The Race Is Over Rails, Not Apps

Tokenized equities are worth somewhere in the high-$2 billions to roughly $3 billion on-chain in early September 2026, depending on which tracker you read — and the number misses the actual contest. What is being built is not a category of retail products. It is a composable financial primitive: an equity exposure that can sit in a wallet, post as collateral in a lending market, settle against a perpetual, or form one leg of a multi-asset DeFi position.

The race for on-chain stocks is therefore a rails race — decided at the level of issuance models, custody structures, chain selection, attestation standards and oracle depth, not at the level of which app wins users. And that layer is fragmenting by design: there is no canonical on-chain NVDA. NVDAx (xStocks), NVDAon (Ondo), NVDAc (Coinbase) and NVDAB (Binance bStocks) are four separate legal wrappers on four chains with distinct rights. Telling them apart is the analysis, not a footnote.

The Issuer Landscape: Seven Distinct Models

Seven issuers matter, and they do not build the same instrument. The same ticker exists as several non-fungible tokens with different legal claims, different chains and different rights.

Issuer	Legal shape	Backing claim	Chains	Rights / notes
xStocks (AAPLx)	Jersey tracker certificate (Backed Assets JE Ltd)	1:1 shares in custody	Solana primary by value; also ETH / BNB / X Layer / HyperCore spot	Price exposure, no vote. Payward/Kraken framework; Bybit/OKX alliance venues. Sep 1: Payward–LSEG plan to tokenize the 100 largest London names, subject to approval; UK still excluded
Ondo Global Markets (NVDAon)	Total-return note	1:1 + buffer at a US broker-dealer (Alpaca cited)	Ethereum, Solana, BNB	Dividends reinvested, so the token drifts slightly above one share economically. Usually the value leader; Ondo reports >$1B TVL, 260+ names, >$18B cumulative volume
Binance bStocks (NVDAB)	ADGM certificate (BTech Holdings)	1:1 at a US brokerage	BNB Chain + Binance CEX	Claim on the issuer's interest, not a DTC position. Not the same as "Binance Stocks," the separate off-chain brokerage book of 7,000+ real names
Dinari dShares	Broker-dealer / SEC-registered transfer-agent claim	1:1 brokerage	Ethereum, Avalanche, Arbitrum, Base; Sei/Solana planned	Aug 4 US launch, ~724 names incl. the S&P 500, bought with USDC into self-custody. Most regulation-proximate model; composability more gated by design
Robinhood Stock Tokens	Jersey debt security (current program)	Program-backed	Robinhood Chain (Arbitrum Orbit, mainnet live)	Distinct from Robinhood's earlier EU derivatives. Holder-heavy, value-light; raw ERC-20 supply overstates AUM (RWA.xyz). Limited withdraw-to-DeFi
Coinbase (AAPLc, NVDAc)	Beneficial claim, bankruptcy-remote custody	1:1 shares in custody	Base	Mega-cap catalog for now; built for Base composability. Chainlink market-data partnership announced Aug 24
Securitize (SECZ)	Issuer-sponsored share	The issuer's own listed stock	Avalanche, Solana	Began NYSE trading Jul 2, 2026 and tokenized its own common stock the same day (~$295M on-chain at launch). The issuer tokenizes itself, not a third-party wrapper

OKX Unified Tokenized Stocks belongs in the picture but not in that table: it is a venue product, not an issuer, with an xStocks backend. Mechanics below.

On issuer share: third-party trackers and Ondo's own figures disagree on how much of on-chain issuer value Ondo holds, so this piece cites neither as a current share.

Where Liquidity Actually Sits

On-chain tokenized-stock market cap ran from the high-$2 billions to roughly $3 billion depending on the tracker in the week of 5–7 September 2026 — Asortino put it at $2.93 billion on September 7, while Token Terminal recorded an all-time high of $3.1 billion on September 5. Chain concentration, same week: BNB Chain around $1.0 billion, Ethereum around $770 million, Solana around $716 million. Treat all of these as moving figures — pull fresh at publish.

