Aave Horizon RWA

Identification — Aave Horizon RWA

• Name / website / docs: Aave Horizon RWA Market; official product site and documentation are Aave’s Horizon pages. It is an institutional RWA-collateral lending market: permissioned tokenized RWAs are supplied as collateral, while stablecoin liquidity is permissionless.
• Launch: Aave announced launch on August 26, 2025; the market launched on Ethereum and was built on Aave V3.3.
• Chains: Ethereum mainnet only confirmed in the supplied scope and official documentation.
• Native token: None identified. AAVE is the broader Aave ecosystem/governance token, not a Horizon-specific native asset; Horizon uses assets including USDC, RLUSD, GHO and permissioned RWA tokens. Contracts / verification
• Horizon Pool proxy: 0xAe05Cd22df81871bc7cC2a04BeCfb516bFe332C8 — independently identified in Aave documentation and Etherscan. Etherscan labels it “Aave: Horizon RWA Market Pool V3,” shows a proxy, and identifies implementation 0xcD8B9aC3…97E48A23a; proxy source/ABI is verified, while Etherscan reports no audit submitted on the address page.
• Other core addresses: Not verifiable as of October 10, 2026 with the required two-source cross-check. Dune MCP was unavailable in this run; therefore no on-chain/Dune verification is claimed. > Data limitation: Dune cross-check, latest block/timestamp, market TVL, and full address registry: Not verifiable as of October 10, 2026. Fork lineage
• Yes: Horizon is a separate, licensed instance initially forked from Aave V3.3.
• Material changes vs upstream: permissioned RWA collateral; issuer allowlisting; non-transferable RwaATokens; issuer/admin recovery and position-management functions; RWA-specific NAV oracle safeguards; stablecoins remain permissionless.
• Audits/reviews: The public Horizon repository lists Certora and StErMi reviews for “Aave Horizon (v3.3),” while the upstream V3.3 codebase also has separate security reviews. Exact coverage and remediation status for every Horizon-specific modification are Not verifiable as of October 10, 2026.
• Malicious-modification history in similar forks: Not verifiable as of October 10, 2026; no reliable, directly comparable incident record was established within the search budget.

Aave Horizon looks like a real product, not just a landing page: Aave’s docs include a dedicated Horizon section with integration examples, a named Ethereum pool address, and references to API methods and SDKs. Aave also describes Horizon as a lending market for qualified/institutional users, and third-party coverage says the platform launched on Ethereum for borrowing stablecoins against tokenized assets. On maturity, the strongest signal is that Aave documents programmatic integration and an API surface, and says Horizon is compatible with AaveKit API methods and can be embedded via ERC-4626 vaults.

That supports an operational product layer beyond marketing pages, but the web evidence does not independently confirm live deposits/withdrawals or current user-facing transaction flows today; those are Not verifiable as of 2026-10-10. I did not find clear evidence of broken links, fake metrics, or obvious template-site signs in the material reviewed. The presence of a dedicated docs page, changelog entries, a GitHub repository, and multiple launch/update posts argues against a placeholder site.

Open API: Yes, at least at the integration level. Aave states Horizon is compatible with AaveKit API methods, exposes SDKs including React, TypeScript, and GraphQL API, and can be integrated programmatically. The exact public endpoints and whether they are open without restrictions is Not verifiable as of 2026-10-10.

Aave Horizon itself does not have a separately verifiable bug bounty page in the gathered sources; the repository notes it “will be subjected to this bug bounty” once governance upgrades the relevant production instances. The Aave-wide Immunefi program is active and covers Aave v3 and related systems, with Critical rewards at 10% of funds directly affected from $50,000 to $1,000,000; High rewards at 10% of funds directly affected from $10,000 to $75,000; Medium at $10,000; and Low at $1,000. The program was approved in governance proposal 325 and was live on Immunefi in September/October 2023.

Reported payout activity shows at least one valid submission in October 2026, with a $5,000 bounty plus a $500 Immunefi fee for a low-severity GHO issue, and earlier February 2025 governance payouts totaling $61,000 plus $6,100 fee for four submissions.

