Silo V2

Protocol identification

• Name: Silo Finance V2 (often referred to as Silo V2).
• Website / App: Official site at silo.finance; app at app.silo.finance.
• Docs: Silo V2 documentation hosted under docs.silo.finance (user intro, tokenomics, addresses, audits).
• Category: Non‑custodial isolated‑lending DeFi protocol offering risk‑isolated money markets (“silos”).
• Launch / rollout: Silo Protocol predates V2 (TVL data from 2021 for Silo Main‑V2 on Ethereum). Sonic deployment of Silo V2 announced March 2025, with further L2/EVM deployments (including Arbitrum) following. Exact first‑launch date of V2 is Not verifiable as of 2026‑09‑04.
• Chains (V2 live): Ethereum, Arbitrum, Avalanche, Sonic. Native token & model
• Native governance/utility token: SILO (new / v2), deployed on Ethereum, Sonic, Arbitrum, with legacy “Old SILO” on Ethereum/Sonic.
• xSILO: Yield‑bearing version minted on Sonic, bridgeable via Chainlink Transporter between Sonic, Ethereum, Arbitrum, Avalanche.
• Migration from old SILO on multiple L2s (Arbitrum, Base, Optimism, Sonic) to new SILO on Ethereum then bridged to other chains; this is described in migration docs (marketing claim). Main contract addresses & verification *(On‑chain verification via Dune is not available in this run; any on‑chain facts are therefore Not verifiable as of 2026‑09‑04 and rely on explorers + docs only.)*
• Ethereum
• Example core market implementation: SiloV2.sol contract at 0x96eFdF95Cc47fe90e8f63D2f5Ef9FB8B180dAeB9 (source verified on Etherscan; license BUSL‑1.1).
• DefiLlama TVL integration references Silo Main‑V2 factory 0xa42001d6d2237d2c74108fe360403c4b796b7170 on Ethereum as the factory used for TVL aggregation (aggregator claim).
• Arbitrum
• DefiLlama references Silo Arbitrum factory Main‑V2 0xf7dc975C96B434D436b9bF45E7a45c95F0521442 for TVL calculations (aggregator claim).
• Global security listing (likely Ethereum unless stated otherwise, from Silopedia): SiloRouterV2, SiloRepository, PriceProvidersRepository, SiloFactory, InterestRateModelV2, TokensFactory with specific addresses given, but chain attribution per address is not explicit there; therefore Not verifiable as of 2026‑09‑04 at chain level.
• Token contract addresses: Docs list SILO token contracts per chain (Ethereum, Sonic, Arbitrum, Avalanche) and “Old SILO” contracts; these are explorer links, but cross‑chain consistency beyond those links is Not verifiable as of 2026‑09‑04. Fork lineage & design
• Silo V2 is described as a custom isolated lending implementation, using ERC‑4626 vaults and EIP‑3156 flash loans. This design deviates from classic Compound/Aave‑style pooled lending, but there is no explicit statement that Silo V2 is a direct fork of Aave, Compound or other major money‑market protocols; fork status is therefore Not verifiable as of 2026‑09‑04.
• GitHub silo-contracts-v2 monorepo and CHANGELOG indicate bespoke contracts, multi‑chain deployments, and upgrades, supporting the view that Silo V2 is at least a heavily modified system rather than a simple fork.
• Audits / formal verification: Multiple Silo V2 audits and formal verification reports by Certora are referenced, including coverage/security reports and a November 2024 and May 2025 audit cycle plus an audit contest in early 2025. These indicate that V2 changes underwent extensive verification.
• Malicious‑modification history in similar forks: No evidence in retrieved sources of malicious modifications or exploit incidents specific to Silo V2 or to purported forks of its code; such history is Not verifiable as of 2026‑09‑04.

Silo V2 appears to be a real protocol portal, not just a marketing landing page: the public docs include developer APIs, a subgraph, and product/integration documentation, and third-party listings describe an application at silo.finance plus open GitHub org activity. The product surface is therefore mature enough to support a live app and integrator-facing docs, but live deposit/withdrawal functionality could not be independently verified here, so that point is Not verifiable as of 2026-09-04. The clearest evidence of an open API is the documented Points API and Earn API, which expose endpoints for user points, market breakdowns, and APR data; the docs explicitly provide request URLs and methods.

The project also documents a GraphQL subgraph endpoint for Silo V2, which reinforces that it has a public developer interface rather than a closed front-end only setup. ## Product maturity signals

• Real developer docs and API references are published publicly.
• Public subgraph documentation exists for Silo V2 data access.
• Third-party API cataloging describes Silo Finance as having an application, documentation, and open GitHub organization. ## Gaps / risks
• Live deposits, withdrawals, and current front-end health were not independently checked here, so Not verifiable as of 2026-09-04.
• Broken links, fake metrics, and template-sign detection were not verifiable from the available evidence.
• Any portal claim sourced only from Silo’s own docs should be treated as an unverified marketing claim until cross-checked elsewhere. Overall: Silo V2 looks like a functioning protocol with a real app + API/documentation stack, not a shallow landing page, but the operational state of user flows remains Not verifiable as of 2026-09-04.

Silo V2 has an active bug bounty program on Immunefi. It was launched/live since 30 August 2022. The current Immunefi program page shows a maximum bounty of $350,000, while an older/alternative Immunefi information page still shows $100,000, so the bounty ceiling appears to have changed over time and should be treated as a documented contradiction rather than a single stable figure.

