Center for Social Research and Data Archives

The SSJDA Seminar

We are inviting interested participants to the SSJDA Seminars. The SSJDA Seminars provide young researchers from overseas the opportunity to exchange research ideas and present their latest empirical research results in the field of social sciences, thereby building an international network of young researchers. Please check the SSJDA Seminar Guidelines for details. We also welcome you to watch the video below for an introduction to the seminar and how to make the most of it.

The SSJDA Seminar [November 24, 2026] "Connectedness, Response Loading, and Network Resilience under Financial Uncertainty″

Date & Time
November 24(Tue), 2026, 15:00～16:00（JST）
Title
Connectedness, Response Loading, and Network Resilience under Financial Uncertainty
Speaker
Hiroto Imai* and Etsusaku Shimada(Iwate Prefectural University)　*Main presenter
Discussant
Keisuke Kawata (The University of Tokyo)
Venue/Tool
Online (Zoom)
Language
English and Japanese
Abstract
This study examines whether greater financial connectedness necessarily implies greater network fragility. Using Fama-French 49 industry returns, we develop an auditable state-conditioned diagnostic in which resilience depends jointly on network topology and model-implied response loading. A fixed recent benchmark establishes internal coherence, while a recovered January 1990-March 2021 panel supports Low- and High-VIX comparisons with 2,000 moving-block-bootstrap replications. High-VIX states exhibit substantially denser industry-return networks, yet state-specific loadings yield a larger stability margin, a higher model-implied local boundary, and a smaller diffuse cumulative response. Exploratory fixed-loading and market-factor-residual counterfactuals show that this density-fragility reversal reflects offsetting topology and loading channels. Rolling associations with subsequent volatility, industry dispersion, and drawdown magnitude are specification-sensitive, limiting a forecasting interpretation. More broadly, the analysis highlights how network structure and the allocation of common responses should be considered jointly when evaluating resilience, and provides a basis for extending the framework to computational environments involving common algorithmic signals.

PAGE TOP