Paper: arXiv 2512.23386

Authors: Sunghun Ko, Jinsuk Park

Abstract

We study Arbitrum’s Express Lane Auction (ELA), an ahead-of-time second-price auction that grants the winner an exclusive latency advantage for one minute. Building on a single-round model with risk-averse bidders, we propose a hypothesis that the value of priority access is discounted relative to risk-neutral valuation due to the difficulty of forecasting short-horizon volatility and bidders’ risk aversion. We test these predictions using ELA bid records matched to high-frequency ETH prices and find that the result is consistent with the model.

Complexity vs Empirical Score

  • Math Complexity: 8.5/10
  • Empirical Rigor: 7.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs advanced stochastic calculus and theoretical derivations for arbitrage profit models and bidder valuation under risk aversion, while backing these claims with high-frequency ETH price data and auction bid records from Arbitrum’s ELA.

Research Flowchart

  flowchart TD
  A["Research Goal<br>Quantify how volatility & risk aversion<br>affect the value of priority access<br>in Arbitrum's Express Lane Auction"] --> B["Theoretical Model<br>Single-round, second-price auction<br>Risk-averse bidders<br>Hypothesis: Value ≈ f(risk, volatility forecast)"]
  B --> C["Data & Methods<br>Match ELA bid records to<br>high-frequency ETH price data"]
  C --> D["Computation<br>Estimate volatility & test<br>bid sensitivity to price uncertainty"]
  D --> E["Key Findings<br>Priority access value is discounted<br>relative to risk-neutral valuation<br>Consistent with model predictions"]

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes a hypothesis that the value of priority access in the Express Lane Auction is discounted relative to risk‑neutral valuation because short‑horizon volatility is hard to forecast and bidders are risk‑averse.

Source sentence
we propose a hypothesis that the value of priority access is discounted relative to risk-neutral valuation due to the difficulty of forecasting short-horizon volatility and bidders’ risk aversion.

Assumes 5

  • The log price of the risky asset follows a semi‑martingale process with stochastic volatility. model
    Source sentence
    asset at CEX, which is a true price, follows a semi-martingale process with stochastic volatility, d (log Pt) = µt dt + σt dWt .
  • The time interval T is small enough that the price is approximately constant (P0 ≈ Pt). model
    Source sentence
    with assumption that T is small enough so that P0 ≈Pt for all t ∈[0, T].
  • Bidders are risk‑averse and participate in a single‑round second‑price auction. model
    Source sentence
    Building on a single-round model with risk‑averse bidders, we propose a hypothesis...
  • Gas fees and trading costs are ignored in the theoretical revenue proportionality. execution
    Source sentence
    In practice, a block is not generated continuously, and the arbitrageur has to bear the gas fees and trading costs. However, as shown
Show 1 more
  • Integrated variance (IV) is estimated from realized variance of a single historical price path. data
    Source sentence
    We adopt a standard approach, treating the realized variance (RV), the sum of squares of log returns rt, as ...

Data

  • ELA auction bid records · Arbitrum crypto-onchain proprietary
    May to October 2025 · All ELA rounds; focus on bidders 0x8c6f and 0x95c0
    Source sentence
    For the auction bids, we obtained the data directly from Arbitrum’s database [32]. Our sample spans bids from May to October 2025.
  • High‑frequency ETH price data · Binance market-prices public
    same as auction sample (May‑Oct 2025) · ETH spot price
    Source sentence
    We proxy market‑wide volatility using high‑frequency ETH price data from Binance.

Recognised sources: Binance

Costs

Unclear - sources disagree The platform classifier and the paper-text extraction give different answers, so no value is asserted.

Platform classifier: Not applicable

Paper text extraction: Partial - Gas fees and trading costs are mentioned as real costs but are not quantified in the headline revenue results.

Source sentence (the authors' words)
In practice, a block is not generated continuously, and the arbitrageur has to bear the gas fees and trading costs.

Finds 4

  • Bids increase with expected volatility and decrease with volatility uncertainty, confirming the hypothesized sign restrictions. statistical
    Source sentence
    Under (11), the theory predicts θ1,i > 0 and θ2,i < 0: bids should increase with expected volatility (more MEV to extract) and decrease with uncertainty about volatility (stronger risk discount).
  • A statistically significant negative variance risk premium is estimated, indicating a discount for volatility uncertainty. statistical
    Source sentence
    The empirical findings reveal a statistically significant discount attributed to the inherent uncertainty (i.e., risk) associated with forecasting future volatility and the generated arbitrage revenue.
  • The full model with a variance risk premium fits the data better than the restricted model without it. performance
    Source sentence
    This study shows that variance risk premiums are statistically significant and negative. Specifically, it demonstrates that the full model provides a better explanation of the findings compared to the restricted model.
  • Robustness checks across time and volatility regimes preserve the sign pattern of coefficients despite magnitude changes. statistical
    Source sentence
    Although the magnitude of the estimated coefficients varies across subsamples, the sign pattern is preserved.