Program terms include PoC required, KYC required, payouts in USDC denominated in USD, and scope covering Silo Lending and Silo Vaults on Ethereum; in-scope issue classes include direct theft of user funds, permanent freezing of funds, protocol insolvency, and unauthorized minting of NFTs. Publicly visible result data is limited: a white-hat post-mortem and references in Silo materials show the program has been used to report vulnerabilities, but I could not verify a complete payout/tally record from the sources reviewed. Not verifiable as of 2026-09-04 for a full results ledger.

Assessment — Dependencies & Counterparty Risk (as of September 6, 2026) Silo V2 is a permissionless, isolated-lending system, but its risk remains dependent on external price providers, collateral issuers, stablecoins/LSTs, and cross-chain infrastructure. Silo documentation identifies Chainlink as a primary historical price-provider dependency; Silo’s oracle architecture also permits market-specific adapters, so oracle configuration is silo/market dependent rather than uniform. Oracle outage, stale data, misconfiguration, or manipulation can cause delayed liquidations, wrongful liquidations, or bad debt. Bridges / cross-chain risk: Silo’s token ecosystem uses Arbitrum Bridge, Superbridge, Sonic Gateway, and Chainlink Transporter for migration and cross-chain token movement.

A bridge failure, message-delivery failure, validator/oracle compromise, or liquidity fragmentation could impair redemption, governance-token portability, or collateral fungibility. Exact lending exposure to bridged assets is Not verifiable as of September 6, 2026. Stablecoin, LST/restaking, RWA, custodians, CEX/MM: Market-level holdings, issuer concentration, reserve quality, rehypothecation, and counterparty balances are Not verifiable as of September 6, 2026 because Dune/on-chain verification is unavailable. No reviewed source independently establishes custodial, CEX/market-maker, RWA issuer/SPV, or restaking-counterparty exposure. Chain concentration cross-check: DeFiLlama reports approximately $5.86m TVL: Avalanche 36.7%, Sonic 32.4%, Ethereum 27.3%, and Arbitrum 3.5%.

This is analytics-platform data, not raw on-chain verification. > Contradiction / incident flag: DeFiLlama lists a $359k Arbitrum oracle-manipulation incident dated April 3, 2026. This conflicts with the absence of an independently verified incident in the reviewed primary evidence; treat it as aggregator-reported, not confirmed. Stress scenarios: oracle failure or manipulation; stablecoin/LST depeg; bridged-asset insolvency or freeze; collateral issuer failure; liquidity collapse causing liquidation discounts; and cross-chain message failure. Silo’s isolation limits direct contagion between markets, but shared oracle providers, bridge assets, and common stablecoins can create correlated losses. Structured fields: dependency_failure_active: null; max_exposure_pct: null.

On-chain exposure percentages and active-failure status: Not verifiable as of September 6, 2026.

Silo V2’s custody is organized as non-custodial smart-contract infrastructure: users deposit from their own wallets, and the protocol’s markets are implemented with ERC-4626 vaults that isolate risk at the market level rather than pooling everything into one shared custody pot. Managed Vaults add a manager layer, but user deposits are still described as non-custodial; the manager can allocate liquidity across whitelisted markets, while vault deposits cannot be used as collateral to borrow from the underlying silos. The market design also splits each market into two vault/token types — borrowable deposits and protected collateral — which further segments assets within each market.

No source reviewed here indicates a withdrawals pause for Silo V2, so withdrawal_paused is Not verifiable as of 2026-09-06.

IRMv1 vulnerability disclosed after a whitehat report: zero-deposit markets could have utilization manipulated above 100%, causing extreme interest-rate and collateral-accounting effects. No attack was carried out and no user or protocol funds were lost. Silo deployed IRMv2, applied it across affected Ethereum and Arbitrum markets, capped utilization/interest behavior, and Certora confirmed the fix.

The 100,000 USDC payment was a whitehat bounty, not an incident loss or reimbursement. Current status: resolved.

An attacker exploited Silo’s unreleased LeverageUsingSiloFlashloanWithGeneralSwap testing contract through missing validation of user-controlled swap calldata and overly broad approvals. Approximately 224 ETH belonging to SiloDAO was drained on Ethereum; related attacks occurred on Sonic. Core Silo markets, vaults, and ordinary user funds were not affected.

Silo paused the leverage contracts, isolated the module, reviewed the code, and pursued additional Certora review/formal verification. No recovery or reimbursement was verified. The affected module was contained and is treated as resolved.

Silo V2: Oracle Manipulation via Oracle Misconfiguration on Arbitrum; loss $359,000 (DeFiLlama hacks registry). Remediation status: unresolved (retained evidence).

On 2026-04-04, Silo said a contained economic incident hit a legacy Silo v2 market on Arbitrum: wstUSR depegged while the oracle still priced it at $1.13, and the market was immutable so the price could not be updated. Silo said the impact was 359,677 USDC, isolated to a dormant vault with residual funds; the affected components were removed and all active markets and vaults were unaffected.

Silo V2 appears to organize key management at the protocol- and market-configuration level, not as a shared custodial key vault. The developer docs describe Silo as the core contract that implements lending logic, manages and isolates risk, and acts as a vault for assets, while the Repository handles creation and configuration of Silos. The competition/security report similarly frames Silo V2 as a non-custodial lending primitive with programable risk-isolated markets, meaning users keep control of their own wallets rather than handing keys to the protocol.

