mev-beta/docs/ALTERNATIVE_MEV_STRATEGIES.md

Alternative MEV Strategies - Implementation Guide

Date: October 26, 2025 Purpose: Expand beyond atomic arbitrage to profitable MEV extraction

---

🎯 Overview

Based on profitability analysis, atomic arbitrage alone is insufficient. This guide covers three high-profit MEV strategies:

1. Sandwich Attacks - Front-run + back-run large swaps
2. Liquidations - Liquidate under-collateralized lending positions
3. JIT Liquidity - Just-in-time liquidity provision

---

🥪 Strategy 1: Sandwich Attacks

What is a Sandwich Attack?

User submits: Swap 100 ETH → USDC (0.5% slippage tolerance)

MEV Bot sees this in mempool:
1. Front-run: Buy USDC (pushes price up)
2. User's swap executes (at worse price due to #1)
3. Back-run: Sell USDC (profit from price impact)

Profit: Price impact captured = User's slippage

Profitability

Target: Swaps > $10,000 with slippage > 0.3%
Profit per sandwich: $5-$50 (0.5-1% of swap size)
Frequency: 5-20 per day on Arbitrum
Daily profit: $25-$1,000

Implementation Steps

1. Mempool Monitoring

// pkg/mev/sandwich/mempool.go
package sandwich

type MempoolMonitor struct {
    client     *ethclient.Client
    logger     *logger.Logger
    targetChan chan *PendingSwap
}

type PendingSwap struct {
    TxHash          common.Hash
    From            common.Address
    To              common.Address
    TokenIn         common.Address
    TokenOut        common.Address
    AmountIn        *big.Int
    AmountOutMin    *big.Int // Minimum acceptable output
    Slippage        float64  // Calculated from AmountOutMin
    GasPrice        *big.Int
    DetectedAt      time.Time
}

func (mm *MempoolMonitor) MonitorMempool(ctx context.Context) {
    pendingTxs := make(chan *types.Transaction, 1000)

    // Subscribe to pending transactions
    sub, err := mm.client.SubscribePendingTransactions(ctx, pendingTxs)
    if err != nil {
        mm.logger.Error("Failed to subscribe to mempool:", err)
        return
    }
    defer sub.Unsubscribe()

    for {
        select {
        case tx := <-pendingTxs:
            // Parse transaction
            swap := mm.parseSwapTransaction(tx)
            if swap != nil && mm.isSandwichable(swap) {
                mm.targetChan <- swap
            }
        case <-ctx.Done():
            return
        }
    }
}

func (mm *MempoolMonitor) isSandwichable(swap *PendingSwap) bool {
    // Criteria for profitable sandwich:
    // 1. Large swap (> $10,000)
    // 2. High slippage tolerance (> 0.3%)
    // 3. Not already sandwiched

    minSwapSize := big.NewInt(10000e6) // $10k in USDC units
    minSlippage := 0.003 // 0.3%

    return swap.AmountIn.Cmp(minSwapSize) > 0 &&
           swap.Slippage > minSlippage
}

2. Sandwich Calculation

// pkg/mev/sandwich/calculator.go

type SandwichOpportunity struct {
    TargetTx        *PendingSwap
    FrontRunAmount  *big.Int
    BackRunAmount   *big.Int
    EstimatedProfit *big.Int
    GasCost         *big.Int
    NetProfit       *big.Int
    ROI             float64
}

func CalculateSandwich(target *PendingSwap, pool *PoolState) (*SandwichOpportunity, error) {
    // 1. Calculate optimal front-run size
    //    Too small = low profit
    //    Too large = user tx fails (detection risk)
    //    Optimal = ~50% of user's trade size

    frontRunAmount := new(big.Int).Div(target.AmountIn, big.NewInt(2))

    // 2. Calculate price after front-run
    priceAfterFrontRun := calculatePriceImpact(pool, frontRunAmount)

    // 3. Calculate user's execution price (worse than expected)
    userOutputAmount := calculateOutput(pool, target.AmountIn, priceAfterFrontRun)

    // 4. Calculate back-run profit
    //    Sell what we bought in front-run at higher price
    backRunProfit := calculateOutput(pool, frontRunAmount, userOutputAmount)

