[RFC] IL Hedge Hook — Automated Impermanent Loss Protection for Uniswap v4 LPs

Requests for Comment
1

[RFC] IL Hedge Hook — Automated Impermanent Loss Protection for Uniswap v4 LPs

Category: Request for Comment
Author: JJ (BELTA Labs)
Status: Draft — seeking community feedback
No governance action required at this time.

TL;DR

BELTA is a Uniswap v4 Hook that automatically hedges LP impermanent loss. LP provides liquidity to a BELTA-hooked pool — everything else is automatic. Premium collection, IL calculation, payout settlement. No manual hedging. No trading desk. No additional UX.

5.1-year backtest with real ETH prices (2020–2025, including COVID/LUNA/FTX):

Without Hedge With BELTA
LP annual return +4.0%/yr +7.8%/yr
Net benefit — +3.7%/yr

LP pays 12% of fee income as premium. Receives 35% of IL as reimbursement. Pays less than they get back.

1. The Problem

49.5% of Uniswap V3 LPs underperform HODL (Loesch et al., 2021). IL is the #1 reason LPs exit:

IL → LP exits → TVL drops → slippage ↑ → volume ↓ → fee revenue ↓

This directly impacts Uniswap’s competitive position. And existing solutions don’t solve it:

Approach Why It Fails
Gamma / Arrakis Auto-rebalances range, but IL still occurs
Manual Perps hedge Requires active monitoring, rebalancing, funding management — trading desk level ops
Options (Opyn / Lyra) Low liquidity, expensive, expiry management
Wide range Fee efficiency drops dramatically

The core issue: no solution converts IL hedging into zero-friction passive infrastructure.

A $50K LP who wants to hedge with Perps today needs to: (1) monitor position in real-time, (2) calculate hedge ratio, (3) execute Perp trades, (4) manage funding rates, (5) rebalance periodically. This is operationally impractical for most LPs. BELTA automates the entire process inside a V4 Hook.

2. Why Now — UNIfication + V4 Hooks

Two recent developments make this the right time:

V4 Hooks make it technically possible. In V3, IL hedging required external monitoring of LP position NFTs, off-chain calculations, and manual Perp trades. V4’s Hook system allows direct on-chain detection of position creation, swaps, and exits — enabling fully automated IL settlement within a single contract. This protocol could not have existed before V4.

UNIfication makes LP retention more urgent. With the recent UNIfication proposal turning on protocol fees (1/4 to 1/6 of LP fees), LPs now keep less of each trade. This compresses LP margins and increases the risk of attrition — making IL hedging infrastructure more valuable than ever. A protocol that improves LP net returns by +3.7%/yr directly counteracts the margin compression from protocol fees.

3. How It Works

3.1 V4 Hook Callbacks

Single Hook contract. Four callbacks. LP does nothing.

Callback What Happens
afterAddLiquidity Position registered, entry price snapshotted, underwriting check applied
beforeSwap Dynamic fee calculated (volatility → fee multiplier 1–4x)
afterSwap Premium accumulated from fee delta, epoch IL tracker updated
afterRemoveLiquidity IL calculated, 35% coverage applied, payout settled from Underwriter Pool

3.2 Underwriter Pool — Who Pays the IL

BELTA is not a magic trick — someone absorbs the IL. That someone is the Underwriter Pool: external capital providers who deposit into a vault and earn premium income + Aave yield in exchange for taking first-loss IL risk.

The pool is protected by three independent layers:

  • Layer 1 — Dynamic Fee (V4 native): beforeSwap tracks EWMA volatility and adjusts fee multiplier (up to 4x). Premium income rises exactly when IL risk is highest. Self-balancing.
  • Layer 2 — Aave Yield: Idle pool capital earns ~5%/yr on Aave/Morpho. Baseline income regardless of LP activity.
  • Layer 3 — Perps Hedging: Protocol maintains ETH shorts during down markets. Backtest: 36% of raw IL absorbed by short profits. Funding cost: ~2%/yr.

3.3 On-Chain Underwriting — Adverse Selection Defense

Not every LP position is covered. The Hook enforces:

  • Narrow range (<±10%): rejected — IL multiplier too extreme
  • Minimum position: $10,000
  • Minimum hold: 72 hours
  • Position age: 1 hour before coverage activates

This structurally excludes high-risk profiles that would drain the pool. Similar to how insurance companies don’t cover pre-existing conditions.

3.4 Smart Contract Architecture

Contract Role
BELTAHook.sol V4 Hook main — IL calculation, settlement logic, Dynamic Fee
UnderwriterPool.sol ERC-4626 Vault — pool management, premium distribution, cooldown
EpochSettlement.sol 7-day epoch IL settlement — Keeper automation, daily payout cap
VolatilityOracle.sol EWMA volatility tracking — Dynamic Fee curve
TreasuryModule.sol Treasury buffer management + Aave yield stacking
HedgeManager.sol Perps delta-hedging — dYdX/GMX/Hyperliquid adapter (Phase 2+)
BELTAToken.sol Governance + revenue participation token
xBELTA.sol Staking vault — exchange rate appreciation model

All contracts built with Foundry. Scaffold complete. Source: GitHub.

