In the era of green transition and financial innovation, the stability of the financial system is a critical prerequisite for providing the long-term funding that financial institutions require to navigate transformation and undertake innovative projects. However, the frequent occurrence of tail events in recent years has severely impacted this stability, greatly undermining investor confidence and leading to a withdrawal of patient capital. This erosion of the funding base directly impairs the financing capacity of financial institutions, thereby adversely affecting the progress of green transition and the investment in financial innovation. To deeply analyze the contagion paths that exacerbate such vulnerabilities, this research explores the interactive mechanism between financial institutions’ tail risk and investor sentiment from a complex network perspective, employing the Tail-Event Driven Network Risk (TENET) model and a generalized predictive variance decomposition method. Our empirical findings reveal that: First, risk spillovers are primarily intra-sectoral, with investor sentiment exhibiting a more pronounced clustering effect, particularly in the securities sector. Second, despite differences in individual institution centrality, the overall structure of the tail risk and investor sentiment networks is highly similar. Third, and most critically, significant cross-contagion exists between the two layers, with investor sentiment acting as a key information transmission channel for tail risk. Finally, during tail events, the synchronic increase in interlayer connections confirms the time-varying and highly correlated nature of these spillover effects. The findings of this study are crucial for understanding how tail risk contagion can trigger the withdrawal of patient capital and disrupt financing channels. They provide valuable insights for policymakers aiming to mitigate such systemic vulnerabilities, thereby safeguarding the financial foundations necessary to sustain both green transition and financial innovation.
Liu et al. (Thu,) studied this question.