Dependency & counterparty assessment — Aave Horizon RWA (Ethereum)

• Core design: Aave Horizon is a separate, permissioned Aave v3.3 instance. RWA collateral is issuer-allowlisted and non-borrowable; liquidity is supplied permissionlessly in USDC, RLUSD and GHO. This creates dependence on issuers’ transfer controls, legal structures, NAV reporting and redemption processes.
• Oracles/manipulation: RWA prices rely on issuer-reported NAV rather than liquid market prices. Chainlink DON feeds use configured upper/lower bounds and stale-price continuation; LlamaRisk/Chainlink infrastructure and an emergency multisig are therefore critical dependencies. Main residual risks are stale NAV during issuer/custodian stress, incorrect issuer data within bounds, oracle/configuration failure, and delayed emergency action.
• Issuers/SPVs/custody: Initial collateral included Superstate USTB/USCC and Centrifuge JRTSY/JAAA, with USYC planned/added through the market’s expansion. USTB/JRTSY/JAAA are primarily Treasury or investment-grade-credit exposures; USCC adds crypto-carry/CME-futures and Treasury counterparty risk. Exact custodian, SPV bankruptcy-remoteness, redemption waterfall and issuer concentration for the live reserve set are Not verifiable as of October 10, 2026 without on-chain and asset-level legal records.
• Stablecoin exposure: Lenders face USDC, RLUSD and GHO depeg, redemption, issuer insolvency and liquidity risk. GHO additionally depends on Aave facilitators/GSM arrangements and may transmit risk from USDC/USDT backing. No LST/restaking collateral is identified in the Horizon design reviewed.
• Bridges/CEX/MM: Ethereum-only deployment means no protocol-level bridge dependency was identified. Direct CEX or market-maker exposure is Not verifiable as of October 10, 2026. Issuer off-ramps, secondary-market liquidity and authorized liquidators remain practical exit dependencies.
• Failure scenarios: NAV stale/overstated → under-collateralized loans and bad debt; issuer/custodian/SPV insolvency → delayed or impaired redemption; stablecoin depeg → lender losses or liquidity run; oracle outage → frozen borrowing/liquidations; concentrated borrower default → loss socialization to stablecoin suppliers. Structured fields
• dependency_failure_active: null
• max_exposure_pct: null On-chain exposure percentages and current failure status are Not verifiable as of October 10, 2026 because Dune MCP/on-chain verification was unavailable.

Aave Horizon is described as non-custodial lending infrastructure: users keep control of funds, and Aave Labs says it cannot move funds; borrowing and lending are executed by smart contracts. For Horizon’s RWA assets, custody is organized at the issuer/asset level rather than by Aave itself: issuers onboard investors, run KYC/allowlisting, and manage who can hold the token, while the protocol enforces those rules automatically. The documentation also describes a permissioned model in which only qualified, allowlisted users can interact with the RWA market, and RWA aTokens are non-transferable to respect issuer transfer restrictions.

On the custody question for the underlying real-world assets, Aave governance discussion states that Aave DAO cannot legally custody real-world assets, implying custody must sit with an issuer or other licensed custodian rather than the protocol.

Key management is layered, not fully decentralized:

• Protocol ownership and upgrades: Aave DAO retains ownership of the Horizon contracts and controls upgrades. Day-to-day Horizon administration is delegated to an Aave Labs Executive role, which can manage risk parameters, asset listings, oracle configuration, and supply/borrow caps without a separate AIP for each change.
• Emergency controls: An emergency multisig can pause or freeze Horizon reserves, particularly after an RWA oracle price-bound breach. The broader Aave Protocol Emergency Guardian is documented as a 4-of-7 multisig; the Ethereum Guardian address is 0x2CFe3ec4d5a6811f4B8067F0DE7e47DfA938Aa30.
• RWA issuer / manager keys: Each RWA uses a specialized, non-transferable RwaAToken. An ATOKEN_ADMIN_ROLE may be assigned to the RWA issuer or its designated manager. Through the RwaATokenManager, authorized accounts can perform controlled transfers for wallet recovery or position migration. This means issuers retain meaningful operational control over collateral positions.
• User-key recovery: If an institutional user loses its private key, the issuer can migrate the collateral and borrowing position to a replacement wallet. Issuers may also restrict additional borrowing when legally or operationally required.
• Underlying asset permissions: Allowlisting is enforced by each RWA issuer’s token contract, separate from the permissionless stablecoin liquidity pools. Thus, the main custodial/key risk is concentrated in issuer permissioning and recovery keys, rather than in stablecoin suppliers’ access. Risk assessment: Horizon is institutionally workable but has material centralized-key dependencies: Aave Labs’ delegated executive role, the emergency multisig, and issuer-controlled RWA administration. Current per-contract role holders, signer custody procedures, HSM/MPC use, and issuer key-rotation policy: Not verifiable as of October 10, 2026.

