Posts

Essays, analysis, and insights on private equity, tokenization, and decentralized finance.

Where the Gate Sits

The Token Chose the Gate examined a tokenized credit fund whose token came with a private-placement gate. This piece looks at a different wrapper, one issued around a fund's token that anyone can hold, and asks what it can do, what it costs and when it is worth having. Tokenized money-market funds rely on allow-listing of wallets to keep peer-to-peer trading within the rules, but the authors of BIS Bulletin No 115 write that "such allow lists only constrain direct holding of TMMFs." An allow-list tells a fund who its holders are and enforces its eligibility rules, and its reach stops at direct holding. The Bulletin counts wrapping into stablecoins as one way investors can "bypass allow-lists in practice". Re-Plumbed put tokens like these on the side of tokenization where tokens can move, within the allow-list. The harder case is a token that passes the fund's own return and losses to whoever holds it, with no gate behind it at all.

Fixed Against Floating

Anthropic filed its S-1 on 26 October and an amendment on 2 November that adds the third quarter of 2026. The price range, share count and offering size are still blank, so there is no valuation to test. What the filing does give is nine months of results under the company's own name, enough to score claims made before those numbers were public. In August, Benzinga asked Tessera's founder, Chan Ahn, what Anthropic would need to justify a $2 trillion valuation, and reported his answer as roughly $725 billion of revenue by 2036. Against the amended filing, revenue has run ahead of the figures he used: $33.2 billion in the first nine months of 2026, against $2.5 billion a year earlier, with $17.0 billion in the third quarter alone. His full-year estimate of $50 billion to $55 billion would need a fourth quarter of roughly $16.8 billion to $21.8 billion, so it looks more likely to prove low than high. The piece then turns to margins, stock-based compensation and about $895.4 billion of commitments, and sets out where the filing counts against him and where it does not.

The Token Chose the Gate

Apollo Diversified Credit Fund, an interval fund registered with the SEC, sells Class A shares from $2,500 with a sales load of up to 5.75%, and Class I shares from $1,000,000. Its tokenized feeder, ACRED, run by Securitize, holds Class I shares, and RWA.xyz lists a $50,000 minimum limited to US accredited investors and non-US professional investors. Measured against the class it holds, the token lowered the bar. Measured against what a US buyer can already get, it raised one: an investor who can buy Class A from $2,500 cannot hold ACRED without being accredited. The gate follows from the feeder's legal route. ACRED's feeder is a private fund relying on Rule 506(c) and section 3(c)(1), and its latest Form D records $110.9 million sold to 26 investors. Franklin Templeton's tokenized money-market fund takes $20, which shows a registered wrapper is possible, though not that one would be as simple for private credit. The piece is written as of 28 October 2026.

Re-Plumbed

Set stablecoins aside and this month's largest tokenization numbers came from places no retail holder can reach. Broadridge's Distributed Ledger Repo averaged $359 billion a day in September, BNY completed a triparty repo workflow with Citadel Securities using a tokenized U.S. Treasury, and JPMorgan put Kinexys at over $4 trillion in cumulative transactions since inception. Those are activity figures. The other side of tokenization, where tokens can move between wallets, is measured as a balance: on RWA.xyz, tokenized US Treasuries that can move outside their issuing platform were worth $14.65 billion on 25 October. Even there, none of the five largest products is open to US retail buyers. RWA.xyz splits tokenized assets into "distributed" at $39.00 billion and "represented" at $389.25 billion. Broadridge's repo ledger accounts for $343.02 billion of the represented figure, or 88.1%, and RWA.xyz labels that series "Daily Repo Turnover." Most of the distance between the two totals is one platform's daily repo figure. The piece is written as of 26 October 2026.

Scoring Our Own Test

Tessera's piece Something That Moves argued in September that retail adoption of tokenized assets was arriving through convenience rather than rights-first design, and it ended with the evidence that would show it wrong. One of its two tests was that if RWA.xyz's holder count fell back as launchpad activity quieted, it was a trade rather than adoption. Three weeks of data give a first answer, and it is not the one we would like to claim. Launchpad activity has certainly quieted. DefiLlama's daily figures for StonkFun peaked at $63.0 million on 21 September, stood at $24.6 million when the test was set on 28 September, and reached $1.7 million on 20 October, about 93% below the day the test was set and 97% below the peak. By week, the seven days to 20 October came to $11.6 million against $148.9 million three weeks earlier. That is half the test met, and only for part of the period. The piece sets out what the holder count did alongside it, and what the result says about the test as much as about the argument.

