Accepted papers and working drafts from Hayashi.

Adrasteia: AI Agent Workflow for Economic Paper Writing

T.Y. Hayashi, Huanxi Zhang (University of Wisconsin-Madison), Shawn Yu (Boston College), Ruoran Lai (Sun Yat-sen University), and Jiazhuo Li (University of Michigan)

Accepted at the ICML 2026 Workshop on AI for Math

Presented at: EC'26 Workshop on AI-Driven Research in EconCS (oral contributed talk); Can AI Do Theory? Workshop at STOC TheoryFest 2026 (Google Research); AEA Annual Meeting 2027 (scheduled)

Adrasteia presents a human-in-the-loop AI agent workflow for economic paper writing, spanning assumption search, model structuring, argument revision, and referee-style critique.

How Much Must a Private Mempool Hide? Exact Leakage Thresholds for Sandwich Attacks

T.Y. Hayashi, Jiazhuo Li (University of Michigan), and Ruoran Lai (Sun Yat-sen University)

Accepted at NeurIPS 2026

Private and encrypted mempools hide pending transactions to stop sandwich attacks and other forms of maximal extractable value (MEV), but what they hide is rarely everything: a transaction's pair, direction, and a coarse range for its size can still leak. How much leakage makes sandwiching pay? We answer exactly for a fee-free constant-product automated market maker, the pricing rule behind Uniswap v2.

When One Leak Pays Forever: Context Binding and the Price of Deterring Collusion

T.Y. Hayashi, Shawn Yu (Boston College), Ruoran Lai (Sun Yat-sen University), and Huanxi Zhang (University of Wisconsin-Madison)

Accepted at NeurIPS 2026

A coalition that deviates once can profit many times when what it sells keeps working. In a threshold-encrypted mempool, a leading defense against maximal extractable value (MEV), a quorum of the decryption committee that sells its decryption capability to a front-runner exposes every later block that the capability still decrypts. We ask how large a penalty, such as slashable stake, deters this kind of collusion.

Smart Contracts, Dumb Money: Who Pays to Read the Code?

T.Y. Hayashi, Jie Liu (Wuhan University), and Huanxi Zhang (University of Wisconsin-Madison) (authors in alphabetical order)

Presented at: Bank of Italy ToDeFi 2026*; AFA PhD Poster Session (scheduled) (* presented by coauthor)

The project examines information quality, adverse selection, and investor behavior in open-source smart-contract environments and digital asset markets.

Smoothing the Cliff: Fair and Incentive-Compatible Priority Allocation via Randomized Mechanisms

T.Y. Hayashi, Shawn Yu (Boston College), and Huanxi Zhang (University of Wisconsin-Madison) (authors in alphabetical order)

Presented at: EC'26 Workshop on Online Learning and Economics; 12th Tsinghua University SEM Graduate Workshop (Outstanding Presentation Paper Award); EEA-ESEM*; ESIF Economics and AI+ML Meeting (Cornell University)*; KOF Swiss Economic Institute at ETH Zurich* (* presented by coauthor)

This paper studies randomized allocation mechanisms designed to smooth cliff effects while preserving fairness and incentive compatibility in priority-based environments.

The Private Enforcer: Algorithmic Deterrence and the Shadow Tax on Insider Trading

T.Y. Hayashi, Ruoran Lai (Sun Yat-sen University), Weiyu Qi (University of Chicago), and Shawn Yu (Boston College) (authors in alphabetical order)

Working Paper

Presented at: Federal Reserve Board Summer Workshop on Money, Banking, Payments, and Finance (Gerzensee, Switzerland); China Financial Research Conference (Tsinghua PBCSF); Asia Meeting of the Econometric Society, China (HKU Business School); 38th Asian Finance Association Conference* (* presented by coauthor)

In DeFi markets, MEV acts as a private enforcement mechanism that penalizes informed trading and creates a shadow tax on strategic trading behavior.

A Blessing in Disguise? DeFi Exploits and Short-Horizon Responses in U.S. Commercial Paper Spreads

T.Y. Hayashi and Tse Hou (University of Minnesota and Federal Reserve Bank of Minneapolis) (authors in alphabetical order)

Dormant working paper

Accepted for presentation at: Yale Pre-Doctoral Economics Conference (Yale Tobin Center); Asia Meeting of the Econometric Society, China (HKU Business School); 5th Annual Hong Kong Conference on FinTech and AI in Finance (City University of Hong Kong)

DeFi exploits generate short-run spillovers into traditional financial markets via money market fund reallocations, creating a measurable cross-market transmission channel.

Why dormantEarly hand-coding of the dataset introduced errors in date matching. An LLM check in 2026 caught them, and with the dates corrected the estimated effect is no longer robust, so the paper has been put on hold in the interest of rigor. The episode shows why LLMs matter as research assistants: the errors had passed through manual coding unnoticed.