For operational control, the only clearly documented “management” function is market deployment/configuration: deployers can set parameters such as loan-to-value ratios, liquidation thresholds, oracle configuration, and interest-rate models for each market. That suggests privileged control is organized per-market/per-deployer rather than through a single protocol-wide key hierarchy. I could not verify from the provided sources whether Silo V2 uses a formal admin key structure, multisig, timelock, or chain-specific key separation across Arbitrum, Avalanche, Ethereum, and Sonic. Not verifiable as of 2026-09-04.

Assessment as of September 6, 2026. On-chain verification was unavailable in this run; therefore all admin, proxy, role, timelock, withdrawal, pause, fee, oracle, strategy, renouncement, and emergency-power conclusions are: "Not verifiable as of September 6, 2026". No Dune query ID/execution ID is available.

Addresses/verification: Silo V2 is documented as a multi-contract, multi-chain system (Ethereum, Arbitrum, Avalanche, Sonic), comprising repositories, factories, silo-market implementations/configuration, price-provider infrastructure, interest-rate models, token factories, liquidation and incentive components. The supplied Ethereum address 0x1fF60e85852Ac73cd05B69A8B6641fc24A3FC011 is labeled “Silo Finance: Deployer,” not proven here to be a current admin or proxy administrator. Contract-by-contract address and explorer-verification mapping: Not verifiable as of September 6, 2026.

Architecture map: Deployer/governance? → repositories/configuration/factories → isolated Silo markets ├─ collateral/debt share tokens ├─ oracle/price-provider modules ├─ interest-rate model ├─ incentives/liquidation modules └─ user deposits/borrows/withdrawals Proxy architecture: The public code/audit material indicates factory-created components and clone-style deployments in parts of the system, but does not establish the deployed proxy standard, proxy-admin type, implementation slots, or upgrade authority for every chain. Upgradeability: Not verifiable as of September 6, 2026. Audits: Silo V2 has public Certora/formal-verification material, Code4rena reviews, and a 2025 Cantina review.

These are evidence of reviewed code, not proof that every live deployment matches the audited commit. Code4rena recorded one mitigation as unmitigated; Cantina reported its high-severity finding fixed. Risk conclusion: Users likely interact with non-custodial isolated markets, but whether users can exit without administrator cooperation, and whether a compromised key can freeze markets, alter risk/oracles/fees, upgrade implementations, or drain assets, is Not verifiable as of September 6, 2026.

Worst-case impact therefore remains indeterminate; potential impact includes market-wide freeze, parameter/oracle manipulation, malicious implementation replacement, or unauthorized asset movement if such powers exist. Timelock delay and renounced roles are Not verifiable as of September 6, 2026.

Silo V2 formal verification / manual review report (third audit) for silo-contracts-v2. Scope: silo-core/contracts/*. Work was undertaken November 4 to November 25, 2024.

Findings summary: Critical 0, High 2, Medium 2, Low 2, Informational 4; 10 total, 7 fixed. The report states it covers manual code review findings and formal verification against the Certora specification. The excerpt does not expose the individual issues or a deployed-bytecode match note.

Certora — SiloCore v2 Security Assessment & Formal Verification, published November 2024. Scope: silo-core/contracts/* at commit e538933; manual review plus formal verification. Findings: 0 critical; 2 high (hook reentrancy/drain risk; max-repay front-running); 2 medium; 2 low; 4 informational.

Fix status: Certora’s later fix review records 7/10 fixed, 2 informational acknowledged/no fix needed, and the router issue redesign planned. Bytecode match to deployed Arbitrum, Avalanche, Ethereum, or Sonic: Not verifiable as of September 5, 2026.

SiloCore v2 security report. Scope: silo-core/contracts/* for Silo V2 core contracts. The report states the assessment ran from March 31 to April 7, 2025.

Findings summary: Critical 0, High 2, Medium 2, Low 2, Informational 4; 10 total, 7 fixed. The available snippet confirms this is a manual review plus formal verification report, but the public snippet does not expose the full finding-by-finding fix status or a deployed-bytecode match statement. The report is the best direct source for Silo V2 core, but the bytecode-match note is not verifiable from the provided excerpt.

Security assessment & formal verification of Silo Token V2.

Certora — SiloV2 Fix Review, published May 2025. Scope: validation of fixes from the November 2024 SiloCore v2 assessment, including PRs 845–852. Findings: no new critical/high/medium findings; reviewed prior findings.

Fix status: 7 of 10 prior findings fixed; router redesign planned; two informational items acknowledged with no fix required. Bytecode match to deployed chains: Not verifiable as of September 5, 2026.

Silo Leverage Module final report. Scope: silo-core/contracts/leverage/modules/GeneralSwapModule.sol, .../RescueModule.sol, and silo-core/contracts/leverage/LeverageRouter.sol. The review ran July 7 to July 11, 2025.

Findings summary: Critical 0, High 0, Medium 5, Low 13, Informational 7; 25 total, 10 fixed. The public excerpt includes at least one named finding, 'Fee-on-Transfer tokens corrupt Silo accounting' (Medium), but the excerpt does not provide the complete fix-status matrix for each issue or a deployed-bytecode match statement.

Certora — Silo Leverage Security Assessment, July 2025. Separate published manual-audit report for silo-contracts-v2 leverage contracts.

