Transaction Costs and Speed in the Ethereum Ecosystem: Scalability of the Mainnet and Layer 2s

Battle Count: 74

The paper is relevant to quantitative trading in the DeFi and blockchain space. Transaction costs and speed are critical inputs for algorithmic trading strategies on DEXs, cross-chain arbitrage, and market-making on blockchain networks. The forecasts of when Ethereum Mainnet and L2 fees converge with Solana levels (August 2027 and October 2026 respectively) directly impact strategy selection across chains. However, the paper focuses on infrastructure-level metrics rather than asset pricing, portfolio construction, or trading signal generation, limiting its direct applicability to traditional quantitative trading.

The methodology consists of standard OLS linear regressions with HAC standard errors, time trend extrapolation, and simple algebraic forecasting equations. All models are single-variable or two-variable linear specifications. Implementation requires only basic econometric software (e.g., Python statsmodels, R, Stata). The main complexity lies in data collection and aggregation from Dune and Google BigQuery, not in the modeling itself.

The paper provides public Dune query URLs (e.g., https://dune.com/queries/7347447, https://dune.com/queries/6946173, https://dune.com/queries/6962020, https://dune.com/queries/6946454) and references Google BigQuery public datasets (bigquery-public-data.crypto_ethereum, bigquery-public-data.goog_blockchain_polygon_mainnet_us). ETH and POL prices are sourced from CoinGecko. All regression specifications, equations, and forecast models are fully documented. However, no code repository is provided, and some data processing steps (e.g., aggregation methods, fee conversion) are described in text rather than code.