Let the Safest Assets Bear Losses First? A Counterintuitive New Approach to On-Chain Distribution

Allow assets to be used as collateral in as many places as possible while minimizing the risk for lenders.

Written by: Blockworks Advisory

Compiled by: AididiaoJP, Foresight News

In most structured products, the safest assets are at the front of the repayment order. Priority holders receive lower returns in exchange for principal protection; the riskiest portion of funds bears losses first. This article argues the opposite: placing tokenized government bonds in a subordinate asset layer, beneath higher-yielding credit assets. It sounds counterintuitive, but this may be the most promising path for regulated issuers to open channels on-chain.

The reason lies in the changing customer acquisition methods. Early DeFi protocols bought liquidity with their own tokens—using APRs and rewards above market rates to build lending markets and DEX pools, and then gradually reducing rewards once organic rates could sustain the market.

New-generation on-chain issuers cannot replicate this strategy. Whether it’s tokenized government bond funds or private credit instruments, they lack governance tokens to issue; regulated issuers cannot attract holders through subsidies. They can only open channels through product structure—designing assets in a way that other protocols are willing to borrow, trade, and repackage.

Assets like OnRe's ONyc and Figure's PRIME have over 30% of their market value related to leverage in the lending market.

In contrast, the layered structure specifically designed to allocate risk among holders is pitifully small: layered positions account for less than 1% of lending deposits. Circular lending and layering are nearly the same in risk allocation, yet the scale differs significantly, indicating a design issue.

Circular Lending: The Proven Method

Circular lending is recursive borrowing. You deposit interest-bearing assets as collateral, borrow stablecoins, and then buy more of that asset, repeating the process. Each round of circulation increases demand for that asset, which is why lending integration can account for such a large proportion of RWA supply.

One cycle distributes risk between two parties. The lender provides stablecoins and only incurs losses if the collateral falls below the liquidation threshold and the liquidator does not close the position in time. The borrower bears all risks above this line: price fluctuations of the collateral and the risk of stablecoin borrowing rates exceeding asset yields.

Liquidation does not happen quickly. The premise of liquidation is that someone can quickly sell the collateral. Redemption of tokenized government bonds or private credit funds can take days to settle, sometimes with notification periods and limits. Liquidators receiving RWA collateral may not be able to sell it at face value, so during periods of stress, on-chain prices may fall below net asset value. Liquidating at this discount makes the perceived safety margin, calculated at net asset value, actually thinner. The lending market's response is to give RWA a lower loan-to-value ratio, which limits the leverage assets can bear and the scale of channels that circular lending can bring.

Borrowing rates fluctuate. As long as asset yields exceed stablecoin borrowing costs, circular lending is profitable. When the market leverages in other directions, stablecoin borrowing rates soar; they rise fastest when many circular loans are closed simultaneously. An asset with a 4% yield can still be profitable at a 3% borrowing rate; once utilization pushes the rate to 8%, the spread turns negative.

Yield Splitting: Making Circulation More Predictable

Platforms like Pendle address the interest rate side of trading. They split interest-bearing assets into two tokens: principal tokens lock in fixed returns until maturity, while yield tokens carry floating rates. Principal tokens are more suitable as circular collateral because their value converges to face value on a known date—this is also why Pendle's usage is so closely tied to the lending market.

Borrowers using principal tokens for circular lending know exactly how much they can earn on that side. The borrowing side remains floating unless the borrower also fixes it; and at each maturity date, they must roll it over again.

Layering: Same Roles, Tiny Scale

Layering splits the risk and return of a single asset into two tokens. Priority tokens receive lower returns and do not bear losses within a certain threshold (the coverage). Subordinated tokens receive higher returns and bear losses first.

The risk allocation method of circular lending is similar. The lender acts as the priority, receiving moderate returns and only losing in case of liquidation failure. The borrower acts as the subordinate, enjoying leveraged returns while also bearing the first loss before liquidation.

