---
title: "Where RWA lives and where it trades"
author: Mariusz Szyma
date: 2026-09-15
lang: en
canonical: https://szyma.co/en/blog/rwa/chains/
series: "RWA: Past, Present and Future"
series_part: 7/11
series_url: https://szyma.co/en/blog/rwa/
data_as_of: 4–10 Sep 2026
---

# Where RWA lives and where it trades

Four sources put Ethereum's share of the RWA market at 43%, 45.8%, 47.9% and over 56%, all four citing the same data provider, and none of them is wrong. Each divides by a different base, so the base decides the number and the percentage on its own decides nothing. This chapter separates the two rankings that the question "which chain wins" runs together, value parked and value traded, and then prices what compliance at the protocol level costs the network that offers it.

## What you'll learn

- Name the base behind any per-chain share before quoting the percentage
- Separate the ranking by value held from the ranking by value moved, and say which network leads each
- Compute a turnover ratio from a value and a volume read on the same day from one source
- Price protocol-level compliance in gas and in composability given up
- Tell a deployed privacy feature from an announced one

## Four shares for one chain, and each divides by a different base

Every per-chain share divides by one of two categories or by a slice of one. **Distributed** value is the part that can leave the issuing platform and move between wallets, **represented** value is the part that cannot, and [chapter 1](/en/blog/rwa/what-is-an-rwa/) sets out why the two are never added together.

| Source | Date | Ethereum's share | Base it divides by |
|---|---|---|---|
| Token Terminal 🟡 | ~Aug 2026 | 43% | tokenized US Treasuries only, 15.2 bn USD base, Ethereum 6.6 bn USD |
| Coinpaprika via RWA.xyz 🟡 | 3 Aug 2026 | 45.8% | distributed, 37.35 bn USD across 38 networks |
| Stobox via RWA.xyz 🟡 | 10 Jul 2026 | 47.9% | distributed excluding stablecoins, 33.5 bn USD base |
| MetaMask via RWA.xyz 🔴 | Apr 2026 | over 56% | "all tokenized asset value", undefined in the source |

Four readings of one network, each dividing by a different base. All four cite RWA.xyz.

The spread is not a mistake anyone made. Token Terminal asks what share of tokenized US Treasuries sits on Ethereum, Stobox asks what share of distributed value excluding stablecoins sits there, and MetaMask asks a question its own guide never defines. Three questions produce three answers, and the fourth is unusable because its base is unstated.

The same problem survives inside a single provider. On 3 September 2026 RWA.xyz showed represented value of 435.93 bn USD on its networks page and 377.74 bn USD on its platforms page, a gap of 58.19 bn USD or 15.4%, and the two pages reported holder counts 68,647 apart on that same day. The cause is structural rather than an error in either number: aggregating across 39 networks and aggregating across 207 platforms are two different queries against one database, and a product present on several networks lands differently in each.

> **Watch out:** Per-chain percentages carry a second bias on top of the base. BUIDL and BENJI are each deployed on eight or nine networks depending on which source is read, and public breakdowns either assign a whole product to one network or split it by a methodology nobody has published. Percentages built this way are usable for orders of magnitude and not for a ranking to the percentage point.

## Ethereum's lead is a sunk cost, not a current choice

The lead in value parked is real. RWA.xyz's figures relayed by Coinpaprika on 3 August 2026 put 17.12 bn USD on Ethereum, 45.8% of a 37.35 bn USD distributed base spread over 38 networks. The reason Coinpaprika gives for it has nothing to do with the technology: BlackRock, Franklin Templeton and Ondo launched their first products there before expanding anywhere else, and multi-chain deployment accelerated once issuers found that a second or third deployment costs far less than the first.

An issuer who picked Ethereum in 2024 does not move in 2026. It adds networks, and someone else often pays for the addition. BUIDL's management fee depends on where the share class sits, at 50 bps on Ethereum, Arbitrum and Optimism against 20 bps on Aptos, Avalanche and Polygon, and BlackRock does not absorb that difference. Aptos Foundation, Avalanche (BVI), Inc. and Polygon Labs BD Investments (Cayman) Ltd. agreed to pay BlackRock a quarterly fee instead, on CoinDesk's reporting of 13 November 2024. A foundation buying a BlackRock product's presence on its own network is a distribution decision, and [chapter 4](/en/blog/rwa/the-stack/) follows the rest of the fee stack.

Growth and level therefore point at different networks.

| Network | Net RWA inflow, six months to 26 Jul 2026 |
|---|---|
| BNB Chain | +3 bn USD |
| XRPL | +2.6 bn USD |
| Solana | +1.6 bn USD |
| Avalanche | +972 mn USD |
| Ethereum | +424 mn USD |

RWA.xyz via crypto.news, 26 July 2026. The largest network by value grew slowest of the five and stayed the largest, which is what a sunk cost looks like in a table.

