An Argument for Protocol Owned Liquidity

Monte Carlo Randomized Buybacks vs. A single Wide Uni v3 Position

January 2026 OP Mainnet OP/ETH 0.3% pool Data: Flipside Crypto and Dune Analytics
Author: Carlos Mercado GitHub Repo

Key Takeaway

A single month of OP Mainnet transaction fees (Jan 2026) would 20%+ increase the liquidity of OP/WETH on mainnet, make OP a market leader in protocol owned liquidity, and return MORE OP than randomized buybacks.

218,923
Monte Carlo Mean
232,617
LP Result (OP-equiv)
+6.3%
Single Month LP Excess Returns (OP)
~166K
Annualized Excess Returns (OP)

The Theory

Buybacks can be dangerous. They invite discounted cashflow modeling that can result in high short-term volatility of token price. Protocol Owned Liquidity takes a cut of this kind of arbitrage and generates a flywheel of activity that compounds over time.

Looking at Jan 2026 OPM swap data for the Uniswap v3 OP/WETH 0.3% pool on OP Mainnet (the most liquid OP/WETH pool of any platform on OPM) to compare randomized OP buys vs. a single wide concentrated liquidity position for that pool, with daily compounding of fees.

The LP strategy would have collected a position worth ~232,617 OP (Jan 31 closing price) compared to the average of 1,000 Monte Carlo simulations of randomized buybacks ~218,923 OP. Notably, it also beat the maximum outcome across the 1,000 random simulations — showing the value of fee compounding.

Simplifying Assumptions

The actual buyback proposal allocates 50% of Superchain sequencer revenue (OP Mainnet + Base + Unichain + other OP Stack chains). For simplicity, this analysis uses 100% of OP Mainnet transaction fees only, ignoring other Superchain revenue. This keeps the data simple while still demonstrating the strategy comparison.

Process & Strategy

Data Processing

Script: 01_processing.py

Input: Raw swap data (opweth03-swaps-jan2026.csv) with sqrtPriceX96, amounts, timestamps

Output: Hourly Open-High-Low-Close price data from the swaps(hourly_ohlcv.csv)

  • OHLC prices (OP per ETH) derived from sqrtPriceX96
  • Buy/sell volumes for OP and ETH
  • LP fees earned (0.3% of sold amounts)
  • VWAP (volume-weighted average price) per hour

Strategy 1: Monte Carlo Random Purchases

Script: 02_monte_carlo_buys.py

Every day (Jan 2 - Jan 31) spend the previous day's tx fees (Jan 1 - Jan 30) buying OP at random times in random splits.

  • Daily Budget: Day T budget = Day T-1 transaction fees (in ETH)
  • Execution: Randomly select 1-10 purchase hours each day
  • Price Sampling: Execute buys at prices sampled uniformly from each hour's low-high range
  • Simulations: 1,000 runs to capture distribution of outcomes

Other potential strategies including RSI/MACD (technical indicators) purchasing, Bayesian Updating & Kelly Criterion for sizing, etc.

Strategy 2: Simple Wide LP

Script: 03_simple_lp.py

Every day (Jan 2 - Jan 31) deposit the previous day's tx fees (Jan 1 - Jan 30) AND any accumulated swap fees into a fixed wide range LP position.

  • Purposefully Wide Position: 90000 to 94980 (~8,099 to ~13,327 OP/ETH) - a wide range capturing likely price movement
  • Daily Budget: Day T budget = Day T-1 transaction fees + Day T-1 earned LP fees (compounding)
  • Token Split: Split ETH budget into ETH + OP at the required ratio using match_tokens_to_range()
  • Liquidity: Add liquidity to the position using get_liquidity()
  • Fee Calculation: Swap level fee_share = position_liquidity / (active_liquidity_in_pool + position_liquidity)

Wrote & tested custom Uniswap v3 Python package

  • sqrtpx96_to_price() - Convert on-chain price format to human-readable OP/ETH
  • match_tokens_to_range() - Determine ETH/OP split needed for a given tick range
  • get_liquidity() - Calculate liquidity units from token deposits
  • get_position_balance() - Get current ETH and OP in the position at any price
  • tick_to_price() - Check if swap prices fall within the LP range

Final position value is converted to OP-equivalent by pricing the ETH component at end-of-period price.