Circular lending is larger than layering for two reasons:

First is maturity. Circular lending has years of accumulated infrastructure, liquidator networks, and risk parameters behind it.

Second is correlation. In circular lending, the assets held by the lender are not the same as those protecting it, and they are safer—you are borrowing RWA with USDC. In the same asset structure, priority and subordinate share portions of the same asset pool. If the underlying suffers losses that breach coverage, both priority and subordinate lose together. The portion of yield given up in exchange for protection is not worth it. Subordinated tokens also have their issues as collateral: they correspond to inherently illiquid assets, and they are the most volatile and hardest to liquidate, so the lending market has little reason to accept them. Under the same asset structure, both priority and subordinate are difficult to promote.

Reasons for Non-Correlated Layering

Cross-asset layering seeks inferior funds from another type of risk pool, bypassing the correlation issue.

Using a higher-yielding credit asset (like Maple's private credit product) as priority and tokenized government bonds as subordinate. Government bond holders provide first-loss coverage for the credit asset, while the priority uses the excess yield to pay this fee.

A simple example can clarify the economic account. A funding pool has $80 million in credit products with a 10% yield; there is also $20 million in government bonds with a 4% yield. Priority holders only take 8%, not 10%, effectively giving up $1.6 million annually to the subordinate. The subordinate earns $800,000 from government bond yields, plus this $1.6 million, resulting in a 12% return on the $20 million. In exchange, the subordinate absorbs the first $20 million of losses from the credit portion, providing 25% coverage to the priority.

In this structure, government bond holders are selling credit protection for priority assets, and this protection is fully collateralized by government bonds. In traditional markets, this is called financing credit default swaps or credit-linked notes. On-chain, it helps open channels for both sides.

Priority becomes a better collateral. A credit token with 25% first-loss protection can achieve a higher loan-to-value ratio than the original asset, allowing for more leverage and access to more lending markets.

Subordinate becomes a yield-enhanced government bond product. As long as no credit events occur, its value relies on government bonds, which are stable in price and quick to redeem. This characteristic makes it more suitable for the lending market than subordinates under the same asset structure.

Both issuers benefit. The credit issuer obtains a better version of an asset to borrow against, while the government bond issuer gets a higher-yielding product based on their fund. Neither side needs to issue tokens for incentives.

To Be Established, Three Conditions Must Be Met

The structures mentioned above are based on one assumption: when positions go wrong, someone must be able to exit quickly enough to prevent losses for the protected party. For cross-asset layering, this assumption must meet three conditions.

Subordinates must maintain liquidity under stress scenarios. The key is whether government bonds can be redeemed or sold at face value within the week when credit assets start recording losses. Government bonds are the most likely assets to pass this test—during a credit crisis, investors typically buy government bonds, not sell them.

Coverage pricing must be realistic. The premium for subordinates must reflect the expected losses and severity of losses for priority assets. If actual losses exceed coverage, priority is re-exposed; if the premium is too thin, no one will take the subordinate.

Losses must be recorded on-chain in a timely manner. Whether the repayment order can proceed depends on whether losses for priority assets can be recorded promptly. If credit products report net asset values monthly or if there are delays in confirming defaults, subordinates may appear normal on-chain after coverage has been exhausted, leading the lending market for subordinate tokens to provide incorrect pricing.

Conclusion

Relying on incentives to open channels essentially means paying people to participate. Relying on structure to open channels makes assets easier to hold by other protocols. Circular lending is the proven method: effective, large in scale, and it represents the current standard for RWA integration. Yield splitting makes these circulations more predictable.

Layering and circular lending bear similar risk roles, but the design of the same asset has kept it from scaling. Regulated issuers without tokens need a structure that allows assets to be used as collateral in as many places as possible while minimizing the risk for lenders. Placing government bonds in a subordinate position answers this question for both credit issuers and government bond issuers—provided that subordinates maintain liquidity, coverage pricing is realistic, and losses are recorded on-chain in a timely manner.