Centrifuge (CFG): Bridging Real-World Assets to DeFi

Introduction

Centrifuge holds about $1.635B in total value locked, and 95.6% of it sits in two funds run by the same asset manager (DefiLlama and rwa.xyz, 2026-08-24). That single fact is the most useful thing to know about the protocol in 2026, because it explains both what worked and what did not. The origination model Centrifuge was built on — tokenizing invoices and receivables as NFT collateral, tranched into senior and junior claims — still exists but no longer carries the balance. What carries the balance is institutional fund tokenization: Janus Henderson's treasury and CLO strategies, New York Life's high-yield bond fund, Apollo's credit strategy. This article covers what Centrifuge actually holds today, how the newer deRWA wrappers work, what the concentration risk means, and why the CFG token's value-capture question has escalated into a proposal to convert it into equity.

What Is Centrifuge and How Did It Become a Leading RWA Protocol?

Centrifuge is infrastructure for issuing tokenized funds and credit products onchain. It began as a securitization protocol for small originators and became, over 2025 and 2026, the plumbing that large asset managers use to put regulated funds on public blockchains.

What Is Centrifuge

The protocol does not originate credit and does not manage the underlying assets. It provides the issuance stack: the vault contracts that take investor deposits, the token standard that represents a position in a fund, the cross-chain routing that puts the same fund on nine networks, and — since 2026 — the registered transfer-agent function that lets it operate inside US securities plumbing. The asset manager runs the portfolio; Centrifuge runs the rails. That division is why a firm like Janus Henderson, with roughly $500B under management, or New York Life Investment Management, with roughly $807B, can issue onchain without becoming a crypto company.

Pool Architecture Origins

The original system was Tinlake, launched on Ethereum in 2021: originators tokenized real-world debt as NFTs encoding the asset's value, maturity and a hash of its legal documentation, locked those NFTs as pool collateral, and the pool issued DROP (senior) and TIN (junior) tranche tokens. In April 2021 a New Silver real-estate renovation pool became the first real-world-asset-backed DeFi loan funded by MakerDAO, establishing the DAO-treasury-into-real-world-credit pattern that still drives most of Centrifuge's liquidity. That architecture worked and is still documented, but it is worth being clear that it is not where the money is now. The migration to V3 moved the protocol off Polkadot onto Ethereum and its layer-2s, and the products that scaled on V3 are institutional funds, not originator pools.

Data current as of August 2026.

How Does Centrifuge's Asset Origination Process Actually Work?

There are two answers, and the honest version distinguishes them: the originator model the protocol was designed around, and the fund-issuance model that now holds almost all the assets.

Asset Origination Process

In the originator model, a business assembles documentation for a real-world debt instrument — an invoice, a mortgage, a trade receivable — and the protocol mints an NFT encoding the principal, maturity, rate and a cryptographic hash of the legal documents. That NFT moves into a pool as collateral, the pool validates it against eligibility parameters such as asset type and per-borrower concentration, and capital is drawn against it up to face value. Repayments flow back through the pool contract and distribute according to the tranche waterfall. Every step is onchain, so investors can see pool composition and outstanding exposure in real time.

How Fund Issuance Works Now

The institutional path is different and simpler. An asset manager already runs a portfolio off-chain under conventional fund law, with a custodian, an administrator and in some cases a credit rating. Centrifuge issues a token representing a share of that fund, handles subscriptions and redemptions through an asynchronous vault, and distributes the token across chains. There are no NFTs, no tranches and no originator underwriting, because the risk work happens inside a regulated fund structure rather than inside the protocol. This is the model that scaled: JTRSY holds $881.3M and JAAA $691.9M, together 95.6% of distributed assets, with Apollo's ACRDX third at $43.4M and everything else combined under 2% (rwa.xyz, 2026-07-29).

What Is the Senior/Junior Tranche Structure and Why Does It Matter?

The tranche structure matters less than it used to, and understanding why is more useful than memorising the mechanics.