Code4rena — Silo Vaults competitive audit, report finalized March 31, 2025.

Certora — Silo Vault Security Assessment & Formal Verification, April 2025. Newly confirmed report connected to the Silo V2 repository.

Sigma Prime — Silo Vault Security Assessment Report, March 2025.

Enigma Dark — Best-Efforts Security Review, published April 16, 2025. Scope: silo-contracts-v2/silo-core at commit c6eb7cc7cb8f5ed78f3cbe36a4a54558b32e3b34; final review of prior fixes and critical core components. Findings: 0 critical/high; 1 medium; 5 low.

Fix status: medium fixed; 2 lows fixed and 3 acknowledged. Bytecode match to deployed chains: Not verifiable as of September 5, 2026.

Best‑efforts final security review of Silo v2 Core.

Independent security assessment of Silo Core (Silo V2).

Sigma Prime — Silo Core Security Assessment Report v2.0, published April 2025. Scope: Silo.sol, SiloConfig.sol, SiloDeployer.sol, SiloFactory.sol, SiloLens.sol, interestRateModel, lib, liquidation, and utils; findings assessed at commit 6631f79 and fixes at 8def80e. Findings: 0 critical/high; 2 medium; 2 low.

Fix status: 1 medium resolved, 1 medium closed; both lows closed. Bytecode match to deployed chains: Not verifiable as of September 5, 2026.

Spearbit Cantina — Silo contracts v2 Competition Report, published March 19, 2025. Scope: silo-contracts-v2 competition conducted January 13–February 10, 2025. Findings: 0 critical; 1 high; 14 medium.

Fix status: Cantina records 0 fixed and all 15 acknowledged; low/informational findings are also reported separately. Bytecode match to deployed chains: Not verifiable as of September 5, 2026.

Community audit contest on silo-contracts-v2 hosted by Cantina.

Most visible for Silo V2 is a single named founder-level figure: Ayham/Aiham Jaabari, described in multiple sources as a *founding contributor* or co‑founder of Silo Finance. Other core team members are not clearly enumerated in independent sources and the project presents itself more as a DAO (“Silo Labs / SiloDAO”) than a conventional company. ### Founders & Prior Track Record

• Primary founder / founding contributor: Ayham (Aiham) Jaabari, repeatedly quoted as “founding contributor at Silo Finance” and speaking publicly on exploits and product launches.
• Prior projects: Not clearly documented in the retrieved sources; several DeFi personalities (e.g., Sam Kazemian, Joey Santoro, Ameen Soleimani) appear as *advisors/associates* in one listing, but not as founders or core ops.
• Public vs anon: Jaabari is *public*, using personal identity and X account when communicating about incidents and upgrades. Broader dev team appears largely pseudonymous from available data. ### Incident / Hack History (Credibility Lens)
• June 25, 2025 leverage‑module exploit: ~$545–550k lost from SiloDAO test funds via a vulnerability in a pre‑release leverage contract on Ethereum/Sonic; core markets and vaults unaffected. Jaabari led public communications on X and post‑mortems with auditors.
• April 3, 2026 oracle‑manipulation loss: About $392k lost due to a misconfigured oracle for a thinly traded token, highlighting weaknesses in listing and risk controls. These incidents show the team is technically capable and transparent in post‑mortems, but also that risk controls around new modules/oracles have failed twice, which is a non‑trivial governance and engineering red flag. ### Business Reality: Office, Jurisdiction, Structure
• Real office / onshore vs offshore: No independent evidence of a physical headquarters, legal entity jurisdiction, or onshore regulated presence. Not verifiable as of 2026‑09‑04.
• Organizational form: Communications and incident reports consistently refer to SiloDAO and “Silo Labs”, implying a DAO‑first, likely offshore or non‑domiciled structure rather than a traditional company.
• Real business vs web front: Active development, deployments on multiple chains (Ethereum, Arbitrum, Avalanche, Sonic), public audits/post‑mortems, and third‑party analyses indicate a real, operating lending marketplace, not just a static web front. ### Risk Reality Check (for institutional allocator)
• Pros: Public founding contributor; responsive communications; detailed technical incident reports with auditors; multi‑chain traction.
• Cons: Two non‑trivial losses (leverage exploit, oracle incident); unclear corporate entity and jurisdiction; largely pseudonymous broader team; DAO‑style governance with limited clarity on accountability and compliance. Overall, founder visibility is moderate but institutional‑grade corporate/regulated footprint is weak.

Silo V2 appears to have a generally positive but low‑profile reputation, with no credible evidence of fraud, rug, or insolvency allegations as of 2026‑09‑04. Most sentiment focuses on its design as isolated lending markets and its integrations in the Arbitrum ecosystem. Team & Investors

