Developing an Early Warning System for Financial Networks: An Explainable Machine Learning Approach.

Identifying the influential variables that provide early warning of financial network instability is challenging, in part due to the complexity of the system, uncertainty of a failure, and nonlinear, time-varying relationships between network participants. In this study, we introduce a novel methodology to select variables that, from a data-driven and statistical modeling perspective, represent these relationships and may indicate that the financial network is trending toward instability. We introduce a novel variable selection methodology that leverages Shapley values and modified Borda counts, in combination with statistical and machine learning methods, to create an explainable linear model to predict relationship value weights between network participants. We validate this new approach with data collected from the March 2023 Silicon Valley Bank Failure. The models produced using this novel method successfully identified the instability trend using only 14 input variables out of a possible 3160. The use of parsimonious linear models developed by this method has the potential to identify key financial stability indicators while also increasing the transparency of this complex system.

Patients with Type 2 Diabetes, Higher Blood Pressure, and Infrequent Fundus Examinations Have a Higher Risk of Sight-Threatening Retinopathy.

Background: Diabetic retinopathy (DR) is the most common cause of preventable blindness among working-age adults. This study aimed to evaluate the impact of the regularity of fundus examinations and risk factor control in patients with type 2 diabetes (T2DM) on the prevalence and severity of DR. Methods: One hundred and fifty-six T2DM patients were included in this cross-sectional study. Results: In this sample, the prevalence of DR was 46.2%. Patients with no DR mainly did not examine the fundus regularly, while most patients with mild/moderate nonproliferative DR (NPDR) underwent a fundus examination regularly. In 39.7% of patients, this was the first fundus examination due to diabetes, and 67% of them had sight-threatening DR (STDR). Diabetes duration (p = 0.007), poor glycemic control (HbA1c) (p = 0.006), higher systolic blood pressure (SBP) (p < 0.001), and diastolic blood pressure (DBP) (p = 0.002) were the main predictors of DR. However, the impact of SBP (AOR 1.07, p = 0.003) and DBP (AOR 1.13, p = 0.005) on DR development remained significant even after adjustment for diabetes duration and HbA1c. The DR prevalence was higher in patients with higher blood pressure (≥130/80 mmHg) than in those with target blood pressure (<130/80 mmHg) (p = 0.043). None of the patients with target blood pressure had STDR. The peaks in SBP and DBP were observed in T2DM with DR and the first fundus examination due to diabetes. Conclusions: In this T2DM sample, DR prevalence was very high and strongly related to blood pressure and a lack of regular fundus examinations. These results indicate the necessity of establishing systematic DR screening in routine diabetes care and targeting blood pressure levels according to T2DM guidelines.

Systemic Importance and Risk Characteristics of Banks Based on a Multi-Layer Financial Network Analysis.

Domestic and international risk shocks have greatly increased the demand for systemic risk management in China. This paper estimates China's multi-layer financial network based on multiple financial relationships among banks, assets, and firms, using China's banking system data in 2021. An improved PageRank algorithm is proposed to identify systemically important banks and other economic sectors, and a stress test is conducted. This study finds that China's multi-layer financial network is sparse, and the distribution of transactions across financial markets is uneven. Regulatory authorities should support economic recovery and adjust the money supply, while banks should differentiate competition and manage risks better. Based on the PageRank index, this paper assesses the systemic importance of large commercial banks from the perspective of network structure, emphasizing the role of banks' transaction behavior and market participation. Different industries and asset classes are also assessed, suggesting that increased attention should be paid to industry risks and regulatory oversight of bank investments. Finally, stress tests confirm that the improved PageRank algorithm is applicable within the multi-layer financial network, reinforcing the need for prudential supervision of the banking system and revealing that the degree of transaction concentration will affect the systemic importance of financial institutions.

Research on systemic risk of China's bank-asset bipartite network.

Systemic risk caused by banks due to common asset holdings serve as a significant contagion channel. In this study, we use empirical data from Chinese banks to construct a bank-asset bipartite network, employ the DebtRank algorithm for risk measurement, and incorporate asset price correlation into the DebtRank algorithm. Then we show the changes of the systemic risk in the Chinese banking system from 2018 to 2021. Furthermore, we analyze the systemic risk triggered by different types of banks and different industry assets and quantify the impact of each asset under different stress scenarios. We also conduct a validity analysis of asset price correlation, finding that the systemic risk considering asset price correlation is higher than that without considering asset price correlation. This study of financial systemic risk under the bank-asset bipartite network provides a new perspective for the regulation of systemic risk and is of significant importance for the prevention of systemic risk.

