In this short paper we briefly introduce some of the works presented during the conference “Network models and stress testing for financial stability” which was held in Mexico City hosted by Banco de Mexico on September 26-27, 2017. The papers ranged from new applications of network models for financial stability analysis to stress testing at central banks and common assets contagion. The novelty and originality of most of the works make this special issue a good read for the specialists in the Financial Stability field, including academics as well as practitioners and obviously financial authorities. We hope that you find this short introduction to the special issue and the special issue itself, interesting and useful for your everyday work.

Network models, stress testing methods and early warning systems are attracting growing interest both among scholars and practitioners. In this short paper, we illustrate some of the insights they have to offer both in terms of new fundamental scientific understanding of the emergence systemic risk and in terms of concrete applications to the policy areas of financial stability and macro-prudential policy. Finally, we discuss new research pathways to address the challenging questions still open, including multiplex networks, big financial data, and climate-finance.

In this cover paper, we introduce a Special Issue (SI) published after the fourth edition of a series of financial stability conferences organized by Bank of Mexico, CEMLA, Bank of Canada, Zurich University and the Journal of Financial Stability in November 2021. Before providing our perspective on why the research papers included into the SI are of great relevance, we give a brief and personal overview of recent directions in financial stability research in general, esp., related to topics accentuated by the COVID-19 pandemic or post-pandemic economic and financial conditions and their complexity. Papers published in the SI cover four topics of research in the financial stability field, featuring some outstanding and innovative projects presented during the conference. The first topic is on interconnectedness and shock transmission in the financial system, diving deep into asset fire sales, interconnectedness of various segments of the financial system, in addition to banks, on the optimality of systemic risk capital buffers, and on how risks are priced in the interbank market network. The second one touches upon climate change risks looking at investors’ reactions to international climate policy developments, in particular on the Paris Agreement front and how to jointly model physical and transition risk in the banking system, including the important concept of double materiality. The third topic is represented by projects focused on policy analysis for systemic risk mitigation, specifically dealing with macroprudential policy instruments and crisis mitigation policies. Finally, research papers in the last topic on big data and market data focus on the innovative ways to explore the growing body of data sources, such as data collected by regulators, including credit register data, supervisory data and market data on financial transactions, to better understand sources and implications of systemic risk.

We analyze the effects on financial stability of the interplay between climate transition risk and market conditions, such as recovery rate and asset price volatility. To this end, we extend the framework of the climate stress-test of the financial system by including an ex-ante network valuation of financial assets which accounts for asset price volatility as well as for endogenous recovery rate on interbank assets. Moreover, we also consider the dynamics of indirect contagion of banks and investment funds, which are key players in the low carbon transition, via exposures to the same asset classes. We derive some analytical results and we apply the model to a unique supervisory dataset in a range of climate policy scenarios and market conditions. In the event of a disorderly low-carbon transition, stronger market conditions allow to reach more ambitious climate policies at the same level of financial risk.

We examine the role of imposing tighter limits on interbank exposures in reducing contagion and aggregate losses. In our model contagion risk arises as a result of the individual idiosyncratic failure of each bank in the banking system. Following Guerrero-Gomez and Lopez-Gallo (2004), we use a sequential default algorithm that is useful for tracing the path of contagion from a trigger bank to other banks during several contagion rounds. We test different types of limits on inter-SIB (systematically important banks) exposures, SIB to non-SIB exposures, and non-SIB to all other banks; and we study three different assumptions about banks’ behavioural responses under a stricter regulatory lending regime. We also “stress test” all banks within the banking system and extend the analysis on the benefits of using tighter limits in a fragile banking system. Calibrating the model to Mexican banking sector data, this network model shows that tighter limits for inter-SIB exposures are a useful tool for reducing contagion risk. Moreover, we find that tighter limits may lead to an increase in contagion risk under specific allocation assumptions.

In financial stability, it is essential to know the determinants of interest rates in interbank markets because they are important vehicles for liquidity allocation among banks and are relevant for monetary policy transmission. Recent research indicates that banks with excess liquidity exercise their market power by rationing liquidity during periods of financial stress. This confirms the value of knowing the banks connections and identifying liquidity spreaders in such markets to manage contagion risk, liquidity hoarding and to preserve financial stability. In addition to well studied bank features such as size, liquidity and credit risk, we study which network metrics relate to interest rates during different periods. Using transaction level data on unsecured and secured lending, we apply an approach that employs network theory, econometric models and machine learning to analyze the structural properties of the secured and unsecured interbank markets in Mexico. Our findings support the “too-interconnected-to-fail” hypothesis. In the secured interbank market, PageRank shows a relationship with interest rates, while metrics associated with the notion of influence and systemic risk (Katz and DebtRank) are relevant in the unsecured interbank market. In general, a bank with high centrality lends at higher rates and gets funding at lower rates.

Financial markets create endogenous systemic risk, the risk that a substantial fraction of the system ceases to function and collapses. Systemic risk can propagate through different mechanisms and channels of contagion. One important form of financial contagion arises from indirect interconnections between financial institutions mediated by financial markets. This indirect interconnection occurs when financial institutions invest in common assets and is referred to as overlapping portfolios. In this work we quantify systemic risk from indirect interconnections between financial institutions. Complete information of security holdings of major Mexican financial intermediaries and the ability to uniquely identify securities in their portfolios, allows us to represent the Mexican financial system as a bipartite network of securities and financial institutions. This makes it possible to quantify systemic risk arising from overlapping portfolios. We show that focusing only on direct interbank exposures underestimates total systemic risk levels by up to 50% under the assumptions of the model. By representing the financial system as a multi-layer network of direct interbank exposures (default contagion) and indirect external exposures (overlapping portfolios) we estimate the mutual influence of different channels of contagion. The method presented here is the first quantification of systemic risk on national scales that includes overlapping portfolios.

The inability to see and quantify systemic financial risk comes at an immense social cost. Systemic risk in the financial system arises to a large extent as a consequence of the interconnectedness of its institutions, which are linked through networks of different types of financial contracts, such as credit, derivatives, foreign exchange, and securities. The interplay of the various exposure networks can be represented as layers in a financial multi-layer network. In this work we quantify the daily contributions to systemic risk from four layers of the Mexican banking system from 2007 to 2013. We show that focusing on a single layer underestimates the total systemic risk by up to 90%. By assigning systemic risk levels to individual banks we study the systemic risk profile of the Mexican banking system on all market layers. This profile can be used to quantify systemic risk on a national level in terms of nation-wide expected systemic losses. We show that market-based systemic risk indicators systematically underestimate expected systemic losses. We find that expected systemic losses are up to a factor of four higher now than before the financial crisis of 2007–2008. We find that systemic risk contributions of individual transactions can be up to a factor of one thousand higher than the corresponding credit risk, which creates huge risks for the public. We find an intriguing non-linear effect whereby the sum of systemic risk of all layers underestimates the total risk. The method presented here is the first objective data-driven quantification of systemic risk on national scales that reveal its true levels.

Capturing financial network linkages and contagion in stress test models are important goals for banking supervisors and central banks responsible for micro- and macroprudential policy. However, granular data on financial networks is often lacking, and instead the networks must be reconstructed from partial data. In this paper, we conduct a horse race of network reconstruction methods using network data obtained from 25 different markets spanning 13 jurisdictions. Our contribution is two-fold: first, we collate and analyze data on a wide range of financial networks. And second, we rank the methods in terms of their ability to reconstruct the structures of links and exposures in networks.