Posts

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.

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.

On 28 September Reuters reported that two customers accounted for nearly a quarter of Anthropic's revenue in 2025. The reporting that followed points at something larger. Anthropic's filing is still confidential, and a search of SEC EDGAR on 5 October found no public registration statement, but Reuters, which has seen it, reports the share of revenue sold through Amazon and Google rising from 11% in 2023 to 32% in 2024 and 47% in 2025, about $2.16 billion, through their cloud marketplaces. The reading here is that the larger concentration is the channel rather than the customer list. Nearly half of last year's revenue arrived through two marketplaces, and the companies that own them are, by Reuters' account of the filing, also Anthropic's investors, its compute suppliers, its competitors and its customers. Reuters still does not name the two customers who each generated 12% of revenue in 2025, and this piece does not guess. If either of them is one of those same two companies, the two concentrations fall on the same names and add up to more than either figure shows on its own.

On 25 September the Sixth Circuit became the second federal appeals court to rule against Kalshi, and the reasoning matters more than the result. Having found that Kalshi had not shown its sports contracts were swaps, the court added that Kalshi would still be unlikely to succeed if they were. The opinion, by Judge Julia Smith Gibbons for a three-judge panel with no separate opinion, decided the Ohio and Tennessee appeals together. It upheld Ohio's refusal of a preliminary injunction, vacated the injunction Kalshi had won in Tennessee, and sent both cases back. This is a preliminary ruling that binds one circuit, and the full court could still be asked to rehear it. The reading here is that the second holding changes what the Commodity Futures Trading Commission can do for Kalshi's sports business. The Commission keeps the power to refuse. Within the Sixth Circuit, and against general gambling laws, approval is now an argument it has to win in court rather than a power the statute hands it. A national answer has to come from the Supreme Court or from Congress.

RWA.xyz counted 4,010,763 addresses holding tokenized stocks on 28 September, up 67% in thirty days. The industry took two routes to get there. One camp put the registered share itself on-chain, with rights verified, holders checked in advance and transfers restricted to approved wallets; the other built tokens that give economic exposure and that eligible holders can keep in any wallet and move freely. The holder counts on the two sides are not close. bStocks reports 1,402,291 and xStocks 666,831, both classified as transferable. Superstate Opening Bell, which puts the registered shares themselves on-chain under token-level permissioning and KYC'd wallets, reports 84. The caveats are real: these are addresses rather than people, Opening Bell lists only three assets, and two large platforms fit neatly on neither side. The gap that remains is read here as a preference for tokens that are easy to hold and move. Tessera sits in the second camp and has a commercial interest in that reading being right. The data counts addresses; the preference is an interpretation of them.

On 27 August, Databricks told the Securities and Exchange Commission that it had sold $4,999,997,255 of stock to 136 investors. The filing reports $5,000,000,000 offered and $2,745 remaining, and all three figures are quoted as filed. It is one document, it is public, anyone can read it, and it does not contain the number everybody wants. A Form D is the notice a company files when it sells securities without registering them, due within fifteen days of the first sale. It collects the issuer's details, the exemption claimed, the date of first sale, the amounts offered, sold and remaining, and the number of investors who have already invested. It does not collect the price, the valuation, or the identity of a single buyer. That is not an oversight and it is not evasion. The form is a notice that an exempt offering has happened, filed so that the Commission knows it happened. This piece reads what that record does and does not establish.

On the afternoon of 15 September the Senate declined to begin debating the CLARITY Act. The result was on the wire within minutes of the vote, and the result is the least interesting thing in the record. What is interesting sits in the roll call, and the roll call only means something next to a second one taken the same afternoon and a third taken fourteen months earlier. The vote was cloture on the motion to proceed, which decides whether the Senate may begin debating a bill. It is not passage and it enacts nothing. A bill that fails it has not been voted down on its merits. It has failed to reach the point at which its merits get debated. The threshold is three-fifths of senators sworn, which is sixty. The 7 September piece argued that the question of whether Kalshi's sports contracts are federally regulated derivatives or unlicensed gambling had three live routes to an answer, judicial, legislative and administrative. This is what the legislative test produced, and it produced more than a number.

For years the standard defense of private markets ran like this: illiquidity is a feature rather than a bug, lock-ups force discipline, and the absence of a daily price is what lets managers think long-term. If an investor cannot sell for ten years, that is the price of admission to the best returns in finance. It was always a convenient story, and the people telling it were the ones who benefited from investors being unable to leave. This year the people who built that system started dismantling it. Morgan Stanley, with roughly $8 trillion under management, is building blockchain settlement infrastructure in-house rather than renting it. At an industry summit in March, its head of digital asset strategy said the firm needs to build this internally because it cannot simply rent the technology. The bank is preparing an institutional digital wallet for tokenized assets, support for tokenized equities on its own trading venue and, most tellingly, it is advancing tokenized private equity secondary markets. That last item deserves a second reading. A bulge-bracket bank is building secondary markets for private equity. The entire point of a secondary market is to let people sell something before it matures. That is liquidity, the thing private markets spent a generation insisting investors did not need. Morgan Stanley is not alone, and none of this is new in principle. Hamilton Lane and KKR have offered tokenized feeder funds through Securitize for several years, cutting minimums on institutional-only funds from millions to tens of thousands. What changed in 2026 is that the work stopped being a fintech experiment at the edge and became core infrastructure strategy at the center of Wall Street. When the incumbents start building the exit that investors were told they did not need, the original argument is finished.