Computing the impact of central clearing on systemic risk.

The paper uses a graph model to examine the effects of financial market regulations on systemic risk. Focusing on central clearing, we model the financial system as a multigraph of trade and risk relations among banks. We then study the impact of central clearing by a priori estimates in the model, stylized case studies, and a simulation case study. These case studies identify the drivers of regulatory policies on risk reduction at the firm and systemic levels. The analysis shows that the effect of central clearing on systemic risk is ambiguous, with potential positive and negative outcomes, depending on the credit quality of the clearing house, netting benefits and losses, and concentration risks. These computational findings align with empirical studies, yet do not require intensive collection of proprietary data. In addition, our approach enables us to disentangle various competing effects. The approach thus provides policymakers and market practitioners with tools to study the impact of a regulation at each level, enabling decision-makers to anticipate and evaluate the potential impact of regulatory interventions in various scenarios before their implementation.

Systemic risk factors of retinopathy in patients with systemic lupus erythematosus.

OBJECTIVE: To investigate the risk factors of lupus retinopathy (LR) in patients with systemic lupus erythematosus (SLE). METHODS: This is a retrospective, cross-sectional study. LR patients admitted at Peking Union Medical College Hospital from June 2013 to April 2023 were reviewed. Age- and gender-matched SLE patients without retinopathy were selected as controls. Medical records including clinical manifestations, laboratory data and ophthalmic examination were collected. Univariate and multivariate logistic regression analyses were conducted. RESULTS: One hundred and twelve LR patients (198 eyes) were included, with 12 cases (14 eyes) presenting with retinal macrovascular obstruction, and 100 cases (184 eyes) only exhibiting microvasculopathy. Multivariate analysis indicated the presence of haemolytic anaemia, decreased haemoglobin (HGB) and higher relative percentage of neutrophils were independent risk factors for LR (p < 0.05). The first two were also risk factors for retinal microvasculopathy, whereas secondary antiphospholipid syndrome (APS) was for macrovascular obstruction. In male group, LR had significant associations with decreased HGB, no matter which types of retinopathy (p < 0.05). In female group, LR was significantly associated with haemolytic anaemia, presence of antiphospholipid antibodies, decreased white blood cells and relative high percentage of neutrophils. Specifically, haemolytic anaemia (p = 0.002) was significantly associated with retinal microvasculopathy, and APS (p = 0.003) was significantly associated with macrovasculature obstruction. CONCLUSION: LR was related to haemolytic anaemia, decreased HGB levels and higher percentage of neutrophils. Retinal microvasculopathy accounted for most cases and macrovasculature obstructions were rare. Male and female patients have distinct risk factors. Early ophthalmic screening is recommended especially for those with risk factors of LR.

How real estate bubbles affect the systemic risk of financial institutions in the United Arab Emirates.

The paper addresses a crucial gap in the literature by examining the interplay between real estate price bubbles and systemic risk in the United Arab Emirates (UAE) from 2006 to 2022. The paper employs a three-step testing procedure: bubble detection using the bootstrapped GSADF test, measuring systemic risk using Delta-CoVaR and MES measures, and assessing the impact of real estate bubbles on bank risk through panel data regression. Utilizing a sample of 17 conventional banks operating in the UAE, the study demonstrates that the interplay between real estate price bubbles and systemic risk is influenced by the specific characteristics of banks. Higher levels of loan growth, leverage, and bank size heighten the systemic risk faced by banks during asset price bubbles. Interestingly, the results also indicate that banks with a greater degree of income diversification contribute less to systemic risk during periods characterized by real estate bubbles. The results from this study are useful for policymakers in designing and implementing regulations to stabilize and prevent the UAE's banking sector from being affected by real estate price bubbles.

Growth differentiation factor-15 and circulating biomarkers as predictors of periodontal treatment effects in patients with periodontitis: a randomized-controlled clinical trial.