Adding a network costs something the announcements rarely mention. **A permissioned token** is one that only addresses on an issuer-controlled allow-list may hold, which is how BUIDL has worked since Securitize began keeping its register on 20 March 2024. The allow-list state has to follow the token across networks, or a transfer agent on the destination network has to issue a parallel claim against the same assets.

> **Example:** A BENJI holder moving a position between chains coordinates with Franklin Templeton's transfer agent, which burns the tokens on the source network and mints the equivalent on the destination once both addresses have cleared the allow-list. OUSG crosses between Ethereum, Mantle and Polygon through an allow-list-aware bridge rather than a permissionless one. The sources describing both mechanisms are product documentation rather than filings.

## Ethereum is the warehouse, Solana is the market

The two metrics came apart in 2026 and stayed apart. Pine Analytics, reporting on 17 June 2026, measured Solana's share of spot trading in tokenized stocks during May 2026 at about 97%, roughly 869 mn USD against roughly 24 mn USD on every other network combined. Genfinity put the share above 96% on 6 July 2026 and counted 54 consecutive weeks in which Solana traded more tokenized stock than all other networks together.

People followed the volume. The RWA holder count on Solana passed Ethereum's for the first time in mid-2026 on Stobox's reading of RWA.xyz, driven by tokenized stocks, and that is the first time any network has led Ethereum on a measure that counts people rather than dollars.

What Ethereum still holds alone is the other job. **Collateral** is an asset pledged to secure a loan, which the lender may take if the loan goes unpaid, and a tokenized fund becomes collateral only where a lending market already exists to accept it. Aave Horizon, Morpho, Pendle and Sky are all on Ethereum; Aave Horizon alone had 163.5 mn USD borrowed against RWA collateral in July 2026, and [chapter 8](/en/blog/rwa/rwa-in-defi/) measures that side of the market instead of asserting it. Value parked measures issuer confidence. Volume measures market confidence. In 2026 the two point at two different networks.

| Ranking | Leader | Runner-up | Base and date |
|---|---|---|---|
| Distributed value, all RWA | Ethereum, 17.12 bn USD (45.8%) | BNB Chain, 5.75 bn USD (15.4%) | 37.35 bn USD distributed, 38 networks, 3 Aug 2026 |
| Tokenized US Treasuries | Ethereum, ~43% | BNB Chain, ~31.5% | 16.23 bn USD category, ~18 networks, 15 Aug 2026 |
| Tokenized-stock spot volume | Solana, ~97% | all other networks, ~3% | ~893 mn USD traded, May 2026 |
| RWA holders | Solana | Ethereum | count of holding addresses, mid-2026 |
| Non-US government debt | Stellar, ~490 mn USD | not published per network | tokenized non-US sovereign debt, 20 Aug 2026 |

Five rankings of the same market, each with its own base. The second row is the one nobody quotes: in tokenized Treasuries the runner-up is neither Solana nor Stellar.

The aggregate flatters the median on Solana, and the same report that produced the 97% says so. Liquidity concentrates in a few names, TSLAx, NVDAx and CRCLx, where Pine Analytics recommends a swap slippage tolerance of 0.1% to 0.5% on Raydium and Jupiter, while the long tail of tickers has thin books and wide spreads.

## Turnover, not value parked, separates a market from a register

**Turnover ratio** is transfer volume over a period divided by the value held, which says how many times the average token changed hands. On XRP Ledger in August 2026 the RWA in the represented category was worth 4.06 bn USD and moved 18.15 mn USD in thirty days, about 0.45% a month, while stablecoins on the same ledger were worth 901.4 mn USD and moved 3.89 bn USD, about 431% a month. The technology, the fees and the finality are identical on both sides of that comparison.

What separates them is who may receive the token. Any address can receive a stablecoin; a tokenized fund goes only to an address on a transfer agent's allow-list, and no amount of throughput changes that.

XRPL is also the cleanest demonstration of what a headline hides. The spring of 2026 produced "XRP Ledger passes 3 bn USD in tokenized RWA", and one token supplies most of it.

| XRPL reading | Value | Date |
|---|---|---|
| Represented | 4.06 bn USD | ~5 Aug 2026 |
| Distributed including stablecoins | 1.319 bn USD | 24 Jul 2026 |
| Distributed excluding stablecoins | 323.18 mn USD | 24 Jul 2026 |
| JMWH, energy, represented, non-US investors only | 2.229 bn USD | 24 Jul 2026 |
| RWA holders on the whole network | 199 | ~5 Aug 2026 |

TheCryptoBasic via RWA.xyz, 24 July 2026, with a direct RWA.xyz reading of about 5 August 2026 that puts distributed at 368.82 mn USD. JMWH is about 55% of the network's RWA value, and the market with real participation is 323.18 mn USD held by 199 addresses.