X Layer is an OKX expansion rail, not the current depth centre — a real but small sleeve at roughly $87 million, about 3% of chain share (Asortino, September 7), with Solana and Ethereum carrying the bulk of xStocks value.

Two methodological points matter to DeFi readers. First, DEX liquidity is per issuer-token, not per ticker: the same NVDAx pool can be routed by several frontends, so pool volume is not venue volume, and a 24-hour AMM snapshot can crown a "leading venue" while describing one issuer-token on one chain. An August 23 snapshot by Aerodrome core contributor Alexander Cutler (@wagmiAlexander) shows the trap — four names (NVDA, GOOGL, AAPL, META), one 24-hour window, unpublished query filters, Robinhood at roughly 48%, a redacted issuer at roughly 26%, bStocks at roughly 22%, xStocks at roughly 3%. That is a methods demonstration, not market-share data.

Second, OKX Unified Tokenized Stocks is venue unification at the order-book level, not cross-issuer on-chain standardization. OKX runs one book per underlying — XAAPL, not one market per issuer-token — with an xStocks backend, deposits and withdrawals on Solana and X Layer, and USDT settlement around the clock. It improves the trading experience. It is not a single canonical on-chain token across issuers, and not a global settlement standard.

"1:1 Backed" Is Not One Legal Structure

Every issuer describes its product as "1:1 backed," and none of those phrases carry the same legal meaning. As the issuer table shows, the wrappers span a tracker certificate (a collateral claim on custodied shares, not a DTC shareholding), a total-return note (a debt instrument tracking price plus reinvested dividends), a beneficial claim in bankruptcy-remote custody, a broker-dealer and transfer-agent claim, a Jersey debt security, and an issuer-sponsored share. In none of these cases does tokenization override the legal, contractual or transfer restrictions on the underlying instrument.

Attestation — issuer attestations, security interests, and proof-of-reserve-style disclosures where published — proves the committed backing statement is unchanged at a point in time. It does not establish continuous real-time 1:1 correspondence, does not confer rights absent from the underlying instrument, and does not certify that secondary volume is wash-free. A protocol accepting an equity token as collateral is accepting one specific wrapper and one specific attestation standard, not a homogeneous "backed share."

The regulatory frame shifted on the same day as the LSE announcement. On September 1, 2026, the SEC proposed to modernize the rules for registered transfer agents — the first substantive update since the late 1970s and early 1980s — in a release that directly addresses transfer agents interacting with tokenized securities, distributed ledger technology and smart contracts, and that contemplates blockchain-native transfer agents in US markets. January 2026 SEC staff guidance had already held that tokenized stocks fall under existing securities and derivatives rules regardless of the ledger.

Ecosystem Awareness and Chain Scope

There is no single canonical NVDA token on-chain: NVDAx, NVDAon, NVDAc and NVDAB are four non-fungible instruments on four chains, so a protocol that integrates one has not integrated the others. Most usable depth currently sits on Solana, Ethereum, BNB Chain and CEX order books — not on X Layer. That is the honest state of the market in early September 2026.

The CoAgentic build is X Layer-native. CoAgentic Markets currently tracks 16 X Layer xStocks as part of its on-chain equity infrastructure work: it screens verified xStock contracts on X Layer, measures observable pool liquidity and 24-hour volume, and hashes research snapshots to a non-custodial registry on X Layer testnet, with a mainnet registry as future work. It does not track the global tokenized-equity market, all issuers or all chains. The oracle is designed to be cross-venue-depth-aware — able to acknowledge where deeper books exist on other chains and venues rather than reading a single pool — and multi-issuer and cross-chain coverage are research and roadmap priorities, not shipped features.

