE go to 23.5 hours on December 6

On Morgan Stanley's Thoughts on the Market, Michael Cyprys, who covers brokers, asset managers and exchanges, said Nasdaq and NYSE move to 23.5-hour trading on December 6, with a path to 24/7. He explained why tokenization matters here:

"If securities can trade 24-7 or derivatives, you may also need the cash leg of that transaction to keep pace. Right now, there are certain futures contracts that do trade over a weekend, but those positions do need to be pre-funded on Friday."

• Morgan Stanley's base case has tokenized real-world assets growing from about $40 billion today to about $2.3 trillion by 2030. Most of that comes from cash, Treasury and collateral uses, not private markets, which Cyprys sees as "a little bit further out."
• Global assets under management are expected to grow from about $160 trillion to about $250 trillion by 2030, roughly 9% a year. About three-quarters of that growth comes from rising markets, leaving only about 2.5% from new money. So the winners will be firms with the best distribution and products, not just exposure to the market.

(Thoughts on the Market, "How AI and Tokenization Could Reshape Wealth Management", Oct 1, 2026)

Bill Ackman is buying exchanges and data franchises cheaply

On Bloomberg Talks, Ackman said his new closed-end fund, Pershing Square USA, is fully invested after raising $5 billion in its IPO. He named two capital-markets holdings. S&P Global, which "you couldn't touch at something less than 30 times earnings," he bought "at 21 times." Intercontinental Exchange: "You could own it 24 times earnings today. You buy it at 17 times."

The fund has had a rough start. Its shares are down about 25%, and it trades at a big discount to the value of its holdings. A closed-end fund has a fixed number of shares, so its share price can drift away from what its portfolio is worth. Ackman says the fund's net asset value is down only about 3% since the IPO, and about 2.4 points of that is the cost of the IPO itself. "It's a marketing issue," he said. (Bloomberg Talks, "Pershing Square founder and CEO Bill Ackman Talks Howard Hughes Plans, IPOs and Anthropic", Sep 30, 2026)

Index news

• Moderna may join the Nasdaq 100. Biotech Hangout noted the stock was flat on the news, possibly because its weight in the XBI biotech ETF matters more to specialist investors. (Biotech Hangout, Oct 2)

Prediction markets: Kalshi's volume problem

The sharpest prediction-market story of the week came from Unchained's The Chopping Block.

• The numbers didn't add up. Kalshi's new perpetual futures product showed huge trading volume against tiny open interest. Open interest is the value of positions actually left open. One host put it plainly: "We have 3 million of open interest and 500 million of volume. Like there's nobody trading here except market makers."
• How it happened. As the hosts understood it, makers earned 0.3 basis points and takers paid 0.3 basis points, so trades netted to zero on fees. But "certain market makers were earning equity based on how much volume they were doing." That gave them a reason to trade back and forth with each other to pump up volume.
• The fallout. The Wall Street Journal reported the CFTC, the main US derivatives regulator, was investigating Kalshi's volume claims. Kalshi then ended the liquidity incentive program.
• The real question, as one host put it, is whether you're "scaling it in such a way to create a false impression of usage," especially if those figures are used in fundraising.

(Unchained, "The Chopping Block: Bitget's 387 Million Dollar Hack, Kalshi's Cooked Perps Volume, and Agentic Bank Runs", Oct 2, 2026)

Separately, Daily Crypto News reported the White House is considering new CFTC rules for prediction markets. (Daily Crypto News, "Oct 1: Bitcoin Holds $84K", Oct 1, 2026)

M&A: deals in AI, fintech, mining and private credit

AI and fintech

• AMD is buying World Labs for $8.2 billion. World Labs is Fei-Fei Li's AI company that builds models of the physical world. On TITV, Celesta Capital's Sriram Viswanathan framed it as part of a broader pattern: chipmakers are moving "up the stack" into AI models, while model builders want more control over their chips. (The Information's TITV, Sep 29; 20VC, Oct 1)
• Nubank is eyeing an $8 billion to $12 billion takeover of UK digital bank Monzo, per the headline of 20VC's weekly news roundup. (20VC, Oct 1)

Gold miners: "M&A season has begun"

On The KE Report, Dave Erfle walked through a cluster of gold-mining deals:

• Artemis Gold is taking over Vista Gold for its Mount Todd project, at about $37 per ounce of gold in the ground.
• Luca Mining is buying mines from Capstone Copper and Agnico Eagle.
• Elemental Royalty is buying a $290 million royalty package.
• A possible Northern Star and Gold Fields deal remains uncertain.