    // 5. Subtract costs
    gasCost := big.NewInt(300000 * 100000000) // 300k gas @ 0.1 gwei

    netProfit := new(big.Int).Sub(backRunProfit, frontRunAmount)
    netProfit.Sub(netProfit, gasCost)

    return &SandwichOpportunity{
        TargetTx:        target,
        FrontRunAmount:  frontRunAmount,
        BackRunAmount:   backRunProfit,
        EstimatedProfit: backRunProfit,
        GasCost:         gasCost,
        NetProfit:       netProfit,
        ROI:             calculateROI(netProfit, frontRunAmount),
    }, nil
}

3. Execution (Bundle)

// pkg/mev/sandwich/executor.go

func (se *SandwichExecutor) ExecuteSandwich(
    ctx context.Context,
    sandwich *SandwichOpportunity,
) (*ExecutionResult, error) {
    // Create bundle: [front-run, target tx, back-run]
    bundle := []common.Hash{
        se.createFrontRunTx(sandwich),
        sandwich.TargetTx.TxHash, // Original user tx
        se.createBackRunTx(sandwich),
    }

    // Submit to Flashbots/MEV-Boost
    bundleHash, err := se.flashbots.SendBundle(ctx, bundle)
    if err != nil {
        return nil, fmt.Errorf("failed to send bundle: %w", err)
    }

    // Wait for inclusion
    result, err := se.waitForBundle(ctx, bundleHash)
    if err != nil {
        return nil, err
    }

    return result, nil
}

Risk Mitigation

1. Detection Risk

• Use Flashbots/MEV-Boost (private mempool)
• Randomize gas prices
• Bundle transactions atomically

2. Failed Sandwich Risk

• User tx reverts → Our txs revert too
• Mitigation: Require target tx to have high gas limit

3. Competitive Sandwiching

• Other bots target same swap
• Mitigation: Optimize gas price, faster execution

Expected Outcomes

Conservative Estimate:
- 5 sandwiches/day @ $10 avg profit = $50/day
- Monthly: $1,500

Realistic Estimate:
- 10 sandwiches/day @ $20 avg profit = $200/day
- Monthly: $6,000

Optimistic Estimate:
- 20 sandwiches/day @ $50 avg profit = $1,000/day
- Monthly: $30,000

---

💰 Strategy 2: Liquidations

What are Liquidations?

Lending Protocol (Aave, Compound):
- User deposits $100 ETH as collateral
- User borrows $60 USDC (60% LTV)
- ETH price drops 20%
- Collateral now worth $80
- Position under-collateralized (75% LTV > 70% threshold)

Liquidator:
- Repays user's $60 USDC debt
- Receives $66 worth of ETH (10% liquidation bonus)
- Profit: $6 (10% of debt)

Profitability

Target: Under-collateralized positions on Aave, Compound
Profit per liquidation: 5-15% of debt repaid
Typical liquidation: $1,000-$50,000 debt
Profit per liquidation: $50-$5,000
Frequency: 1-5 per day (volatile markets)
Daily profit: $50-$500 (conservative)

Implementation Steps

1. Position Monitoring

// pkg/mev/liquidation/monitor.go

type Position struct {
    User              common.Address
    CollateralToken   common.Address
    CollateralAmount  *big.Int
    DebtToken         common.Address
    DebtAmount        *big.Int
    HealthFactor      float64 // > 1 = healthy, < 1 = liquidatable
    Protocol          string  // "aave", "compound"
    LastUpdated       time.Time
}

type LiquidationMonitor struct {
    aavePool     *aave.Pool
    compTroller  *compound.Comptroller
    priceOracle  *oracle.ChainlinkOracle
    logger       *logger.Logger
}

func (lm *LiquidationMonitor) MonitorPositions(ctx context.Context) {
    ticker := time.NewTicker(5 * time.Second)
    defer ticker.Stop()

    for {
        select {
        case <-ticker.C:
            // 1. Fetch all open positions from Aave
            aavePositions := lm.getAavePositions()

            // 2. Fetch all open positions from Compound
            compoundPositions := lm.getCompoundPositions()

            // 3. Calculate health factors
            for _, pos := range append(aavePositions, compoundPositions...) {
                healthFactor := lm.calculateHealthFactor(pos)

