Solana's first LVR-aware vault.
Single-sided USDC. Rotates capital out of pools where arbs are draining LPs.
0% of Solana LPs lose money after LVR · live ticker above
of Solana LPs are losing money supplying liquidity.
Across 500 analyzed pools in the last 30 days, LVR exceeded fee income on 15% of them. The math is deterministic — the losses, quiet.
Don't redesign the pool.
Move the LP capital.
When a pool starts getting picked off, the drainwatch keeper rotates the LP float into whichever pool is still earning more in fees than it’s leaking to LVR.
Every rotation is signed off-chain by an attestor and re-validated on-chain before a single token moves.
See the 11 on-chain gatesFrom on-chain
to verdict.
Four deterministic steps. All the math from Milionis et al., adapted for Solana's microstructure.
Every swap, not samples.
We pull the full trade history from ionic's Solana indexer. Tick-level, no lossy aggregation.
Integrated variance.
Resample to 1-minute bars. Compute realized variance across the window — the textbook estimator, stripped of microstructure noise.
ℓ = σ² · L · √P / 4
Apply the Milionis et al. LVR formula per bucket. Sum across the window. This is what a perfect arbitrageur extracts from the AMM.
LVR versus fees.
Compare extracted LVR against fees earned. When LVR exceeds fees, LPs lose money supplying liquidity — regardless of how the pair performed.
Every LVR number is reproducible.
Here’s the derivation.
Full derivations, empirical validation, and our 5-step refinement chain live in the methodology doc.
Read the methodologyCheck your pool.
Paste a Solana pool address. We’ll compute LVR, fees, and net return from on-chain swap history — the same signal the vault keeper rotates on.
Open vault dashboardLoss-versus-rebalancing, priced in dollars.
Continuous LVR integral from Milionis et al., computed per-pool over Solana swap history.
Fees vs. LVR
Live receipts
- Orca·SOL/USDC-2.4%
- Raydium·JUP/SOL+1.8%
- Meteora·WIF/USDC-5.1%
- Orca·JTO/SOL+0.6%