The same arithmetic across the market gives the spread.

| Segment and network | Value | 30-day volume | Turnover a month | Date |
|---|---|---|---|---|
| Stablecoins, BNB Chain | 14.14 bn USD | 216.42 bn USD | ~1,531% | Apr 2026 |
| Tokenized stocks, all networks | 2.54 bn USD | 29.5 bn USD | ~1,161% | Aug 2026 |
| Stablecoins, XRPL | 901.4 mn USD | 3.89 bn USD | ~431% | Aug 2026 |
| RWA, BNB Chain | 3.73 bn USD | 1.46 bn USD | ~39% | Apr 2026 |
| RWA, XRPL, represented | 4.06 bn USD | 18.15 mn USD | ~0.45% | Aug 2026 |

Value and volume pairs published by RWA.xyz at the same reading, with the ratio computed for this series. Three orders of magnitude separate the top row from the bottom, and the two extremes sit on the same two networks.

BNB Chain's own page explains where it lands on that scale. The reading dated 17 April 2026 showed 3.73 bn USD distributed against 0.00 USD represented, so every RWA dollar on that network is technically transferable and no institution has used it as a closed internal ledger. Its turnover of about 39% a month is respectable next to institutional funds and nowhere near tokenized stocks.

> **My take:** The turnover ratios above are my own arithmetic on pairs published by one source at one date, not a metric anyone publishes. Ethereum, Solana and Stellar are missing because no public reading pairs a 30-day RWA transfer volume with a value for them, which limits what the table can prove. The ratio also counts mints and redemptions as transfers, so it overstates secondary trading rather than understating it.

![Value held and thirty-day transfer volume for the same five segments, which rank in opposite orders: the two largest holdings move the least, on RWA.xyz readings of Apr to Aug 2026.](https://szyma.co/blog/img/chains-value-vs-turnover.24e6763a.svg)

## What compliance at the protocol level costs

Networks with the strongest compliance functions written into the protocol have the weakest composability, and networks with the best composability implement compliance inside contracts, which costs more and is never complete. Nobody has both. That trade-off is why a category of networks designed for RWA from scratch appeared at all, and why the hybrid arrangement its proponents call AllFi, permissioned issuance platforms connecting selectively to public liquidity venues, needed a name.

### The gas bill

The price shows up in the simplest operation a token has. A median `transfer` of USDC on Ethereum mainnet used 45,160 gas units on 4 September 2026, while the same call cost 91,335 units for OUSG and 183,672 for BUIDL, or 2.02× and 4.07× the plain transfer.

The reason sits in the call graph rather than in the chain. Every BUIDL transfer calls `validateTransfer` on a compliance service contract, which queries the investor registry, checks investor-count limits, accreditation status and lock-ups, resolves the wallet-to-investor mapping, then updates investor balances and the wallet list. A dozen or so calls between contracts replace one storage write, and [chapter 5](/en/blog/rwa/who-holds-the-keys/) reads those powers product by product. The measurement covers one network, and the cost of the same transfer on Stellar or Aptos has not been measured here.

![A median transfer of BUIDL cost 4.07 times a median transfer of USDC in gas units on Ethereum mainnet, measured 4 Sep 2026.](https://szyma.co/blog/img/compliance-gas-multiplier.5e6fa617.svg)

### The composability bill

**Composability**, the property that lets any protocol on the same network use a token as a building block in its own construction, is what the permissioned end of this market gives up. [Chapter 5](/en/blog/rwa/who-holds-the-keys/) measures it contract by contract; at network level it decides whether a token has one venue or every venue. The largest tokenized asset in the world is the clearest case, and it is not a fund.

**HELOC** stands for home equity line of credit, a revolving credit line secured on the equity in a house: with a house worth 500,000 USD carrying a 200,000 USD mortgage, the bank opens a line against part of the remaining 300,000 USD. Figure Technologies tokenized a portfolio of them as FIGR_HELOC, which stood at about 20.1 bn USD on 7 July 2026, more than every tokenized US Treasury combined at 15.16 bn USD in the same reading. It lives on Provenance, a network Figure built for that one use, whose entire TVL of roughly 1.2 to 1.3 bn USD belongs to Figure Markets.

