KETJU Research

← The Register

Dollar lending

Curvance

Rejected The evidence weighs against it
Issued
2026-08-14
Last confirmed
2026-08-14
Next check due
2026-11-15
Research basis
Individual research
Chains
Monad · Governed, no freeze

The evidence supports rejection because every tracked Curvance position settles on Monad, which this registry has not approved, and because no named unleveraged market has enough operating evidence to overcome that chain gate. The old thesis incorrectly treated the whole protocol as leveraged looping. Curvance now offers isolated lending markets and routed unleveraged vaults as well as one-click leverage. Those products must be separated. Its official site says markets are isolated, may use two oracles, use dynamic liquidation penalties, and that routed vaults rebalance across approved markets without a lockup. It also reports zero protocol exploits since launch, but that is a project claim from a short record, not an independent finding about incidents. Work with Trail of Bits and a Cantina review provide meaningful assurance. They do not cover Monad, each accepted collateral, curator authority or stressed liquidity. Reopen only for a named unleveraged market after chain approval.

The research file

Mechanism and product separation

Curvance combines isolated lending markets, automated vaults and an interface for looping. A direct supplier deposits a loan asset into a market and earns borrower interest. A borrower posts collateral and can be liquidated if its value breaches the market threshold. Curvance’s routed vaults allocate deposits across approved markets and rebalance for rates, adding risks from allocation and configuration. The one-click leverage path borrows and redeposits to amplify an LST or lending return, adding repeated debt and greater liquidation risk. These are three different decision units. This memo rejects the current Monad deployment and does not claim that an unleveraged direct market has the same payoff as a leveraged loop.

Control, oracle and liquidation design

Curvance says each market is isolated so one market’s failure should not automatically spread to another. Markets may check prices with two independent oracles, and liquidations use dynamic penalties and auction-based order flow. Routed vaults choose an asset and risk level, then move deposits among an approved market set. Their curator or allocator therefore controls that set, caps and rebalance behavior. The public site reviewed did not fully list governance and privileged contract powers. Institutional review must identify every market owner, oracle source, pause role, upgrade key, cap setter and vault allocator with its delay. Isolation limits contagion only when shared assets, oracles, administrators and external strategies are also understood.

Assurance and operating record

Curvance markets independent audits and names Trail of Bits and Spearbit. A public Cantina competition states that the reviewed code was built with Trail of Bits and covered VeCVE and most money-market components. The official homepage reports deposits above the size floor and zero protocol exploits since launch. Those claims are useful but do not prove the exact deployed bytecode, cover every vault module or show an absence of bad debt and withdrawal failures. No consolidated official incident record, market-level series of realized losses or independent solvency report was identified. A young protocol on a young chain has not yet passed a full credit and liquidity cycle. Audit scope reduces code uncertainty. It does not price collateral, oracle, liquidation or chain failure.

Exit and liquidity

A direct lender can redeem only from unborrowed cash in the selected isolated market. High utilization can delay withdrawal until borrowers repay or new liquidity arrives. A routed vault adds allocation timing and may need to withdraw from several markets before paying. “No lockups” does not guarantee immediate cash if the underlying markets are utilized. A leveraged user must unwind repeated borrows and collateral, and price movement or oracle updates can liquidate the position before an orderly exit. All settlement, oracles and liquidators depend on Monad availability and finality. Proposed-size liquidity must be measured in the exact market and asset. Aggregate protocol deposits and the number of pools do not establish a client exit.

Comparison and decision

For unleveraged stablecoin supply, an approved Aave V3 market is the mature pooled comparison and a named Morpho market or vault is the isolated-market comparison. For LST looping, the right alternative is no leverage or a separately approved leveraged mandate, not another lending brand. Curvance may ultimately compete on capital efficiency and interface quality, but neither feature changes the registry’s settlement-chain requirement. The refusal is not a claim that Curvance was exploited or that isolation is unsound. It says the current product set is only on an unapproved chain, its short operating record lacks market-level loss and exit evidence, and its routed and leveraged products require separate review.

Observable reopening conditions

First complete a Monad chain review that results in approval. Then select one unleveraged Curvance market or vault and publish its collateral, loan asset, oracle sources, LLTV, liquidation penalty, interest model, caps, curator, admin, pause and upgrade authorities. Match the deployed bytecode to the current audit scope and publish twelve months of utilization, bad debt, liquidation, oracle deviation and withdrawal-failure history. For a vault, freeze or delay changes to the market allowlist and disclose every rebalance. At proposed size, complete withdrawal and collateral-sale tests inside written time and slippage limits during stressed utilization. Any looping path receives a separate leverage memo and cannot inherit approval.

Sources

The claims above trace to these. Where a number could not be independently verified, the thesis says so.

Inherited controls

The research above describes the protocol layer. Every position also inherits the asset it holds and the chain it settles on. The layer with the most administrative power sets the position’s effective control; that describes control, not quality or suitability.

ChainVerdictControlControl constraint
MonadApproved with limits Governed, no freeze the L1 has a public validator path, but its short production record, single initial client lineage, and Foundation-directed delegation keep stake and operations concentrated.
The memo is public. Monitoring connects the research to positions clients actually hold and flags evidence changes for advisor review. $49 per advisor per month, first 14 days free. Start the trial.