Putting an asset on-chain is no longer the hard problem. Anyone can mint a token. The harder problem, and the one the rest of the industry is now racing to solve, is telling the chain what that token is actually worth. The distinction gets lost in most tokenization coverage, and it is worth being precise about, because it is also the problem Tessera is working on. There are two separate data problems for any tokenized real-world asset, and they are not the same problem. The first is proving the asset exists. Does the token have something real behind it, and are the reserves backing it actually there? This is what Chainlink Proof of Reserve does. Tessera's Proof of Reserve feed is live on Solana mainnet, verifying on-chain that the reserves backing T-Tokens exist. That is the integrity layer, and it answers the question of whether the thing is real. The second problem is pricing the asset. What is it worth right now? What net asset value does a lending protocol use to compute a loan-to-value ratio? What number does a perpetual futures exchange reference to mark a position? This is a net asset value and price-feed problem, and it is a fundamentally different piece of infrastructure from Proof of Reserve. Proof of Reserve tells you the backing is there. It says nothing about what the asset is worth. Those are two different oracles answering two different questions, and conflating them is one of the most common mistakes in this space. For most of tokenization's history the industry has concentrated on the first problem, meaning issuance and backing. The last few weeks have made it clear that the frontier has moved to the second. In May, DTCC, the backbone of US securities settlement, said it would use Chainlink for a tokenized collateral platform that automates pricing, valuation and settlement, building on an earlier net asset value pilot run with JPMorgan and BNY Mellon. That is the plumbing of traditional finance treating on-chain valuation as a problem worth solving at scale. It is already in production elsewhere. VanEck and Securitize's tokenized Treasury fund became one of the first assets to qualify as collateral on a major decentralized finance lending market, using a verified net asset value feed to price it. A tokenized asset cannot be used as collateral until a protocol can trust what it is worth. The feed is the precondition: no feed, no collateral. That is the whole story in one line. The valuation layer is what turns a token from something a holder can keep into something a holder can use.

Two things happened in capital markets this week that look like routine news taken separately. Taken together they are the same story told from opposite ends. On 14 April, Deutsche Börse, operator of the Frankfurt Stock Exchange and one of the oldest and most systemically important financial market infrastructures in the world, invested $200 million in the crypto exchange Kraken. The stated purpose was to build shared infrastructure spanning trading, custody, settlement, collateral management and tokenized assets. Not a hedge, and not an experiment. A strategic commitment to operating across both worlds at once. On the same day, the US Senate moved materially closer to passing the CLARITY Act, the most significant piece of crypto market structure legislation so far, after a compromise on the stablecoin yield provisions that had blocked it for nearly a year. A Banking Committee markup is expected in the final weeks of April and a Senate floor vote in May. Senator Lummis, the bill's most prominent champion, has described this as the last realistic window before at least 2030. One institution is building the infrastructure without waiting for the rules. The other is finalizing rules for infrastructure that is already being built. That is not a contradiction. It is what a structural shift looks like from the inside. Understanding why this week matters requires going back further than 2026. The regulatory architecture of capital markets, meaning mandatory disclosure, accredited investor restrictions, anti-money-laundering obligations and the public and private market distinction, was not designed in the abstract. It was designed for a world in which every financial transaction required intermediaries, and in which those intermediaries introduced counterparty credit risk at every step. Before electronic clearing, share transactions moved through physical certificates, multiple brokers, clearinghouses and correspondent banks. Each link was a point of counterparty exposure, and the longer the settlement chain the greater the systemic risk. Public market regulation, meaning listing standards, audited accounts and mandatory disclosure, developed specifically to manage the risks created when thousands of anonymous investors transact through intermediaries they cannot individually assess. The private market carve-out follows directly from that problem. The accredited investor framework says that if access is restricted to those sophisticated and well capitalized enough to absorb losses and conduct their own diligence, the disclosure burden can be relaxed. The access restriction is itself the risk management tool. It exists not because ordinary investors deserve returns less, but because the settlement infrastructure assumed they needed protection from intermediaries they could not evaluate. Know-your-customer and anti-money-laundering rules followed the same logic. The global framework placed compliance obligations on financial intermediaries, meaning banks, brokers and exchanges, because those were the natural chokepoints in the flow of capital. Regulating every transacting party is impractical. Regulating the infrastructure they transact through is not. All of that was rational, and all of it was designed for a world of intermediated, paper-based settlement with significant information gaps between counterparties. That is not the world now being built in.