Abstract: This paper examines the interconnectedness of stress-tested banks using financial news coverage. We use the COVID-19 pandemic as an exogenous shock to investigate the behavior of bank networks during periods of stress. We then propose a new measure of systemic risk using text-based eigenvector centrality, a relative metric of influence within a network. We show that this measure provides a valuable complement to existing systemic risk measures. Our findings highlight the importance of soft information in the context of financial stability. Our approach offers a novel tool to study the financial system’s architecture and complements insights from traditional structured data.
Abstract: Using recent advances in network theory, we estimate the intra-industry volatility connectedness for US publicly traded companies going back to the 1920s. Our volatility connections are related to the vertical upstream/downstream potential of peer firms in the supply chain and to their technological closeness. We develop a stock-level composite centrality measure that captures multiple dimensions of a stock’s interdependence with its industry peers. Using our network and composite centrality estimates, we develop "peer momentum" trading strategies, which sort stocks on their industry peers’ past month average returns weighted by the peers’ influence in the industry. A "peripheral peer momentum" strategy that uses only peripheral stocks’ influence for weighting in the signal construction achieves an annualized Sharpe ratio of 0.65, survives a battery of robustness tests, and helps explain industry momentum.
Abstract: This paper explores the link between systemic risk and firm value gain in the context of bank acquisitions. We find that bank acquisitions on average lead to an increase in acquires' systemic risk (a public cost), which is in turn associated with an increase in firm value (a private benefit) for non-distressed acquisitions. Indeed, acquisitions can create value for acquirers' shareholders by forming larger firms with more diversified interconnections or asset correlations making the acquirer “too-interconnected-to-fail" and increasing the potential for government forbearance in case of failure. Our results are even stronger for diversifying acquisitions and robust to controlling for size of consolidated firm, “too-big-to-fail" mechanism, and type of target.
Work in Progress
Unveiling Novel Activities and Banking Dynamics: Harnessing the Potential of Financial News (with John Wu, Bryson Alexander, and Ethan Buttler)
Beyond Categorization: Advancing Sentiment Polarity with Transformer-Based Models (with Brian Ferrell)
CDS Networks (with Ping McLemore)
High frequency Volatility and Financial Networks (with Cooper Killen)