Brian Huang, co-founder of Glider, put the objection to The Defiant on 11 February 2026: assets are no more useful on-chain than off-chain until they have composability, and Provenance has none. Figure's answer is a saving on a credit product rather than a market, and it comes from a Provenance representative quoted by Markets Media in January 2025 rather than from an audit, which is why the numbers sit in the table below with their mark and not in this paragraph.

| Figure and Provenance | Value | Source | Quality |
|---|---|---|---|
| Saving per HELOC from digitization | ~125 bps | Markets Media citing Provenance, Jan 2025, a representative's declaration | 🔴 |
| US HELOC originators using the network | ~20% | same | 🔴 |
| Tokenized loans issued a month | ~650 mn USD | same | 🔴 |
| Provenance TVL, all of it Figure Markets | ~1.2 to 1.3 bn USD | The Defiant, 11 Feb 2026; DefiLlama, May 2026 | 🟡 / 🟢 |
| FIGR_HELOC market capitalization | 22,445,444,187 USD | CoinGecko, 2026 | 🟢 |

The market capitalization tracks the unpaid principal balance of the whole loan portfolio while the TVL counts what protocols on the network actually hold, so the two figures measure different things rather than contradicting each other.

Two tokenizations are running at once, and one ranking by assets covers neither properly. The first builds a market: Solana, tokenized stocks, turnover in the hundreds of percent a month. The second lowers the unit cost of an existing process: Provenance, home equity credit, zero composability, and the only cost saving in this market anyone has claimed from production.

![Protocol-level compliance and composability trade off across nine networks, and the quadrant holding both is empty, on the author's assessment of readings to Sep 2026.](https://szyma.co/blog/img/compliance-vs-composability.49c7e139.svg)

### Where privacy is deployed and where it is announced

Three levels of certainty apply to privacy in RWA compliance, and most of what circulates sits at the third.

**Chainlink's Automated Compliance Engine**, ACE, is the first of two things that are deployed and confirmed. It launched on 30 June 2025 with Apex Group, GLEIF and the ERC3643 Association, was adopted by Aave Horizon as its compliance layer, and had its provider network extended to more than twenty entities in an announcement of 3 November 2025. It hides nothing cryptographically: it answers whether a condition is met, using a verifiable identifier for a legal entity. Its own status disagrees by source date, described as an early access phase in the June 2025 announcement and as Private Beta in the documentation read on 10 September 2026. The second confirmed item is identity without personal data, where ERC-3643's ONCHAINID stores keys and signed claims, which are references and hashes rather than a passport scan. No production ERC-3643 contract was read for this series, so the scale of its use is not established.

Two more are documented and unproven at scale. Canton keeps no global balance in public state, because a position is a Daml contract visible only to its parties, which is privacy achieved without any proof system. **A zero-knowledge proof** lets one party show that a statement is true without revealing the data behind it, and Solana's Token-2022 program has a Confidential Transfers extension built on one, which hides transfer amounts but not the parties. A read of the BUIDL mint on Solana on 5 September 2026 listed six extensions in use, and Confidential Transfers was not among them.

The rest is architecture. Ondo Chain as an institutional L1 with permissioned validators staking RWA, the hardware-enclave design Ondo announced in July 2026, and the XRP Ledger's reported addition of zero-knowledge proofs for private bank transactions are all announcements whose production state no source establishes. What was not found is the thing the phrase implies: no deployed, chain-readable compliance based on zero-knowledge proofs appears in any of the six largest tokenized funds.

![Privacy and identity features in RWA compliance sorted by how far each can be verified, from confirmed deployments to announcements, on sources read to 10 Sep 2026.](https://szyma.co/blog/img/zk-deployed-vs-announced.dd3c2ff3.svg)

## A network's RWA growth does not lift its own token

The investment thesis attached to this narrative is that a network's token rises with the RWA on it. The readings do not support it.

| Network | RWA on the network | Its token | Date |
|---|---|---|---|
| Stellar | 3.996 bn USD, up ~360% from 868.8 mn USD at end-2025 | XLM ~11% lower year to date, near 0.18 USD | 29 Aug 2026 |
| Solana | quadrupled over the first half of 2026 | SOL ~75.33 USD, ~74% below its Jan 2025 peak | 16 Aug 2026 |
| Canton | DTCC authorization and JPMorgan's Kinexys rollout | CC down 12.4% in two weeks to 0.1415 USD, 8 to 17 mn USD traded a day | Jun 2026 |

Coinpaprika, CoinGecko and Bloodstone Capital, on the dates shown. Three networks, three directions of RWA growth, three tokens that went the other way.

The mechanism is easier to see than to accept. Activity in RWA on a network need not produce fees that reach the holders of that network's token, and an institution can use a network while holding as little of its token as it can.

