skillbase/mev-awareness

## MEV attack types

Type, Mechanism, Victim, Typical cost to victim

Sandwich attack, Attacker places buy before and sell after victim's swap, Large DEX swaps, 0.5-3% of trade value

Frontrunning, Copy profitable tx and submit with higher gas, Arbitrage / liquidation bots, Lost opportunity (100% of profit)

JIT liquidity, Add concentrated LP just-in-time for a swap then remove, Existing LPs in the pool, Reduced fee income for passive LPs

Backrunning, Arbitrage immediately after victim's price impact, Any price-moving swap, Indirect — victim doesn't lose extra

Liquidation MEV, Front-run liquidation calls for the bonus, Borrowers near liquidation, Faster liquidation = less time to add collateral

Time-bandit, Reorg blocks to capture past MEV, All users in reorged blocks, Rare on Ethereum post-merge

## MEV risk assessment

For any on-chain action, evaluate:

```

## MEV Exposure Assessment

**Action:** [what you're doing]

**Size:** [USD value]

**Chain:** [Ethereum, Arbitrum, etc.]

**Risk factors:**

| Factor | Value | Risk level |

|--------|-------|------------|

| Trade size vs pool liquidity | $X / $Y TVL = Z% | High if >1% |

| Slippage tolerance set | X% | High if >1% |

| Public mempool? | Yes/No | High if yes |

| Number of hops (routing) | N | Higher = more MEV surface |

| Token liquidity | Thin/Deep | Thin = worse execution |

**Estimated MEV cost:** $X (Y% of trade)

**Protection recommendation:** [specific action]

```

## Sandwich attack math

```

A sandwich attack is profitable when:

attacker_profit > 2 * gas_cost

Attacker profit ≈ victim_slippage * trade_size * pool_fee_adjustment

Example:

Victim swaps $100K USDC → ETH on Uniswap V3

Pool TVL in active range: $5M

Price impact: ~2%

Sandwich extracts: ~50-70% of price impact = $1,000-$1,400

Gas for 2 txs: ~$50

Attacker profit: $950-$1,350

Victim loss: $1,000-$1,400 (worse execution price)

```

**Rule of thumb:** if your trade is >0.5% of pool liquidity in the active range, expect sandwich risk.

## Protection strategies

### Flashbots Protect

```

// Add Flashbots Protect RPC to wallet

// Ethereum Mainnet: https://rpc.flashbots.net

// Sends txs directly to block builders, skipping public mempool

// In MetaMask: Settings → Networks → Add Network

// Or programmatically:

const provider = new ethers.JsonRpcProvider("https://rpc.flashbots.net")

```

- Txs are hidden from public mempool — sandwichers can't see them.

- Tradeoff: slightly slower inclusion (builder must win the block auction).

- Free. Default choice for any mainnet Ethereum swap >$1K.

### DEX aggregator settings

```

1inch / CoW Swap / Paraswap settings:

- Slippage: set to minimum viable (0.3-0.5% for stables, 0.5-1% for volatile)

- MEV protection: enable (1inch Fusion, CoW Swap batch auctions)

- Deadline: 10-20 minutes (shorter = less time for MEV)

```

### Splitting large trades

```

For trades >$50K:

- Split into 3-5 chunks over 5-15 minutes

- Use different DEX routes for each chunk

- Or use TWAP order (CoW Swap, 1inch Fusion)

Example:

$200K USDC → ETH

Instead of 1 tx: split into 4 × $50K, 5 minutes apart

Estimated savings: $800-$2,000 vs single large swap

```

### L2 MEV landscape

Chain, MEV risk, Sequencer model, Notes

Ethereum L1, High, PBS (builder market), Use Flashbots Protect

Arbitrum, Low-Medium, Centralized sequencer (FCFS), No public mempool currently; sequencer orders by arrival

Optimism, Low-Medium, Centralized sequencer, Similar to Arbitrum; MEV will increase with decentralized sequencer

Base, Low-Medium, Centralized sequencer (Coinbase), Same model as Optimism

Polygon PoS, Medium-High, Public mempool, Sandwich attacks common; use Flashbots on Polygon

## JIT liquidity impact on LPs

JIT liquidity providers add concentrated liquidity right before a large swap, capture most fees, then remove it immediately after:

- **Impact:** passive LPs earn 20-40% less fees on large swaps.

- **Defense:** use tighter ranges (more fee share per $ of liquidity) or move to pools with JIT protection.

- **Quantify:** compare your expected fee share (your liquidity / total active liquidity) vs actual received. If consistently lower, JIT is the cause.