Scope: Ethereum; reviewed October 10, 2026. Dune MCP was unavailable, so on-chain role-holder, proxy-admin, event-history, and timelock checks are Not verifiable as of October 10, 2026. Known address: Horizon Pool: 0xAe05Cd22df81871bc7cC2a04BeCfb516bFe332C8 (Ethereum). The PoolAddressesProvider, PoolConfigurator, ACLManager, oracle, aToken implementations, RWA token-manager contracts, and emergency multisig addresses: Not verifiable as of October 10, 2026. Architecture: ``text Governance/owner [holder & timelock: unverified] | PoolAddressesProvider -- pointers/proxy implementations | +------+--------+----------------+ | | | Pool proxy PoolConfigurator Oracle/Sentinel | | | aTokens/RWA aTokens ACLManager/roles | RWA Token Manager -> issuer-authorized transfers `` Aave V3’s provider is the admin/factory for protocol proxies and can replace Pool/Configurator implementations or change oracle, ACLManager, and related pointers. The PoolConfigurator controls reserve configuration through ACL roles. Privilege surface: pause/freeze, reserve activation, caps, LTV/liquidation parameters, aToken/debt-token implementation changes, oracle/sentinel changes, and rate-strategy configuration are privileged functions.

Horizon’s RWA aToken manager uses AccessControl; its authorized-transfer role can call authorizedTransfer, while the default admin grants/revokes those roles. Exit: users can generally withdraw permissionless stablecoins through the Pool without an admin transaction, subject to liquidity, debt, reserve pause/freeze, and RWA-specific transfer restrictions. Horizon RWA receipt tokens are non-transferable; emergency freezing and issuer/NAV-oracle failures can block or impair exits. Worst case: compromise of the governance/provider authority could upgrade Pool logic, replace the oracle/ACL manager, alter risk controls, freeze withdrawals, or install draining logic. Compromise of RWA issuer/transfer roles could forcibly transfer RWA aTokens.

Direct admin fund-drain capability without governance: Not verifiable as of October 10, 2026. Timelock delay and renounced roles: Not verifiable as of October 10, 2026. Assessment: upgradeable, governance-dependent architecture with material oracle, issuer, allowlist, and emergency-freeze risk. No claim that the deployed Horizon instance matches the cited audits can be verified.

Fields: admin_can_drain=null; audited_deployment=null; upgradeable=true; unresolved_critical=null; unresolved_high=null.

Horizon-specific StErMi report for the Aave v3.3 Horizon codebase. The repository lists StErMi as a Horizon security reviewer, but no deployed-bytecode match was independently verified.

Horizon-specific Certora report for the Aave v3.3 Horizon codebase. The repository lists Certora as a Horizon security reviewer, but no deployed-bytecode match was independently verified.

## Founders and team Aave Horizon is not a standalone protocol with a separately identified founding team; it is an Aave Labs initiative and an Ethereum-based lending market built on Aave V3.3. The publicly identifiable founder is Stani Kulechov, founder of Aave and Aave Labs; independent coverage identifies him as the founder who announced Horizon. The available sources do not establish a complete Horizon-specific executive roster, named operating team, or whether all contributors are public. Public vs. anonymous: Aave’s principal founder is public, but Horizon’s full team composition is not verifiable from the reviewed sources. ## Prior projects and outcomes Aave Labs has a substantial prior operating history through the Aave lending protocol.

Horizon followed earlier Aave RWA work, including Aave Arc and a 2021 RWA lending-market collaboration involving Centrifuge. Horizon launched in August 2025 following Aave governance approval, with collateral support associated with Superstate, Circle and Centrifuge and a broader institutional network including Securitize, VanEck, WisdomTree, Ripple and others. These partnerships indicate a real product and institutional counterparties rather than a purely informational website, but partner participation does not independently verify financial performance or solvency. ## Hacks and credibility No Horizon-specific hack, exploit, or loss is identified in the reviewed sources. Not verifiable as of October 10, 2026 whether Horizon has had zero security incidents across all contracts and counterparties.

Horizon inherits Aave V3 infrastructure, but that does not eliminate risks from its new deployment, permissioning logic, RWA token issuers, NAV/oracle feeds, legal structures, or stablecoin borrowers. The sources reviewed do not provide an independent Horizon security track record sufficient to rate it as incident-free. ## Legal presence, offices, and business reality The product is described as an institutional RWA lending market on Ethereum, combining permissioned collateral with permissionless stablecoin supply. That supports the conclusion that Horizon is a functioning financial product rather than merely a web front.

However, the reviewed materials do not establish a confirmed Horizon-specific legal entity, registered office, jurisdiction, or onshore/offshore operating structure. Not verifiable as of October 10, 2026. Governance material explicitly stated that precise contractual and legal arrangements remained under review. ## Risk conclusion Credibility: medium-to-high at the sponsor level, unproven at the Horizon-entity level. Aave’s public founder, established protocol history, governance process, Ethereum deployment, and named institutional partners are positive signals. The main diligence gap is organizational and legal: Horizon-specific ownership, liability, office, regulated entities, and operating personnel remain insufficiently documented in independent sources. On-chain verification was unavailable for this run; therefore TVL, exposure, holders, and contract activity are Not verifiable as of October 10, 2026.

