Loopscale

To fulfill the vision of onchain finance, we don't need higher yields or more composability. We need to eliminate the spectre of "full loss of funds" hanging over the heads of everyone participating. It's an impossible risk to price in and makes the "cost" of DeFi far too high to be worth the endgame. In light of recent events, I wanted to share more on how Loopscale's architecture itself introduces structural security advantages over the pool model for onchain lending in addition to the robust technical and operational security processes we've implemented over the past year. It starts with understanding the constraints of the current model. Pool-based lending (e.g. Drift, Aave, Morpho, Fluid, Kamino) relies on shared state: global oracles, global LTVs, and co-mingled deposits. With one bad parameter change or one bad asset, everyone, lenders *and* borrowers alike, are exposed. Loopscale meaningfully addresses this at a primitives level: 1. Collateral isolation: No collateral on Loopscale is auto-rehypothecated. Your collateral for each loan or loop sits in a standalone PDA. If anything happens to the vault you're borrowing from, borrowers have zero risk of their collateral being compromised as a result. In the recent exploit, much of what was taken was idle collateral sitting in shared pools, assets like JLP and jitoSOL that no one was borrowing. Markdowns on bad debt also impacted borrowers directly, which cannot occur with segregated, non-rehypothecated collateral accounts. If bad debt occurs elsewhere on the platform, borrowers and loopers cannot be impacted. Your position is your position. 2. Deposit isolation: Unlike Morpho or Fluid, deposits into one USDC vault are never co-mingled with deposits from other vaults into shared sub-markets. All deposits remain in their own isolated accounts, with liquidity unified on the Creditbook. If one market has issues, you're not competing for withdrawals with other vaults, and you're not subject to parameter changes lobbied for by curators you never deposited with. 3. Lender control: There is no concept of a global oracle, global LTVs, or a universal approved asset list. Lenders can create positions and control the specific oracles, terms, and assets at all times. There is no scenario where we, or anyone attempting to compromise us, could change your positions. 4. Vault guardrails: Even where depositors delegate to vault curators, multiple layers of protection exist independently of trust in the curator: • New asset-oracle combinations require approval from our administrative multisig before a curator can add them. • All risk-increasing decisions (LTV/LQT increases, cap increases, new assets, oracle changes) are automatically timelocked. • Each vault has customized caps on individual asset allocations and all asset flows, limiting exposure to any single collateral issue or curator-level compromise. All of this follows from the architecture itself. We believe order book-based lending is the only model that can support the scale and seriousness of capital that DeFi needs to fulfill its promise.

To fulfill the vision of onchain finance, we don't need higher yields or more composability. We need to eliminate the spectre of "full loss of funds" hanging over the heads of everyone participating. It's an impossible risk to price in and makes the "cost" of DeFi far too high to be worth the endgame. In light of recent events, I wanted to share more on how Loopscale's architecture itself introduces structural security advantages over the pool model for onchain lending in addition to the robust technical and operational security processes we've implemented over the past year. It starts with understanding the constraints of the current model. Pool-based lending (e.g. Drift, Aave, Morpho, Fluid, Kamino) relies on shared state: global oracles, global LTVs, and co-mingled deposits. With one bad parameter change or one bad asset, everyone, lenders *and* borrowers alike, are exposed. Loopscale meaningfully addresses this at a primitives level: 1. Collateral isolation: No collateral on Loopscale is auto-rehypothecated. Your collateral for each loan or loop sits in a standalone PDA. If anything happens to the vault you're borrowing from, borrowers have zero risk of their collateral being compromised as a result. In the recent exploit, much of what was taken was idle collateral sitting in shared pools, assets like JLP and jitoSOL that no one was borrowing. Markdowns on bad debt also impacted borrowers directly, which cannot occur with segregated, non-rehypothecated collateral accounts. If bad debt occurs elsewhere on the platform, borrowers and loopers cannot be impacted. Your position is your position. 2. Deposit isolation: Unlike Morpho or Fluid, deposits into one USDC vault are never co-mingled with deposits from other vaults into shared sub-markets. All deposits remain in their own isolated accounts, with liquidity unified on the Creditbook. If one market has issues, you're not competing for withdrawals with other vaults, and you're not subject to parameter changes lobbied for by curators you never deposited with. 3. Lender control: There is no concept of a global oracle, global LTVs, or a universal approved asset list. Lenders can create positions and control the specific oracles, terms, and assets at all times. There is no scenario where we, or anyone attempting to compromise us, could change your positions. 4. Vault guardrails: Even where depositors delegate to vault curators, multiple layers of protection exist independently of trust in the curator: • New asset-oracle combinations require approval from our administrative multisig before a curator can add them. • All risk-increasing decisions (LTV/LQT increases, cap increases, new assets, oracle changes) are automatically timelocked. • Each vault has customized caps on individual asset allocations and all asset flows, limiting exposure to any single collateral issue or curator-level compromise. All of this follows from the architecture itself. We believe order book-based lending is the only model that can support the scale and seriousness of capital that DeFi needs to fulfill its promise.

Loopscale has no direct exposure to the ongoing situation with Drift protocol and funds in the majority of Loopscale products, including the USDC Genesis and USDC OnRe Vaults, are safe. A small share of the SOL Genesis Vault deposits (~$170K) is allocated to MarginFi, a fraction of which may have indirect exposure through Project0 borrowers collateralized by Drift positions. Should any losses materialize, Loopscale will make affected users whole. We take the technical and operational security of our infrastructure seriously. Our thoughts are with the Drift team and all users affected.