Job Market Paper
Strategic Firm Capacity Selection under the Network Effects: A Computational Model of Capacity Investment Dynamics in the ICT Industry
Abstract: Traditional industrial organization theory often interprets R&D investment as a driver of innovation and technological progress. In contrast, this paper focuses on a different but equally critical dimension of R&D—the physical investment in production and service capacity, which determines a firm’s ability to deliver high-quality services. In many ICT markets, profitability is not limited by the absence of innovation but by the constraints of infrastructure and capacity utilization. When firms exceed their optimal service capacity, product quality deteriorates and consumer satisfaction declines. Moreover, consumer behavior is increasingly shaped by digital community interactions and network effects, which can amplify or dampen the market value of these physical investments. To capture these mechanisms, we develop an agent-based model of competing ICT firms that strategically select their capacity levels while interacting with consumers embedded in a social network. The model uncovers how network effects, congestion, and capacity adjustments jointly shape market outcomes, profitability, and consumer engagement. By emphasizing the strategic role of physical R&D investments, this paper contributes to a deeper understanding of capacity management and network-driven competition in digital service industries.
Publications
Polytope Fraud Theory, International Review of Financial Analysis, Volume 97, 2025.
Coauthored by: Dongshuai Zhao, Florian Schweizer, Didier Sornette
Abstract: Polytope Fraud Theory (PFT) extends the existing triangle and diamond theories of accounting fraud with ten abnormal financial practice alarms that a fraudulent firm might trigger. These warning signals are identified through evaluation of the shorting behavior of sophisticated activist short sellers, which are used to train several supervised machine learning methods in detecting financial statement fraud using published accounting data. Our contributions include a systematic manual collection and labeling of companies that are shorted by professional activist short sellers. We also combine well-known asset pricing factors with accounting red flags in financial features selections. Using 80 % of the data for training and the remaining 20 % for out-of-sample test and performance assessment, we find that the best method is XGBoost, with a Recall of 72 % and F1-score of 82 %. Other methods have relatively lower performance, demonstrating the robustness of our results. This shows that the sophisticated activist short sellers, from whom the algorithms are learning, have excellent accounting insights, tremendous forensic analytical knowledge, and sharp business acumen. Our feature importance analysis indicates that potential short-selling targets share many similar financial characteristics, such as bankruptcy or financial distress risk, clustering in some industries, inconsistency of profitability, high accrual, and unreasonable business operations. Our results imply the possible automation of advanced financial statement analysis, which can both improve auditing processes and effectively enhance investment performance. Finally, we propose the Unified Investor Protection Framework, summarizing and categorizing investor-protection related theories from the macro-level to the micro-level.
Link: https://doi.org/10.1016/j.irfa.2024.103734
Papers in progress
Heterogeneous Consumers, Firm Competition and the Data Value Chain: An Agent-based Approach
Abstract: The advancement of Information and Communications Technologies (ICTs) and the trend towards digitalization have significantly fueled the growth of software applications. These applications are distinct from traditional goods due to their reliance on information systems, the intangible nature of their products, and the critical role of licensing. The pricing and availability of these applications differ internationally due to varying local data restrictions. However, the question arises: how does the level of data restriction influence the geographical distribution of economic activities over time? To decipher the impact of data restriction levels on the software market, we employ a game theoretical approach and an agent-based approach involving entities such as software owners and developers, consumers, and governments. Our model includes the interactions between software owners and developers, the selection process of diverse consumers, the tactics of software owners, and the fluctuation of prices over time. The findings from our computational experiments indicate that the intensity of data restrictions can influence the process of price formation. We recommend that governments in different regions should work collaboratively to minimize data regulation and better understand consumer preferences, aiming to benefit both consumers and local businesses.
Link: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5235196
Parameter Estimation Methods in an Agent-based Computational Model: Insights on ICT Industrial Dynamics and Policies
Abstract: Despite the potential of the agent-based modeling (ABM) to uncover emergent phenomena and provide policy insights, the estimation and calibration of model parameters remains a significant challenge. This paper addresses this challenge by proposing a computational model tailored to the ICT industry, where strategic firm behavior and network effects are central. We apply recent advances in parameter estimation—specifically the simulated method of moments and neural network-based surrogate models—using financial data from U.S. telecommunication firms (AT&T and T-Mobile) to calibrate and validate the model, and to identify parameter values and distributions. By systematically comparing these calibration methods, we assess their suitability and practical implications for agent-based economic modeling. We aim to advance the methodological rigor in ABM parameterization and contribute to the broader discussion on improving the validity and applicability of computational economic models. Our findings provide insights into overcoming industry-specific challenges, enhancing the reliability of ABMs, and bridging the gap between theoretical modeling and real-world applications.
Agentic Workflows for Economic Research: Design and Implementation
Coauthored by: Hankui Wang, Herbert Dawid, Philipp Harting, Jiachen Yi
Abstract: This paper introduces a methodology based on agentic workflows for economic research that leverages Large Language Models (LLMs) and multimodal AI to enhance research efficiency and reproducibility. Our approach features autonomous and iterative processes covering the entire research lifecycle--from ideation and literature review to economic modeling and data processing, empirical analysis and result interpretation--with strategic human oversight. The workflow architecture comprises specialized agents with clearly defined roles, structured inter-agent communication protocols, systematic error escalation pathways, and adaptive mechanisms that respond to changing research demand. Human-in-the-loop (HITL) checkpoints are strategically integrated to ensure methodological validity and ethical compliance. We demonstrate the practical implementation of our framework using Microsoft's open-source platform, AutoGen, presenting experimental examples that highlight both the current capabilities and future potential of agentic workflows in improving economic research.
Conferences, workshops and seminars
Oct. 2025 Organizer and speaker of the “AI agents in science” workshop at Wissenswerkstadt Bielefeld
Sept. – Oct. 2024 Tenth Meeting of the German Network for New Economic Dynamics (GENED)
Aug. 2024 CHIMSPAS 2024 International Conference on Challenges In Managing Smart Products And Services
July 2024 27th Annual Workshop on Economics with Heterogeneous Interacting Agent; Economic Policy in Complex Environments Conference in Milan
Mar. 2024 Organization Team Member of the International EPOC Doctoral Workshop at Ca’Foscari University of Venice
Jan. 2024 Participant of Graz Schumpeter Summer School 2024
Jan. 2024 Participant of the 9th World Congress of the International Microsimulation Association (my presentation)
Oct. 2023 Participant of GENED (German Network for New Economic Dynamics) 2023
June 2023 Participant of ISEO Summer School
Apr. 2022 – Dec. 2022 Organization Team Member of the 17th BiGSEM Workshop
Ideas and toolkits