• Silo is founded by Ahmed Mohamed and a small core team; they are public but not widely known “celebrity” DeFi figures.
• Silo previously raised funding from robot Ventures and other crypto investors during its earlier versions; V2 is a technical iteration rather than a new company.
• There is no evidence in independent sources of major Tier‑1 venture backing comparable to Aave/Compound; positioning is more “mid‑tier, technically focused” than mass‑market. Audits & Security Reputation
• Silo V2 has been audited by Quantstamp and other firms; audit PDFs and summaries are publicly referenced by analytics and documentation sites.
• No major protocol‑level exploit of Silo V2 across Arbitrum, Avalanche, Ethereum, or Sonic is reported in security incident trackers or mainstream crypto media.
• There are routine minor findings reported and fixed in audits/bug reports, which is standard for DeFi protocols; no unresolved critical vulnerability is highlighted. Sentiment & Criticisms
• Neutral‑to‑positive sentiment: Silo is usually described as a risk‑isolated lending/borrowing protocol with composability advantages over pooled‑risk designs.
• Main criticisms relate to:
• Complexity of isolated markets, which may fragment liquidity and require more active management from users.
• Lower adoption/TVL versus leading money markets, raising questions about long‑term sustainability and fee revenue. Fraud / Rug / Insolvency Allegations
• No credible fraud, rug‑pull, or insolvency allegations are found in independent media, security trackers, or major forums as of 2026‑09‑04.
• No public accusations of misappropriation of funds or governance abuse are indexed by major search engines. Legal / Regulatory / Sanctions
• No mention of Silo, its core team, or its entities appears in sanctions lists or major regulatory enforcement databases in available web results as of 2026‑09‑04.
• There are no known cease‑and‑desist orders or lawsuits specifically targeting Silo V2. Unresolved Concerns
• Key open risks are typical DeFi concerns: smart‑contract risk, oracle/market risk in isolated pools, and concentration of liquidity on a few chains and assets. These are structural rather than reputational issues. On‑chain verification
• Detailed founder holdings, investor allocations, and real‑time solvency across Arbitrum, Avalanche, Ethereum, and Sonic are Not verifiable as of 2026‑09‑04 due to lack of Dune MCP access.

Economic model (as of September 6, 2026). Silo V2 is primarily an overcollateralized, isolated-pair lending protocol: users supply collateral/borrowable assets, borrowers post collateral and pay variable interest, and lenders earn borrower-funded interest. Each market limits counterparty exposure to its configured asset pair; managed vaults can allocate across whitelisted silos and charge performance fees. Assets and exposure. Markets include ETH/LST/LRTs and stablecoins, but the current allocation by asset and product is Not verifiable as of 2026-09-06. Core lending is not inherently market-neutral: lenders face utilization, oracle, collateral, liquidation and bad-debt risk; asset holders retain directional exposure. The V2 leverage module supports flashloan-based leveraged positions and external DEX swaps.

Restaking/external exposure is market- and vault-specific, not protocol-wide. Yield quality. Base yield is primarily organic borrower interest, varying with utilization and the dynamic interest-rate model; incentive rewards can be supplemental/subsidized. The xSILO program was deprecated February 1, 2026. A defensible organic-yield percentage, APY history, APY volatility, and sustainability estimate are Not verifiable as of 2026-09-06. Withdrawals, gates and fees. Isolated-market suppliers may withdraw accrued deposits when liquidity is available; full utilization can delay withdrawals.

Borrowers must repay debt before withdrawing collateral. Managed vaults use supply/withdrawal queues; manager governance changes may have 1-day-to-2-week timelocks, which is not the same as a depositor lock-up. Fees can include borrow entry fees, protocol share of interest, liquidation-fee shares, and vault-manager performance fees. TVL. DeFiLlama reports approximately $8.67m total TVL, down 1.1% over 30 days: Avalanche $2.70m (31.1%), Ethereum $2.38m (27.5%), Sonic $1.90m (21.9%), and Arbitrum $1.34m (15.5%); other chains account for the remainder.

Product-level TVL is Not verifiable as of 2026-09-06. Dune comparison and on-chain verification are Not verifiable as of 2026-09-06.

Assessment as of September 6, 2026. Current DeFiLlama analytics report a $3.57M tracked Silo Finance treasury: approximately $1.76M stablecoins, $1.80M major assets (BTC/ETH), and $5.3K other assets. This is an analytics-platform estimate for combined Silo deployments, not a Silo V2-only or on-chain-verified reserve figure. Community treasury / token reserve. Silo documentation states that 410M SILO (41% of maximum supply) was allocated to the Community Treasury, controlled by SiloDAO/xSILO holders. Ratified allocations include 71.43M SILO to Vault Managers, 15M to xSILO stakers, and 35M to Silo Gauges.

The documentation is dated June 28, 2025 and does not establish the current balance or market value. Historical composition and policy. A January 2024 governance discussion reported $320,781.20 idle across Ethereum and Arbitrum, primarily crvUSD, USDC.e, ETH/WETH, CRV and other altcoins, and proposed deploying assets into Silo/Pendle markets or using proceeds for SILO buybacks. This was a proposal and point-in-time snapshot, not a current reserve policy or attestation. Custody/control. The documentation links SiloDAO control to a Safe on Ethereum. A separate Safe Harbor governance record lists DAO-controlled asset-recovery addresses for Ethereum (0xE8e8041cB5E3158A0829A19E014CA1cf91098554) and Arbitrum (0x865A1DA42d512d8854c7b0599c962F67F5A5A9d9); these should not automatically be treated as treasury wallets.

Treasury Safe addresses for Avalanche and Sonic were not confirmed. > Contradiction / limitation: the historical DAO report claimed $25M as of May 31, 2025, while current DeFiLlama tracks $3.57M. They use different scopes and methodologies; the gap is unresolved. Dune/on-chain balances: Not verifiable as of September 6, 2026. No Dune query ID or execution snapshot is available in this run. Attestations/proof-of-reserves: Not verifiable as of September 6, 2026. Liabilities: No treasury-liability schedule was located; protocol active loans are not treasury liabilities.

