Banking on Instability: How Shocks Ripple Through the U.S. Financial System

by

Author: Denis Avetisyan

New research reveals that a select group of U.S. banks act as critical conduits for systemic risk, amplifying the impact of economic and policy changes.

This study analyzes dynamic connectedness between U.S. banks, identifying key transmitters and absorbers of systemic shocks using a time-varying parameter vector autoregression (TVP-VAR) model, and assessing the roles of monetary policy and market sentiment.

Despite increasing regulatory scrutiny, the 2023 U.S. banking crisis revealed vulnerabilities beyond traditional asset-by-asset analysis. This paper, ‘Dynamic Risk in the U.S. Banking System: An Analysis of Sentiment, Policy Shocks, and Spillover Effects’, investigates how information-based contagion-driven by perceived similarities in bank business models-amplified systemic risk during a period of rapid interest rate increases. Our analysis demonstrates that a subset of institutions acted as primary transmitters of risk, while others largely absorbed these shocks, a dynamic significantly intensified by market sentiment and policy uncertainty. Can real-time monitoring of these shifting network connections provide a more robust early warning system for future financial instability?

The Inevitable Cascade: Recognizing Fragility in Financial Systems

The recent series of bank failures served as a stark reminder that conventional risk assessment models often fall short in capturing the subtle, yet critical, vulnerabilities within the financial system. These failures weren’t simply isolated incidents; they exposed a systemic weakness where traditional metrics – focused largely on accounting data and static balance sheets – failed to adequately anticipate rapidly evolving market perceptions and liquidity pressures. Consequently, a shift towards more dynamic monitoring practices is essential, focusing on real-time indicators that reflect investor confidence and potential contagion effects. This necessitates moving beyond retrospective analysis to embrace proactive surveillance of financial institutions, allowing for earlier detection of distress signals and, crucially, more effective intervention strategies before localized problems escalate into broader systemic crises.

Bank stock returns function as a vital barometer of systemic risk due to their capacity to rapidly incorporate and disseminate information regarding a financial institution’s perceived health. Unlike lagging indicators such as balance sheet data, stock prices reflect immediate market assessments of solvency, stability, and future performance, effectively acting as an early warning system. Declines in stock value can signal mounting investor concerns about asset quality, liquidity, or earnings potential, prompting preemptive action before these concerns escalate into broader financial instability. This sensitivity stems from the fact that equity markets aggregate the collective intelligence of numerous actors, translating complex financial data and anticipating potential vulnerabilities with remarkable speed, making stock returns a continuously updated, forward-looking measure of risk within the banking sector.

The interconnectedness of financial institutions means that distress at one bank can rapidly propagate throughout the system, creating a cascade of failures; therefore, quantifying these relationships is paramount to proactive risk management. Recent analysis reveals a Total Connectedness Index (TCI) of 67% amongst major banks, demonstrating that nearly two-thirds of a bank’s risk profile stems not from its own internal vulnerabilities, but from external exposures and spillover effects. This high degree of interconnectedness implies that even seemingly isolated problems can quickly amplify, as negative signals – reflected in declining stock returns – transmit across the network. Understanding these return correlations allows for the identification of systemically important institutions and vulnerable pathways, enabling regulators and institutions to implement targeted interventions and bolster overall financial stability before localized issues escalate into widespread crises.

Mapping the Web: A Network Perspective on Systemic Interdependence

Connectedness Analysis, as applied to the US banking system, utilizes a variance decomposition approach to quantify the extent of interdependence among financial institutions. This methodology determines the proportion of each bank’s shock-changes in its financial state-that is transmitted to and impacts other banks within the system. The analysis considers both direct linkages – relationships between specific banks – and indirect connections, capturing systemic risk transmission through multiple pathways. Results are presented as Total Connectedness Indices, measuring the overall interconnectedness, as well as pairwise connectedness measures detailing the degree to which one bank influences another. This allows for identification of systemically important financial institutions and vulnerabilities within the network structure, providing a more comprehensive assessment of financial stability than traditional correlation-based approaches.

Cointegration analysis, a time-series econometric technique, was used to determine if a long-run equilibrium relationship exists between the stock returns of individual banks within the US banking system. This analysis goes beyond simple correlation by assessing whether deviations from this equilibrium are mean-reverting, indicating a tendency for prices to return to a stable, shared relationship. Identification of cointegrated relationships suggests that shocks to one bank’s stock price are likely to be transmitted to, and influence, the stock prices of other cointegrated banks over time, providing evidence of systemic linkages beyond short-term correlations. The presence and strength of these cointegrated relationships were quantified to better understand the potential for risk transmission within the network and to inform systemic risk assessments.

Correlation analysis served as the foundational step in constructing the network model by initially quantifying pairwise relationships between US banks. This analysis determined the strength and direction of these relationships, providing the raw data for network construction. The resulting Total Connectedness Index (TCI) of 67% indicates a high degree of interconnectedness within the US banking system; this value represents the average percentage of a bank’s shock that is transmitted, directly and indirectly, to all other banks in the network, thereby confirming substantial systemic risk transmission potential.

Tracing the Flow: Identifying Institutions as Shocks’ Transmitters and Absorbers

Analysis identified several financial institutions functioning as consistent ‘Net Transmitters’ of systemic shocks within the observed network. Specifically, Silicon Valley Bank (SIVB), First Republic Bank (FRC), and Washington Federal exhibited positive NET values, indicating their roles in propagating risk to other entities. These banks consistently contributed to overall financial instability, as opposed to absorbing shocks originating from elsewhere in the system. The magnitude of their positive NET values, quantified through the TVP-VAR model, demonstrates the extent of their influence as primary sources of risk transmission during the period analyzed.

