What Avana Oracle is asking
An LP's value depends on the prices of the underlyings, the structure of the pool, the current inventory split, accrued fees, and for concentrated liquidity the relationship between current price and the active range. For a lending system, the important question is not only what a dashboard mark might say. It is what can realistically be recovered if the position has to cover a loan. If Avana lends against best case assumptions, it is financing collateral inflation. Avana Oracle is built around recoverable value, not optimistic net asset value.
How valuation is built
The oracle is a multi layer collateral valuation engine. First, Avana prices the underlying assets from robust external references rather than leaning on raw AMM spot state. Second, it reconstructs the LP position deterministically: pool balance reconstruction for fungible LPs, or direct decomposition of concentrated positions by range, liquidity, and current price. Third, it discounts that reconstructed value with recovery haircuts that reflect liquidation slippage, unwind friction, and impermanent loss risk. Impermanent loss is the difference in value between holding assets in a pool and holding them separately. It grows when the paired assets diverge in price. Borrowing power on Avana comes from the discounted value, not from the most flattering mark.
Why simple marks fail
LP collateral is one of the easiest places for a lending system to fool itself. AMM pool state can look healthy while being locally distorted. A concentrated position can look strong while sitting in a riskier inventory mix than users assume. Thin depth, same transaction manipulation, and stale inputs can all produce numbers that look precise without being trustworthy. Avana separates apparent value from reliable collateral value by combining external asset pricing, deterministic LP reconstruction, and protocol level safeguards.
The oracle interface can also distinguish reconstructed principal value, fee value, and reserved value used for protocol risk buffers. Those components do not deserve identical treatment inside a lending market. By standardizing them across LP types, the spoke layer can reason about fungible LPs, NFT based LPs, and multi asset positions consistently while still respecting how different venues expose state.
Checks and safeguards
Different AMM families need different verification paths. Curve stable LPs, Uniswap pairs and NFTs, Balancer weighted pools, and Aerodrome pairs all expose collateral state differently. On Avana, external prices remain the main anchor. Pool specific data and time weighted checks help verify whether the reconstructed collateral state is coherent and harder to game. Those time weighted checks are not the primary price Avana relies on for loans. They are an extra consistency check so Avana can tell whether the collateral state still lines up before that state is treated as credit.
Manipulation resistance on Avana includes deviation thresholds, pool price difference constraints, recovery haircuts, open interest caps, and an oracle sentinel that reacts when feed health degrades or verification inputs stop agreeing. LP collateral is unusually exposed to same transaction abuse, thin market distortions, and stale or contradictory signals because the collateral is more structurally complex than a simple token balance. A protocol that supports LP credit has to assume the pricing surface will eventually be tested. Avana Oracle is written for that assumption.