Erfle's read is that this signals the start of a new M&A season among miners. (The KE Report, "Dave Erfle - Gold & Gold Stocks: Beaver Creek Recap, Metals Technicals, Recent M&A", Sep 29, 2026)

Biopharma partnerships

Biotech Hangout's deal segment covered a new Regeneron and Sanofi collaboration, building on a 20-year partnership that produced the blockbuster Dupixent. It also covered two Summit and AstraZeneca deals and Novo Nordisk's new partnership with a Chinese biotech. (Biotech Hangout, Oct 2)

Private credit: the next consolidation wave

James Elbaor of Marlton LLC told Other People's Money with Max Wiethe that the private credit boom is over, "This is not a growth story any longer," and that mergers are coming:

• The old playbook is dead. Investors used to buy into a private fund at net asset value and wait for it to list at a premium. "That window is absolutely closed." He pointed to BlueRock, whose fund "immediately traded to a 38% discount" to net asset value when it listed.
• Some funds are winding down. He said Blue Owl's OBDC II has stopped offering quarterly redemptions and will instead return cash as it winds down the portfolio.
• Mergers at, or above, asset value. Mount Logan bought TURN at 110% of net asset value, and Source Capital made an unsolicited bid at 101% of NAV. Elbaor argued that big managers like Aberdeen will keep paying premiums, because "bigger is better" and more assets means more management fees. He expects "more mergers, tender offers... and consolidation within these private credit vehicles over the next five years."
• Redemptions are outrunning liquidity. In the same conversation, he said Blackstone's $82 billion BCRED fund received withdrawal requests for about 10% of shares but capped payouts at 5%.

(Other People's Money with Max Wiethe, "The Private Credit Boom is Over: Redemption Requests Exceed Liquidity | James Elbaor | Marlton LLC", Sep 27, 2026)

How bad are defaults? It depends who you ask.

• Hamilton Lane CEO Erik Hirsch told The Compound and Friends that private credit defaults are running around 2%, with no meaningful rise in bankruptcies. (The Compound and Friends, "Erik Hirsch, CEO of Hamilton Lane, on the Explosive Growth of Private Markets", Sep 28, 2026)
• Unf*cking The Republic cited a Fitch estimate of 6.3% and a PIMCO estimate as high as 19%. It argued that loans where interest is added to the balance instead of paid in cash, and quiet restructurings, are hiding the true picture. The episode also noted SoftBank's latest long-dated dollar bonds came with a 9.75% yield, the highest it has paid. (Unf*cking The Republic, "Private Credit: Just Because We're Not Talking About It Doesn't Mean It's Not a Problem.", Oct 5, 2026)

Regulation and crypto capital markets

• The SEC is down to two commissioners. On The Brink with Castle Island reported Commissioner Hester Peirce is retiring, leaving Chair Atkins and Commissioner Uyeda. The SEC has proposed crypto custody rules that would let state trust companies hold crypto for registered investment advisers and funds. It also clarified that token buybacks on functioning networks don't automatically make a token a security. (On The Brink with Castle Island, "Weekly Roundup 10/02/26", Oct 2, 2026)
• More rules are coming. Former CFTC Chairman Chris Giancarlo told CRYPTO 101 that Atkins' team is working on a "regulation crypto assets" proposal and an innovation exemption for tokenized asset marketplaces. (CRYPTO 101, "Ep. 757 Crypto Is Becoming the Architecture of Finance", Oct 5, 2026)
• The first 3x leveraged Bitcoin and Ether ETFs were approved. Thinking Crypto also reported BNY Mellon is in talks with Kraken's parent on a broad custody, trading and payments partnership. (Thinking Crypto News & Interviews, "MAJOR PUMP COMING FOR BITCOIN & ALTCOINS NEXT WEEK!?", Oct 4, 2026)
• Wall Street keeps moving onto crypto rails. Coinbase and Citi expanded a stablecoin payments partnership, and Goldman Sachs launched a $100 billion Treasury fund distributed through a crypto settlement network via SEC-registered broker tZERO. Stablecoins are digital tokens pegged to the dollar. (Thinking Crypto News & Interviews, "HUGE! COINBASE CITIBANK STABLECOINS, GOLDMAN SACHS $100 BILLION CRYPTO", Sep 29, 2026)
• Crypto ETF flows held up. Bitwise CEO Hunter Horsley told The Wolf Of All Streets his firm saw $1.8 billion of ETF inflows in the first half despite the downturn, and about $1 billion in Q3. (The Wolf Of All Streets, "How Billion-Dollar Funds Actually Invest In Crypto", Oct 4, 2026)

What to watch this week

• Whether Oura, Holtec or SB Energy refile with lower prices. A quick return at a reset valuation would show the window is open for the right price.
• The 10-year around 5.3%. TreppWire flagged 5.3%, 5.39% and 5.53% as the next levels to watch. (The TreppWire Podcast, Oct 2)
• Any public S-1 from Anthropic, especially third-quarter numbers, the figure 20VC called "90% of the data" that matters.
• Bank third-quarter earnings, which will show whether the slowdown in IPOs has dented underwriting fees.
• Kalshi and the CFTC, and whether the White House follows through on new prediction-market rules.