Reputation assessment: mixed, with no verified fraud, rug-pull, insolvency, sanctions, or enforcement finding identified in the reviewed material. Aave Horizon is described as a separate, licensed Aave instance for institutional RWA lending on Ethereum, launched in August 2025; the licensing and legal structure are relevant positives, but some contractual details were still under review in the governance record. Key concerns

• Governance criticism has focused on Aave Labs’ accountability, financial disclosure, wallet transparency, and the economics of Horizon. A reported ACI analysis alleged approximately $466 million of supply, with about 69% in stablecoins and 31% in RWA collateral, concentrated largely in USCC; it also cited roughly $216,000 of cumulative collector revenue versus about $4.2 million in Merkl incentives and related costs.
• The same criticism alleged concentration in three positions representing 59% of the pool and questioned whether headline supply overstated economically active lending. These figures are reported third-party/on-chain analysis, not independently verified here because on-chain tooling was unavailable.
• Community sentiment was reportedly supportive of the RWA opportunity but strongly opposed to a proposed new token and concerned about reduced DAO revenue share and governance centralization.
• A GitHub review records a security issue in which stablecoins could potentially be borrowed by non-whitelisted RWA holders; the repository shows a subsequent audit-fix documentation change, but the reviewed evidence does not establish the full remediation or residual risk. Auditors, founders, and investors: The available results identify Aave Labs and founder Stani Kulechov, but do not provide sufficient independent evidence to assess Horizon-specific investors or a complete auditor roster. Not verifiable as of October 10, 2026. Legal, regulatory, sanctions, and allegations: No substantiated Horizon-specific fraud, rug, insolvency, sanctions, or criminal allegation was found in the reviewed sources. The product’s separate licensed structure is presented as a response to regulatory requirements, while the precise legal arrangements remained under review in the governance process. Aave’s broader regulatory history should not be conflated with Horizon-specific risk. Because no primary regulatory database or on-chain verification was available in this run, adverse-event absence is not proof of absence. Overall institutional view: reputable parent brand and formal compliance-oriented structure, offset by unresolved transparency, concentration, incentive-efficiency, governance-control, and collateral-data concerns. The strongest open diligence item is independently validating Horizon’s liabilities, active borrowers, collateral concentration, fee flows, and audit remediation on Ethereum.

Aave Horizon RWA — Economic Model (Ethereum only) Strategy / asset flows. Licensed, separate Aave V3 instance. Allowlisted institutions supply tokenized RWAs—currently including Treasury, credit and fund tokens—as collateral; permissionless liquidity providers supply USDC, RLUSD and GHO. Qualified borrowers draw stablecoins against RWA collateral.

RWA collateral is not borrowable, and its receipt aTokens are non-transferable. Yield source. Stablecoin supplier yield is primarily borrower-paid interest, not a native RWA yield stream. RWA suppliers generally show 0% Aave supply APR; their economic return remains that of the underlying tokenized security/fund, outside Aave. Current indicative rates: RLUSD 5.00%, USDC 5.22%, GHO 1.49%; rates are utilization-driven and variable. Organic vs subsidized. Organic yield exists where borrowers pay interest.

The launch plan also included up to $1 million of AAVE incentives ($500k Aave Labs plus $500k DAO), so historical or realized APY may have been subsidized. Exact current incentive-adjusted APY and organic share: Not verifiable as of October 10, 2026. Risk posture. Pool liquidity is broadly market-neutral lending, but collateral introduces directional exposure to RWA NAV, issuer/default, oracle, legal-enforcement, redemption and stablecoin risks. No restaking or external leverage is embedded in the base strategy.

Borrower looping is possible; governance documentation specifically records profitable GHO/RWA looping when Horizon borrowing costs were below market. Withdrawals / gates / limits. No fixed protocol lock-up identified. Withdrawals require available liquidity; RWA withdrawals additionally depend on issuer allowlists and token transfer/redemption rules. Borrowing is constrained by LTV, liquidation thresholds, asset caps and permissioned liquidation workflows.

Variable borrow interest, reserve-factor capture, gas and any issuer-level fees apply; no separate Horizon withdrawal fee was identified. Protocol revenue. Proposed economics: 50/50 split between Aave Labs and Aave DAO from Horizon reserve-factor and GHO revenue. TVL / trend. DeFiLlama reports $257.6m TVL, 100% Ethereum, down 0.6% over 30 days; active loans $124.51m. Aavescan reports $377.8m deposits and $122.0m active loans. This is a methodology contradiction: DeFiLlama excludes borrowed coins and counts collateral locked for lending; Aavescan reports reserve deposits. Dune TVL, product split and trend: Not verifiable as of October 10, 2026. APY sustainability. Sustainable component: borrower-paid interest and underlying RWA income.

APY is volatile with utilization, caps, GHO policy and incentives; zero-yield RWA collateral means headline deposits should not be interpreted as RWA yield. Structured fields: organic_yield_pct: null; leverage_ratio: null

Scope: Ethereum only. Aave Horizon is a lending market, not a standalone DAO treasury. The documented Horizon Pool address is 0xAe05Cd22df81871bc7cC2a04BeCfb516bFe332C8. Treasury / reserve size: liquid_reserves_usd and current treasury balances are not verifiable as of October 10, 2026.

Dune MCP was unavailable for this run, so no current raw on-chain balance, reserve, or liability calculation was performed. No durable Dune query ID or execution ID exists for this assessment. A historical governance-post analysis reported approximately $216,000 cumulative DAO revenue in the Horizon collector, composed of roughly 55k aGHO, 57k aUSDC, and 104k aRLUSD; this is stale and is not a current treasury balance. Historical composition snapshot — stale: A public archive-node analysis reported approximately $466 million total supplied assets, comprising about $322m stablecoins—RLUSD $214m, GHO $80m, USDC $28m—and $144m RWA collateral, mainly USCC at $135m, with VBILL $6.2m, JAAA $2.5m, and USTB $0.4m.

It also reported zero supply for USYC and JTRSY at that snapshot. These figures are not current and were not independently reproduced with Dune. Treasury address / custody: The specific DAO collector address receiving Horizon fees is not verifiable from the reviewed sources as of October 10, 2026. Governance states that Horizon fees were redirected to a DAO-owned collector via a Guardian-controlled multisig.

RWA assets remain issuer-controlled/off-chain in their underlying custody arrangements; Horizon holds tokenized claims and collateral positions, not the underlying securities directly. Control / policy: Initially, Aave Labs held the delegated executive role for listings, caps, risk parameters, and oracle configuration; Governance Framework v2 moved new asset onboarding to DAO governance, while the DAO retains contract ownership and upgrade control. Attestations / valuation: Horizon uses issuer-reported NAV with Chainlink DON bounds. The mWIN onboarding references Chainlink CRE proof-of-reserve using claims supplied by Midas and Northern Trust NAV; this is asset-specific, not a protocol-wide reserve attestation.

Conclusion — Aave Horizon RWA has no native token. The 2025 proposal contemplated a Horizon token, but subsequent reporting says Aave DAO consensus rejected issuing one; therefore Horizon-specific supply, market cap, FDV, emissions, unlocks, allocations, staking and liquidity are not applicable. Related parent token: AAVE (not Horizon-native)

• Ethereum contract: 0x7fc66500c84a76ad7e9c93437bfc5ac33e2ddae9.
• Aggregator snapshot: 16.0M total/max supply; 15.436M circulating; market cap approximately $2.675B; FDV approximately $2.772B. These are AAVE figures, not Horizon figures, and are analytics-platform data rather than raw on-chain verification.
• Utility/governance: AAVE, stkAAVE and aAAVE provide Aave governance voting power; governance controls protocol changes, funding and deployments.
• Horizon value accrual: the approved/announced framework describes a Horizon profit share to the Aave DAO of 50% in year 1, 30% in year 2, 15% in year 3 and 10% thereafter. A DAO buyback program has also been discussed for AAVE, but this is not a Horizon-specific holder distribution.
• AAVE emissions, vesting/unlock history, team/investor/treasury/community allocation, actual unlock execution, holder concentration/insider attribution, current DEX depth and main-listing depth: Not verifiable as of October 10, 2026 (Dune unavailable; no reliable current primary dataset located).
• AAVE mint, blacklist, fee-switch and control-authority analysis: Not verifiable as of October 10, 2026 under this run’s on-chain-data constraint. Horizon governance is delegated operationally to Aave Labs in some areas, while Aave DAO retains contract ownership/upgrades; later asset onboarding was moved toward DAO approval. Contradiction / finding: an earlier Horizon proposal specified a possible 15% Horizon-token allocation to Aave DAO, but the later position was that no new Horizon token would be created. The latter is the operative conclusion; no Horizon token contract or tokenomics market exists.

If Bitcoin falls below $10,000, the main stress on Aave Horizon RWA on Ethereum is likely macro-driven risk aversion, not direct collateral contagion from BTC itself. Horizon’s design is centered on RWA collateral with borrowers taking stablecoin loans against tokenized real-world assets, and its risk controls are based on asset-specific LTV, liquidation thresholds, oracle bounds, and emergency freeze mechanisms rather than crypto-collateral price shocks. For Horizon, a BTC crash would most plausibly affect the protocol through second-order effects: lower DeFi risk appetite, reduced institutional borrowing demand, and potential pressure on the broader Aave ecosystem’s revenue and token sentiment.

