Disagreement and Asset Pricing: Evidence from Prediction Markets [Job Market Paper]
with Simeng Li
Draft available upon request
How beliefs about future events-and the disagreement among them-shape asset prices is a central question in finance, but existing empirical measures of investor beliefs are indirect, low-frequency, or aggregated. We address this using the universe of transaction-level data from Polymarket, a money-backed prediction market, over June 2024 through March 2026. From the same transaction tape, in the common metric of probabilities, we construct two high-frequency primitives: a price-entropy measure of consensus uncertainty and a dollar-volume-weighted measure of investor disagreement. We establish three results. First, the prediction market and the U.S. equity market are informationally integrated in both directions: stocks lead Polymarket within the trading day, while overnight Polymarket innovations predict the next morning's equity open. Second, disagreement positively predicts next-day stock returns-opposite in sign to analyst-forecast dispersion-consistent with compensation for unspanned macro state risk rather than mispricing. Third, consensus uncertainty predicts lower near-term equity volume and realised volatility, with the volatility released as uncertainty resolves, so it governs the timing of repricing rather than its level. A parsimonious heterogeneous-beliefs model rationalises the full pattern: disagreement is priced, while uncertainty is traded.
Presentations and scheduled presentations: AFA 2027 (Poster), AFBC 2026, SBFC 2026, FMA Asia/Pacific Conference 2026, FMA 2026 (Doctoral Consortium), SFA 2026, HEC Paris, Finance Research Revolution 2026, China Fintech Research Conference 2026, QMUL*, Warwick*, Future Finance Fest*, Bayes Business School* (Poster).
When Supply Becomes a Demand Signal: Evidence from UK Gilt Issuance
Draft available upon request
For decades, research on sovereign-debt issuance has centred on whether uniform- or discriminatory-price auctions best minimise government borrowing costs. This paper argues that the debate may have overlooked subtle design features that can meaningfully affect outcomes. I study the United Kingdom’s Post-Auction Option Facility (PAOF)—introduced in 2009—which allows the Debt Management Office to sell an additional tranche to winners at the average accepted price. Using the universe of gilt auctions from 1998 to 2017, I find three main results. First, primary dealers bid more aggressively after the PAOF’s introduction. Second, the mechanism reduced total fiscal cost by roughly £1.2 billion. Third, the DMO’s activation decision—announced with results—moves secondary prices in opposite directions: activation raises prices (lowers yields), while non-activation depresses them, consistent with the model’s signaling channel. Together, these findings show that even small, responsive-issuance mechanisms can enhance information revelation and materially lower sovereign borrowing costs without altering the core auction design.
Presentations and scheduled presentations: SFA 2026, SWFA 2026 (Doctoral Consortium Award), LSE, International Finance Society Annual Meeting (HKU iCube, Poster), University of Naples Federico II.
The Incidence of Interchange Regulation
with Gilles Chemla, Marco Di Maggio, and Simeng Li
Draft available on https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7410038
We study two opposite changes in card-interchange regulation observed at the same European issuer. In October 2021, the post-Brexit removal of European caps from UK-EEA card-notpresent transactions generated a positive issuer-revenue shock. In November 2025, a Swiss competition-authority settlement cut domestic debit interchange and generated a negative shock. The price response is sharp and corridor-specific in both directions. The Brexit repricing raises issuer interchange yield by 1.45 log points per unit of corridor exposure, implying a 55 percent increase at the United Kingdom's exposure; the estimate has flat pre-trends, a null placebo year, and survives exact wild-cluster and randomisation inference with thirteen countries. The Swiss cap lowers the effective domestic rate by 10.75 basis points relative to EEA controls and does not transmit to adjacent corridors. Downstream incidence is asymmetric. Following the Swiss cut, the free plan loses 13 percent of its users, its share migrates toward paid plans, and remaining users spend more; the user base adjusts within a quarter while the rate adjusts over roughly a year. Following the Brexit increase, spending per user, transactions per user, and ticket size show no statistically detectable change, while aggregate users and spending are too imprecisely estimated to distinguish acquisition from no response. A simple plan-pricing framework rationalises the pattern: when a mass-market account is already priced at zero, an interchange increase need not be passed through, whereas a decrease forces adjustment through prices, benefits, acquisition, or upgrades. The evidence identifies the payment corridor as the unit of regulation and shows that a cap's first-stage effect can be close to mechanical while its consumer incidence is state-dependent.
Debt Demand Shocks and the Eroding Government Bond Convenience Yield
with Philippe Mueller, Andreas Schrimpf, and Dora Xia
Draft available upon request
This paper utilizes high-frequency transaction data around U.S. Treasury auctions to isolate pure demand shocks and analyze their impact on the sovereign convenience yield. Building on inelastic market theory and intermediary constraints, we document a "loss of uniqueness" channel: weak auction demand systematically compresses the safety premium of U.S. debt, both domestically and across international synthetic markets. These findings demonstrate that high-frequency demand flows, rather than just aggregate macroeconomic supply, play a persistent and crucial role in the pricing of global safe assets, particularly outside of Quantitative Easing regimes.
Presentations and scheduled presentations: HKMA.
(*presented by co-author)