4. How This Differs from DeFi Insurance

This is not Nexus Mutual or InsurAce. Those protocols failed at IL hedging because:

Traditional DeFi Insurance BELTA
Covered event Vaguely defined (hacks, depegs) → claim disputes IL — mathematically exact, calculated on-chain
Claim process Governance vote, subjective Automatic — no claims, no disputes
Adverse selection No defense → pool drained On-chain underwriting rules
Pool backing Token-denominated → crash = insolvency Stablecoin-based + Aave yield
Actuarial basis No historical data 5.1 years of backtested IL data

The fundamental difference: IL is a deterministic function of price change. It’s not a subjective event — it’s math. IL = 2√r/(1+r) - 1. This means coverage can be priced, backtested, and settled entirely on-chain with zero human judgment.

5. Backtest — Real Data, Real Crashes

270 weekly ETH/USD close prices (Etherscan/CoinGecko, Jan 2020 – Mar 2025). Not synthetic. Includes:

  • COVID crash (ETH -38.5% in one week)
  • LUNA collapse (-31.9%)
  • FTX (-22%)
  • Aug 2024 Yen carry trade (-22.2%)

LP Economics ($10,000 position, 5.1 years):

Without Hedge With BELTA
Fee income (gross) $11,536 $11,536
Premium paid (12% of fees) — -$1,384
IL suffered -$9,464 -$9,464
IL reimbursed (35%) — +$3,312
Net P&L +$2,072 +$4,000
Annual return +4.0%/yr +7.8%/yr

Why does this work? IL concentrates in black swan weeks. LP pays $5.17/week in premium during quiet periods. Gets $50–$216 back in a single crash week. Over 5.1 years, the payouts exceed the premiums by $1,928.

Protocol sustainability (Underwriter Pool):

At scale ($20M pool, $120M hedged TVL), with all three layers active:

  • Pool grows at +9.3%/yr (median, sliding window stress test)
  • IL/Income ratio: 0.89x — income exceeds IL payouts
  • Stress tested through 2022 bear market (ETH -80% from ATH)

Full backtest methodology in the whitepaper.

6. Why This Matters for Uniswap

LP retention is Uniswap’s core growth constraint.

If LPs earn more and stay longer → TVL deepens → slippage decreases → volume increases → protocol revenue grows. BELTA is direct infrastructure for this flywheel.

Specifically:

  • V4 Hook native — no governance changes, no core protocol modifications needed
  • Dynamic Fee — one of the most practical beforeSwap use cases, directly aligned with Hook Design Lab focus areas
  • Measurable LP improvement — +3.7%/yr net benefit, backed by 5.1 years of real data
  • Protocol fee synergy — as UNIfication protocol fees compress LP margins, IL hedging becomes more important for LP retention
  • Infrastructure, not a competitor — BELTA makes Uniswap LPs more profitable. It doesn’t fork, replace, or compete with Uniswap. It’s a Hook.

7. Security Considerations

  • No external oracle dependency — IL calculated from V4 internal tick data only. No Chainlink or TWAP manipulation risk for core IL settlement.
  • Circuit breaker — 24h IL payouts capped at 30% of pool. Single payout capped at 5%.
  • Multisig — protocol parameter changes require 3/5 signatures.
  • Audit plan — 2 independent professional audits budgeted. First audit after testnet Phase 1 completion, second before mainnet pilot.
  • Epoch-based settlement — IL settled in 7-day batches, not real-time. Prevents flash-loan-style exploits and reduces gas costs.
  • Cooldown periods — Treasury: 30-day unstaking + 2–5% early withdrawal fee. Senior: 7-day cooldown + daily 10% TVL withdrawal cap. Structurally prevents bank runs.

8. Timeline

Milestone Target Deliverable
M1 — Hook MVP Month 1–3 BELTAHook.sol testnet deployment (Sepolia/Unichain), 12+ epoch executions, IL accuracy report
M2 — First Audit Month 4–5 Security audit report, vulnerability fixes, grant applications submitted
M3 — Mainnet Pilot Month 5+ Pool $100K live operation, 24 epochs real settlement data, VC pitching
M4 — Open Market Month 11+ Pool $10M, Treasury/Senior dual pool, BELTA token + xBELTA launch
M5 — DEX Payment Month 36+ Pool $20M+, DEX partnership negotiations

9. Current Status

Item Status
Whitepaper v5.6.0 — 33 pages, full mechanism design + backtest (EN)
Smart contracts 8 .sol files (Foundry), scaffold complete (GitHub)
Backtest engine v5.0.1 — real ETH prices, sliding window stress test, sensitivity analysis
Website official-belta.github.io/belta-labs
Next milestone Testnet deployment (Sepolia / Unichain)
Entity BELTA Labs Pte. Ltd. (Singapore)

10. Looking for Feedback

This is an RFC — we’re here to get challenged, not to sell.

Specific questions for the community:

  1. Mechanism design: Do you see a flaw in the premium/coverage structure? Is 12%/35% the right ratio?
  2. Underwriter economics: Would you deposit into the Underwriter Pool at these expected returns?
  3. LP adoption: As an LP, would you use a BELTA-hooked pool? What would change your mind?
  4. Technical: Any concerns about the Hook callback design or gas implications?
  5. UNIfication context: With protocol fees now live, does IL hedging infrastructure become more or less important for LP retention?
  6. Missing risks: What failure modes are we not seeing?

11. Links

We’d rather hear “this won’t work because X” now than discover X after mainnet. All feedback welcome.