Confidence-Weighted Prices: What Prediction Markets Elicit That Surveys Cannot

A prediction-market price is usually read as an average of traders' beliefs, weighted by wealth or risk tolerance. A well-scored survey of the same traders should then do as well, and in recent comparisons it does. We show that a market collects something such a survey cannot, traders' confidence in their own forecasts. No payment based on the outcome of a binary event gives an expected-utility respondent a strict incentive to report her confidence. In a call market, an ambiguity-averse trader who chooses her demand price by price reveals both her forecast and its weight in the price. She reveals them exactly, for any second-order belief and any smooth-ambiguity preference with constant absolute attitudes, and the firm can undo strategic shading. The reason is that ambiguity aversion makes a trader's willingness to trade away from her forecast depend on how sure she is of it. Two further results follow. To a second-order approximation the price is a forecast combination weighted by self-assessed precision. With its weights fixed, it beats the survey mean exactly when those weights are closer to the optimal weights than equal weights are. It loses when the least accurate traders are the most overprecise, and without position limits it can overshoot the truth. With dispersed information it gains much only when traders draw on few independent sources. In simulations, outcome-scored track records reveal little about accuracy, so the schedules become the firm's main source. A firm should therefore collect the schedules, not only the price. Working paper, not peer reviewed. Includes the electronic companion (pages ec1-ec7). Code: https://doi.org/10.5281/zenodo.24369926

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-19
DOI
https://doi.org/10.5281/zenodo.24370671
Primary Topic
Forecasting Techniques and Applications
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Confidence-Weighted Prices: What Prediction Markets Elicit That Surveys Cannot

Gwang-Syung George Hwang
Zenodo (CERN European Organization for Nuclear Research)
Forecasting Techniques and Applications
preprint

Confidence-Weighted Prices: What Prediction Markets Elicit That Surveys Cannot

Gwang-Syung George Hwang
preprint en

Abstract

A prediction-market price is usually read as an average of traders' beliefs, weighted by wealth or risk tolerance. A well-scored survey of the same traders should then do as well, and in recent comparisons it does. We show that a market collects something such a survey cannot, traders' confidence in their own forecasts. No payment based on the outcome of a binary event gives an expected-utility respondent a strict incentive to report her confidence. In a call market, an ambiguity-averse trader who chooses her demand price by price reveals both her forecast and its weight in the price. She reveals them exactly, for any second-order belief and any smooth-ambiguity preference with constant absolute attitudes, and the firm can undo strategic shading. The reason is that ambiguity aversion makes a trader's willingness to trade away from her forecast depend on how sure she is of it. Two further results follow. To a second-order approximation the price is a forecast combination weighted by self-assessed precision. With its weights fixed, it beats the survey mean exactly when those weights are closer to the optimal weights than equal weights are. It loses when the least accurate traders are the most overprecise, and without position limits it can overshoot the truth. With dispersed information it gains much only when traders draw on few independent sources. In simulations, outcome-scored track records reveal little about accuracy, so the schedules become the firm's main source. A firm should therefore collect the schedules, not only the price. Working paper, not peer reviewed. Includes the electronic companion (pages ec1-ec7). Code: https://doi.org/10.5281/zenodo.24369926

Zenodo (CERN European Organization for Nuclear Research)
Forecasting Techniques and Applications
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon