Electric Capital: Tracking 501 types of yield-generating RWA assets, we discovered these patterns

Author: Maria Shen

Compiled by: Jiahua, ChainCatcher

We have sorted through 501 sources of real-world yield and cross-referenced them with the current widely adopted RWAs on-chain, leading to the following conclusions:

1. The demand for stablecoins will first bring government bonds on-chain, while the high concentration of government bonds is attracting higher-yielding assets on-chain.

2. Mainstream strategies that make high-yield assets feel "instant" will drag returns toward risk-free interest rate levels.

3. The sources of tokenizable yields are abundant. Seven opportunity clusters reveal which assets can be unlocked on-chain.

4. The biggest challenge lies in distribution. Among the 35 non-stablecoin yield-generating RWAs with a market cap over $50 million, only two have more than 2,000 holders. This is largely a result of intentional product design.

5. Despite friction, on-chain real-world assets will continue to grow. Two-thirds of the stablecoin supply had remained stable even before risk-free rates landed on-chain. Today, the foundational scale has exceeded $280 billion, and structural demand is continuously attracting new RWAs on-chain.

1. Stablecoin demand brings government bonds on-chain, and concentration is attracting high-yield assets on-chain

The supply of stablecoins has historically had an inverse relationship with the federal funds rate—when rates were near zero, the supply exceeded $180 billion, and as rates broke above 5%, the supply began to decline. In January 2024, this pattern turned: even with rates maintaining above 5%, the supply began to grow again, surpassing $280 billion since then.

The supply of stablecoins had an inverse relationship with the federal funds rate until January 2024 when the two diverged. Data source: Electric Capital Stablecoin Dashboard, Federal Reserve Economic Data (FRED).

What changed this pattern was the appearance of risk-free rates on-chain for the first time.¹ From Ondo's initial layout in 2023 to BlackRock and Centrifuge's subsequent expansions in 2024, issuers began to offer government bond yields in the form of on-chain tokens. Stablecoin holders can earn risk-free rates for the first time without leaving the crypto ecosystem.

Yield-generating RWAs are arranged by launch date, with bubble size reflecting total asset value. The largest assets are concentrated in the 3% to 5% yield range. Newer, smaller projects continue to extend above the yield curve. Data source: rwa.xyz 7-day tracking annualized yield and launch date, DeFiLlama Yields API, protocol documentation.

Government bonds are now the largest category of RWA, with a scale of about $11 billion. The same dynamics are bringing other debt instruments on-chain: private credit ($2.8 billion), corporate bonds ($1.9 billion), and non-U.S. government debt ($1.1 billion). The market is highly concentrated: the top 10 assets account for 64% of the total value, with the yields of the 18 largest assets between 3% and 5%.

This concentration is driving stablecoin reserves to seek diversification and attracting higher-yield assets on-chain. However, high-yield assets are more challenging to bring on-chain than government bonds.

2. Every tokenized asset faces a timing mismatch, and every solution comes at the cost of yield

On-chain capital operates 24/7, settles in seconds, and can be redeployed within the same block. Off-chain assets cannot do this. Every tokenized real-world asset has this timing mismatch, and the gap manifests in two dimensions:

Deployment lag. Funds deposited on-chain can only generate returns once they are truly invested in the underlying assets. Private credit loans take weeks to issue, and real estate transactions take months to complete. Until the underlying assets are fully deployed, funds earn nothing.

Redemption lag. When holders want to exit, the underlying assets cannot be liquidated immediately. BUIDL achieves daily settlements through BlackRock, but the feeling of instant redemption is because Circle pre-funded with USDC. ACRED's underlying fund offers quarterly redemption windows.³ Real estate funds require locking in for several years.

Government bonds are the fastest off-chain asset class, but to feel smooth on-chain, various workaround solutions are still needed. The demand for high-yield assets is even greater. The cost amplifies as liquidity decreases: the slower the underlying asset, the more yield is consumed by the workaround solutions.

Currently, there are three strategies bridging this timing gap, but all transfer the costs of illiquidity to those willing to bear them:

A. Park idle funds in liquid assets. Keep a portion of funds in lower-yield but instantly accessible positions. New deposits can start earning immediately while waiting for the underlying asset deployment to complete; redemptions also do not have to wait for short-term loans to mature. There are two specific variants:

• Deposit into DeFi lending protocols. Example: Maple's syrupUSD fund parks unallocated funds in multiple DeFi protocols (including Sky and Aave) as a liquidity buffer.⁴ New deposits can earn from the buffer immediately while waiting for loans to be issued, and withdrawals do not have to wait for short-term loans to mature. The cost is that every dollar in the buffer earns less than the yield directly used for loans, compressing the overall yield of the fund.
• Use government bonds as a buffer. Example: USDai's sUSDai uses government bonds as the base yield through M0 while issuing GPU collateral loans on top of that.⁵

B. Spread yields across the entire fund pool. New deposits are merged into an already operating fund pool, so no individual depositor experiences deployment delays. New funds dilute the overall yield, but for a sufficiently large fund pool, this dilution is negligible. The cost is that existing depositors subsidize new depositors. Example: The vault on Morpho.

C. Obtain liquidity from third parties. Holders do not redeem from the fund but obtain funds from others, and the fund itself does not need to sell any assets. This strategy is only applicable for accelerating redemptions, not for accelerating deployments.

• Pre-fund a stablecoin fund pool that buys tokens at net asset value (NAV). Example: Circle pre-funded up to $20 million USDC in a smart contract for BUIDL,⁶ providing an instant stablecoin exit channel for the largest tokenized government bond product. When holders redeem, BUIDL transfers to Circle, and USDC is simultaneously transferred to the holders. Circle then completes the redemption with BlackRock off-chain. If redemption demand exceeds the pre-funded amount, holders will revert to the fund's standard daily redemption process.
• Market makers buy tokens at NAV. Example: Centrifuge's Anemoy Liquid Network has professional counterparties (Wintermute, Keyrock, Arbelos)⁷ providing instant redemptions for Centrifuge fund tokens, paying holders in stablecoins the same day (up to $125 million, 24/7). Market makers bear the waiting costs: holding tokens, earning yields, and then completing redemptions through the fund's regular slow channel.
• Pledge RWA tokens in DeFi lending markets for borrowing. If tokens are listed as collateral in lending markets, holders can borrow stablecoins without redeeming— even on weekends or outside the fund's redemption window. The fund itself does not participate. This mechanism also supports looping, which will be discussed later as a catalyst for stimulating new demand for RWAs.

The timing mismatch exists because one leg is on-chain, and the other leg is off-chain. Bridging this gap is key to making high-yield assets practically feasible on-chain.

3. The sources of tokenizable yields are extremely rich, with seven clusters revealing the asset space that can be unlocked on-chain

The 34 yield sources already deployed on-chain are concentrated in familiar areas: government bonds, private credit, corporate bonds. The vast majority of the remaining sources have not reached meaningful scale. Seven different barriers lie ahead.

501 sources of yield are distributed across 15 categories (34 sources were excluded for methodological reasons, see the end of the document for explanation). 93% of the concentrated analysis has not formed a significant scale on-chain. Data source: Electric Capital

Download the complete classification table of 501 sources (CSV)

433 off-chain yield sources can be divided into seven groups based on the conditions required to enter on-chain. Detailed information on each source can be found in the CSV file.

This dataset contains categories, source names, descriptions, example tools, and on-chain status for all 501 sources of yield. It is recommended to enhance the CSV using large language models (LLMs) to add additional fields such as yield ranges, risk factors, liquidity status, access requirements, or regulatory jurisdictions.

Macro forces can accelerate the development of specific clusters. The rising insurance losses from climate events are expanding the catastrophe bonds and insurance-linked securities (ILS) market. Parametric models—triggering payouts automatically based on measurable events like wind speed or earthquake magnitude rather than relying on loss assessments—are naturally suited for on-chain settlement. The surge in AI infrastructure spending is creating demand for financing GPU clusters, data centers, and energy contracts on-chain.

These seven clusters cannot encompass yield sources that do not yet exist. It took a century for oil to evolve from a physical commodity to a mature derivatives market. GPU computing may complete a similar transformation within a few years, as its secondary market has been electronic from the start. Subscription revenue from Twitch live streaming did not exist in financial markets five years ago, but today, the infrastructure to tokenize it is already in place. The 467 sources we have sorted out are a lower limit, not an upper limit.

4. The biggest challenge lies in distribution

New types of yields and strategies only matter if they truly reach capital. However, today, distribution channels are extremely limited.

Each bubble represents a yield-generating asset, sized according to on-chain value, filtered to actual earning holders of savings/staking. Data source: rwa.xyz, Etherscan, and other on-chain scanning tools.

Most yield-generating assets are concentrated on the left side of the chart: among the 35 non-stablecoin RWAs with a market cap over $50 million, 33 have fewer than 2,000 holders. This is largely due to structural reasons: BUIDL is a fund limited to accredited investors with a minimum subscription of $5 million,⁸ and about 100 holders are a product design outcome. But for products that could reach a broader user base, this low number reflects the current RWA's reliance on partner distribution channels. Exceptions appear in the upper right corner—such as sUSDe, sDAI, and sUSDS, which have far more holders than other products.

This comparison points to three distribution strategies:

A. Collaborate with deployers and curators.

Large deployers like Sky and Ethena allocate funds to RWAs, with a single deployment decision potentially moving hundreds of millions overnight. Centrifuge's JAAA—a tokenized AAA-rated CLO that reached $743 million at the time of data collection—had almost all AUM from a single allocation by Sky through Grove.⁹ On March 9, 2026, Grove redeemed $327 million in a single transaction, causing JAAA to lose 44% of its value in one day. Even the largest tokenized RWA, BUIDL, has its value highly concentrated in the protocol: the top 10 holders control 98% of the shares, and these holders are Ethena (through USDtb), Ondo (through OUSG), and Sky (through Spark).¹⁰

Curators like Steakhouse and Gauntlet decide which assets can serve as collateral for their Morpho vaults,¹¹ thereby opening distribution channels for thousands of depositors. Losing a curator means closing a distribution channel.

BlackRock and Apollo have the negotiating power to establish these partnerships, while smaller issuers need to compete for positioning opportunities.

B. Control stablecoins. Embed yield-generating assets within stablecoins and then seek distribution channels for the stablecoins themselves.

• Diversified yield stablecoins, like Sky, allocate to multiple tokenized RWAs simultaneously.
• Specialized yield stablecoins, like USDe (Ethena), reUSD (Re), and sUSDai (USD.ai), each embed a single strategy. The choice between diversification and specialization relates to holders' risk tolerance and expectations; single-strategy stablecoins can be a pathway into the market and expand to more yield types over time.

C. Embed into applications that already have users. The Morpho vault curated by Steakhouse powers USDC lending for Coinbase, Gauntlet provides the same service for Wirex business accounts, and Kraken's DeFi Earn routes through the Veda vault curated by Chaos Labs and Sentora.¹² Curators handle risk and on-chain complexity, while applications manage compliance and user acquisition. Entrepreneurs do not directly interact with end users, but yields ultimately reach them. Today, every major case still needs to route through vault curators, making it essentially a variant of the first path—but that does not mean it must always be so.

Two paths have long-term competitiveness: controlling distribution or becoming irreplaceable infrastructure for the distribution layer.

5. RWA will continue to grow

On-chain capital is sticky. When interest rates rise and risk-free rates have not yet widely appeared on-chain, the supply of stablecoins may decline, but it has not collapsed— a significant scale remains even without access to risk-free rates. Today, real-world yields exist on-chain, and the supply has grown from $130 billion to over $280 billion.¹³

Five forces are reinforcing RWA demand:

A. A larger stablecoin base means a broader distribution of yield preferences.

Protocols managing billions in funds have vastly different demands compared to retail holders who park savings there temporarily. Not everyone is satisfied with the 3% yield from government bonds: some want 8% from private credit, while others pursue 15% from leveraged strategies. These demands are currently flowing toward the same batch of products, creating a strong pull for more yield types.

B. The concentration of highly correlated underlying assets and user competition has created a thirst for diversification.

The concentration of low-yield, highly correlated assets is the demand engine attracting the next wave of RWAs on-chain.

BlackRock's BUIDL serves as the backing asset for Ethena's USDtb, the primary underlying asset for Ondo's OUSG, and a direct holding in Sky's Grove protocol.¹⁴ The three largest protocols in the on-chain yield space have correlations with the same government bond fund.

When underlying assets come under pressure, the urgency for diversification increases. The default rate for private credit has risen above 5%, and the redemption event of Grove-JAAA reveals how quickly concentrated positions can unravel.

Risk curators and stablecoins also need to attract more users through differentiated products. If each protocol can only access the same limited products, there can be no differentiated competition. The pressure to compete for users drives the intrinsic need to bring more assets on-chain.

C. Vaults allow curators to absorb duration and liquidity risks that individual assets cannot bear.

Vaults lower the entry barrier for new assets: multi-asset vaults do not require every holding to be fast or highly liquid.

Morpho's scale exceeds $6 billion precisely because curators like Steakhouse and Gauntlet have built portfolios that mix liquid and illiquid positions.

Apollo has signed a partnership agreement to acquire up to 90 million MORPHO tokens within 48 months,¹⁵ indicating its intention to use Morpho as a distribution channel for tokenized credit. Traditional asset management firms view vault infrastructure as a distribution channel.

Vaults are still in a maturation phase. Current yields are partially subsidized by token incentives—a vault promoting a 12% APY may only have an actual organic yield of 4%, with the remainder coming from token rewards. Additionally, curators lack standardized ratings, and depositors have limited tools to assess risk.