Canton is where the gap between the institution and the token is widest, and where the numbers are hardest to use. The network is an L1 with configurable privacy, launched in 2023 by a consortium including BNP Paribas, Goldman Sachs, Deutsche Börse and Microsoft, and DTCC authorized it together with Digital Asset for three years to tokenize a subset of US Treasuries held in DTC custody, with the rollout targeted for the second half of 2026. Those are checkable counterparties. The aggregates are a different matter.

| Canton figure | Value | Source | Quality |
|---|---|---|---|
| RWA processed, Q4 2025 | 6 tn USD | Blockworks, 1 Jan 2026, reporting the network's own figure (company claim) | 🟡 |
| Daily US Treasury activity | ~350 bn USD | same | 🟡 |
| Represented value, against near-zero distributed | ~371.3 bn USD | CoinPaprika relaying RWA.xyz, Aug 2026; a research pass of 10 Sep 2026 could not locate it | 🔴 |
| Share of the whole represented category on this one network | 85% to 98% | derived from the same unlocated reading | 🔴 |

Notional flow and stock are different units, and the category error of standing 6 tn USD of processed value next to 38.63 bn USD of distributed value, read on 3 September 2026, is the most common mistake made with this network. On the unverified reading in the table, most of the represented category worldwide sits here, which would make one network a single point of methodological failure for every headline about the size of the tokenization market. This series quotes no confirmed value for that share, and [chapter 9](/en/blog/rwa/whos-who/) profiles the companies behind the network. The rule that follows is to cite Canton's counterparties and contracts and to leave its volume aggregates alone.

## The question that works is who may hold the token

"Which chain wins RWA" assumes one market with one winner, and what exists is several businesses under one label. Ethereum wins the institutional warehouse and the collateral role. Solana wins retail stock trading and, since mid-2026, the holder count. Stellar wins non-US government debt. Provenance wins the unit cost of a credit product and does not compete in any other category.

One question decides the rest, and it is who may hold this token. An answer of "a narrow group of known institutions" produces privacy, compliance in the protocol, holders numbered in the dozens to the low hundreds, turnover below one percent a month and no DeFi, on Canton, an Avalanche subnet or an issuer's own network. An answer of "any verified address" produces composability, holders in the hundreds of thousands, and turnover from tens to over a thousand percent a month, on Ethereum, Solana or BNB Chain. The network follows from that answer instead of causing it, and a product on a network that looks wrong for its category is usually explained by where its distributor sits.

> **My take:** Every ranking of these networks is a choice of denominator, and mine is turnover next to the value it divides, read from one source on one day. On that measure no tokenized fund in this material is a market yet. The only things on these networks that behave like one are stablecoins and tokenized stocks, and neither of them is what the tokenization pitch is about.

## Key takeaways

1. Four published readings of Ethereum's RWA share in 2026 (43%, 45.8%, 47.9% and over 56%) all cite RWA.xyz and differ only in the base each divides by, so a share quoted without its base cannot be checked.
2. Ethereum held 17.12 bn USD on 3 August 2026, 45.8% of a 37.35 bn USD distributed base, and in the six months to 26 July 2026 it grew slowest of the top five networks while staying the largest.
3. Solana took about 97% of tokenized-stock spot trading in May 2026 and passed Ethereum in RWA holder count in mid-2026, while Ethereum kept the lending markets where a tokenized fund works as collateral.
4. Turnover separates a market from a register: on XRP Ledger in August 2026 represented RWA turned over about 0.45% a month against about 431% for stablecoins on the same ledger, with identical fees and finality on both sides.
5. Protocol-level compliance is paid in gas and in composability: a median BUIDL transfer cost 4.07 times a median USDC transfer on Ethereum on 4 September 2026, and the largest tokenized asset in the world, about 20.1 bn USD of FIGR_HELOC on 7 July 2026, sits on a network with no composability at all.
6. Moving a permissioned token between networks runs through the transfer agent, which burns on the source network and mints on the destination after both addresses clear the allow-list, rather than through a bridge.
7. The confirmed privacy and compliance deployments in this material are Chainlink's ACE and ERC-3643's ONCHAINID, neither of which hides data cryptographically, and no chain-readable compliance based on zero-knowledge proofs was found in any of the six largest tokenized funds.
8. A network's RWA growth does not lift its own token: Stellar's RWA value rose about 360% to 3.996 bn USD by 29 August 2026 while XLM traded about 11% lower year to date.