Senior Tranche (DROP)

In a tranched pool, the senior tranche has priority claim on all cash flows: senior holders are repaid before anything reaches the junior tier. A minimum subordination ratio requires that a set percentage of pool capital stays in the junior tranche; if defaults or redemptions erode the junior tier below that floor, new senior investment and new originations pause until it is restored. That is a real structural buffer and it did what it was designed to do when Centrifuge pools took losses. The mechanism is standard structured finance with a special-purpose vehicle replaced by immutable contracts.

Why the Tranches Matter Less Now

The reason to be careful with tranche yields is that the products holding almost all of Centrifuge's assets are not tranched. JTRSY is a short-duration Treasury fund reporting an annualised yield near 3.24%; JAAA holds AAA-rated CLO tranches, where the credit protection comes from the CLO structure itself rather than from a junior buffer inside a Centrifuge pool. Anyone still quoting 4–8% senior and 10–20% junior for Centrifuge is describing the Tinlake-era originator pools, which now hold under 2% of platform assets, in a rate environment that no longer exists. The relevant credit question for a JTRSY or JAAA holder is the quality of the underlying portfolio and its manager, not where they sit in a pool waterfall.

Data current as of August 2026.

How Does Centrifuge's Sky (MakerDAO) Integration Provide Sticky Liquidity?

Sky's capital is the reason Centrifuge has the scale it has, and it is also the protocol's single largest concentration.

Sky Integration History

The relationship began in April 2021 when New Silver established a real-estate renovation pool on Tinlake backed by MakerDAO credit — the first case of a DAO deploying stablecoin reserves into tokenized real-world credit. MakerDAO scaled that across multiple pools through 2023, became the dominant source of pool liquidity, and carried the integration forward when it rebranded to Sky in 2024. The important structural change was the shift from pool-by-pool governance approvals to programmatic allocation, where governance sets a target and the allocator rebalances toward it as reserves fluctuate. That makes the capital stickier than a series of individual votes would be, because it flows on a formula rather than on repeated proposals.

What Grove Actually Holds

Grove, the allocator deploying on behalf of Sky and its USDS stablecoin, holds roughly $769.8M of JTRSY and $124.8M of JAAA, for about $895M of exposure across Centrifuge products (rwa.xyz via Sandmark, 2026-07-29). Grove's deployment followed extensive risk assessment and due diligence, which is genuine validation of the infrastructure. It also means more than half of Centrifuge's platform value is connected to a single allocator's decisions. Separately, and worth not confusing with the allocation, Centrifuge partnered with Grove Basin on a $1B redemption facility for JTRSY that commits daily liquidity for round-the-clock USDC redemptions — that headline number is committed liquidity, not deployed capital. The next phase of the story is whether distribution broadens, because measured on breadth rather than total value, the platform is narrower than its TVL suggests.

What Are Centrifuge's Flagship Institutional Funds and Who Backs Them?

Four funds are worth knowing, and 2026 added the most institutionally significant of them.

JTRSY and JAAA

JTRSY is the Janus Henderson Anemoy Treasury Fund, holding short-duration US Treasury bills. It crossed $1B in Q1 2026 and stands at roughly $881.3M, carries an S&P fund credit-quality rating, reports a 30-day annualised yield near 3.24%, and requires a $500,000 minimum with a 0.25% management fee. JAAA is the Janus Henderson Anemoy AAA CLO strategy at roughly $691.9M, offering the yield pickup of AAA CLO tranches over Treasuries with CLO diversification rather than sovereign credit as the backstop. In June 2026 Ethena selected Centrifuge as a strategic tokenization partner after a competitive RFP and allocated $250M to JAAA — the first collateral diversification in USDe's history, and a meaningful endorsement from one of the largest allocators in crypto (Centrifuge, 2026-06-09).