D. Tranching and yield decomposition expand the buyer pool for every on-chain asset.

A single yield stream only serves a single risk preference. Tranching breaks it into multiple products, each calibrated for different buyers: priority shares offering 4% yield with priority loss protection are suitable for conservative DAO vaults; subordinated shares offering 12% yield but bearing concentrated downside risk attract high-yield seekers willing to bear losses. The same underlying asset can reach two types of buyers simultaneously. Royco Dawn and Strata are early projects building universal tranching layers.

Pendle does something different. It does not tranche by credit risk but splits any yield-bearing token into principal tokens and yield tokens, allowing holders to lock in fixed rates or speculate on variable yields. When Pendle launches a tokenized RWA, it creates a new demand side for traders and hedgers who never directly purchase that RWA.

The range of underlying assets covered by tranching and yield decomposition remains limited today, but as they mature, every new asset brought on-chain can be sliced into products aimed at multiple buyer groups without needing to re-initiate.

E. Leverage amplifies the demand for every asset already on-chain.

Once tokenized RWAs are listed as collateral in lending markets, holders can engage in looping: deposit RWAs, borrow stablecoins, buy more of the same RWA, and repeat. A tokenized asset with a 5% yield can reach 8% to 10% after deducting borrowing costs under 2 to 3 times leverage. Gauntlet has managed leveraged sACRED strategies on Morpho, while Centrifuge's JAAA has been listed on Aave Horizon, and Resolv has proposed investing up to $100 million in similar strategies. Each loop creates incremental demand for the same asset without needing new initiations.

Looping has structural limitations: on-chain borrowing is instant, but JAAA's subscriptions require T+3 settlement, so the speed of liquidation under pressure does not align. Emerging infrastructures like 3F Labs are dedicated to narrowing the settlement gap between on-chain and off-chain. As this gap narrows and more RWAs qualify as collateral, leverage will continuously amplify demand for every asset already on-chain.

Each force reinforces the others. Once an asset is on-chain, has composability, and qualifies as collateral, vaults, tranching, and leverage looping can each amplify demand beyond what the asset itself could attract individually.

Opportunities abound from infrastructure to application layers

• New assets for tokenization
• Infrastructure that makes new assets easier to bring on-chain (e.g., aggregation platforms that consolidate fragmented markets into investable products)
• Strategies that bridge the deployment and redemption gap without compressing yields, or strategies that completely eliminate the settlement gap between on-chain and off-chain
• Synthetic products that introduce yield exposure on-chain before the underlying assets are fully tokenizable
• New distribution methods that break current dependencies on large deployers and curators, or open new channels to institutional capital
• New infrastructure that structures each RWA to serve more types of users

Every asset brought on-chain makes it easier to introduce the next asset and adds value to the infrastructure supporting it.

Research Methodology

We compiled 501 independent sources of yield covering 15 asset categories by reviewing financial industry databases, academic literature, regulatory documents, and practitioner guides. We excluded 34 sources that depend on non-transferable legal relationships (FDIC-insured time deposits, 401(k) self-loans), specific jurisdictional tax incentives (Low-Income Housing Tax Credits LIHTC, Canadian rate-reset preferred shares, 1031 exchange tools), or lack active markets (bonds from the American Recovery and Reinvestment Act maturing in 2010, trust preferred shares banned after the Dodd-Frank Act). The remaining 467 sources constitute the analysis sample. The complete classification table can be found in the downloadable CSV, with each excluded source accompanied by an explanation for its exclusion.

The determination of on-chain status is as follows: mapping the 727 distributed assets tracked by rwa.xyz to the corresponding yield sources in our classification table, supplemented by on-chain products not tracked by rwa.xyz (DePIN protocols, carbon credit platforms, music royalty tokens). If a source has at least one mapped product with an on-chain scale of $50 million or more, it is marked as "tokenized"; if on-chain products exist but all are below $50 million, it is marked as "partially tokenized"; if no on-chain products exist, it is marked as "not tokenized." Tokenized commodities (gold, silver) and private equity are excluded from the yield source mapping as they do not generate yields. Tokenized stocks are limited to "partially tokenized" due to the lack of automated dividend transmission on-chain.

Each launch date in the timeline chart has been verified by at least one independent source: protocol press releases, on-chain contract deployment timestamps, SEC filings, or financial news reports. Three assets were removed due to unverified launch dates: PKH2 (a mining note from the Liquid Network without a public block explorer), BELIF (a fund with limited media coverage), and CFSRS (a newly established fund in Hong Kong without independent reporting at the time of release). These three remain in the underlying dataset and appear in other charts.