Dominik Svoboda

Working Paper

Bot-Induced Social Media Manipulation and Stock Market Distortion

with Daniel Bradley, Jan Hanousek Jr., and Jan Hanousek

Using an exogenous shock to bot activity, we show that bot presence and their impact rise with the ease of infiltrating social platforms. Bots significantly manipulate social media attention and are most likely to target firms that are more volatile with recent momentum and around earnings releases. When bot penetration is high, stock price synchronicity increases and post-earnings announcement drift is stronger, indicating that bots lead to less efficient prices. Bot accounts promoting stocks generate sharp positive returns followed by partial reversals and retail order imbalances likely exacerbating trading frenzies. Overall, bots create information frictions that distort market prices. Our evidence reinforces regulators' concerns about AI-driven stock manipulation.

Working Paper

Information Gaps and Price Efficiency: The Role of Social Media

with Jonathan Brogaard, Jan Hanousek Jr., and Velma Zahirovic-Herbert

We study the role of social media in financial markets in a setting with well-defined information frictions. Using exogenous shocks from natural disasters and a delay in official reporting, we identify periods in which investors lack reliable information about local fundamentals. We show that in the absence of coverage, municipal bond yields exhibit an initial reaction followed by a predictable correction as information becomes available. Exploiting the staggered introduction of location-based social media, we find that these corrections disappear when decentralized information is present. Our results demonstrate that social media improves price efficiency by expanding information availability when traditional sources are limited.

Published · The Journal of Real Estate Finance and Economics

Beyond the Flood: The Pricing of Media Silence in Housing Markets

with Jan Hanousek Jr. and Velma Zahirovic-Herbert

We examine how housing markets respond to local shocks when public information is asymmetrically distributed. Using major flooding events in Florida, we compare transactions in affected counties with and without local disaster-related media coverage. We find that properties in affected locations without coverage sell at significantly lower prices than comparable properties in covered areas, with media silence adding roughly 1% point to the post-flood discount. This additional discount is concentrated among buyers with in-state mailing addresses and institutional buyers, consistent with greater access to local or alternative information channels, while out-of-state buyers show no systematic price adjustment in media-silent areas. The pricing effect appears stronger in higher-income neighborhoods, a pattern we interpret as suggestive that information frictions persist even in financially resilient markets. Taken together, the results are consistent with media silence shaping post-disaster valuations and with stronger relative advantages for buyers who have access to local or alternative information channels.

Published · The North American Journal of Economics and Finance · 2026

Systematic signals of short squeezes: insights from rare events

with Svatopluk Kapounek and Peter Albrecht

Our paper investigates the determinants of short squeeze occurrences from rare events using a unique hand-collected dataset. Our results show that elevated short interest and spikes in investor attention significantly increase the likelihood of a short squeeze, while institutional ownership has a stabilizing effect. These findings suggest that short squeezes are not purely random events but can be systematically anticipated based on observable market signals. The study offers implications for traders, portfolio managers, and regulators seeking to better understand and monitor the conditions under which short squeezes are likely to occur.