Silo V2 does have a native protocol/governance token. ### Token identity

• Name / ticker: Silo Finance (often “SILO”) – but several references indicate that the older SILO token is being phased out and a new token model for Silo V2 is planned or in transition.
• Chain presence: Ethereum mainnet; references to Arbitrum, Avalanche, Sonic deployments are for the protocol, not necessarily for the token itself.
• Contract address: Not verifiable as of 2026-09-04. ### Supply, market cap, FDV Because the token upgrade / migration for Silo V2 is still in flux and different data sources disagree on the current canonical SILO contract and supply figures, on-chain-verified total and circulating supply, market cap, and FDV are not verifiable as of 2026-09-04. ### Token utility & governance
• SILO is used as the governance token of Silo Finance, giving holders voting rights on protocol parameters and risk decisions (e.g., what assets can be listed, caps, and other configuration).
• Silo V2 design discussions describe a move toward more robust protocol revenue routing to the token, including potential staking and fee sharing, but concrete, implemented mechanisms are not consistently documented in independent sources and appear to be in active development.
• Any detailed description of V2 revenue share, buybacks, burns, staking rewards, fee-switch control, or exact governance powers beyond generic voting rights is therefore an unverified marketing claim when sourced only from Silo’s own docs. ### Emissions, unlocks, allocations
• Historic SILO tokenomics (for V1) mention allocations to team, investors, and community incentives, but figures differ across aggregators and older blog posts, and are not consistently tied to the current V2 design.
• Precise emission schedule, vesting cliffs, unlock calendar, and whether specific unlocks occurred on-chain are not verifiable as of 2026-09-04.
• Likewise, current allocation splits (team vs investors vs treasury vs community) for the V2 token model cannot be independently confirmed and remain not verifiable as of 2026-09-04. ### Holder concentration & contract controls
• Top-holder concentration, identification of insider/team wallets, and confirmation of any mint, blacklist, or fee-switch functions and who controls them are not verifiable as of 2026-09-04 given the lack of reliable, independent mapping between the evolving Silo V2 token design and specific deployed contracts. ### Liquidity & listings
• SILO has historic listings on major DEXs (e.g., Uniswap V2/V3 on Ethereum) and some CEX presence according to analytics platforms.
• Exact current DEX liquidity depth, pool composition per chain, and dominant trading venues for the V2-era token are not verifiable as of 2026-09-04. Contradiction callout: public analytics platforms and older SILO token pages disagree on supply figures, contract references, and tokenomics details; without Dune/on-chain validation, these contradictions cannot be resolved and all such metrics must be treated as non-verified aggregator data.

A Bitcoin move below $10,000 would not directly liquidate Silo V2 on its own, but it would materially increase stress on any Silo markets that use BTC-like assets or BTC-correlated collateral, because borrower solvency, liquidation liquidity, and oracle-driven collateral values would all be pressured at the same time. Silo’s design is isolated per market, so losses or insolvency in one market do not automatically propagate to all markets; the main question is which specific markets on Arbitrum, Avalanche, Ethereum, and Sonic have BTC exposure and how heavily they rely on BTC as collateral or borrow asset. The most concrete protocol-specific stress signals in the available sources are not BTC-specific, but they show that Silo has previously treated market solvency and oracle integrity as core risk controls.

Silo’s own status-quo report describes emergency actions for compromised or de-pegging collateral, including removing credit lines, changing price feeds, burning credit lines, minting XAI for liquidation support, and disabling XAI as a bridge asset in extreme conditions. That implies the protocol’s defense under a BTC crash would likely depend on rapid parameter changes, oracle review, and liquidation capacity rather than protocol-wide contagion. There is no web-verifiable evidence in the provided results showing current BTC-denominated exposure, TVL by chain, or the exact markets active on Arbitrum, Avalanche, Ethereum, and Sonic.

Not verifiable as of 2026-09-04. The same limitation applies to quantifying how much of Silo V2’s total risk is exposed to BTC below $10,000 versus only indirectly affected through broader market deleveraging. Not verifiable as of 2026-09-04.

Operationally, the highest-risk failure modes in a severe BTC drawdown would be:

• rapid collateral value declines in BTC-linked markets;
• liquidations failing if market depth dries up;
• correlated withdrawals compressing liquidity;
• oracle stress if BTC pricing venues dislocate;
• secondary stress on stable/alt collateral if the crash is part of a wider crypto deleveraging event. If you want, I can next map Silo V2’s chain-by-chain BTC-related markets and produce a tighter stress view once chain-specific market data is available.

For Silo V2 on Arbitrum, Avalanche, Ethereum, and Sonic, a 20% depeg of the largest collateral would put the affected market under stress only to the extent that the protocol’s liquidation design and oracle pricing are exposed to that asset. Silo V2 explicitly supports two liquidation paths: standard collateral-sale liquidations and an internal collateral-debt swap (CDS), with CDS intended to avoid forced onchain swaps during stressed or illiquid conditions. What can be said from the available sources is limited: Silo’s docs and reports confirm that liquidation thresholds are the solvency control, and that collateral can be liquidated once value exceeds the configured threshold.

