What’s Actually Inside Tokenized Private Credit

The global private credit market is estimated at $2 to $3.5 trillion, depending on who is measuring, and projected to reach $5 trillion by 2029. The tokenized portion currently tracks at least $27.3 billion across products indexed on the rwa.xyz credit dashboard. Apollo, Coinbase, and Securitize are building on-chain credit products. Figure went public on Nasdaq. The narrative is that private credit has found its on-chain moment.

That $27.3 billion is not one number. In April 2026, rwa.xyz introduced a framework that classifies tokenized assets into two categories based on whether tokens can be moved to wallets outside the issuing platform and transferred peer-to-peer. Distributed assets ($5.14 billion) can be moved and transferred freely, with the blockchain serving as a channel for capital formation. Represented assets ($21.89 billion) cannot leave the issuing platform or be transferred between wallets, with the blockchain serving as a recordkeeping and reconciliation layer. These are different categories, not a subset relationship.

Each side contains structurally different products. The represented side is dominated by Figure’s $17.6 billion HELOC token, traditional consumer lending on blockchain rails. The distributed side includes DeFi credit protocols, tokenized fund wrappers, hashrate-backed credit, and a long tail of 2,328 smaller items. These products have different risk profiles, different counterparty dependencies, and different investor experiences. Lumping them together under a single number obscures more than it reveals.

This analysis breaks down what’s inside each measurement, then evaluates the sector through the altrntv labs five-part framework: structure, economics, operations, distribution, and liquidity.

What the sector actually looks like

Represented value: $21.89 billion

Represented assets are tokens that cannot be moved to wallets outside the issuing platform or transferred between wallets. The blockchain functions as infrastructure for recordkeeping and reconciliation, not as a distribution channel. The metric is dominated by a single product.

Figure’s HELOC token ($17.6 billion) accounts for 80% of represented value. Figure originates home equity loans on Provenance Blockchain and trades them through Figure Connect, a marketplace used by Goldman Sachs, Jefferies, and Deutsche Bank. It went public in September 2025 at a $7.6 billion valuation and settles $600 million in loans monthly. The HELOC token is a real credit product backed by real home equity. But it is classified as represented because it cannot currently be moved outside Figure’s platform or transferred between wallets. rwa.xyz notes that Figure is working to make its tokenized assets compatible with DeFi protocols, at which point they would be reclassified as distributed.

The remaining 20% of represented value ($4.3 billion) is more diverse than the Figure dominance suggests. Tradable ($2.2 billion) has tokenized over 48 institutional-grade private credit positions on ZKsync, from fintech senior secured loans to legal receivables and music royalties. CRX Digital Assets ($594 million across 17 products) and PACT ($576 million across 7 products, structured credit origination on Aptos) are growing platforms. Bitbond ($460 million), Intain ($395 million), and VERT Capital ($379 million, Brazilian credit markets) round out the next tier. Mercado Bitcoin contributes $274 million across 795 assets, reflecting the depth of Latin American tokenized credit activity. These platforms share the represented classification but vary widely in structure, geography, and asset type.

Distributed value: $5.14 billion

Distributed assets are tokens that can be moved to external wallets and transferred peer-to-peer. The blockchain serves as a channel for capital formation, enabling direct access to a global investor base. This is where the sector’s real diversity lives, and where the subcategories matter most.

DeFi-native credit protocols originate or facilitate loans on-chain with on-chain liquidity. Maple’s syrupUSDC ($1.34 billion) and syrupUSDT ($406 million) are institutional lending to crypto-native firms through a Pool Delegate model. Centrifuge ($476 million) has built genuine on-chain securitization with SPV-per-pool structures and tranche subordination. Pareto operates its own credit marketplace with $179.7 million in outstanding loans and $2.12 billion in cumulative credit extended, using on-chain structured credit vaults for institutional borrowers like FalconX ($135.3 million, managed by M11 Credit), RockawayX ($19.8 million), and Fasanara Digital ($11 million). Clearpool runs single-borrower permissionless pools across seven chains. Deploi, a Latvia-based SPV, issues ERC-3475 tokenized notes across five segregated credit pools (consumer, SME, mortgage, litigation funding, distressed debt) under EU Prospectus Regulation, with yields ranging from 6.5% to 17.5%. Goldfinch’s legacy pools extended undercollateralized loans to emerging-market borrowers, with three defaults totaling ~$18 million. This subcategory is where the most structural innovation and the most defaults have occurred.

Tokenized fund wrappers take existing institutional credit strategies and add blockchain distribution. Apollo’s ACRED ($133 million across 63 wallets) is a tokenized feeder fund for Apollo’s Diversified Credit Fund, issued through Securitize across six chains. Midas’s mF-ONE (which scaled to ~$190 million on Morpho, now ~$71 million in total asset value) is a tokenized certificate tracking Fasanara Capital’s F-ONE credit strategy, issued under a Liechtenstein FMA-approved prospectus. Goldfinch Prime pivoted from permissionless lending to aggregated access to institutional credit managers (Apollo, Ares, Golub Capital). Centrifuge’s institutional partnerships (Janus Henderson, Apollo via ACRDX) also fit this model. The credit decisions are entirely traditional. Blockchain adds distribution and composability, not underwriting.

Hashrate-backed credit is a smaller but distinctive subcategory. STOKR’s Blockstream Mining Note 2 ($905 million) and PKH Mining Note 2 ($559 million) are credit instruments where capital is deployed into bitcoin mining operations and obligations are serviced by future mining revenue. BMN1 returned 103% in BTC terms. These are real credit (the borrower receives capital, the investor receives yield from productive operations), but the collateral is hashrate and future mining output rather than receivables, real estate, or corporate cash flows. Their $1.46 billion combined value represents 28% of distributed value.

Infrastructure layer

Underneath both represented and distributed products sits an infrastructure layer that the sector depends on. This layer spans tokenization platforms, transfer agents, custody providers, oracle networks, cross-chain bridges, and lending protocols. None of these are credit products themselves, but credit products cannot function without them.