DeFi Mechanics: Usable vs Nominal Liquidity

Token existence is not market usability. Depth is fragmented per issuer-token, per chain, per pool: a protocol integrating tokenized equities as collateral is not getting "the NVDA market," it is getting one wrapper's pool, which may be thin at the size that matters.

Integrations are appearing at various stages of development, mostly via secondary reporting rather than primary confirmation. Venus Protocol has reportedly added Binance bStocks — Tesla, Nvidia and SpaceX-exposure tokens — to its Core Pool on BNB Chain as collateral; Coinbase and Chainlink are reported to have paired market-data infrastructure with Base tokens in late August; Backpack has described a unified portfolio-margin product that would let holders of real US shares post them as perpetuals margin; and Coinbase has filed for 24/7 stock perpetuals. The pattern points to three uses — leverage the token, borrow stablecoins against it, or trade the basis between token and perpetual — and, as CryptoSlate has noted, that collateral is arriving before the off-hours oracle and liquidation infrastructure is settled.

The comparison with tokenized Treasuries is instructive. Tokenized T-bills are the mature RWA collateral: large, with a well-understood peg to an external benchmark and roughly 4–5% native yield. Tokenized equities are earlier — most carry no native yield unless dividends are reinvested into the wrapper (Ondo's note structure), and the price drifts against the NYSE tape when US markets are closed. Both compete for the same DeFi collateral slot; equity tokens bring more exposure and more peg, rights and oracle risk for the protocol that accepts them. Pricing infrastructure is consolidating — CF Benchmarks launched an xStocks index in Q2 2026, Chainlink data now covers Coinbase's Base tokens — but the missing piece is a cross-venue-depth-aware oracle that reads liquidity across chains and venues rather than anchoring to one pool's mid-price.

Tokenized FX: Related Category, Not This Market

Currencies are already on-chain — as multi-issuer stablecoins and tokenized deposits, not an xStocks-style framework for EUR or JPY. USD coins dominate supply; non-USD stablecoins stay small. On-chain FX perpetuals trade tens of millions a day against a roughly $7–10 trillion/day OTC market. Institutional work (JPMorgan Kinexys, BIS Project Agorá) targets tokenized reserves and payment-versus-payment settlement, not retail wrappers.

Forward Look: What Has to Happen to Scale

Three things stand between here and tokenized equities functioning as a DeFi primitive at scale.

Onshore regulatory access. The SEC's September 1 transfer-agent proposal is the strongest US signal yet — it addresses blockchain recordkeeping and on-chain transfer agents directly. With Dinari's onshore broker-dealer stack, Securitize's issuer-sponsored model and Nasdaq's issuer-centric token design pointing the same way, US access is closer to resolution than it has been, and the offshore wrappers are building volume while the perimeter widens.

Liquidity consolidation and oracle infrastructure. Per-issuer-token fragmentation is the core usability problem: NVDAx depth on Solana does nothing for a BNB Chain protocol that needs NVDAon depth. Fixing it means cross-chain routing and depth-aware oracles that abstract the fragmentation, plus off-hours liquidation infrastructure so a protocol can hold equity collateral through a weekend or market holiday.

Cross-chain coordination beyond CEX books. OKX's Unified Tokenized Stocks shows venue-level unification is achievable — one book per underlying, one settlement flow — but only at the CEX layer. On-chain, NVDAx, NVDAon and NVDAc stay separate, and the DeFi use cases that matter — composable collateral, programmable settlement, agent-driven portfolios — need that layer resolved, not just the interface on top.

At a market cap still short of $3 billion, with integrations only beginning to appear across lending, margin and perpetuals, the category is nowhere near the scale or composability of tokenized Treasuries. The LSE partnership and the SEC transfer-agent proposal landing on the same day mark 2026 as an infrastructure year, not a maturity year.

CoAgentic Dev researched and drafted this analysis. Reviewed and approved by OrionJVale. Corrections and verifiable additions via the CoAgentic contact page.

Tokenized Equities: The Infrastructure Race for On-Chain Stock Markets