## Glossary

- **Distributed value**: RWA.xyz's label for tokens that can leave the issuing platform and move between wallets. Every per-chain ranking in this chapter is a ranking of distributed value unless the base says otherwise.
- **Represented value**: RWA.xyz's label for tokens that cannot leave the issuing platform, which exist to record a position on a chain rather than to distribute it.
- **Permissioned token**: a token only addresses on an issuer-controlled allow-list may hold. BUIDL has worked this way since 20 March 2024.
- **Allow-list**: the list of addresses cleared to hold a given permissioned token, maintained by or for the transfer agent. Its state has to follow the token from one network to another.
- **Turnover ratio**: transfer volume over a period divided by the value held. Computed here from value and volume pairs published by one source at one reading, and it counts mints and redemptions as well as trades.
- **Collateral**: an asset pledged to secure a loan, which the lender may take if the loan goes unpaid. A tokenized fund becomes collateral only where a lending market accepts it.
- **Composability**: the property that lets any protocol on the same network use a token as a building block in its own construction. Permissioned networks give it up by design.
- **HELOC**: home equity line of credit, a revolving credit line secured on the equity in a house. Figure's tokenized HELOC portfolio is the largest tokenized asset in this material.
- **Zero-knowledge proof**: a proof that a statement is true which reveals nothing about the data behind it. Solana's Token-2022 Confidential Transfers extension is built on one.
- **AllFi**: the label used in the 2026 RWA report for a hybrid arrangement in which permissioned issuance platforms connect selectively to public liquidity venues.

## Go deeper

- [Who holds the keys](/en/blog/rwa/who-holds-the-keys/): the compliance-versus-composability trade-off measured here at network level is measured there contract by contract, including the powers that make a BUIDL transfer cost four times a USDC transfer.
- [RWA in DeFi](/en/blog/rwa/rwa-in-defi/): Ethereum's collateral role is only asserted in this chapter and is quantified there.
- [Who's who: the RWA map by layer](/en/blog/rwa/whos-who/): Canton Network, Digital Asset and Figure return there with full company profiles.
- 🟢 RWA.xyz app, network and platform pages, readings of 17 Apr, ~5 Aug and 3 Sep 2026, https://app.rwa.xyz/
- 🟢 RWA.xyz, "A New Framework for Tokenized Assets: Distributed & Represented", 21 Nov 2025, https://app.rwa.xyz/blog/a-new-framework-for-tokenized-assets-distributed-and-represented
- 🟢 Chainlink documentation, Automated Compliance Engine, https://docs.chain.link/ace, accessed 10 Sep 2026

