Intro
LP collateral is one of the hardest risk problems in DeFi. A simple token can usually be priced with a direct market reference and liquidated with relatively standard logic. An LP position cannot. Its value depends on the underlying assets, the structure of the pool, the relationship between those assets, fee accrual, liquidity depth, and, in concentrated liquidity systems, the active range of the position itself. In stressed markets, those variables do not just move together. They can compound one another. That means a protocol supporting LP backed borrowing has to do much more than assign a price and set a loan to value ratio.
Avana is being designed around that reality. Its security model starts from the assumption that LP positions are path dependent instruments and that any serious lending framework for them must be conservative by design. The goal is not simply to let users borrow against LPs. The goal is to do so through a system that isolates risk, values collateral defensively, monitors health continuously, and liquidates with precision when needed. That is why Avana’s model is not built around one protective mechanism, but around several layers working together.
Context
The first layer is architectural. Avana is designed on top of Aave v4’s hub and spoke structure because LP collateral needs both shared liquidity and isolated market logic. The hub is intended to hold the common liquidity layer, manage reserve accounting, and enforce broader solvency constraints. The spoke is where LP specific collateral logic lives. This includes collateral admission, valuation rules, pool specific risk parameters, and liquidation behavior. That separation matters because LP collateral should not be allowed to impose its assumptions on the rest of the system. If an LP market becomes stressed, the protocol needs a structure in which that stress is contained rather than allowed to bleed across unrelated collateral types. In Avana, the spoke is not just a convenience layer. It is part of the security boundary.
The second layer is oracle and valuation design. LP positions cannot be safely supported through a single flat price feed because the protocol has to understand what the position actually contains before it can determine what it is worth. Avana’s valuation approach is built around a dual source framework. External price feeds provide the primary reference for underlying assets, while AMM derived pricing acts as an independent verification layer. The point of this structure is not merely redundancy. It is manipulation resistance. If an LP position can only be valued safely when both the asset prices and the pool state are coherent, then the protocol should require that coherence before extending credit. In practice, that means borrowing power should exist only when pricing remains inside acceptable tolerance bands, rather than whenever one source alone prints a convenient number.
Model
But price verification alone is not enough. Avana’s risk model also needs to account for how different LP structures behave under stress. Stable pools, correlated asset pools, blue chip volatile pairs, and concentrated liquidity positions do not deserve the same treatment simply because they are all AMM liquidity. That is why risk in Avana is assigned at the market level rather than the asset level. The pool structure itself matters. Asset correlation matters. Liquidity depth matters. Range concentration matters. In practice, that means collateral factors, liquidation thresholds, and borrowing conditions are derived from the actual nature of the pool rather than from a broad category label. A system that ignores these distinctions is not simplifying risk. It is mispricing it.
Value
This is also where Avana’s position valuation model becomes important. The protocol is not asking only what an LP position is worth in normal market conditions. It is asking what that position should count for as collateral inside a lending system. Those are different questions. Raw market value has to be translated into collateral value through a more conservative framework. One of the core principles here is that the stronger side of an LP pair should not be allowed to disguise the weakness of the other. In practical terms, safer collateral treatment comes from anchoring borrowing power to the weaker aspects of the position and then applying pool level risk adjustments that reflect volatility, liquidity quality, and stress assumptions. This is one of the central differences between a protocol designed for LP collateral and one that merely accepts LP tokens.
Risk
The next layer is health monitoring. Because LP positions remain active after being posted as collateral, risk does not stand still once a loan is opened. The value of the position can change with price moves, fee generation, range migration, and shifts in pool quality. A healthy position at deposit can become unhealthy later without any action from the borrower. That is why Avana is designed around continuous per position health assessment rather than periodic or static checks. The protocol must monitor whether the relationship between collateral value and debt remains inside the safety envelope, and that assessment must be sensitive to how the underlying LP actually behaves. In a system like this, health factor is not a decorative metric. It is the real time expression of whether the collateral model is still holding.
Flow
Liquidation design is the point where all of these assumptions are tested. LP collateral cannot be liquidated like a normal token balance because the position itself is structured liquidity. It may include uncollected fees, concentrated exposure, and venue specific unwind paths. Avana is being built around the idea that liquidation should restore solvency with the minimum necessary disruption. That means collecting value already embedded in the position before destroying principal where possible, removing only the amount of liquidity required to repair the loan, and routing the unwind through execution paths that prioritize deterministic settlement over blunt extraction. Precision matters here because overly aggressive liquidation does not just punish the borrower. It also degrades capital efficiency and can turn a manageable problem into unnecessary value destruction.
System
This is one reason the Smart Agents architecture matters so much in the Avana design. Liquidation is not just a permissionless trigger. It is an operational process that has to monitor positions, reconcile data, simulate execution, and interact with onchain liquidity under adverse conditions. A robust liquidation system for LP collateral therefore needs more than smart contracts alone. It needs reliable external execution infrastructure that can observe the right signals, respond quickly, and unwind positions in a way that is economically sound. In Avana’s model, these agents are part of the broader safety system, not an afterthought bolted onto the end of the protocol. Their job is not merely to liquidate. Their job is to help ensure that liquidation remains orderly, timely, and minimally destructive when it becomes necessary.
Users
The operational layer matters just as much as the market layer. Avana’s security model assumes that governance and risk configuration must be controlled with the same discipline as collateral treatment. That means risk parameters should change through structured processes, sensitive changes should sit behind timelocks and review, emergency controls should exist for genuinely adverse conditions, and monitoring should extend beyond individual positions to the behavior of whole market categories. A protocol that supports LP collateral cannot rely on static configuration and wishful thinking. It needs an operating model that expects markets, correlations, and liquidity conditions to change over time.
Outlook
All of this adds up to a simple principle: LP collateral is only viable if security is treated as a system, not a feature. Architecture isolates risk. Oracles validate reality. Valuation translates complexity into conservative borrowing power. Health checks monitor drift over time. Liquidation restores solvency with precision. Operational controls govern how the whole framework evolves. None of these layers is sufficient by itself. Together, they form the basis of a protocol that can take LP positions seriously as collateral.
Takeaway
That is the broader meaning of Avana’s security model. The protocol is not being designed to make LP borrowing look simple by ignoring what makes LP positions difficult. It is being designed to make LP borrowing possible by respecting that difficulty directly. In that sense, security is not something added after the fact. It is the organizing principle behind the entire architecture.