NYLIM, Apollo and the 2026 Partnerships

The other 2026 additions matter for what they signal. New York Life Investment Management, managing roughly $807B, tokenized its US High Yield Corporate Bond Strategy on Centrifuge — its first tokenized fund, and one of the first high-yield corporate bond strategies to come onchain. Apollo's diversified credit strategy, ACRDX, is the third-largest product at $43.4M. Coinbase named Centrifuge a preferred tokenization infrastructure and backed it with an equity investment in CFG, bringing the framework to Base; Kraken Institutional and OKX followed with partnerships of their own. A 180-year-old asset manager and a major US exchange both choosing the same infrastructure is a stronger signal about where tokenization is consolidating than any TVL number.

Data current as of August 2026.

How Does Centrifuge V3 Upgrade Enable Cross-Chain RWA Composability?

V3 is the part of the Centrifuge story that has delivered most cleanly, because it solves a problem the institutional funds actually have: a permissioned fund token cannot circulate in DeFi.

V3 Architecture and the deRWA Wrapper

V3's structure is pool, vault, token. The pool is created by an issuer and mapped to off-chain assets; the vault, built on the ERC-7540 asynchronous standard that Centrifuge co-authored, handles deposits, redemptions and yield; the token represents the investor's position. Asynchronous settlement matters because redemption cannot always be instant — the vault queues a request and settles it when liquidity is available. The composability layer is deRWA: it wraps a permissioned fund token and issues a freely transferable ERC-20 carrying the same NAV exposure and yield accrual. That is what lets an asset requiring a $500,000 minimum and a whitelist circulate as ordinary DeFi collateral. deJTRSY, deJAAA and deCRDX all now exist alongside their underlying funds.

Cross-Chain Deployment and Liquidity

The chain footprint is real rather than announced: DefiLlama attributes $1.274B of Centrifuge TVL to Ethereum, $261.5M to Avalanche, $58.13M to Base, $21.33M to Plume, $15.12M to Monad and smaller amounts to Pharos, BSC, Optimism and Arbitrum. deSPXA gained traction on Uniswap and was accepted as collateral on Euler; JAAA became the first asset to support leveraged trading on 3F; deJTRSY and deJAAA went live on Sushi on Stellar and on X Layer as part of a distribution partnership with OKX. The most consequential addition came on 19 August 2026, when Centrifuge added Symbiotic's Liquid Lane across JAAA, JTRSY and NYLIM's HYB, using an onchain request-for-quote marketplace so market makers can fill redemption requests and holders receive USDC immediately while normal fund redemption settles separately (Cointelegraph, 2026-08-19). Exit liquidity, not issuance, was the binding constraint on tokenized funds, and this is a direct attempt to solve it.

How Fast Has Centrifuge Grown and What Risks Should Investors Know?

Growth was genuine and has flattened. Both halves of that sentence matter.

Growth, Honestly Measured

Distributed assets across Centrifuge products grew from $254.5M in July 2025 to about $1.65B in July 2026, placing it eighth among tokenization platforms with roughly 4.5% of the $37B distributed-asset market — behind Securitize, Ondo and Circle, ahead of Ethena and Maple (rwa.xyz via Sandmark, 2026-07-29). That is a genuine six-fold year. But TVL fell 6% quarter over quarter in Q2 2026 and sits around $1.635B, so the compounding has paused even as the partnership announcements accelerated. CFG token holders grew 16% over the same quarter to 10,988, which is a healthier signal than the TVL line and still a small number in absolute terms.

The Three Risks

The first risk is concentration, and it is structural rather than hypothetical: two funds are 95.6% of assets, one asset manager runs both, and one allocator holds roughly $895M of them. A change of mind at Sky governance would move the platform more than any market event. The second is credit risk in the underlying, which tokenization does not remove. Centrifuge's early pools took real losses — the New Silver pool's non-performing loans in 2022 reduced NAV and triggered the default-management process, with junior holders absorbing losses first and senior holders protected, exactly as designed. The lesson was that the contracts enforce the waterfall correctly while credit assessment, legal recovery and collections stay entirely off-chain. The third is that on-chain transparency shows you a loss faster; it does not help you recover it.

How Does the CFG Token Capture Value and What Is Its Price Outlook?

This is where the Centrifuge story gets uncomfortable, and 2026 brought it to a head.