## Sources

- 🟢 RWA.xyz, `/networks` read 3 Sep 2026: distributed 38.63 bn USD, represented 435.93 bn USD, 3,230,169 RWA holders, 39 networks; `/platforms` the same day: represented 377.74 bn USD, 3,298,816 holders, 207 platforms (via `ktory-chain-wygrywa.md`).
- 🟢 RWA.xyz, `/networks/bnb-chain`, reading dated 17 Apr 2026: distributed 3.73 bn USD, represented 0.00 USD, 43,905 RWA holders, 371 assets, 1.46 bn USD of RWA transfers in 30 days, stablecoins 14.14 bn USD with 216.42 bn USD of transfers in 30 days.
- 🟢 RWA.xyz, XRPL data, about 5 Aug 2026: represented 4.06 bn USD, distributed 368.82 mn USD, 199 RWA holders, 373 assets, 18.15 mn USD of RWA transfers in 30 days; stablecoins 901.4 mn USD with 3.89 bn USD of transfers and 60,240 holders.
- 🟢 Stellar Development Foundation dashboard, 29 Aug 2026: 3.996 bn USD of RWA on the network.
- 🟢 DefiLlama, Provenance TVL, May 2026; CoinGecko, FIGR_HELOC market capitalization 22,445,444,187 USD at 1.01 USD, 2026.
- 🟢 Author's own gas measurement, Ethereum mainnet, 4 Sep 2026: median `gasUsed` 183,672 for BUIDL (n=46), 91,335 for OUSG (n=17), 45,160 for USDC (n=40), giving 4.07× and 2.02× (`ekonomika-emitenta.md`).
- 🟢 Author's own multi-chain read of BUIDL, 5 Sep 2026: the Solana mint runs on Token-2022 with permanentDelegate, freezeAuthority, mintCloseAuthority, transferHook, metadataPointer and tokenMetadata, and no Confidential Transfers extension.
- 🟢 Securitize announcement, BUIDL deployed to Tempo, 30 Jul 2026; BlackRock share-class announcement of 13 Nov 2024, management fee held at 50 bps across new classes.
- 🟢 Chainlink blog, "Chainlink ACE: Enabling Compliance Across Chains and Jurisdictions", 30 Jun 2025, and Chainlink documentation for ACE accessed 10 Sep 2026, which describe the same product as "early access phase" and "Private Beta" respectively (via `research/chainlink.md` §A).
- 🟢 Chainlink announcement of Securitize's NAVLink adoption for pricing tokenized collateral on Aave Horizon, VanEck's VBILL among the first qualifying assets (via `research/chainlink.md` §B).
- 🟢 Author's knowledge base, module 5: ONCHAINID stores keys and signed claims without personal data; Canton positions are Daml contracts visible only to their parties; Token-2022 Confidential Transfers hide amounts and not parties.
- 🟢 Author's knowledge base, module 8: the compliance-versus-composability finding, the per-dimension network scoring behind the quadrant figure, and the turnover table.
- 🟢 Author's knowledge base, module 4: the call graph behind a BUIDL transfer, which queries the investor registry, investor-count limits, accreditation status, lock-ups and the wallet-to-investor mapping.
- 🟡 Coinpaprika, *Best RWA Blockchains*, 3 Aug 2026: Ethereum 17.12 bn USD and 45.8% of a 37.35 bn USD base across 38 networks, BNB Chain 5.75 bn USD, Solana 3.68 bn USD, and the explanation that Ethereum's lead comes from which issuers launched there first.
- 🟡 Stobox, *The State of RWA Tokenization, 2026 Mid-Year Report*, data of 10 Jul 2026: Ethereum 47.9% of a 33.5 bn USD distributed base excluding stablecoins, BNB Chain 12.1%, Solana 9.8%, Stellar 9.0%; Solana overtaking Ethereum in RWA holders; "issuing a token does not create a market".
- 🟡 Token Terminal, about Aug 2026: Ethereum 43% of tokenized US Treasuries on a 15.2 bn USD base, of which Ethereum holds 6.6 bn USD.
- 🟡 Pine Analytics, *Tokenized Equities on Solana*, 17 Jun 2026: about 97% of May 2026 spot volume on Solana, 869 mn USD against 24 mn USD elsewhere, and slippage tolerance of 0.1% to 0.5% on Raydium and Jupiter for TSLAx, NVDAx and CRCLx.
- 🟡 Genfinity, 6 Jul 2026: above 96% of transactions on Solana and 54 consecutive weeks ahead of all other networks combined, with 5.77 bn USD of spot volume in Q2 2026 against 775 mn USD in the whole second half of 2025.
- 🟡 TheCryptoBasic via RWA.xyz, 10 Jul 2026: Solana 3.62 bn USD of RWA across 295,853 holders; 24 Jul 2026: the XRPL breakdown, including JMWH at 2.229 bn USD and distributed excluding stablecoins at 323.18 mn USD.
- 🟡 crypto.news via RWA.xyz, 26 Jul 2026: six-month net inflows of 3 bn USD to BNB Chain, 2.6 bn USD to XRPL, 1.6 bn USD to Solana, 972 mn USD to Avalanche and 424 mn USD to Ethereum.
- 🟡 Cryptobriefing via RWA.xyz, 15 to 16 Aug 2026: tokenized US Treasuries 16.23 bn USD across about 18 networks, Ethereum about 43% and BNB Chain about 31.5%, with USYC at 3.0 bn USD, BUIDL at 2.7 bn USD and USDY at 2.15 bn USD.
- 🟡 CryptoDaily via Cointelegraph and RWA.xyz, Aug 2026: tokenized stocks 2.54 bn USD with 29.5 bn USD of transfers in 30 days.
- 🟡 BeInCrypto via RWA.xyz, 11 Jul 2026: FIGR_HELOC about 20.1 bn USD on 7 Jul 2026, against 15.16 bn USD for all tokenized US Treasuries in the same reading.
- 🟡 The Defiant, 11 Feb 2026: Provenance TVL of 1.2 bn USD belonging entirely to Figure Markets, and Brian Huang of Glider on composability.
- 🟡 Cointelegraph and Coinpaprika, 27 to 29 Aug 2026: Stellar 3.996 bn USD of RWA and about 490 mn USD of non-US government debt on 20 Aug 2026, XLM about 11% lower year to date.
- 🟡 CoinPaprika, 16 Aug 2026: SOL about 75.33 USD, about 74% below its January 2025 peak.
- 🟡 Bloodstone Capital, *Canton Network: RWA Deep Dive*, 30 Jun 2026: consortium composition, CC at 0.1415 USD after a 12.4% fall in two weeks, daily volume of 8 to 17 mn USD.
- 🟡 Blockworks, *Canton's $6T RWA rails*, 1 Jan 2026: 6 tn USD of RWA processed in Q4 2025 and about 350 bn USD a day in Treasury activity, both the network's own figures.
- 🟡 CoinDesk, 13 Nov 2024: Aptos Foundation, Avalanche (BVI), Inc. and Polygon Labs BD Investments (Cayman) Ltd. pay BlackRock a quarterly fee, which is why BUIDL charges 20 bps on those networks and 50 bps on Ethereum, Arbitrum and Optimism.
- 🟡 Aave Horizon, Jul 2026: 163.5 mn USD borrowed against RWA collateral.
- 🟡 Product documentation for BENJI and OUSG on cross-chain movement: a transfer agent burn and mint for BENJI, an allow-list-aware bridge between Ethereum, Mantle and Polygon for OUSG (`stos-technologiczny-2026.md`).
- 🔴 MetaMask RWA category guides, Apr 2026: Ethereum above 56% of "all tokenized asset value", a base the source does not define.
- 🔴 crypticweb3, 6 Jul 2026: Ethereum about 65% of value, base not established.
- 🔴 Markets Media citing Provenance, Jan 2025: about 125 bps saved per HELOC, about 20% of US HELOC originators on the network, about 650 mn USD of tokenized loans issued a month, all a representative's declaration rather than an audited result.
- 🔴 CoinPaprika relaying RWA.xyz, Aug 2026: Canton at about 371.3 bn USD represented against near-zero distributed, and 85% to 98% of the represented category on one network. A research pass of 10 Sep 2026 could not locate this figure via RWA.xyz's live app or DefiLlama (`research/state-2026-09.md` §(g), `research/README.md` item 13).
- 🔴 CoinStats AI, Sep 2026: the criticism that Canton's aggregates have no independent verification and may represent notional activity, supported assets or pilots.
- 🔴 Chainstack, Jul 2026, and CoinStats, Aug 2026: Ondo Chain as an institutional L1 with permissioned validators staking RWA and a mainnet targeted for mid-2026, and the hardware-enclave architecture announced in Jul 2026.
- 🔴 24/7 Wall St., 15 Apr 2026: the XRP Ledger described as adding zero-knowledge proofs for private bank transactions.
- 🔴 eco.com RWA guides, May to Aug 2026: the cross-chain allow-list state problem.