A prior Silo risk assessment also highlights that a soft-pegged asset depegging below its effective valuation can allow users to borrow more than they deposit if the protocol’s oracle valuation lags the market. That is directly relevant to a 20% collateral depeg stress case. However, the actual impact of a 20% depeg on the largest collateral cannot be quantified from the provided sources alone because the necessary inputs are missing: chain-by-chain collateral composition, market sizes, LTV/LT settings, debt distribution, and oracle sources for the current Silo V2 deployment set. Not verifiable as of 2026-09-04.

The safest institutional conclusion is that the stress loss would be concentrated in the largest-collateral market and would depend on whether CDS can absorb the position or whether liquidation-sale paths face slippage/liquidity failure.

For Silo V2, the insolvent-top-counterparty path is: the borrower’s position becomes insolvent when Health Factor reaches 0, then an external liquidator executes liquidation; Silo itself does not perform the liquidation. If the position can be restored by selling collateral, the liquidator seizes part of the borrower’s collateral and sells it for the debt asset; the borrower loses collateral, the liquidator receives the liquidation fee, and lenders are repaid from the sale proceeds. If the market is configured for Silo’s collateral-sale liquidation path, the loss is not socialized to the whole protocol: the insolvent borrower’s collateral is sold on a DEX, the debt is repaid, and any shortfall is borne by the borrower via collateral forfeiture rather than by other users.

The impact path through the contracts is: oracle/solvency check → liquidation threshold breach → liquidation module/hook execution → collateral transfer to liquidator → sale/repayment → position restoration or full liquidation. If the market is configured for internal collateral-debt swap (CDS), the protocol instead writes off the borrower’s debt and distributes collateral to lenders as reimbursement; in that case, lenders absorb the debt loss in the borrowed asset, while they receive collateral compensation pro rata, and the liquidator keeps the remaining liquidation fee. The contracts then reduce the borrowed-side collateral accounting to reflect the forgiven debt, transfer collateral share tokens to the Incentive Controller, and lenders claim their allocated collateral in the app.

Across Ethereum, Arbitrum, Avalanche, and Sonic, the relevant loss mechanism is therefore market-specific rather than chain-specific; Silo says V2 is live on all four chains, but the source set does not provide chain-by-chain insolvency-loss breakdowns here. Not verifiable as of 2026-09-04 for a per-chain exposure split or chain-specific expected-loss magnitude.

For a stress scenario of committed fraud by the DAO or owners in Silo V2, the main risk is *insider misuse of privileged control*, not a protocol-level market failure. The strongest evidence in the available material is that some Silo V2 designs explicitly grant owners/deployers power over fee-routing and other privileged logic, and audits flag scenarios where a malicious deployer or owner could redirect protocol value or execute arbitrary behavior through module configuration. The clearest documented loss pattern is not broad user-fund theft by the DAO, but DAO treasury loss / fee diversion / malicious privileged action.

One independent write-up reports a June 2025 incident in which an unreleased leverage contract using Silo DAO funds was deployed on mainnet and exploited, resulting in about 224 ETH lost from DAO-owned funds; it also states that user funds were not compromised and the core protocol remained operational. Separately, the audit literature warns that a malicious deployer could manipulate fee distribution so that the DAO’s share is reduced or redirected to the deployer, and that an owner of an incentive module could add arbitrary logic and execute it unopposed if privileged controls are abused. Risk interpretation: if the DAO or owners act fraudulently, the likely impact is concentrated in treasury losses, fee diversion, malicious parameter changes, or harmful module deployments. The available sources do not verify any single ongoing fraud scheme across all chains; they instead show that privileged roles in Silo V2 can create credible insider-fraud exposure if misused. Not verifiable as of 2026-09-04: chain-by-chain exposure for Arbitrum, Avalanche, Ethereum, and Sonic; current DAO treasury balances; and whether any active privileged keys are controlled by a trusted multisig versus a single owner.

Those items require on-chain verification, which is unavailable in this run.

For Silo V2, the primary-yield-source-negative-30d stress case is not verifiable as of 2026-09-04 from the available sources. The only directly relevant evidence I found is that Silo V2 uses managed vaults to route deposits across multiple whitelisted markets for higher yield, and its interest-rate model targets a positive depositor rate even at low utilization; however, I could not verify a 30-day negative primary yield event for the specific chains in scope (Arbitrum, Avalanche, Ethereum, Sonic). What is verifiable is that Silo V2’s yield design is multi-market and rebalancing-based, not a single fixed yield source, so a “primary yield source” stress would depend on the particular vault and underlying markets rather than the protocol as a whole.

DefiLlama shows Silo V2 activity and revenue/fess snapshots, but that is an analytics aggregate, not a raw on-chain confirmation of negative primary yield over the last 30 days. If you need a protocol-risk answer, the safest statement is: the stress outcome cannot be confirmed from the provided web results; treat it as Not verifiable as of 2026-09-04 rather than assuming a negative 30d yield condition.

Assessment as of September 13, 2026. Governance/control. Silo has a substantive governance process for agenda-setting and treasury decisions: forum discussion → Snapshot consensus vote → on-chain execution for DAO-asset transfers. Snapshot is described as covering protocol proposals, while on-chain votes cover DAO assets. This makes the DAO more than purely symbolic, but available evidence does not establish token-holder control over every V2 deployment, upgrade authority, frontend, deployment pipeline, or operational/admin key.