Tokenization and issuance platforms handle the creation, compliance, and lifecycle management of tokenized securities. Securitize is the most systemically important, SEC-registered as transfer agent, broker-dealer, ATS operator, investment advisor, and fund administrator, with $4.6 billion in AUM on its platform and a pending Nasdaq listing via Cantor Fitzgerald SPAC at $1.25 billion. Products like Apollo’s ACRED and BlackRock’s BUIDL are structurally dependent on Securitize for token issuance, KYC/AML, transfer processing, and regulatory compliance. If its registrations were challenged, multiple products would be affected simultaneously. STOKR provides equivalent infrastructure on Bitcoin’s Liquid Network as a Luxembourg-regulated VASP, issuing mining notes, tokenized equities (Strategy/MSTR, Metaplanet), and tokenized funds (Aquarius, Goldstream). Tokeny, acquired by Apex Group in May 2025, has tokenized more than $32 billion in assets and expanded its reach through Apex’s global fund administration network. DiGiFT, dual-licensed by the Monetary Authority of Singapore and the Hong Kong SFC, provides end-to-end tokenization, issuance, distribution, and trading for institutional RWAs, with partnerships including UBS, Invesco, Wellington, DBS, and BNY, and a $150 million credit fund tokenization with Flow Capital (April 2026). DTCC Digital Assets (evolved from its Securrency acquisition) provides enterprise-grade infrastructure for tokenizing and managing digital securities, including lifecycle automation and privacy-enhanced collateral management. Tradable has tokenized $2.19 billion in institutional-grade private credit positions on ZKsync, from fintech senior secured loans to legal receivables, making it the largest tokenization platform by represented credit value after Figure. Deploi launched its first issuance (UK consumer credit notes) on Polygon through Assetera in May 2026, with Nasdaq CSD providing ISINs and approximately €100 million in additional issuance expected over six months, representing the EU-regulated approach to tokenized issuance infrastructure.

Custody infrastructure provides the institutional-grade storage and signing capabilities that regulated participants require. Fireblocks, with its NYDFS-chartered Trust Company and clients including FalconX, Galaxy, and Bakkt, provides wallet infrastructure, policy engines, and tokenization tooling across over 80 blockchain networks. Anchorage Digital operates as a federally chartered crypto bank and qualified custodian. BitGo provides custody and settlement infrastructure. These providers sit between the tokenized product and the investor, and their operational reliability is a precondition for institutional participation.

Oracle and cross-chain infrastructure connects tokenized credit products to pricing data, compliance information, and multi-chain distribution. Chainlink’s CCIP processed over $18 billion in cross-chain transfer volume in Q1 2026 and serves as the bridge infrastructure for Maple’s syrupUSDC (over $3 billion in cross-chain deposits) and Coinbase’s wrapped assets. Wormhole provides the cross-chain bridge for Apollo’s ACRED across six chains. These are not background services. When ACRED is leveraged on Morpho across multiple chains, the oracle feeding the price data and the bridge moving the token are load-bearing infrastructure. An oracle delay or bridge exploit would affect every product that depends on them.

DeFi lending protocols have become infrastructure for tokenized credit, not just downstream users of it. Morpho, Aave, and Kamino now host an estimated $700 million in leveraged positions against tokenized private credit tokens. Aave’s Horizon, its RWA-focused money market, has $176 million in loans outstanding. These protocols provide the composability layer that makes tokenized credit positions usable as collateral, but they also extend the risk surface of every product deposited into them.

Origination infrastructure is an emerging layer. Credit Coop provides on-chain structured lending tooling that converts future cash flows into smart contract collateral, enabling crypto-native businesses to access credit without traditional overcollateralization. It has facilitated $1.7 billion in cumulative lending volume across 11 borrowers, though Rain (a digital payment card company) accounts for $1.6 billion of that. Other borrowers include Karta ($47.1 million), SuperStable ($23.9 million), and Coinflow ($10 million), spanning cross-border payments, DeFi index products, and stablecoin infrastructure. PACT Foundation operates a consortium-based microfinance platform for emerging-market credit origination. These are early-stage, but they represent an attempt to build the origination rails natively on-chain rather than tokenizing loans originated through traditional channels.

Provenance Blockchain, DART, and Figure Connect form Figure’s proprietary infrastructure stack. Hastra ($317 million) distributes yields from Figure’s lending operations to DeFi participants on Provenance and Solana.

The infrastructure layer is where the sector’s systemic risk concentrates. A credit product can have sound underwriting, clean structure, and honest economics, and still fail if the infrastructure underneath it breaks. The more products that share the same infrastructure providers, the more correlated the risk becomes.

The holder base tells its own story

The sector reports 186,091 holders across 2,380 assets, but the concentration is extreme. Apollo’s ACRED has 63 holders across six chains. Sixty-three. That is $133 million distributed across 63 wallets, with an average position of roughly $2.1 million and only 35 active addresses in the past 30 days. Maple’s syrupUSDC has approximately 2,989 holders on Ethereum for $1.2 billion in on-chain market cap, an average position around $400,000. These are not mass-adoption products. They are institutional instruments with a thin holder base and concentrated positions. The headline holder count is spread across thousands of smaller assets and legacy positions. The products that actually matter to the sector’s narrative are held by a remarkably small number of participants.

A Figure HELOC, a Maple loan to a crypto trading firm, a STOKR mining note, a tokenized Apollo credit fund, and a Pareto structured credit vault all share a dashboard label, but their structures, risk profiles, and investor experiences diverge significantly. An allocator evaluating “tokenized private credit” needs to know which of these they are actually looking at.

Structure: What are investors actually underwriting?

The structural divergence across the sector is wider than many participants acknowledge.

Figure operates a vertically integrated lending stack. Loans are originated, registered on DART (a blockchain-native lien registry), and traded on Provenance. The investor is underwriting consumer credit (HELOCs backed by home equity) through institutional infrastructure. The structure is recognizable to anyone in mortgage or consumer finance.

Tradable takes a different approach: tokenizing existing institutional credit positions rather than originating new ones. With $2.19 billion in represented value across 48 assets on ZKsync, it is the second largest platform after Figure by represented credit value. The investor is underwriting third-party credit positions (fintech senior secured loans, legal receivables, music royalties) through Tradable’s deal-ownership tokenization layer. The platform has 59 holders and zero monthly transfer volume, which means $2.19 billion in tokenized credit with no secondary market activity.