BACKGROUND: During the last decades, in patients with periodontitis, periodontal treatment has been shown to reduce the potential release of local and systemic biomarkers linked to an early risk of systemic inflammatory disorders. This study evaluated the efficacy of non-surgical-periodontal treatment (NSPT) on growth differentiation factor 15 (GDF-15) and related circulating biomarkers such as glutathione peroxidase 1 (GPx-1), c-reactive protein (hs-CRP), and surfactant protein D (SP-D) in periodontal patients and explored whether subjects who had high GDF-15 levels at baseline showed increased clinical benefits following NSPT at 6-months follow-up. METHODS: For this two-arm, parallel randomized clinical trial, patients with periodontitis were randomly allocated to receive quadrant scaling and root-planing (Q-SRP, n = 23, median age 51 years old) or full-mouth disinfection (FMD, n = 23, median age 50 years old) treatment. Clinical and periodontal parameters were recorded in all enrolled patients. The primary outcome was to analyse serum concentrations changes of GDF-15 and of GPx-1, hs-CRP, and SP-D at baseline and at 30, 90, and 180-days follow-up after NSPT through enzyme-linked immunosorbent assay (ELISA) and nephelometric assay techniques. RESULTS: In comparison with FMD, patients of the Q-SRP group showed a significant improvement in clinical periodontal parameters (p < 0.05) and a reduction in the mean levels of GDF-15 (p = 0.005), hs-CRP (p < 0.001), and SP-D (p = 0.042) and an increase of GPx-1 (p = 0.025) concentrations after 6 months of treatment. At 6 months of treatment, there was a significant association between several periodontal parameters and the mean concentrations of GDF-15, GPx-1, hs-CRP, and SP-D (p < 0.05 for all parameters). Finally, the ANOVA analysis revealed that, at 6 months after treatment, the Q-SRP treatment significantly impacted the reduction of GDF-15 (p = 0.015), SP-D (p = 0.026) and the upregulation of GPx-1 (p = 0.045). CONCLUSION: The results evidenced that, after 6 months of treatment, both NSPT protocols improved the periodontal parameters and analyzed biomarkers, but Q-SRP was more efficacious than the FMD approach. Moreover, patients who presented high baseline GDF-15 and SP-D levels benefited more from NSPT at 6-month follow-up. TRIAL REGISTRATION: NCT05720481.

What (If Anything) is Wrong with High-Frequency Trading?

This essay examines three potential arguments against high-frequency trading and offers a qualified critique of the practice. In concrete terms, it examines a variant of high-frequency trading that is all about speed-low-latency trading-in light of moral issues surrounding arbitrage, information asymmetries, and systemic risk. The essay focuses on low-latency trading and the role of speed because it also aims to show that the commonly made assumption that speed in financial markets is morally neutral is wrong. For instance, speed is a necessary condition for low-latency trading's potential to cause harm in "flash crashes." On the other hand, it also plays a crucial role in a Lockean defense against low-latency trading being wasteful developed in this essay. Finally, this essay discusses the implications of these findings for related high-frequency trading techniques like futures arbitrage or latency arbitrage-as well as for an argument as to why quote stuffing is wrong. Overall, the qualifications offered in this essay act as a counterbalance to overblown claims about trading at high speeds being wrong.

Carbon footprints of lending and bank performance: international evidence from panel data.

The paper studies the interaction between a set of bank performance indicators (concentration, profitability, and risk) and the carbon footprint of bank loans. Our research builds on the panel data analysis for 37 countries during 2010-2018, adopting local projections proposed by Jordá (Am Econ Rev 95(1):161-182, 2005), a feasible alternative to panel VAR estimation in case of short time series. In order to account for potentially different patterns in the relationship among the indicators, we split the whole panel into two sub-panels, using K-means clusterization based on income level, resource abundance, and overall environmental performance. For the whole panel, the carbon footprint is driven by systemic risk, while leading the non-performing loans (NPL) ratio and Z-score. Thus, curbing systemic risk matters to reduce the carbon footprint of bank loans. Otherwise, it may amplify the effects of the latter on the NPL ratio and Z-score. Interestingly, the effect of systemic risk on the carbon footprint stems from the sub-panel consisting of developed countries, while the effect of the carbon footprint on the NPL ratio and Z-score is mainly shaped by developing and emerging market economies. The relationships between the carbon footprint of lending, concentration, and profitability are much less pronounced both for the whole panel and for the sub-panels.

Centrality measures of financial system interconnectedness: A multiple crises study.

We explore how asset returns could be a good proxy to detect interlinkages in the financial system. This paper employs a US dataset for the 2002-2021 period. Pairwise returns correlation indicate the interconnectedness at the preliminary stage. The Principal Component Analysis captures a significant portion of variance and detects the co-movement and highly connected state of the financial market during crises. Granger centrality tested with pairwise directional variance decomposition indicates the importance of banks and insurance companies in the US financial system. This paper recommends policymakers use multiple network models to validate and calibrate the SIFIs list.

Interest Rate Normalization in a World with High Indebtedness.

