Dynamic Fee Models Create Unintended Market Behaviors

Priority gas auctions (PGAs) turn block space into a winner-take-all game, where >90% of MEV is captured by a few sophisticated searchers. This creates a feedback loop where high fees for arbitrage become the primary cost for users.

• Game: Searchers overpay for gas to front-run, burning value.
• Consequence: Base fees spike, punishing all other network activity.
• Example: Ethereum's 1559 base fee is a direct tax from this game.

Dynamic LP fees on DEXs like Uniswap V3 create a prisoner's dilemma. Rational LPs undercut each other to near-zero fees to capture volume, destroying the sustainable yield model.

• Game: Race to the bottom on fee tiers to attract arbitrage flow.
• Consequence: TVL becomes mercenary, fragmenting liquidity and increasing slippage.
• Result: Protocols like Trader Joe's LB and Maverick innovate with concentrated liquidity that gamifies fee positioning.

Proposer-Builder Separation (PBS) and MEV-Boost on Ethereum create a new game where validators outsource block building to maximize extractable value. This centralizes power in a few builder relays.

• Game: Validators choose the highest-paying block, not the most neutral.
• Consequence: Risk of censorship and builder/relay oligopolies.
• Counterplay: Projects like Flashbots SUAVE and EigenLayer attempt to re-decentralize this supply chain.

Dynamic models like EIP-1559's base fee can be gamed. MEV bots spam transactions during predictable spikes (e.g., NFT mints, airdrops), creating artificial congestion and pricing out real users. This turns a stability mechanism into a volatility amplifier.

• Result: User TX costs can spike 1000%+ in seconds.
• Systemic Risk: Creates predictable, profitable attack vectors for sophisticated actors.

Decouple fee payment from immediate execution. Protocols like Fuel Network and Starknet's fee model explore averaging. Users pay a predictable rate for a time window, while the protocol manages the volatile spot market.

• Key Benefit: User experience shifts from volatility to predictability.
• Key Benefit: Removes the incentive for bots to front-run fee updates, smoothing network demand.

Sequencers running Arbitrum or Optimism bundles have a profit motive. A dynamic fee model allows them to capture value during high demand, but their optimal strategy may not align with minimizing user cost or maximizing chain throughput.

• Result: Potential for rent extraction during network stress.
• Conflict: Sequencer as profit-seeker vs. sequencer as public good.

Implement a PBS (Proposer-Builder Separation) style auction for block space at the sequencer level. This is being explored by Espresso Systems and Astria. Fees are set by competitive bidding, and a portion of sequencer/MEV revenue is redistributed back to users or the DAO.

• Key Benefit: Aligns sequencer revenue with market efficiency, not congestion.
• Key Benefit: Can fund gas rebates or public goods, turning a cost into a yield mechanism.

dApps with hardcoded gas limits or oracle price feeds (e.g., Chainlink) can fail when fees dynamically spike. Liquidations on Aave or Compound may not execute, and arbitrage bots for Uniswap pools become unprofitable, breaking core DeFi mechanics.

• Result: Protocol insolvency risk during market stress.
• Systemic Risk: Creates correlated failure points across the ecosystem.

Move beyond users paying gas directly. Systems like UniswapX, CowSwap, and Across use solver networks. Users submit intents ("I want this swap"), and solvers compete to fulfill them, bundling and optimizing gas payment on the backend.

• Key Benefit: User never sees a gas fee; pays in output token.
• Key Benefit: Solvers absorb volatility and optimize execution across layerzero and other domains, achieving better net prices.