Centrifuge has built the closest thing to genuine on-chain securitization. Each pool operates through a special purpose vehicle. Legal ownership of assets transfers to the SPV. Tranches provide structural subordination: junior takes first loss, senior gets protected yield. This is the traditional securitization stack rebuilt with public-blockchain settlement. The structure is sound in principle, but each pool depends on the quality of its specific asset originator.

Deploi takes a similar securitization approach from the EU-regulated side. A Latvia-based SPV issues ERC-3475 tokenized notes across five segregated pools (consumer, SME, mortgage, litigation funding, distressed debt), with Assetera GmbH (MiFID II) handling distribution and Nasdaq CSD Riga providing custody and settlement. Originators retain 5% skin-in-the-game per EU Securitisation Regulation. The structure provides direct noteholder claims against specific originator assets, but each pool’s risk is isolated: losses in one pool cannot be covered by another. The €1 billion programme limit is ambitious for a platform without a live secondary market.

Maple uses a Pool Delegate model where credit professionals underwrite borrowers and manage loan terms. After its 2022 defaults, Maple pivoted to overcollateralized loans to institutional borrowers. The structure is cleaner than before, but the risk still concentrates in the Delegates’ judgment.

Goldfinch’s legacy pools are undercollateralized loans to borrowers in emerging markets: Kenya, Southeast Asia, Latin America. The structural ambition was to bring DeFi capital to real-world credit needs. Three defaults totaling ~$18 million have shown how that plays out in practice. Goldfinch Prime is a fundamentally different structure: aggregated access to institutional credit managers with $1 billion+ AUM, 90%+ senior secured portfolios, and SEC registration.

Apollo’s ACRED is a tokenized feeder fund. The investor is underwriting Apollo’s Diversified Credit Fund through a Securitize-issued token. The credit decisions are entirely traditional. Blockchain adds distribution, composability, and cross-chain access, not underwriting. Midas’s mF-ONE operates on a similar principle: a tokenized certificate tracking Fasanara Capital’s F-ONE credit strategy.

STOKR’s Blockstream Mining Notes represent a different kind of credit. Investors are providing capital to bitcoin mining operations in Georgia, Montreal, and Texas, with obligations serviced by future mining revenue. BMN1 returned 103% in BTC terms. This is credit (capital extended, yield from productive activity), but the collateral is hashrate, not receivables or real estate. The structure is a Luxembourg-regulated security token on Bitcoin’s Liquid Network, issued through STOKR’s tokenization infrastructure.

Securitize is not a credit product. It is the regulated infrastructure layer that multiple credit products are built on, and it is the most systemically important single entity in the sector. SEC-registered as transfer agent, broker-dealer, ATS operator, investment advisor, and fund administrator, it is the only vertically integrated tokenization provider with all five registrations. Apollo’s ACRED, BlackRock’s BUIDL, and Hamilton Lane’s SCOPE are all structurally dependent on Securitize for token issuance, KYC/AML, transfer processing, and regulatory compliance. The platform manages $4.6 billion in AUM. It signed a memorandum of understanding with the New York Stock Exchange to support tokenized securities and is pursuing a Nasdaq listing via Cantor Fitzgerald SPAC at $1.25 billion. If Securitize’s registrations were challenged or its operations disrupted, the impact would cascade across every product built on its rails. No other entity in the sector carries this level of cross-product structural dependency.

The structural question for the sector is not whether blockchain can represent credit. It clearly can. The question is whether investors understand which layer is doing the underwriting, which layer is doing the tokenization, and how much of the sector’s structural integrity depends on a small number of shared infrastructure providers.

Economics: How does value reach the investor?

Stated yields across the sector range from 5% to 17.5%, but that spread is not a menu of options at different price points. It is a map of fundamentally different economic engines, different credit risks, and in some cases, different currencies.

Consider three products that all appear under the same category label. Figure’s returns come from US consumer credit: home equity lines backed by residential property, generating traditional lending yields through institutional infrastructure. STOKR’s mining notes generate returns in BTC from bitcoin mining operations in Georgia, Montreal, and Texas, not USD from credit payments. Deploi’s five segregated pools span EU consumer loans, SME working capital, mortgage bridge loans, litigation funding, and distressed debt, with yields ranging from 6.5% to 17.5% depending on the pool. These are not variations on a theme. They are entirely different businesses producing entirely different risk exposures.

Higher yields generally mean higher credit risk, not better deals. Goldfinch’s legacy pools offered 10-17% on undercollateralized loans to emerging-market borrowers. Three defaults totaling ~$18 million showed what that premium was pricing. Credix claims 14-15% on Latin American credit, reflecting both the LatAm risk premium and the thinner transparency around originator quality. Deploi’s distressed debt pool targets 15-17.5% on non-performing loans, the highest stated yield in the sector and the most explicit about the risk it carries. When a product offers double-digit yields on private credit, the question is not whether the return is attractive but what is generating it and what happens when the underlying credit deteriorates.

BTC-denominated returns are a different economic proposition entirely. STOKR’s BMN1 returned 103% in BTC terms. Whether that translates to a strong USD return depends on bitcoin’s price at the time of distribution. An investor underwriting a mining note is taking credit risk on the mining operation and commodity risk on bitcoin simultaneously. This is not comparable to a USD-denominated credit product on any standard risk-adjusted basis, yet both appear on the same dashboards under the same category label.

Across much of this sector, the cost path from gross yield to net investor outcome cannot be fully traced from public materials. Fund wrappers like ACRED layer institutional fund management fees, tokenization platform fees, and transfer agent fees. The total drag on returns is difficult to determine from publicly available documents. Deploi’s information memorandum discloses pool-level yields but the fee split between Deploi, Assetera, and the loan originators is not fully transparent. DeFi protocols vary: Maple’s fee structure is partially visible; Centrifuge fees are set per pool; Goldfinch has protocol-level fees that are not prominently disclosed. Credit Coop’s origination fees are embedded in the borrower rate but the protocol’s share is not clearly broken out. This is not an accusation of hidden fees. It is an observation that the sector has not yet developed the disclosure norms that would let an allocator do basic gross-to-net analysis from public information.