                // Under-collateralized?
                if healthFactor < 1.0 {
                    lm.logger.Info(fmt.Sprintf("🎯 Liquidation opportunity: %s", pos.User.Hex()))
                    lm.executeLiquidation(ctx, pos)
                }
            }
        case <-ctx.Done():
            return
        }
    }
}

func (lm *LiquidationMonitor) calculateHealthFactor(pos *Position) float64 {
    // Get current prices
    collateralPrice := lm.priceOracle.GetPrice(pos.CollateralToken)
    debtPrice := lm.priceOracle.GetPrice(pos.DebtToken)

    // Calculate values in USD
    collateralValue := new(big.Float).Mul(
        new(big.Float).SetInt(pos.CollateralAmount),
        collateralPrice,
    )

    debtValue := new(big.Float).Mul(
        new(big.Float).SetInt(pos.DebtAmount),
        debtPrice,
    )

    // Health Factor = (Collateral * LiquidationThreshold) / Debt
    liquidationThreshold := 0.70 // 70% for most assets on Aave

    collateralValueFloat, _ := collateralValue.Float64()
    debtValueFloat, _ := debtValue.Float64()

    return (collateralValueFloat * liquidationThreshold) / debtValueFloat
}

2. Liquidation Execution

// pkg/mev/liquidation/executor.go

type LiquidationExecutor struct {
    aavePool    *aave.Pool
    flashLoan   *flashloan.Provider
    logger      *logger.Logger
}

func (le *LiquidationExecutor) ExecuteLiquidation(
    ctx context.Context,
    position *Position,
) (*ExecutionResult, error) {
    // Strategy: Use flash loan to repay debt
    // 1. Flash loan debt amount
    // 2. Liquidate position (repay debt, receive collateral + bonus)
    // 3. Swap collateral to debt token
    // 4. Repay flash loan
    // 5. Keep profit

    // Calculate max liquidation amount (typically 50% of debt)
    maxLiquidation := new(big.Int).Div(position.DebtAmount, big.NewInt(2))

    // Execute flash loan for liquidation
    userData := le.encodeLiquidationData(position, maxLiquidation)

    result, err := le.flashLoan.ExecuteFlashLoan(
        ctx,
        position.DebtToken,
        maxLiquidation,
        userData,
    )

    if err != nil {
        return nil, fmt.Errorf("liquidation failed: %w", err)
    }

    return result, nil
}

3. Flash Loan Callback

// contracts/liquidation/LiquidationBot.sol

function receiveFlashLoan(
    IERC20[] memory tokens,
    uint256[] memory amounts,
    uint256[] memory feeAmounts,
    bytes memory userData
) external override {
    require(msg.sender == BALANCER_VAULT, "Only vault");

    // Decode liquidation data
    (address user, address collateralAsset, address debtAsset, uint256 debtToCover)
        = abi.decode(userData, (address, address, address, uint256));

    // 1. Approve Aave to take debt tokens
    IERC20(debtAsset).approve(AAVE_POOL, debtToCover);

    // 2. Liquidate position
    IAavePool(AAVE_POOL).liquidationCall(
        collateralAsset,  // Collateral to receive
        debtAsset,        // Debt to repay
        user,             // User being liquidated
        debtToCover,      // Amount of debt to repay
        false             // Don't receive aTokens
    );

    // 3. Now we have collateral + liquidation bonus
    uint256 collateralReceived = IERC20(collateralAsset).balanceOf(address(this));

    // 4. Swap collateral for debt token
    uint256 debtTokenReceived = swapOnUniswap(
        collateralAsset,
        debtAsset,
        collateralReceived
    );

    // 5. Repay flash loan
    IERC20(tokens[0]).transfer(BALANCER_VAULT, amounts[0]);

    // 6. Profit = debtTokenReceived - debtToCover - flashLoanFee
    uint256 profit = debtTokenReceived - debtToCover;

    emit LiquidationExecuted(user, profit);
}

Risk Mitigation

1. Price Oracle Risk

• Use Chainlink price feeds (most reliable)
• Have backup oracle (Uniswap TWAP)
• Validate prices before execution

2. Gas Competition

• Liquidations are competitive
• Use high gas price or Flashbots
• Monitor gas prices in real-time

3. Failed Liquidation

• Position already liquidated
• Mitigation: Check health factor immediately before execution

Expected Outcomes

Conservative Estimate:
- 1 liquidation/day @ $100 profit = $100/day
- Monthly: $3,000

Realistic Estimate (volatile market):
- 3 liquidations/day @ $300 profit = $900/day
- Monthly: $27,000

Optimistic Estimate (market crash):
- 10 liquidations/day @ $1,000 profit = $10,000/day
- Monthly: $300,000

---

⚡ Strategy 3: JIT Liquidity

What is JIT Liquidity?

