Roxana Mihet, University of Lausanne and Swiss Finance Institute
Kumar Rishabh, University of Lausanne
Abstract: How and to what extent does cyber risk impact firms in the modern data
economy? When firms face a high risk of losing their data and algorithms, this
unequivocally leads to reduced knowledge stocks, decreased productivity, and
slower overall economic growth. Notwithstanding, the empirical analysis pursued
in this study suggests that cyber risk may also mitigate some of its own adverse
effects as it ex-ante prompts digitally-savvy firms to pursue digital innovation
that can enhance productivity in other domains. We observe increased innovation
rates in response to higher cyber risk, driven primarily by data-intensive firms and
by firms which intensively pursue in-house cyber security protection rather than
third-party cyber security delegation. In a second stage, we develop a structural
heterogeneous-firm growth model of the data economy to illustrate and explain
the channels through which cyber risk exerts influence over firms’ productivity,
profits and growth.
Yan Xiong, Hong Kong University of Science and Technology
Abstract: We study a credit market in which the lender bases its lending decisions on a borrower's digital profile, and the borrower can manipulate its digital profile at a cost. We show that when the extent of data collected by the lender is observable, as the lender utilizes more data in its underwriting models, the borrower is more likely to manipulate their digital profile, which impairs the quality of the lender's data and its lending decisions. Therefore, even if obtaining and analyzing additional data is costless, the lender will voluntarily limit its own data coverage. In contrast, when the data coverage is unobservable, the lender tends to use all available data. Thus, disclosure policy can play a valuable role in allowing the lender to credibly commit to limiting its data coverage. Moreover, in the aggregate, borrowers too prefer that some digital data be collected.
Jeremy Bertomeu, Washington University in St. Louis
Yupeng Lin, National University of Singapore
Yibin Liu, National University of Singapore
Zhenghui Ni, National University of Singapore
Abstract: On March 31, 2023, the Italian Data Protection Authority found ChatGPT in violation of privacy laws and banned the service in Italy, providing a natural experiment to evaluate the economic impact of generative AI. Italian firms with higher exposure to the technology suffered a negative market reaction of about 9% during the ban period compared to firms with lower exposure. The negative impact was greater for smaller and more newly established businesses, supporting a link with creative destruction. We also document that the ban affected the information environment. Italian-based analysts issued fewer forecasts than foreign analysts covering the same Italian firm. Moreover, bid-ask spreads widened during the ban, especially for firms with fewer institutional investors, limited analyst coverage, and fewer foreign investors. Overall, our results suggest a dual role for generative AI in enhancing firm productivity and information processing.
Discussant: Gregor Schubert, University of California-Los Angeles