For many products in this sector, tokenization’s economic value remains unproven. Figure is the clearest case where blockchain reduces origination and settlement costs. For fund wrappers like ACRED, tokenization adds distribution reach and DeFi composability but also adds platform fees on top of existing fund expenses. For DeFi-native protocols, the case depends on whether on-chain origination and global liquidity sourcing produce better borrower access or lower costs than traditional channels. That case may eventually be made. It has not been made yet by the numbers many of these products have produced.

For allocators, the economic question is not just “what is the yield?” It is: where does the yield come from, what risk is it pricing, what fees sit between the gross return and the investor, and can any of that be verified from public materials? Across much of this sector, the full picture is difficult to assemble.

Liquidity: Can investors actually get out?

This is the sector’s weakest dimension and its most persistent framing problem.

The fundamental issue: technical transferability is not the same as real liquidity. A token that can be moved between wallets is not the same as an investment that can be exited at or near fair value without material delay or friction.

Figure Connect has the closest thing to a real secondary market, with institutional counterparties trading loan positions. But it is closed to retail and DeFi participants.

Maple’s syrupUSDC offers near-instant redemptions through a dynamic liquidity buffer. This is the most liquid exit mechanism among DeFi credit products, but it depends on the buffer remaining adequate under stress. What happens during mass redemptions is not fully documented.

Goldfinch provides a cautionary example. FIDU, the token representing Senior Pool positions, is technically transferable. During the protocol’s defaults, it traded at a significant discount to NAV. Holders who needed to exit discovered that a transferable token backed by illiquid, impaired loans is not a liquid position. It became a case study in distressed on-chain debt.

Apollo’s ACRED tokens are transferable across six chains. But fund-level redemption terms govern actual exit. The token moves freely; the capital does not. Adding to the complexity, ACRED positions can now be used as collateral on Morpho for leveraged DeFi strategies, creating liquidity in one sense while adding systemic risk in another.

Centrifuge pools each have their own redemption mechanics, and there is no unified secondary market across the protocol. An investor in one Centrifuge pool cannot assume anything about exit terms from another. Liquidity is a per-pool property, not a protocol-level feature, and it varies significantly depending on the asset originator and tranche structure.

Clearpool’s cpTokens are transferable, but pool liquidity is a function of utilization. When utilization is high (meaning much of the capital is lent out), available exit liquidity is low. The tokens can move; the question is whether anyone is on the other side.

Pareto introduces a different pattern: composable collateralization as an alternative to direct exit. Receipt tokens from Pareto’s Credit Vaults can be posted as collateral on Morpho at up to 77% loan-to-value. This lets holders borrow against their position rather than redeem it. It is a form of liquidity, but not the kind many investors mean when they ask “can I get out?” The underlying credit exposure remains, and the leverage adds liquidation risk on top.

STOKR’s Blockstream Mining Notes trade on the Liquid Network, a Bitcoin sidechain. This is a real secondary market, but a narrow one: the buyer base for tokenized bitcoin mining hashrate exposure on a specialized sidechain is limited. BMN1 returned 103% in BTC terms, which helps, but market depth for these instruments is thin relative to their stated value.

Deploi is structured for a different liquidity model entirely. Settlement runs through Nasdaq CSD Riga, and distribution is handled by Assetera under MiFID II. Deploi’s issuance infrastructure went live on Polygon in May 2026 with its first series (consumer credit notes up to €5M per note) and approximately €100M in pipeline volume. Exit terms will depend on Assetera’s secondary market development and any future exchange listings. No capital has cycled through a full entry-and-exit yet, so the liquidity model remains untested.

The broader traditional private credit market is experiencing its own liquidity stress. In March 2026, Blackstone’s $82 billion BCRED fund gated $3.7 billion in redemption requests. Carlyle’s CTAC saw requests for 15.7% of shares against a cap three times smaller. The idea that tokenization solves private credit’s liquidity problem needs to be measured against the reality that the underlying assets are inherently illiquid. Putting them on-chain does not change that.

What on-chain infrastructure can change is the friction around illiquidity. Traditional private credit locks investors into quarterly redemption windows, opaque gating decisions, and manual transfer processes that can take weeks. On-chain settlement is near-instant. NAV marks can be published transparently rather than disclosed quarterly. Fractional positions can broaden the secondary buyer base beyond the handful of institutional desks that currently trade private credit stakes. Programmable redemption mechanics can replace discretionary gating with rules that investors can evaluate before they commit capital. None of this makes a five-year loan liquid. But it can make illiquid positions easier to price, easier to transfer, and easier to exit at fair value when a willing buyer exists. The gap between “illiquid” and “trapped” is where tokenization’s liquidity case is strongest, and much of the sector has not yet built the infrastructure to prove it.

Operations: Who is actually running these products?

Operational maturity varies as widely as structure.

Figure is a publicly traded company with an integrated technology stack, SEC registration, and institutional banking counterparties. Whatever questions exist about its business model (and a critical short report from Morpheus Research has raised several), the operational infrastructure is real and examinable.

Maple’s operational quality depends on its Pool Delegates. The Orthogonal Trading default in December 2022 was fundamentally an operational failure: M11 Credit, the relevant Pool Delegate, did not detect that Orthogonal was operating while insolvent and misrepresenting its FTX-related losses. Maple’s post-2022 pivot to overcollateralized lending reduces borrower-level credit risk, but the Delegate-dependency remains.

Centrifuge’s operational model distributes risk across individual pool issuers. Each SPV has its own asset originator. This means operational quality is not a protocol-level property; it varies by pool. An investor in a Centrifuge pool backed by a strong trade finance originator has a different operational risk profile than one backed by a weaker consumer lender.

Goldfinch’s legacy defaults exposed the difficulty of monitoring unsecured borrowers in emerging markets from a DeFi protocol. Tugende misrouted funds in violation of its loan agreement. Stratos was simultaneously a borrower and an equity investor in Warbler Labs, the company that built Goldfinch. These are not abstract operational risks. They are the kind of counterparty failures that any credit underwriter should be testing for.