Aave’s own Horizon materials say shared stablecoin liquidity supports institutional borrowing, while LlamaRisk defines per-asset parameters and the system can freeze new exposure if oracle or valuation issues arise. The key protocol-specific risk is not Bitcoin price directly liquidating RWA positions. Aave documentation says Horizon uses bounds-checked NAV oracles and rejects out-of-range updates to prevent faulty liquidations, while liquidation is triggered when positions become under-collateralized according to the RWA asset’s own parameters.

That means a BTC crash only matters to Horizon if it causes broader market dislocation that changes funding conditions, redemption behavior, or governance/risk settings. What is not verifiable as of 2026-10-10: current Ethereum-only Horizon TVL, borrower concentration, reserve utilization, and the exact amount of exposure that would remain resilient under a BTC sub-$10k shock. Without on-chain verification, the size of the impact cannot be quantified from the available sources.

For Aave Horizon RWA on Ethereum, a 20% depeg of the largest collateral is not a direct liquidation test unless that asset’s liquidation threshold is above zero and the market uses it as borrowable collateral; the protocol docs and Horizon config indicate RWA collateral is intended to be collateral-only, with liquidation threshold set to 0 for RWA tokens in the referenced configuration. In that setup, a 20% price drop in the largest RWA collateral would primarily reduce borrower equity and could impair repayment capacity, but it would not mechanically trigger liquidations from that collateral alone if its liquidation threshold remains zero. The practical stress impact is therefore portfolio- and configuration-dependent: if Horizon is operating with RWA-only collateral at zero liquidation threshold, the “largest collateral depeg 20%” scenario is mainly a NAV / credit-quality shock, not a standard Aave liquidation shock.

Aave’s general liquidation rule is that positions become liquidatable when health factor falls below 1, and health factor is computed from collateral value and the applicable liquidation threshold; however, that general mechanism only applies where the asset is configured with a positive threshold. What cannot be verified from the provided sources is the current largest collateral name, its exact share of Horizon exposure, and the live parameter set on Ethereum as of now. Not verifiable as of 2026-10-10.

Scope: Dune is unavailable in this run; current TVL, reserve balances, and actual exposure are Not verifiable as of October 10, 2026. The assessment below is structural, not a quantified loss estimate. 1. RWA issuer / underlying-asset vehicle insolvent Expected loss path: NAV suspension or markdown → collateral health factor falls → permissioned liquidator seizes the RWA token and attempts issuer redemption/transfer.

If recovery value is below stablecoin debt, the shortfall becomes bad debt. Bankruptcy-remote or segregated vehicles may preserve investor recourse, but recovery can be delayed and limited; this is asset-specific, not a Horizon guarantee. Absorber: borrower/collateral holder first; then Horizon stablecoin suppliers for any deficit. Compensation: liquidation bonus and eventual issuer/legal recovery; no protocol principal guarantee. Contract path: oracle update → liquidation eligibility → liquidationCall transfers the permissioned RWA to an allowlisted liquidator against repayment of USDC/GHO; residual deficit is recorded at the reserve level. 2. Institutional borrower insolvent Expected loss path: missed repayment → interest accrues → health factor breaches threshold → liquidator repays debt and receives collateral.

Loss equals debt minus realizable, permissioned-redemption value. Absorber: stablecoin liquidity providers, after collateral realization. Compensation: interest income and liquidation incentive; Umbrella can cover an eligible same-asset deficit after its configured offset, by slashing corresponding staked aTokens/GHO. Contract path: Pool debt grows; oracle determines health factor; permissioned liquidation transfers the RWA; Umbrella/Collector deficit-offset contracts may burn staked coverage assets. 3. Stablecoin issuer or GHO-facilitator failure Expected loss path: USDC/RLUSD impairment, or GHO liquidity/peg stress, reduces the value of supplier claims or repayment proceeds. Absorber: reserve suppliers and potentially Umbrella stakers only for the affected asset/network; Aave does not automatically obtain compensation from the issuer. Contract path: reserve accounting continues in token units; withdrawals/liquidations may remain executable but economically impaired. Umbrella coverage is asset-specific, not cross-asset.

4. Allowlisted liquidator/redemption venue insolvent Expected loss path: seized RWA cannot be redeemed or transferred promptly → liquidation capacity disappears → collateral value becomes stale and bad debt can accumulate. Absorber: ultimately liquidity providers; the liquidator bears its own execution/settlement loss. Compensation: none guaranteed; emergency pause, caps, and oracle controls limit new exposure but do not restore lost value.

For the DAO-or-owner fraud stress scenario, I found no verified evidence that Aave Horizon’s DAO or owners have committed fraud as of October 10, 2026. The available material instead describes Horizon as a permissioned, non-custodial institutional RWA market with onchain users retaining control and Aave Labs explicitly stating it cannot move funds. There are, however, governance and centralization concerns that are relevant to this stress case.

Governance proposals indicate Aave Labs retains permissions to enable or disable assets, work with risk managers on parameters and oracles, and administer supply/borrow caps, which concentrates operational control in a small set of actors. The documentation also notes that issuers can forcibly transfer RWA aTokens in certain cases, which is a meaningful trust and control feature for institutional assets, but it is not itself evidence of fraud. The strongest negative claims in the search results are about economics and incentives, not fraud.

One independent report criticizes Horizon’s ROI, arguing DAO incentives exceeded revenue, while other coverage describes the market as early-stage and dependent on institutional adoption. Those are stewardship and performance questions, not proof of misappropriation or deceptive conduct. So the correct stress-test output is: fraud risk cannot be verified from the available sources; the identifiable issue is governance concentration and operator discretion, not confirmed fraud.

Not verifiable as of 2026-10-10.

For Aave Horizon RWA on Ethereum, the primary yield source is the interest paid by institutional borrowers to stablecoin suppliers in the shared Horizon pool; Aave’s documentation states that anyone can supply stablecoins to the market and earn that borrower-paid yield. Under a negative 30-day primary-yield scenario, the stress implication is that supplier earnings would compress materially if utilization falls, borrow demand weakens, or the spread between institutional borrowing demand and the supplied stablecoin base shrinks. The most relevant downside mechanisms are:

• Lower utilization / weaker borrowing demand: yield to suppliers depends on how much institutional collateral is actively borrowing stablecoins at a given time.
• Rate compression: if benchmark short rates rise or borrower demand slows, the incremental spread can vanish, reducing or eliminating attractive carry for suppliers.
• Institutional flow volatility: Horizon’s design explicitly notes that large inflows and sudden withdrawals can flood or drain GHO/stablecoin liquidity and stress the market’s stability.
• Concentration risk: Aave’s Horizon analysis highlights that a small number of large institutional borrowers can move utilization sharply, which makes 30-day yield outcomes sensitive to single-flow events. In short, the stress case is not primarily credit-loss driven for suppliers; it is a cash-yield squeeze driven by lower utilization and spread compression in the institutional borrow pool. Not verifiable as of 2026-10-10: I could not independently confirm on-chain 30-day yield performance or chain-level exposure for Horizon in this run.

Assessment (Ethereum; as of October 10, 2026). Aave Horizon is a licensed, separate Aave V3 instance. Control is hybrid: the Aave DAO retains contract ownership and upgrade authority, while Aave Labs holds the delegated Executive role for day-to-day asset listings, risk parameters, oracle configuration, and supply/borrow caps. The frontend and development are therefore Aave Labs-controlled, although the frontend does not custody funds under Aave Labs’ Terms of Service. Governance process. Historically, Horizon operated under delegated Aave Labs control without an AIP for each change.

Since August 14, 2026, new asset onboarding follows Governance Framework v2: ARFC discussion, LlamaRisk risk assessment, Aave Labs technical assessment, then a binding Snapshot vote; Aave Labs implements approved listings. This makes the DAO real for ownership/upgrades and now asset admission, but not fully controlling operational parameters. Under the requested definition, DAO governance is false because token-holder governance does not directly control all parameters. Funds and emergency powers. No evidence found that Aave Labs or the Horizon emergency multisig can arbitrarily transfer depositor funds.

The emergency multisig can freeze/stop new exposure, and RWA issuers have bounded recovery/position-management powers; these are material centralized controls but not a documented drain function. Exact Horizon multisig signers, threshold, independence, and any Horizon-specific timelock are Not verifiable as of October 10, 2026. Voting concentration/top holders via Dune are Not verifiable as of October 10, 2026 because Dune was unavailable. Contradiction: Aave’s general security messaging says every change requires community voting and mandatory delays, but Horizon’s governance framework expressly permits delegated Aave Labs parameter changes without an AIP.

Company context: Aave Labs Ltd is a Cayman Islands exempt company, General Registry entity 398971; directors were not verifiable from the reviewed sources.

As of October 10, 2026: Aave Horizon is described as an Aave Labs initiative and a separate “licensed” Aave Protocol instance, but the available governance record states that its precise contractual arrangements and legal structure remained under review. Therefore, “licensed” should be treated as an unverified marketing/description claim, not evidence of a named financial-services licence or regulated legal vehicle. Entity/jurisdiction: Aave Labs Ltd is identified in SEC records as the relevant company and is registered in the Cayman Islands; Aave’s privacy policy applies Cayman Islands data-protection law.

The protocol itself is software/smart contracts, not a legal person or counterparty. KYC/AML and restrictions: RWA collateral is permissioned and limited to issuer-allowlisted, qualified users; onboarding typically includes identity verification, compliance checks and issuer terms. Stablecoin liquidity providers are described as permissionless, so compliance is concentrated at the RWA-issuer/asset layer rather than uniformly across the market.

Aave’s interface terms prohibit unlawful activity, money laundering, terrorist financing, sanctions evasion and use from comprehensively sanctioned jurisdictions, and permit wallet/interface blocking. Classification/legal risk: The tokenized RWAs may carry securities, fund-interest, banking, lending, custody and AML obligations depending on issuer, asset and user jurisdiction. No source reviewed establishes a definitive regulatory classification for Horizon itself.

Legal separation between Aave DAO, Aave Labs, issuers, permissioning providers and smart contracts reduces—but does not eliminate—substantive risk: courts or regulators could look through contractual/software labels to actual control, promotion, revenue-sharing, asset servicing or lending activity. Warnings/enforcement/courts: I found no regulator enforcement action specifically against Aave Horizon or Aave Labs Ltd. The SEC meeting with Aave Labs was policy engagement, not enforcement.

A private civil case, Legacy Investments Holdings v. John Does, names multiple Aave entities including Aave Labs Ltd in an amended complaint; its allegations and outcome do not establish regulatory liability. Data protection: Aave states it may process wallet, device, usage and support data; blockchain records cannot generally be deleted, and international transfers may involve the Cayman Islands, EEA, UK and other jurisdictions.

GDPR-style rights are described for EEA/UK users, subject to legal-retention and blockchain immutability limits. Not verifiable as of October 10, 2026: a Horizon-specific licence, standalone legal entity, definitive securities/lending classification, comprehensive AML programme, sanctions designation, or final resolution of the civil case.

Legal entity per GLEIF: AAVE LABS LTD. (LEI 635400TRMDMOYGUYRZ27; jurisdiction KY; registration ACTIVE). OFAC SDN screening of 'Aave Labs Ltd', 'Aave Horizon RWA': no match. SEC litigation and administrative release feeds: no mention.

Aave Horizon does not appear to issue its own stablecoin; it is a lending market on Ethereum that lets qualified users borrow existing stablecoins such as USDC, RLUSD, and GHO, while stablecoins are supplied permissionlessly as liquidity. I could not verify any depeg event history for the specific stablecoin(s) used in Horizon from the available sources, so the depeg count, last depeg date, and max depeg percentage are not verifiable as of 2026-10-10. The best-supported classification is that Horizon is stablecoin-using, but its own stablecoin status is false/none because GHO is an Aave stablecoin, not a Horizon-issued stablecoin.

Aave Horizon’s main risks arise from combining permissionless stablecoin liquidity with permissioned, issuer-controlled real-world collateral. Oracle/NAV integrity, issuer and legal enforceability, liquidation liquidity, privileged administration, and smart-contract or upgrade risk remain material; current on-chain utilization, concentration, balances, and admin state are Not verifiable as of October 10, 2026.

Aave Horizon’s top strengths are: (1) institutional-grade compliance with DeFi composability—it is designed for qualified participants to post permissioned RWAs as collateral while still borrowing permissionless stablecoins, preserving regulatory controls without giving up onchain liquidity; (2) shared liquidity depth—all supplied stablecoins flow into one unified reserve, so new collateral types can tap the full pool immediately; (3) native RWA collateral design—RWAs are used directly as collateral rather than being wrapped through extra layers, which reduces structural complexity and smart-contract surface area; (4) 24/7 instant liquidity—institutions can borrow stablecoins continuously against tokenized assets instead of waiting for traditional settlement cycles; and (5) strong infrastructure credibility and extensibility—it is built on Aave v3.3/non-custodial smart contracts, with transparent onchain administration and programmatic integration that developers can build on. A few supporting details explain why these are meaningful strengths. Aave’s documentation says Horizon is a specialized marketplace for tokenized RWAs tailored to institutional and qualified users, and that borrowers access permissionless stablecoin pools only after supplying permissioned RWA collateral.

Aave’s launch materials also emphasize non-custodial execution, deterministic smart contracts, limited admin powers, and non-transferable aTokens to respect issuer restrictions. The governance proposal adds that Horizon is a licensed instance of Aave V3 with permissioned collateral, permissionless stablecoin supply, and permissioned liquidation workflows, reinforcing the compliance-first design.