Nine Said No

In a revised notice dated 8 October, a tokenized-stock venue that is not yet trading published a short list of companies that had told it no. OKXICE, a 50-50 joint venture between Intercontinental Exchange Holdings and OKC USA Holding, said it had received a Notice of Issuer Objection from Cerebras Systems, AppLovin, Applied Digital and Rocket Lab. A week later, in a notice dated 15 October, the list had nine names. These notices are the first public record we have found of companies using the veto in the SEC's order of 17 September, which lets a venue trade a company's stock in token form, including tokens made without the company's involvement, but only where the token carries the same rights as the ordinary share. For a token made by a third party the company must be told first, trading cannot begin for thirty days, and an objection inside that window stops it. All nine stocks are already offered offshore as tokens without shareholder rights. The veto reached the version that comes with those rights, on one venue. Over the offshore tokens, as we read the order, the companies have no say at all.

Spread, Not Shed

A CoreWeave co-founder told the Financial Times this week where the next round of AI-chip financing could come from. Brannin McBee, who leads CoreWeave's financing, said that winning investment-grade ratings for more of its chip-backed debt would let it borrow from insurers, whose capital "is unlimited". The FT's insurance correspondent had already supplied the other half, saying many large traditional insurers are at their limit for exposure to AI companies. Both can be true at once. Insurance money is already in this market, in investment-grade paper and below it, and the filings name the insurers: 58 purchasers of IREN's notes, more than half of them US life and annuity insurers, and insurers among the lenders on CoreWeave's GPU facilities. Where they are named, the amounts are redacted or not given. What is new is the scale being discussed and the ratings it would take. For scale, the NAIC estimates US insurers' exposure to all private credit at $1.2 trillion at the end of 2025, about 13% of cash and invested assets. It does not break out data centers or AI, and no official figure does.

The Deadline Was Thursday

Eleven amicus briefs are now on the docket of New Jersey's Supreme Court petition against Kalshi, and the cover of every one says it is in support of petitioners. Ten of them were entered between 2 and 8 October. Eleven to none looks like a verdict, and part of it is a calendar. Under the Court's Rule 37.2, a brief supporting a petitioner at this stage must be filed within thirty days of the case being docketed, and that time will not be extended. The petition was docketed on 8 September, which made the deadline Thursday 8 October. The same rule sets a different clock for the other side: a brief supporting a respondent follows the time allowed for the brief in opposition, and Kalshi's response is currently due on 9 November. The briefs come from gaming regulators, tribes, states, opponents of gambling expansion, the National Football League, the gaming industry, and two former officials, among them a former CFTC chairman. Anyone who wants to stand beside Kalshi files on Kalshi's timetable rather than New Jersey's.

In Exchange for the Guarantees

Tessera's piece of 2 September held back on supplier-side residual value guarantees, saying that whether any given arrangement qualified as circular "depends on documents we have not read". The filings have now been read, and they had been on file all along. AMD disclosed a $4.1 billion guarantee of a partner's data center lease obligations in February, Nvidia disclosed guarantees of partners' facility lease obligations "in exchange for warrants" the same month, Broadcom's backstop appeared on 9 June, and Nvidia's largest guarantee in an 8-K on 17 August. Every one was public before the 2 September piece said otherwise. Checking those sources against Tessera's August work turned up errors in three pieces, corrected here and in a note on each page. They cover the date at which anyone first had to price AI hardware, four telecom figures from the Nortel and Lucent era that were wrong or unsupported, and the shape of AMD's investment in Anthropic. One argument is weaker as a result. Staging was offered as what separates today's commitments from the telecom era, and Lucent's were largely undrawn too. Revenue separates them more.

Who Was Selling

Renaissance Capital counts three sizable US IPOs postponed before pricing in the third quarter, all within a few weeks of each other in September: Oura, Holtec Nuclear and Bamboo Insurance. Postponements are usually read as a verdict on the market. The more useful thing to look at is who was selling. Holtec offered 50 million shares at $15 to $18, all of them sold by the company and none by existing holders, and asked the SEC on 25 September to withdraw its registration. Bamboo offered 35 million at $18 to $20, none sold by the company and all of them by existing holders. Oura offered 50 million at $40 to $44, of which 13.5 million were to come from the company and 36.5 million from existing holders. All three were profitable on the figures in their filings. As of 7 October, neither Bamboo nor Oura had filed to withdraw. The seller column is the part of these deals that gets read last and says the most.