Apollo/ACRED benefits from Apollo’s $700 billion+ AUM and institutional infrastructure. The operational risk here is not in the credit management. It is in the DeFi composability layer. When ACRED tokens are leveraged on Morpho and bridged across chains via Wormhole, the operational dependency chain extends well beyond Apollo’s institutional controls.

Securitize is not a credit product, but it is an operational dependency for several of them. It serves as transfer agent, broker-dealer, ATS operator, investment advisor, and fund administrator across products including Apollo’s ACRED and BlackRock’s BUIDL. If Securitize experienced a regulatory challenge, a technology failure, or an operational disruption, the impact would not be limited to one product. It would ripple across multiple large tokenized fund offerings simultaneously. This is not a criticism of Securitize’s quality (its five SEC registrations are unmatched in the sector), but a concentration risk that allocators should understand.

Clearpool relies on market discipline rather than centralized underwriting. Single-borrower pools mean each pool carries borrower-specific operational risk, and there is no protocol-level credit committee or risk management function. The model works when borrowers are transparent and markets are attentive. Whether that holds under stress is an open question.

STOKR’s operational model depends on Blockstream as the mining operator for its hashrate-backed notes. Mining operations are concentrated in Georgia, Montreal, and Texas. This is infrastructure risk, not credit risk: equipment failures, energy cost spikes, or regulatory changes in those jurisdictions affect the product directly. The operational chain is narrow and geographically concentrated in a way that many private credit products are not.

Deploi’s operational model is built around EU regulatory infrastructure. Assetera handles distribution under MiFID II, Nasdaq CSD Riga handles settlement, and Deploi retains 5% originator risk per EU Securitisation Regulation. The framework is well-defined on paper, but the operational chain is untested. The first series went live in May 2026, and no capital has cycled through the full origination-to-redemption process yet.

Credit Coop has facilitated $1.7 billion in cumulative volume, but 94% of that ($1.6 billion) comes from a single borrower: Rain. The next largest borrower, Karta, accounts for $47.1 million. This is extreme concentration risk. The platform’s operational model converts future cash flows into smart contract collateral, which is structurally innovative, but the practical reality is that Credit Coop’s track record is largely a Rain track record.

A recurring name across the sector illustrates a subtler operational risk. M11 Credit was the Pool Delegate that failed to detect Orthogonal Trading’s insolvency during Maple’s 2022 default. M11 Credit now manages Pareto’s largest Credit Vault, the $135.3 million FalconX facility. This is not necessarily disqualifying (credit professionals operate across multiple platforms), but allocators should understand the counterparty overlap across products that present themselves as independent.

The operational picture gets more complex when composability is factored in. Many of these products are described as “on-chain,” but the operational chain behind a tokenized credit position can involve more intermediaries than its traditional equivalent. A single leveraged position against a tokenized credit fund might depend on: the fund manager for credit decisions, a transfer agent for token issuance and compliance, an oracle for NAV pricing, a cross-chain bridge for token movement, a DeFi lending protocol for leverage, a liquidation engine for risk management, and a KYC verification layer for regulatory compliance. In traditional private credit, an investor holds a fund position through a custodian. The dependency chain is short and well-understood. On-chain composability does not eliminate intermediaries. It replaces familiar ones with new ones, and it connects them in ways that make operational failures harder to isolate. When Midas’s mF-ONE experienced its NAV decline, the operational cascade ran from Fasanara’s credit exposure through Midas’s valuation mechanics into Morpho’s liquidation thresholds. Each layer operated as designed, but the combined effect was a contagion path that no single operator controlled.

The intermediary chain is long today partly because the infrastructure is early, and partly because it needs to be. Tokenized credit is tied to real-world assets with real legal claims. A misconfigured smart contract, a faulty oracle price feed, or a bridge exploit can create losses that cannot be reversed the way a failed wire transfer can be recalled. Many of the intermediaries in the current stack (transfer agents, compliance layers, custodians, oracle providers) are not inefficiencies waiting to be automated away. They are guardrails for a system where on-chain errors have off-chain consequences. Removing them before the underlying infrastructure is provably reliable would expose real capital to risks that traditional finance solved decades ago.

That said, as on-chain standards mature, many of these functions can consolidate. Programmable compliance can embed KYC and transfer restrictions directly into the token, reducing reliance on separate verification layers. Smart contracts can automate NAV calculations, redemption processing, and reporting that currently require dedicated service providers. On-chain transparency makes real-time auditability possible, replacing quarterly reporting cycles with continuous visibility into fund positions, counterparty exposures, and collateral health. Interoperable token standards and cross-chain messaging protocols like Chainlink CCIP can reduce the need for bespoke bridge integrations. The long-term case is that on-chain credit infrastructure compresses the operational stack: fewer intermediaries, faster settlement, and lower costs per transaction. That case is directionally credible. But the sector is in a middle stage where the training wheels are load-bearing. The intermediary chain will compress over time, but the sequence matters: proving reliability first, then removing guardrails, not the reverse.

Distribution and composability: Who can access these products, and what can they do with them?

The buyer base is splitting into four lanes that do not always communicate with each other.

Institutional access(Figure Connect, Apollo/ACRED for qualified purchasers, Credix) serves regulated capital through KYC’d channels. This is the fastest-growing lane and the one that most closely resembles traditional credit distribution. Credix, often grouped with permissionless DeFi protocols, actually requires full KYC/KYB through Securitize and accreditation verification (income, net worth, or professional license). It is functionally an institutional product.

Permissionless DeFi access (Maple’s syrupUSDC, Clearpool) allows deposit into credit pools without identity verification, but “permissionless” requires qualification. Maple’s syrupUSDC blocks US persons and residents of 33 other jurisdictions, including Australia, Venezuela, and the standard sanctions list. These restrictions are enforced at the interface and terms-of-service level, not at the smart contract level. The contracts themselves remain open. Clearpool restricts 20 jurisdictions, primarily sanctions-list countries (Cuba, Iran, North Korea, Russia, China, etc.), but does not block the United States. The practical access picture is messier than “anyone can deposit.”

