<aside> 🚀

Fee Boost is a novel mechanism designed to align incentives between Liquidity Providers, the DAO, and external Searchers.

The core thesis: Static fees lose value during volatility, while current dynamic fees are too slow to capture it.

We propose mechanism for fee optimisation.

</aside>

<aside>

v1 → v2 Evolution

v1 (Harberger Tax): Agents rented fee control by paying continuous taxes to the DAO, earning 100% of admin fees while in control.

Critique: The rent model was detrimental to DAO revenue. The DAO traded stable admin fees for volatile, often lower tax income.

v2 Solution: Eliminates rent entirely. Agents capture only temporary spread, and LPs keep growing baseline revenue. DAO bottom line is protected and enhanced.

</aside>

v2: The Core Fee Boost Mechanism

A Market-Driven Fee Model.

Maintain a competitive base fee (e.g., 1 bps) to maintain market share during calm periods. Empower external Agents to implement surge pricing when market conditions allow.

<aside> ⚙️

Operational Logic

1. Trigger: Agent identifies a revenue opportunity and submits a transaction defining a Target Fee and Duration.
2. Boosted State: The pool fee updates to the Target Fee.
3. Reversion: Upon expiry, fee automatically reverts to the Base Fee. </aside>

Dynamic Revenue Allocation

To ensure Agents optimise for protocol revenue, the reward structure is dynamic:

• Traders Pay: Fixed Target Fee (e.g., 15 bps).
• LP Revenue (Floating Baseline): Increases linearly from Base Fee to 100% of Target Fee.
• Agent Revenue: Agent captures the difference.
  • $t=0$: Agent captures maximum spread.
  • $t = \text{boostEndTime}$: Agent captures 0 bps.
  • This decay can be exponential, as volatility tends to decay exponentially.
  • The exponential decay factor decides how much of the initial volatility is shared to LPs and admin.

<aside> ☝

Key insight: This mechanism incentivises Agents to trigger boosts only when spread is profitable and volume justifies it. However, the decay structure alone doesn't prevent unnecessarily prolonged high fees, so we need anti-spam safeguards. Two approaches are to be considered: conditional liquidity bonds (with soft penalties) or governance-gating. Either should work.

</aside>

Anti-Spam Mechanisms

<aside> 🔒

Governance-Gated Boost Access

A Permissioned, Governance-Aligned Approach.

As an alternative to the permissionless Conditional Liquidity Bond (CLB) model, we can implement a permissioned system that relies on governance alignment to prevent spam.

Instead of requiring a bond from anyone, the ability to trigger fee boosts is restricted to veCRV holders. Only users with long-term locked capital can act as Agents. These are selected by a specific set of veCRV holders: DAOs that build on top (those within the SmartWalletChecker whitelist).

<aside> 🏆

For instance, major DeFi protocols like StakeDAO, Yearn, or Convex could designate agents to optimise fee capture on their behalf. This creates a win-win dynamic: their liquid staking tokens generate higher yields, while simultaneously lifting baseline revenue across the entire ecosystem.

</aside>

</aside>

<aside> 🏦

Volatility Prediction Markets and Conditional Liquidity Bonds

Prevention of spam and inefficient pricing via a prediction market structure. This is a permissionless mechanism.

<aside> 💰

Bond Requirements

Agents must deposit a bond in PegKeeper LP Tokens (e.g., crvUSD/USDC).

This acts as a bet that fee revenue will drive the pool's up-only parameter (e.g. virtual_price) higher by a specific threshold.

</aside>

<aside> 🛡️

Slashing Conditions

If the target is missed (volume drops, inefficient pricing), the bond enters a Soft Slash state.

Funds are locked, and all yield generated by the bond is seized by the DAO. The funds are released after the Soft Slash duration.

</aside>

The Soft Slash mechanism penalises Agents who trigger inefficient or unprofitable fee boosts without permanently destroying their capital.

<aside>

Activation Condition

A Soft Slash is triggered when an Agent's boost fails to meet the predicted revenue indicator threshold at the end of the boost period. This indicates that either:

• Trading volume was insufficient to justify the fee increase
• The Target Fee was set too high, deterring traders
• Market conditions shifted unexpectedly

<aside>

Lock Period

Upon activation, the Agent's bond (deposited in PegKeeper LP Tokens) is locked and non-withdrawable for a predetermined duration (e.g., 30-90 days).

The principal remains intact, but control is temporarily forfeited.

After the Soft Slash duration expires, the bond is automatically unlocked.

</aside>

<aside>

Yield Seizure

During the Soft Slash period, all yield generated by the bonded LP tokens (trading fees, CRV rewards, etc.) is automatically redirected to the DAO treasury.

This creates a real economic cost without permanent capital loss.

</aside>

</aside>

</aside>

<aside>

Example Scenario

Agent Alice triggers a 5 bps boost for 1 hour.

If the conditional liquidity bond anti-spam model is chosen, Alice puts forward a $1,000 bond in crvUSD/USDC LP tokens. She predicts the boost will generate enough volume to increase virtual_price by 0.05%.

<aside>

Happy Path Scenario

This is the common scenario regardless of which anti-spam measure is chosen.

Outcome: Volume surges immediately. Alice captures the bulk of the value because volume is front-loaded while her fee share is highest.

Hour 0-1 (The Boost): $10M Volume (Front-loaded)

Hour 1-2 (Decay): Volume adds $2M.

• Fees decay 5 bps → 1 bps. LPs capture 100%. LP Profit: +$600.

Final Result: Alice nets ~$2,875; LPs/Admin net ~$2,725.

Alice earns more because she successfully predicted the volatility spike at the start.

</aside>

<aside>

Exception Path Scenario

This scenario assumes a Conditional Liquidity Bond. If the governance-gated approach is preferred, the bond and slashing conditions are dependent on the DAO that builds on top of Curve.

Outcome: Volume falls short; virtual_price only increases by 0.02%.

Penalty: Alice's $1,000 bond is locked for 28 days. During this period, the ~$3 in yield is captured by the DAO. After 28 days, Alice receives her $1,000 principal back.

</aside>

</aside>

<aside> 📝

</aside>