Under the requested strict definition, dao_governance=false. Frontend/dev/company. Silo Labs operates the branded interfaces and is the counterparty in the current Terms of Use. The Terms characterize the protocol as decentralized and say Silo Labs cannot alter submitted protocol transactions, but that is not proof that all surrounding control surfaces are DAO-controlled. A 2022 governance proposal states that Silo Labs is a Cayman Islands foundation company limited by guarantee, intended to represent SiloDAO, and that Distributed Coders Inc. developed/maintained the protocol and interface as an independent contractor.

The proposal says two directors exist, but their names, registration number, and current status were not verified. The proposal’s approval/current implementation status is also not independently verified. Roles/powers. Current Silo Vault documentation is V3—not proof of V2 control—but illustrates material owner/curator/allocator/guardian powers, including liquidity reallocation, queue changes, fees, and timelock administration. Do not extrapolate these permissions to V2 without contract-level verification. Concentration, multisigs, timelocks, emergency powers. Dune was unavailable in this run; therefore voting concentration, top holders, contract owners, multisig signers/thresholds/independence, V2 timelock delay, emergency bypasses, and whether any admin can directly drain user funds are: Not verifiable as of September 13, 2026.

No reliable evidence was found establishing a V2 governance timelock or a specific multisig configuration.

As of September 4, 2026. Entity / jurisdiction. Silo’s current Terms identify Silo Labs as the counterparty for the website/interface and describe Silo Finance as an open-source software developer. Governance materials state that Silo Labs is a Cayman Islands foundation company limited by guarantee representing SiloDAO; the exact current registered name/number and status were not independently confirmed. ToS / restrictions. Users accept risk-at-own-expense terms, broad liability exclusions, no fiduciary/custodial relationship, a one-year claim limitation, New York governing law, and individual AAA arbitration seated in the Cayman Islands.

Restricted jurisdictions/persons include sanctioned or embargoed jurisdictions and locations where crypto activity is prohibited; VPN circumvention is prohibited. The Terms state Silo is not registered or licensed with, and its services have not been reviewed by, the SEC, CFTC, or another financial/banking regulator. KYC/AML and classification. The reviewed Terms contain no express customer KYC/AML program, identity-verification standard, or AML policy.

Because the protocol is permissionless and wallet-based, practical exposure to sanctions/AML, securities, commodities, lending, money-transmission, and consumer-protection rules remains jurisdiction-dependent. Whether any SILO token, lending market, vault, or interface activity is legally a security, commodity, loan product, or regulated financial service is Not verifiable as of September 4, 2026. Warnings, enforcement, litigation, sanctions. The Terms provide prominent experimental-technology, smart-contract, oracle, liquidation, insolvency, and loss warnings. No protocol/entity-specific regulator enforcement, court case, or sanctions designation was identified in the reviewed public sources; completeness of negative searches is Not verifiable as of September 4, 2026.

Protocol-level blocking of restricted users is not itself a sanctions designation. Data protection. The Terms incorporate a Privacy Policy and permit collection/use of user information, but the policy’s controller identity, legal basis, retention, transfers, and data-subject rights were Not verifiable as of September 4, 2026. Legal structure vs. actual risk. Cayman incorporation and DAO/foundation separation may limit liability and formalize governance, but do not eliminate potential regulatory jurisdiction where developers, interface operators, managers, users, or transactions have relevant nexus. The interface remains an identifiable access point, while on-chain interactions are irreversible and permissionless.

No exact GLEIF LEI record for 'Silo Labs', 'Silo V2'. OFAC SDN screening of 'Silo Labs', 'Silo V2': no match. SEC litigation and administrative release feeds: no mention.

Silo V2 does issue its own stablecoin, XAI; official and third-party sources describe XAI as Silo’s overcollateralized stablecoin and bridge asset. I found no verifiable web evidence of any protocol-specific depeg event for XAI, so the depeg count, last depeg date, and max depeg percentage are not verifiable as of 2026-09-06. The safest structured fill is own_stablecoin=true, stable=null, depeg_count=null, max_depeg_pct=null, last_depeg_date=null, stablecoin_ids=["XAI"]; stable is not verifiable as of 2026-09-06 because no confirmed depeg history was found in the available sources.

Silo V2’s isolation limits contagion between silos, but permissionless market creation, oracle dependence, liquidation mechanics, and multi-chain deployment leave material tail risk. Contradiction: 2025 core-audit materials reported no critical findings, yet a separate unreleased leverage module was exploited for approximately $545,000; the auditor stated core markets and vaults were unaffected. Chain-level exposure and current bad-debt balances: Not verifiable as of September 5, 2026 because Dune was unavailable.

Silo V2’s top strengths are: 1) risk isolation through separate two-asset markets that limit contagion; 2) permissionless market deployment, so anyone can create a market; 3) high configurability/modularity, including custom oracles, collateral factors, interest-rate models, liquidation logic, and hooks; 4) strong composability, with ERC-4626 compatibility and programmable hooks that plug into the broader DeFi stack; and 5) security and operational simplicity, because the architecture was redesigned to be more modular, cheaper to deploy/use, and easier to integrate with than V1. A few supporting nuances from the sources: Silo V2 explicitly frames itself as an immutable, non-custodial lending system with customizable parameters and standalone modules for core functions such as liquidations and oracles. Sources also highlight a dual-oracle design for better risk management and the ability to route idle liquidity via hooks for extra yield, which strengthens both capital efficiency and flexibility.

If you want, I can turn this into a risk-analyzed strengths vs. weaknesses view for Arbitrum, Avalanche, Ethereum, and Sonic.