Hybrid access (Centrifuge’s institutional partnerships, Goldfinch Prime) tries to combine institutional credit quality with on-chain distribution.

Jurisdiction-regulated access is emerging as a distinct fourth lane, and it is not limited to one region. In the EU, Deploi distributes through Assetera under MiFID II with settlement via Nasdaq CSD Riga, and STOKR operates as a Luxembourg VASP issuing mining notes on the Liquid Network. In Asia, DiGiFT holds dual licenses from the Monetary Authority of Singapore and the Hong Kong SFC, providing end-to-end tokenization infrastructure for institutional credit products like Flow Capital’s $150 million fund. Singapore’s MAS applies a “same activity, same risk, same regulatory outcome” principle to tokenized securities, treating them under existing securities law rather than creating new categories. In the UAE, both ADGM and VARA have introduced frameworks for tokenized securities and real-world assets, with ADGM operating under English common law and attracting institutional tokenization projects. Switzerland’s DLT Act provides a legal basis for tokenized securities, and FINMA’s 2026 custody guidance addresses the infrastructure requirements for holding crypto-based assets. These jurisdictions represent a growing alternative to both the US institutional channel and permissionless DeFi: regulated, jurisdiction-specific, and designed for cross-border reach within their respective legal frameworks.

Credit Coop distributes through direct borrower relationships rather than investor-facing channels. Its 11 borrowers access structured lending facilities on-chain, converting future cash flows to smart contract collateral. Distribution here means borrower acquisition, not investor access, which is a fundamentally different go-to-market from the rest of the sector.

The most interesting tension is the dual-access model that several products are adopting. Maple runs institutional KYC’d pools alongside syrupUSDC, which is permissionless for non-US, non-restricted participants. Goldfinch shifted from permissionless emerging-market lending to institutional credit aggregation via Prime. Apollo’s ACRED is a qualified-purchaser product that can be leveraged permissionlessly on Morpho.

The gap between interface restrictions and smart contract access matters. syrupUSDC’s terms of service prohibit US persons. But the smart contracts are deployed on public blockchains and do not enforce geographic restrictions. A US person interacting directly with the contract, bypassing the Maple interface, is violating terms of service but not encountering a technical barrier. This distinction is central to the regulatory risk question: is the issuer responsible for access controls that exist only at the interface layer? The SEC’s January 2026 guidance did not address this directly, but the logic of “changing the format does not change whether securities laws apply” suggests that interface-level restrictions may not satisfy regulatory expectations when the underlying contracts remain open.

But distribution is only half the story. Composability is changing what happens after purchase.

In traditional private credit, an investor buys a fund position and holds it. The position sits in a custody account. It does not interact with anything else. The investor earns yield and waits for redemption windows.

On-chain, tokenized credit positions become building blocks. They can be posted as collateral on lending protocols, bridged across chains, bundled into yield strategies, or looped for leverage. This is composability: the ability for a tokenized asset to plug into other financial protocols without requiring new agreements, intermediaries, or permissions.

The composability layer is developing fast, and the numbers are already material. An estimated ~$700 million in leveraged positions against tokenized private credit now sits across Morpho, Aave, and Kamino. That is not a theoretical risk. It is a live exposure.

Maple’s syrupUSDC is integrated as collateral on Aave (across Ethereum, Base, and Plasma instances), Morpho, Drift (with a $50 million cap), Fluid, Sky, Kamino, and Jupiter. Over $2 billion in syrupUSDC is in circulation, and looping vault strategies have scaled to hundreds of millions in inflows. Maple’s total loan originations now exceed $20 billion cumulative, with weekly volume climbing toward $2.3 billion.

Apollo’s ACRED is available as collateral on Morpho across Ethereum, Polygon, and OP Mainnet. Gauntlet’s quantitative engine manages risk optimization. The strategy is explicit: deposit ACRED, borrow USDC against it, purchase more ACRED, repeat. The loop captures the spread between ACRED’s yield and the stablecoin borrowing cost.

Pareto’s FalconX Credit Vault receipt tokens can be collateralized on Morpho at up to 77% loan-to-value, with KYC handled through Keyring Network’s zero-knowledge verification. Midas’s mF-ONE scaled to roughly $190 million on Morpho within three months of launch, though current total asset value is closer to $71 million. Aave’s Horizon, its RWA-focused money market, has $176 million in loans outstanding. Hastra’s PRIME token makes Figure’s lending yields composable on Provenance and Solana.

This composability is genuinely new. Traditional private credit has no equivalent. The ability to take a tokenized Apollo credit fund position and use it as collateral in a DeFi lending market within minutes, without a phone call or a custody transfer, is a structural capability that did not exist two years ago.

But composability also means that the risk surface of a tokenized credit product extends far beyond the product itself. When ACRED is leveraged on Morpho, the risk is no longer just Apollo’s credit underwriting. It includes Morpho’s liquidation mechanics, Wormhole’s bridge security, oracle pricing accuracy, and the behavior of other participants in the same vault. The product issuer has no control over these downstream uses. The investor may not fully understand the dependency chain they have created.

The first real test case arrived in early 2026. First Brands Group filed for bankruptcy, and exposure tied to First Brands was marked down inside a Fasanara credit strategy. Midas’s mF-ONE, a tokenized certificate tracking Fasanara’s F-ONE strategy, saw its NAV slip roughly 2%. On its own, a 2% NAV decline is unremarkable. But mF-ONE was being used as collateral on Morpho. Leveraged positions saw their collateral value decline, pushing some closer to liquidation thresholds. Lenders ultimately avoided losses. The contagion path was clear: a traditional corporate bankruptcy flowed through a fund valuation into a tokenized certificate into a DeFi lending protocol, affecting participants who may not have understood the full dependency chain.

The scale was manageable. The mechanism was not. And the same composability infrastructure is now being built around products with billions in AUM.