Large Swap Pending:
- User wants to swap 100 ETH → USDC
- Current pool has low liquidity
- High price impact (1-2%)

JIT Liquidity Strategy:
1. Front-run: Add liquidity to pool
2. User's swap executes (we earn LP fees)
3. Back-run: Remove liquidity
4. Profit: LP fees from large swap - gas costs

Profitability

Target: Large swaps with high price impact
LP Fee: 0.3% (UniswapV3) or 0.05-1% (custom)
Profit per JIT: $2-$50
Frequency: 10-50 per day
Daily profit: $20-$2,500

Implementation (Simplified)

// contracts/jit/JITLiquidity.sol

function executeJIT(
    address pool,
    uint256 amount0,
    uint256 amount1,
    int24 tickLower,
    int24 tickUpper
) external {
    // 1. Add liquidity in tight range around current price
    INonfungiblePositionManager(POSITION_MANAGER).mint(
        INonfungiblePositionManager.MintParams({
            token0: token0,
            token1: token1,
            fee: 3000,
            tickLower: tickLower,
            tickUpper: tickUpper,
            amount0Desired: amount0,
            amount1Desired: amount1,
            amount0Min: 0,
            amount1Min: 0,
            recipient: address(this),
            deadline: block.timestamp
        })
    );

    // Position will earn fees from the large swap

    // 2. In next block, remove liquidity
    // (This happens in a separate transaction after user's swap)
}

Risk Mitigation

1. Impermanent Loss

• Tight price range minimizes IL
• Remove liquidity immediately after swap

2. Gas Costs

• Adding/removing liquidity is expensive (400k+ gas)
• Only profitable for very large swaps (>$50k)

3. Timing Risk

• User tx might not execute
• Mitigation: Bundle with Flashbots

Expected Outcomes

Conservative Estimate:
- 5 JIT opportunities/day @ $10 profit = $50/day
- Monthly: $1,500

Realistic Estimate:
- 20 JIT opportunities/day @ $25 profit = $500/day
- Monthly: $15,000

---

📊 Strategy Comparison

Strategy	Daily Profit	Complexity	Risk	Competition
Sandwiches	$200-$1,000	Medium	Medium	High
Liquidations	$100-$900	Low	Low	Medium
JIT Liquidity	$50-$500	High	Medium	Low
Atomic Arbitrage	$0-$50	Low	Low	Very High

Best Strategy: Start with liquidations (low risk, consistent), then add sandwiches (high profit).

---

🚀 Implementation Timeline

Week 1: Liquidations

• Days 1-2: Implement position monitoring
• Days 3-4: Implement liquidation executor
• Days 5-6: Testing on testnet
• Day 7: Small amount mainnet testing

Week 2: Sandwiches

• Days 1-2: Implement mempool monitoring
• Days 3-4: Implement sandwich calculator
• Days 5-6: Flashbots integration
• Day 7: Testing

Week 3: JIT Liquidity

• Days 1-3: Implement JIT detection
• Days 4-5: Implement JIT execution
• Days 6-7: Testing

---

🎯 Success Criteria

Liquidations

• ✅ Monitor 100+ positions in real-time
• ✅ Execute liquidation in <5 seconds
• ✅ 1+ liquidation/day
• ✅ $100+ profit/day

Sandwiches

• ✅ Detect 50+ viable targets/day
• ✅ Execute 5+ sandwiches/day
• ✅ <1% failed bundles
• ✅ $200+ profit/day

JIT Liquidity

• ✅ Detect 20+ large swaps/day
• ✅ Execute 5+ JIT positions/day
• ✅ No impermanent loss
• ✅ $50+ profit/day

---

📚 Resources

• Flashbots Docs: https://docs.flashbots.net/
• Aave Liquidations: https://docs.aave.com/developers/guides/liquidations
• MEV Research: https://arxiv.org/abs/1904.05234
• UniswapV3 JIT: https://uniswap.org/whitepaper-v3.pdf

---

Created: October 26, 2025 Status: IMPLEMENTATION READY Priority: HIGH - Required for profitability