Global debt has grown to record levels. Government debt, corporate debt and household debt around the world rose to a high of 350% of world gross domestic product in 2022. The systemic risk that has built up during the extended period of low interest rates threatens to materialize now as rates rise worldwide. For countries in which external liabilities are high, debt service costs will increase, and refinancing may become prohibitively expensive. External liabilities and their term structures provide insight into which emerging and developing countries may be most vulnerable in the next months. Supplementary Information: The online version contains supplementary material available at 10.1007/s11293-023-09763-y.

Climate risk and financial systems: A nonlinear network connectedness analysis.

To deeply analyze and understand the macro-financial impact of climate change, this paper investigates the effect of climate risk on systemic financial risks by employing a network approach. The results demonstrate that climate risk not only affects a single financial market but also induces risk co-movement, which aggravates potential systemic financial risks. Specifically, the system-wide connectedness across the financial system respectively increased by 2.52% and 1.76% after the withdrawal of the US from the Kyoto Protocol and the Copenhagen UN Climate Change Conference. The bond and stock markets are the primary transmitters of climate shocks, while the forex and commodity markets appear to be more sensitive to climate-related information. In addition, the vulnerability of financial asset price fluctuations to climate risk changes substantially over time. Quantile regressions reveal the positive impact of climate risk on total connectedness across the financial system. This study provides novel insight into how the financial system responds to climate-related information and how systemic risk dynamics materialize.

Identifying influential financial stocks using simulation with a two-layer network.

Risk spillover from one stock to another tends to create a contagion effect in the stock market. Fire sales due to the overlapping portfolios of mutual funds can amplify the contagion risks, leading to a downward spiral of stock prices. In this paper, we simulate this downward spiral phenomenon for the Chinese financial stocks based on a two-layer network structure and aim to identify the influential financial stocks based on their individual induced systemic risks. Our findings show that stock liquidity and the concentration of funds' holding on stocks play important roles in determining systemically important financial institutions. Our results also confirm the statements of "too-big-to-fail" and "too-interconnected-to-fail" of financial institutions in the Chinese market. Our results show that a more sensitive flow-performance relationship of mutual funds can amplify the contagion risk by 41%. However, the magnitude can be more drastic in a low market liquidity scenario, where the contagion risk is boosted by 160%.

Systemic cascades on inhomogeneous random financial networks.

This article presents a model of the financial system as an inhomogeneous random financial network (IRFN) with N nodes that represent different types of institutions such as banks or funds and directed weighted edges that signify counterparty relationships between nodes. The onset of a systemic crisis is triggered by a large exogenous shock to banks' balance sheets. Their behavioural response is modelled by a cascade mechanism that tracks the propagation of damaging shocks and possible amplification of the crisis, and leads the system to a cascade equilibrium. The mathematical properties of the stochastic framework are investigated for the first time in a generalization of the Eisenberg-Noe solvency cascade mechanism that accounts for fractional bankruptcy charges. New results include verification of a "tree independent cascade property" of the solvency cascade mechanism, and culminate in an explicit recursive stochastic solvency cascade mapping conjectured to hold in the limit as the number of banks N goes to infinity. It is shown how this cascade mapping can be computed numerically, leading to a rich picture of the systemic crisis as it evolves toward the cascade equilibrium.

Smart cities at risk: Systemic risk drivers in the blind spot of long-term governance.

In this article, we analyze "digital massification" in smart cities, that is, an ever-growing number of market participants, consumers, and Internet of Things devices with simultaneous accommodation of users to increasing disturbances and inconveniences due to data congestion-as a driver for systemic risk. We argue that digital massification phenomena largely escape societal awareness due to their protracted evolution and are therefore still in the blind spot of long-term governance. Our analysis makes methodological use of historical and relational analogy, and we introduce and elaborate concepts and terms that allow us to discuss the evolutionary nature of massification, that is, the foreseeable increasing probability of the occurrence of trigger events. Using the analogy to the history of road traffic congestion, we deduce that digital massification will most likely lead to a future "risk transition" where tolerated disturbances and inconveniences of the present will turn into systemic impacts. This insight calls for heightened sensitivity in governance to massification phenomena to ensure the long-term resilience of smart cities.

Exploring State Level Factors Associated with Short-Stays in Child Welfare: The Role of Systemic Risk and Surveillance.

Growing attention has been directed toward children who are placed in out-of-home care by child welfare authorities for less than 30 days, deemed "short-stayers". This exploratory study uses multiple national child welfare and population data sources to identify macro level factors associated with short-stays. Two-level logistic regression modeling was conducted to explore how state-level factors were associated with risk of short-stays. Factors associated with lower odds of short-stays included living in a state with a centralized child welfare reporting structure and with greater food insecurity. Factors associated with greater odds included living in a state with a higher percentage of the state's population enrolled in the Supplemental Nutrition Assistance Program and states with more police per capita. Multiple state level factors were associated short-stay risk, which suggests broader systemic factors contribute to these brief removals. Findings suggest greater surveillance by police and social services increases risk of short-stays, which likely have implications for child welfare policy and practice.