Where distribution and composability are heading. The current landscape is fractured: institutional products that DeFi cannot touch, DeFi products that institutions will not touch, and a narrow overlap where both coexist uneasily. But the trajectory points toward convergence. Programmable compliance (embedding KYC, accreditation, and transfer restrictions at the token level rather than the interface level) could allow a single product to serve institutional and DeFi capital simultaneously without the regulatory ambiguity of today’s interface-only restrictions. Lower minimum investments, enabled by fractional tokenization and pooled vault structures, could open private credit to a broader allocator base that currently lacks access: family offices, smaller institutions, and eventually qualified retail. Geographic reach expands naturally with on-chain distribution, since a MiFID II-compliant token issued in Latvia and settled through Nasdaq CSD can reach investors across the EU without the bilateral custody agreements that traditional fund distribution requires. And composability, if the contagion risks are properly managed through better risk parameters and circuit breakers, could create capital efficiency that traditional private credit cannot match: the ability to deploy, hedge, and rebalance credit exposure programmatically rather than through quarterly fund operations. The version of this sector that works is one where on-chain infrastructure solves private credit’s real distribution problems (high minimums, geographic friction, opaque reporting, illiquid secondary markets) without importing new systemic risks faster than the market can price them. That balance has not been struck yet. But the building blocks are closer to real than they were a year ago.

What the sector gets right

This is not a sector running on narrative alone. Several things are genuinely working.

Figure has proven that blockchain can reduce real friction in lending infrastructure. Replacing MERS (the Mortgage Electronic Registration System, the centralized database that tracks loan ownership across the US mortgage market) with DART (Digital Asset Registration Technologies, a blockchain-native lien registry on Provenance), settling loans on-chain, and enabling institutional secondary trading through Figure Connect is not a narrative. It is a functioning system that moves $600 million monthly. Whether or not one agrees with Figure’s broader business strategy, the infrastructure proof point is real.

Centrifuge’s SPV-per-pool securitization structure is sound. Legal isolation, tranche subordination, and on-chain transparency combine to produce something that traditional securitization markets would recognize, with better auditability. This is the structural approach most likely to survive serious institutional scrutiny.

The institutional convergence is genuine. Apollo, Janus Henderson, and Coinbase AM are not experimenting with tokenization as a PR exercise. They are building products, deploying capital, and integrating with DeFi infrastructure. Securitize, with $4.6 billion in AUM on its platform and five SEC registrations, has built the regulated infrastructure that makes institutional tokenization viable, not just aspirational. The tokenized fund wrapper model, whatever its current limitations, represents a credible path toward institutional-grade credit on-chain.

Maple’s recovery demonstrates that DeFi credit protocols can survive catastrophic defaults and rebuild. After losing $36 million to Orthogonal Trading’s insolvency in December 2022, Maple restructured its risk model, pivoted to overcollateralized lending, and rebuilt to $20 billion in cumulative originations. syrupUSDC now holds over $1.3 billion. This is not a guarantee against future failures, but it is evidence that the model can absorb a major shock and come back stronger. Few traditional credit platforms have been stress-tested this publicly and recovered this completely.

On-chain transparency is a genuine improvement over traditional private credit reporting. In traditional private credit, investors receive quarterly NAV updates, often weeks after the reporting period ends. On-chain products can offer real-time visibility into collateral composition, utilization rates, borrower payments, and pool health. Centrifuge’s pool-level transparency, Pareto’s on-chain facility monitoring, and Morpho’s real-time collateral tracking represent a step change in how investors can monitor credit exposure. This does not eliminate credit risk, but it compresses the information gap between what the manager knows and what the investor can see.

Multiple jurisdictions are proving that compliant tokenized credit issuance is possible. In the EU, Deploi’s structure (MiFID II distribution through Assetera, Nasdaq CSD settlement, 5% originator risk retention per EU Securitisation Regulation, ISIN allocation) demonstrates that tokenized private credit can operate within an existing regulatory framework rather than around it. STOKR’s Luxembourg VASP status shows a similar pattern. In Singapore, DiGiFT’s dual MAS and Hong Kong SFC licenses provide regulated infrastructure for institutional tokenized credit. The UAE’s ADGM framework and Switzerland’s DLT Act offer additional paths. These are not permissionless experiments hoping regulators will not notice. They are products built for regulatory compatibility from the start, across multiple legal traditions, and they represent a growing alternative to the regulatory ambiguity that characterizes much of the US market.

The sector has been stress-tested and the system processed the stress. Goldfinch’s defaults happened. Borrowers misrouted funds, conflicts of interest surfaced, and approximately $18 million in losses were absorbed. Midas’s mF-ONE experienced a NAV decline from a real corporate bankruptcy, and leveraged positions approached liquidation thresholds without causing cascading losses. These are not signs of a healthy sector in the traditional sense, but they are signs of a sector that has encountered real-world failure modes and survived them. A sector with no defaults and no stress events would be either too young or too opaque to evaluate. This one has produced enough signal to learn from.

Where the narrative outpaces the structure

The sector tells a compelling story: institutional-grade private credit, on-chain, accessible, composable. Much of that story is directionally true. But in six places, the narrative is running ahead of what the structures actually support.

1. The headline number hides structurally different products classified in structurally different ways. Represented assets ($21.89 billion) are tokens that cannot leave the issuing platform. Distributed assets ($5.14 billion) can be moved and transferred freely. The represented side is 80% Figure’s HELOC token. The distributed side contains DeFi credit protocols, tokenized fund wrappers, hashrate-backed credit, and infrastructure platforms. Each subcategory has different risk profiles, counterparty dependencies, and investor experiences. Issuers and allocators who treat the $27.3 billion as a single coherent sector are starting from a flawed premise.

2. Token transferability is not liquidity. As the Liquidity section details, the gap between “this token can move” and “this investor can exit at fair value” is the sector’s most persistent framing problem. Goldfinch’s FIDU discount, Blackstone’s and Carlyle’s gating events, and the utilization-dependent exit mechanics across DeFi protocols all point to the same conclusion: putting illiquid loans on-chain does not make them liquid. On-chain infrastructure can reduce the friction around illiquidity, but the underlying assets remain inherently hard to exit.