Checked by others

Supported by

  • Delta-Hedged Gains and the Negative Market Volatility Risk Premium (2001)
    On the question: Does the market volatility risk premium appear negative (negative volatility risk premium effect)?
    Compare the findings

    This paper: A statistically significant negative variance risk premium is estimated, indicating a discount for volatility uncertainty.

    That paper: The evidence supports a negative market volatility risk premium.

    Both findings report a negative volatility risk premium; A estimates a statistically significant negative variance risk premium, and B confirms the same negative market volatility risk premium.

Contradicted by

  • 0DTE Trading Rules (2023)
    On the question: Does the variance risk premium significantly affect returns for volatility‑based trading strategies?
    Compare the findings

    This paper: A statistically significant negative variance risk premium is estimated, indicating a discount for volatility uncertainty.

    That paper: A positive 0DTE variance risk premium exists, but at same‑day horizons its economic magnitude is small and difficult to monetize after realistic frictions.

    A finds a statistically significant negative premium; B finds a positive premium but its economic magnitude is small and hard to monetize.

  • 0DTE Asset Pricing (2024)
    On the question: Does the variance risk premium predict market returns?
    Compare the findings

    This paper: A statistically significant negative variance risk premium is estimated, indicating a discount for volatility uncertainty.

    That paper: The high variance risk premium is mainly driven by compensation for upside risk and negatively predicts market returns.

    A finds a statistically significant negative variance risk premium (discount for volatility uncertainty), while B finds a high variance risk premium that positively compensates for upside risk and negatively predicts mar

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.

Cited References (48)

  1. Variance risk premiums (2009)
  2. Expected stock returns and variance risk premia (2009)
  3. The Express Lane to Spam and Centralization: An Empirical Analysis of Arbitrum’s Timeboost (2025) — link
  4. Buying Time: Latency Racing vs. Bidding in Transaction Ordering (2023) — link
  5. MEV Capture Through Time-Advantaged Arbitrage (2024) — link
  6. Automated market making and arbitrage profits in the presence of fees — → 2305.14604
  7. Loss-Versus-Rebalancing under Deterministic and Generalized block-times — → 2505.05113
  8. Automated market making and loss-versus-rebalancing (2022) — link
and 40 more
  1. A Theory of Stock Exchange Competition and Innovation: Will the Market Fix the Market? (2024)
  2. Heteroscedastic Censored and Truncated Regression with crch (2015) — link
  3. Arbitrum Timeboost (2024)
  4. A gentle introduction: Timeboost (2025)
  5. Pandora’s Box: Cross-Chain Arbitrages in the Realm of Blockchain Interoperability (2025)
  6. Cross-Rollup MEV: Non-Atomic Arbitrage Across L2 Blockchains (2024) — link
  7. TimeBoost: Do Ahead-of-Time Auctions Work? (2025) — link
  8. Measuring CEX-DEX Extracted Value and Searcher Profitability: The Darkest of the MEV Dark Forest — → 2507.13023
  9. Non-atomic arbitrage in decentralized finance — → 2401.01622
  10. Flash boys 2.0: Frontrunning, transaction reordering, and consensus instability in decentralized exchanges (2019) — link
  11. Exploiting the errors: A simple approach for improved volatility forecasting (2016)
  12. Uniswap v3: The Universal AMM (2021)
  13. Usage of Markout to Calculate LP Profitability in Uniswap V3 (2024)
  14. Follow-Up Analyses of LP Profitability in Uniswap V3 (2024)
  15. Priority Is All You Need (2024)
  16. $Proo\varphi$ : A ZKP Market Mechanism (2024) — link
  17. Resonance: Transaction Fees for Heterogeneous Computation (2024) — link
  18. The Crowding-In Effect of Public Information on Private Information Acquisition (2025)
  19. Liquid speed: A micro-burst fee for low-latency exchanges (2023)
  20. The high-frequency trading arms race: Frequent batch auctions as a market design response (2015)
  21. Auction theory (2009)
  22. TimeBoost and Backrunning: Probabilistic Strategies (2025)
  23. Our Express Lane Sub-Auction (2025)
  24. A Guide to Volatility and Variance Swaps (1999) — link
  25. The Model-Free Implied Volatility and Its Information Content (2005) — link
  26. Delta-Hedged Gains and the Negative Market Volatility Risk Premium (2003) — link
  27. Model Specification and Risk Premia: Evidence from Futures Options (2007) — link
  28. Tails, Fears, and Risk Premia (2011) — link
  29. The Price of Variance Risk (2017) — link
  30. The Term Structure of Equity and Variance Risk Premia (2020) — link
  31. Downside Variance Risk Premium (2018) — link
  32. Variance Risk Premiums and the Forward Premium Puzzle (2017) — link
  33. Variance Risk in Commodity Markets (2017) — link
  34. Equity Volatility Term Premia (2018) — link
  35. Arbitrum: Timeboost (2025)
  36. How to use Timeboost (2025)
  37. Arbitrum One - L2BEAT (2025)
  38. Realized Range-Based Estimation of Integrated Variance (2007) — link
  39. Does measurement error matter in volatility forecasting? Empirical evidence from the Chinese stock market (2020) — link
  40. Microstructure noise in the continuous case: The pre-averaging approach (2009) — link

4 of 48 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.