CFG Token Mechanics

CFG trades at $0.1433 with a market cap of $82.7M, roughly 93.9% below its November 2021 high of $2.34, on a circulating supply near 577.2M (CoinPaprika, 2026-08-24). The token grants governance over protocol parameters and was intended to be the value-accrual vehicle for an AUM-based fee. The problem is the gap between fees charged and fees reaching holders. DefiLlama records $59.81M in annualised fees across Centrifuge's chains and $5.44M in annualised protocol revenue — and holders' revenue of exactly zero. A fee-switch proposal was put to governance and failed to reach quorum. Nothing in the current design routes the business to the token.

The Proposal to Convert CFG Into Equity

On 17 August 2026 the team posted CP172, proposing that CFG holders be offered the chance to subscribe for equity in Centrifuge, Inc. at one share per token, with holders of 100,000 CFG or more entered directly on the register of members and smaller holders routed through a trust structure with CoinList (Centrifuge governance forum, 2026-08-17). Participation would be optional; non-participants keep or sell their tokens. The rationale offered is that CFG carries crypto volatility and legacy tokenomics despite strong protocol performance, and that shifting regulatory attitudes plus Centrifuge's own registered transfer-agent status make an equity route credible. Forum responses have pushed back hard, arguing that a well-designed fee share or buyback would deliver more upside than illiquid private equity, and asking why the earlier fee-switch failure was treated as a mandate for restructuring rather than as a turnout problem. Open questions include whether converted shares will have any secondary market, what price the first equity round would be struck at, and how dilution is addressed. This is a live proposal, not a decision.

Who Should Use Centrifuge and Which Products Fit Which Investor?

The products and the token suit different people, and conflating them is the most common mistake readers make with this protocol.

Matching Products to Investors

An institution or DeFi treasury wanting rated, short-duration Treasury exposure onchain with committed exit liquidity maps to JTRSY, subject to the $500,000 minimum. An allocator wanting a credit spread over Treasuries without single-name exposure maps to JAAA, which is the product Ethena chose after an RFP. A DeFi user who cannot clear the minimums but wants the same underlying exposure as collateral maps to the deRWA wrappers, which is the genuinely novel capability here — a permissioned fund circulating as an ordinary ERC-20. And an investor wanting high-yield corporate credit onchain now has NYLIM's HYB, which did not exist before June 2026.

Who Should Not Buy CFG

CFG suits someone making a deliberate bet on the equity of a tokenization infrastructure company, and after CP172 that description is closer to literal than it used to be. It does not suit anyone expecting the token to track platform growth, because the mechanism connecting the two does not exist: $59.81M in fees, $5.44M in revenue, nothing to holders. Buying CFG as a proxy for Centrifuge's institutional momentum has been a losing trade for the entire period during which that momentum was real.

What Are the Key Risks of Using Centrifuge for Real-World Asset Exposure?

Three risks are worth separating, because they hit different holders.

Concentration and Counterparty Risk

For a fund holder, the largest risk is not Centrifuge but the fund. JTRSY's credit risk is US Treasury risk plus manager and administrator risk; JAAA's is AAA CLO risk, which is remote but not zero and behaves badly in exactly the scenarios where everything else does too. Layered on top is platform concentration: if Grove reduces its roughly $895M position, the funds themselves are unaffected but Centrifuge's economics are not. For anyone underwriting the platform rather than the product, that single-allocator dependency is the first thing to model.

Liquidity and Structural Risk

Redemption is the constraint that tokenization does not automatically fix. A tokenized fund still settles at NAV on the fund's schedule, which is why Centrifuge has layered on liquidity arrangements — Wintermute for JTRSY since February 2025, a $1B committed facility with Grove Basin, and Symbiotic's RFQ marketplace across three funds since August 2026. Those help and they are also new, so they have not been tested by a market where every holder wants out at once. The structural risk sits with CFG holders specifically: a token with no revenue claim, a failed fee-switch vote, and a live proposal to convert it into an asset class with no secondary market. Whichever way CP172 resolves, holding CFG through it is a governance bet, not a yield position.

FAQ

References / Sources