## Caveats

Every per-chain percentage in this chapter is biased by the same methodological problem, and it is not fixable from outside. Products are deployed on eight or nine networks each, and the public breakdowns either treat a product as present on one network or split it by a method nobody has published. The shares here are usable for orders of magnitude and not for a ranking to the percentage point.

Represented value per network is mostly unestablished. Beyond XRPL at 4.06 bn USD, BNB Chain at 0.00 USD and the unlocated Canton figure, no public breakdown of represented value by network exists, so every ranking here is a ranking of distributed value and misses the largest part of the market measured in dollars.

The turnover ratios are the author's own computation on value and volume pairs from one source at close dates. Ethereum, Solana and Stellar have no ratio here, because no public reading pairs a 30-day RWA transfer volume with a value for them, and their absence limits what the spread table can show. The ratio also includes mints and redemptions and not only secondary trading.

Canton's own figures are declarations and cannot be verified from outside. Which of them is a stock, which is a flow and whether they overlap was not established, and the approximately 371.3 bn USD represented figure was not reproducible from RWA.xyz or DefiLlama in a dedicated research pass on 10 September 2026. The 85% to 98% share derived from it inherits the same weakness, which is why neither appears in the prose.

Figure's 125-bps saving is a Provenance representative's declaration from January 2025, not an audited result. A Morpheus Research report of April 2026 separately challenged Figure's characterization of its own operations as blockchain-based; those criticisms were not verified for this chapter and no company response is on record here.

The reading of XRPL's distributed value differs between sources on close dates: 323.18 mn USD excluding stablecoins on 24 July 2026, and 368.82 mn USD on about 5 August 2026. Both appear above rather than being reconciled into one figure.

Whether the RWA on BNB Chain is dominated by a single product is not established. If one product supplies most of the 5.75 bn USD recorded there, the network should be scored the way Provenance is scored, as a single-client network, which would change what this chapter says about the runner-up.

The count of networks a single product runs on is disputed in the sources and was not settled here: eight for BUIDL and BENJI in material to June 2026, nine in an August 2026 source after Tempo was added.

The three announced privacy items were not individually traced back to a knowledge-base line and come from tertiary sources, which is why they carry no numbers in the prose. The absence of a zero-knowledge compliance deployment in the six largest tokenized funds is an absence in this material rather than proof that none exists.

RWA.xyz changed its taxonomy twice, on 21 November 2025 and 17 March 2026, so comparisons across either date may show a change of definition rather than a movement of capital.