TX Fees (Budget)

TVL of the OP/WETH 0.3% pool on OP Mainnet is ~1.5M OP & 57 ETH as of Feb 2, 2026. So POL would have a significant impact extremely quickly.

Jan 2026: OP Mainnet tx fees averaged ~0.7 ETH/day.
Jan 1 (grey): No Dec 31 budget available | Jan 31 (red): Would be Feb 1's budget, no impact here.

Monte Carlo Simulation Distribution

Total OP purchased across 1,000 simulations of randomized buybacks.

Random timing barely matters: 1,000 simulations clustered within 0.8% of each other
All outcomes between 218K-220K OP — execution timing variance is minimal

LP Strategy Performance

LP position grew to 12.7 ETH + 64.7K OP deposited, earning 1,738 OP in fees
Fee compounding drove daily growth; median 7.8% liquidity share captured trading activity

Summary Comparison

LP Strategy Outperformed ALL 1,000 Monte Carlo Simulations

The LP strategy benefits from:

  1. Fee income: Earns ~1,738 OP + 0.17 ETH in trading fees over 30 days
  2. Compounding: Daily fees roll into the next day's deposit, growing the position faster
  3. Dual exposure: A wide positions maintain exposure to both ETH and OP which reduces the risk of impermanent loss and falling out of range. At the price of less efficient liquidity (in terms of fees accumulated).

Trade-offs

LP Advantages
  • Passive fee income from trading activity
  • Compounding effect over time
  • Dual-asset exposure (ETH + OP)
  • 6% higher returns in this simulation
LP Considerations
  • Impermanent loss risk if price moves outside range
  • More complex to manage than simple buys
  • In a strongly trending market, direct buying may outperform
  • Requires monitoring and potential rebalancing

Conclusion

Final LP Position Composition

At the end of January 2026, the LP position contains:

  • Total ETH deposited: ~12.7 ETH (cumulative daily deposits)
  • Total OP deposited: ~64,700 OP (cumulative daily deposits)
  • Fees earned: 1,738 OP + 0.17 ETH
  • Position OP-equivalent value: ~232,617 OP (at end-of-period price ~10,327 OP/ETH)

Annualized Perspective

The LP strategy generated ~13,694 additional OP over 30 days compared to the Monte Carlo mean. Annualized, this represents:

  • Monthly advantage: ~13,694 OP (+6.3%)
  • Annualized advantage: ~166,000 OP per year at this rate
  • Annualized % gain: The compounding fee effect would accelerate over time as the position grows

Key Takeaway

The tight distribution in Monte Carlo results (std dev ~282 OP on mean of ~218,923) shows that execution timing within each day has minimal impact on DCA outcomes. The real differentiator is the strategy choice itself. LP's median liquidity share of ~7.8% enabled meaningful fee capture from pool trading activity, and the fee compounding effect - while modest over 30 days - would accelerate over longer horizons.

Key Takeaway

In One Sentence

LP beats DCA by 6% because fee compounding outweighs execution timing variance.

In One Paragraph

This analysis demonstrates that for OP buybacks using protocol transaction fees, a simple wide-range LP position on Uniswap V3 systematically outperforms random-timed market purchases. The Monte Carlo simulations revealed that intraday timing matters very little (all 1,000 outcomes clustered within 0.8%), but strategy choice matters a lot. The LP approach earned 1,738 OP + 0.17 ETH in trading fees over 30 days, and the compounding effect — rolling daily fees back into the position — created a 6.3% advantage that would accelerate over longer time horizons. At this rate, LP would generate ~166,000 additional OP annually compared to random DCA.