Analysis identified several financial entities functioning as ‘Net Receivers’ of systemic shocks within the modeled network. These institutions consistently absorbed risk originating from other banks. Signature Bank demonstrated a particularly high negative NET value of -22.0%, quantitatively establishing its position as a primary recipient of these shocks. This metric indicates the overall negative contribution of Signature Bank to systemic risk propagation, meaning it experienced a disproportionately high influx of risk from other institutions within the analyzed period. Silvergate Bank also exhibited characteristics consistent with a Net Receiver role, though its specific NET value was not detailed in this analysis.

The Time-Varying Parameter Vector Autoregression (TVP-VAR) model proved essential in quantifying dynamic systemic risk transmission pathways within the financial network. This methodology enabled the tracing of shock propagation from originating institutions to those absorbing the impact. Analysis using the TVP-VAR model identified the Secured Overnight Financing Rate (SOFR), Silicon Valley Bank (SIVB), and First Republic Bank (FRC) as exhibiting the largest positive NET values – indicators of their roles as primary transmitters of systemic risk. Specifically, SOFR registered a NET value of +11.3%, further substantiating its position as a key originator of shocks within the observed system.

The Wider Forces: Macroeconomic Drivers of Financial System Fragility

Systemic risk within the financial system extends beyond the vulnerabilities of individual banks, being demonstrably shaped by broader macroeconomic forces, most notably fluctuations in Economic Policy Uncertainty (EPU). Research indicates that periods of heightened EPU – reflecting ambiguity surrounding governmental fiscal, monetary, and regulatory actions – correlate strongly with increased systemic risk, as investor anxiety and risk aversion escalate. This suggests that systemic fragility isn’t simply a microprudential concern, rooted in the balance sheets of financial institutions, but a macroeconomic phenomenon driven by external factors. Elevated EPU can trigger correlated failures, reduce market liquidity, and amplify shocks across the financial network, even among institutions that appear fundamentally sound in isolation. Consequently, monitoring and mitigating EPU is crucial for preemptive systemic risk management, necessitating a holistic approach that integrates macroeconomic surveillance with traditional bank supervision.

Fluctuations in the Secured Overnight Financing Rate (SOFR), a key benchmark for short-term borrowing, and shifts in broader market sentiment, as measured by the CBOE Volatility Index (VIX), play a crucial role in how systemic risk propagates through the financial system. Increases in SOFR can quickly tighten financial conditions, increasing the cost of funding for banks and potentially triggering liquidity strains, especially for institutions heavily reliant on short-term wholesale funding. Simultaneously, a rising VIX-often interpreted as a gauge of investor fear-reflects heightened uncertainty and risk aversion, which can exacerbate these pressures by reducing market confidence and encouraging a flight to safety. This interplay creates a feedback loop where SOFR increases and VIX spikes amplify each other, accelerating the transmission of shocks across financial institutions and increasing the overall level of systemic risk; understanding this dynamic is critical for policymakers aiming to proactively mitigate potential crises.

Recent research utilizing a Time-Varying Parameter Vector Autoregression (TVP-VAR) model substantiates the intricate relationship between bank stock returns, broader macroeconomic conditions, and the amplification of systemic risk. This analysis reveals a dynamic feedback loop where shocks originating in the macroeconomic environment, and reflected in bank valuations, contribute to, and are simultaneously influenced by, fluctuations in systemic risk. Notably, the Secured Overnight Financing Rate (SOFR) emerges as a critical conduit for transmitting these shocks throughout the financial system; quantitative results indicate a net transmission value of +11.3%, signifying that SOFR plays a disproportionately large role in propagating financial instability and demonstrating its sensitivity to, and influence over, both bank performance and overall systemic fragility.

The study of dynamic connectedness within the U.S. banking system reveals a truth applicable to all complex structures: some elements are inherently more vulnerable, acting as conduits rather than buffers. This echoes a fundamental principle of decay-systems do not fail due to isolated incidents, but because the inevitable flow of time exposes inherent weaknesses in their architecture. As Leonardo da Vinci observed, “There is no passion or affection in painting, nor in any other art, without some mixture of knowledge.” Similarly, a thorough understanding of how shocks propagate-identifying those banks acting as key transmitters-is crucial for anticipating and mitigating systemic risk. The research demonstrates that stability can be, at times, merely a temporary delay of disaster, especially when certain institutions consistently absorb or amplify external pressures.

What Lies Ahead?

This analysis of dynamic connectedness within the U.S. banking system reveals a familiar asymmetry: certain institutions consistently function as conduits for systemic stress, while others seem designed, by virtue of structure or circumstance, to bear the brunt of it. The study offers a snapshot of this distribution, but time, as always, will reshape the landscape. Identifying these pivotal banks is not a solution, merely a refined understanding of where weaknesses will first manifest. The system doesn’t avoid decay; it redistributes it.

Future work must address the evolving nature of these roles. The observed patterns are, undoubtedly, a product of the regulatory environment and prevailing market conditions at this moment. Any simplification of risk transmission-any attempt to categorize banks as simply ‘transmitters’ or ‘absorbers’-carries a future cost. The inherent technical debt of such categorization will accumulate as the system adapts and new vulnerabilities emerge.

Furthermore, the interplay between monetary policy and sentiment remains a complex, and likely non-linear, relationship. This study acknowledges the influence of policy shocks, but a deeper understanding of how expectations-and the mispricing of risk built upon them-amplify or dampen these effects is crucial. The system’s memory is long, and past interventions inevitably shape future responses, creating a feedback loop that is difficult to fully untangle.

Original article: https://arxiv.org/pdf/2601.01783.pdf

Contact the author: https://www.linkedin.com/in/avetisyan/

See also:

2026-01-06 08:19