3. “Permissionless” access to institutional credit is neither fully permissionless nor fully restricted, and that ambiguity creates unpriced regulatory risk. Maple’s syrupUSDC blocks US persons and 33 other jurisdictions at the interface level, but the underlying smart contracts enforce no geographic restrictions. Clearpool restricts 20 sanctioned jurisdictions but not the US. Credix requires full KYC and accreditation, making it institutional despite its DeFi framing. Apollo’s ACRED is a qualified-purchaser product that can be leveraged on Morpho without the qualified-purchaser restrictions that apply to the underlying fund. The access picture across the sector is a patchwork of interface-level restrictions, terms-of-service prohibitions, and open smart contracts. The SEC’s January 2026 guidance was clear: “Changing the format of a security does not change whether, or how, the federal securities laws apply.” Whether interface-level restrictions satisfy regulatory expectations when the contracts remain open is an untested question. The sector is moving fast. The regulatory framework has not caught up.

4. The holder base is far thinner than the TVL numbers suggest. ACRED holds $133 million across 63 wallets. syrupUSDC holds $1.2 billion across roughly 3,000 wallets on Ethereum. The sector reports 186,091 total holders, but that number spans 2,380 assets, many of which are small or legacy positions. The products driving the sector narrative are held by a remarkably concentrated group. This matters because concentrated holder bases amplify liquidity risk: a small number of large redemptions can stress exit mechanisms that appear adequate under normal conditions. It also means the “adoption” story is really an institutional positioning story. That is not necessarily a weakness, but it should be described honestly.

5. Yields are quoted prominently. Fee stacks are not. Across much of the sector, the cost path from gross yield to net investor outcome cannot be fully traced from public materials. Fund wrappers like ACRED layer management fees, performance fees, tokenization platform fees, and potentially bridge or DeFi protocol fees, but the aggregate cost is not disclosed in a single place. DeFi protocols vary widely in fee transparency. When a product quotes a 10% yield but the fee stack consumes 200-300 basis points that investors cannot easily identify, the gap between stated return and actual return is a narrative problem. Allocators evaluating products in this sector should be asking for the full gross-to-net waterfall.

6. DeFi leverage on private credit is a contagion channel, not a feature. An estimated ~$700 million in leveraged positions against tokenized private credit sits across Morpho, Aave, and Kamino. The mF-ONE / Morpho episode demonstrated the mechanism: a traditional corporate bankruptcy flowed through a fund valuation into a tokenized certificate into DeFi liquidation thresholds. Lenders avoided losses. But the composable risk chain is growing faster than the frameworks to manage it. When looped strategies manufacture 15-20% returns from 7% underlying credit, the difference is leverage, and that leverage is governed by oracle price feeds and automated liquidation mechanics that the product issuer does not control.

What this means

For issuers in the tokenized private credit space: the competitive landscape is more segmented than headline coverage suggests. Positioning against “the sector” is less useful than positioning within a specific subcategory. Allocators who understand the distinction between represented and distributed value, and the subcategories within each, will evaluate products accordingly. Issuers who cannot clearly articulate where their product sits and why their structure fits that subcategory will face harder questions earlier.

Fee transparency is an underused competitive advantage. In a sector where many products do not disclose the full gross-to-net waterfall, an issuer that publishes clear, complete fee documentation signals confidence in its economics. The bar is low enough that basic transparency becomes a differentiator.

Liquidity honesty matters more than liquidity claims. Issuers who overstate exit mechanics will lose credibility the moment an allocator tests the claim. Issuers who describe liquidity constraints clearly, explain what they are building toward, and set realistic expectations will earn more trust than those who conflate token transferability with investor liquidity.

Jurisdiction-regulated paths offer a real differentiation opportunity for issuers willing to build within existing frameworks. Products structured under MiFID II, MAS securities law, ADGM’s tokenized asset framework, or Switzerland’s DLT Act face a higher compliance burden upfront but avoid the regulatory ambiguity that hangs over permissionless models. As institutional capital grows more cautious about unresolved regulatory questions, compliant-by-design products become easier to underwrite regardless of the specific jurisdiction.

Composability adds a new layer of positioning complexity. If your product is designed to be used as collateral, bridged across chains, or integrated into DeFi strategies, that is a distribution advantage and a risk surface expansion at the same time. Issuers need to decide whether composability is a feature they are actively building for or a downstream use they cannot control. Either way, allocators will ask about it.

For allocators evaluating products in this space: the five framework dimensions (structure, economics, operations, distribution, and liquidity) still apply, but composability means they do not stop at the product boundary.

Structure determines what you are actually underwriting, but composability can change the effective structure of your position after purchase. An ACRED token held in a wallet is exposure to Apollo’s credit fund. The same token leveraged 3x on Morpho is a different risk profile entirely. Economics determines whether the yield justifies the risk, but when a strategy quotes 15-20% returns from 7% underlying credit, the difference is leverage, not alpha. Ask where yield comes from before evaluating whether it compensates for the risk.

Liquidity, operations, and distribution each extend beyond the product in a composable environment. Your exit may depend on liquidity conditions in protocols you did not choose. Your operational dependency chain may include oracles, bridges, and liquidation engines the issuer does not control. And the product’s access model may expose it to participants the issuer did not intend to reach, creating regulatory risk that neither party has priced.

The practical implication: evaluate the product on its own merits, then evaluate what composability does to the position you actually hold. These are two different assessments, and much of the coverage treats them as one.

For the sector as a whole: the institutional convergence is real and accelerating. But the gap between how products are structured, how they are measured, how they are accessed, and how they behave in composable environments is wide enough to create serious problems if it is not closed deliberately.

The sector does not need more capital. It needs better shared definitions, honest measurement, and risk frameworks that account for what composability actually does to private credit. The products that get this right will attract the most durable capital. The ones that do not will discover, as Goldfinch and Maple already have, that the market’s memory for structural failures is long.

“Tokenized private credit” will always contain structural nuance. A Figure HELOC, a Maple institutional loan, and a tokenized Apollo fund share a label but not a risk profile. That ambiguity is not a failure of the sector. It reflects a category still early enough that origination, underwriting, servicing, and settlement are migrating on-chain at different speeds across different products. As more of that stack moves on-chain, the definitions will sharpen. Until then, the distinction between what the label promises and what each product actually delivers is the most important thing issuers and allocators can get right.

Sources and methodology