Examining the Time Varying Spillover Dynamics of Indian Financial Indictors from Global and Local Economic Uncertainty.

The research aims to excavate the role of global (Fed Rate, Crude, Real Dollar Index) and endogenous economic variables (GDP and Consumer Price Index) in shaping the spillover amongst the major Indian Financial indicators, viz. Nifty Index, MCX Gold, USDINR, Govt. Bond 10Y maturity and agricultural index N-Krishi. To facilitate cross-comparison decomposition of time-varying spillover output generated from Time-Varying Vector Autoregression (TVP-VAR) with aggregation at three layers is performed. The research finds that Indian Financial Indicators are vulnerable to spillover shocks from global variables predominantly driven by Fed Rate and Real Dollar Index. USDINR turns out to be most sensitive to global shocks and transgresses the shock to other financial indicators. Importantly, persistently high inflation has brought volatility spikes in the directional spillover to financial indicators. Though spillover subsidence is observed post-2014, with an all-time high during GFC, a sudden spurt in all financial indicators has been observed post-Covid-19, with Govt. bonds showing a sporadic rise. An important observation relates to staunch spillover from GDP during GFC with reoccurrence post-Covid. Additionally, a closely knit spillover tie is observed among USDINR, N-Krishi, and Crude. The study is beneficial to RBI to proactively monitor the weakening rupee along with Fed tapering to manage the rising spillover post-Covid-19. The effort of RBI has to be reciprocated by the government in inflation targeting to reinforce the curbing efforts of rising shock spillover.

Quantifying interconnectedness and centrality ranking among financial institutions with TVP-VAR framework.

Financial risk is spread and amplified through the interconnectedness among financial institutions. We apply a time-varying parameter vector autoregression model to analyze the dynamic spillover effects in the Chinese financial system. We find that the 2017 house price control policies have significantly increased the risk of China's financial system. Before 2017, with the prosperity of the real estate market, the interconnectedness of the Chinese financial system continued to decline, while after 2017, with the slowdown of house price growth and the downturn of the real estate market, the interconnectedness turned to increase. For different sectors, the trends and the magnitudes of the spillover effects are diverse, and any sector can contribute to systemic risk in a dynamic way. Finally, we rank 20 systemically important financial institutions according to two centrality measures. The stable institution ranking provides less noisy information for regulators to formulate a policy and intervene in the market effectively.

Dynamic Banking Systemic Risk Accumulation under Multiple-Risk Exposures.

Much of the existing research on banking systemic risk focuses on static single-risk exposures, and there is a lack of research on multiple-risk exposures. The reality is that the banking system is facing an increasingly complex environment, and dynamic measures of multiple-risk integration are essential. To reveal the risk accumulation process under the multi-risk exposures of the banking system, this article constructs a dynamic banking system as the research object and combines geometric Brownian motion, the BSM model, and the maximum likelihood estimate method. This article also aims to incorporate three types of exposures (interbank lending market risk exposures, entity industry credit risk exposures, and market risk exposures) within the same framework for the first time and builds a model of the dynamic evolution of banking systemic risk under multiple exposures. This study included the collection of a large amount of real data on banks, entity industries, and market risk factors, and used the ΔCoVaR model to evaluate the systemic risk of the China banking system from the point of view of the accumulation of risk from different exposures, revealing the dynamic process of risk accumulation under the integration of multiple risks within the banking system, as well as the contribution of different exposures to banking systemic risk. The results showed that the banking systemic risk of China first increased and then decreased with time, and the rate of risk accumulation is gradually slowing down. In terms of the impact of different kinds of exposures on system losses, the credit risk exposure of the entity industry had the greatest impact on the banking systemic risk among the three kinds of exposures. In terms of the contribution of the interbank lending market risk to the systemic risk, the Bank of Communications, China Everbright Bank, and Bank of Beijing contributed the most. In terms of the contribution of the bank-entity industry credit risk to the systemic risk, the financial industry, accommodation and catering industry, and manufacturing industry contributed the most. Considering the contribution of market risk to the systemic risk, the Shanghai Composite Index, the Hang Seng Composite Index, and the Dow Jones Index contributed the most. The research in this paper enriches the existing banking systemic risk research perspective and provides a reference for the regulatory decisions of central banks.