Complex Social Systems: A Primer

Philosophers, sociologists, and economists have long observed how a seemingly chaotic web of social interactions among individuals, institutions, and their environment coalesces into an order (Sawyer Reference Sawyer2005). Moreover, studies on the genesis and evolution of living organisms (Varela, Maturana, and Uribe Reference Varela, Maturana and Uribe1974) have shed new light on how social institutions and organized social systems come into existence (Teubner Reference Teubner1988; Luhmann Reference Luhmann and Bednarz1995; Capra Reference Capra1997). Similarly, compelling economic research equates finance with natural ecosystems (Haldane and May Reference Haldane and May2011; Arthur Reference Arthur2015, 151). The common denominator of these studies is that social systems of different scales—ranging from small communities to sophisticated economies—stem from complexity. Like natural systems, such as a cell or the biosphere, they display a series of common properties. Specifically, three general features define any complex social system and are particularly relevant to this analysis. First, complex must not be confused with complicated. A system is complicated when its features can be inferred by examining its components. Complicated systems—like computers, nuclear power plants, or even financial transactions—are governed by linear dynamics, whereby each input corresponds to a proportional output (Cilliers Reference Cilliers1998, 3). Differently, when the dynamics within a system give rise to larger patterns that do not respond to direct cause-and-effect relationships, the system is complex (Kauffman Reference Kauffman1993, 29). This property, also observed in living organisms, is referred to as “nonlinearity” and indicates that outputs are not proportional to the corresponding inputs (Cilliers Reference Cilliers1998, 5–7). For instance, small changes in the genome of a pathogen or its environment can lead an isolated outbreak to become a large-scale epidemic (Kucharski 2020, 23–26). Nonlinear dynamics can also generate positive feedback loops; as further examined below, high investment returns stem from small actions that are amplified through iterations between market participants.

In essence, nonlinearity explains why complex social systems entail unexpected outcomes. Due to multiple causal links, recursive exchanges, and feedback loops interrupting the linear sequence of events, changing one part of a complex system (input) might result in an exponential change in the other part (output) at a different time. Perhaps the most iconic metaphor to explain nonlinearity is the one offered by Edward Lorenz (Reference Lorenz1963) and termed the “butterfly effect.”Footnote ⁹ Lorenz observed that the formation of a hurricane might depend on whether a butterfly located on a distant part of the globe had flapped its wings weeks earlier. In more prosaic terms, this property is known as “sensitivity to initial conditions” and was first noted in the nineteenth century by French mathematician and engineer Henri Poincaré to indicate that a minor cause in a complex physical system determines a much larger effect.Footnote ¹⁰

Second, complex social systems are not transient phenomena. In the natural and social world alike, through reiterated interactions, no matter how small, different elements generate a self-organized whole. This property is referred to as “emergence” and indicates that the nonlinear dynamics among various components spontaneously result in a new organizational structure: a system. Many small interactions between individuals produce larger patterns that ultimately result in a social system (Sawyer Reference Sawyer2005, 5), similar to a V-shaped flock of birds where each bird sets its course based on the relative position of nearby birds rather than lining up to follow a leader. The result is not merely a material structure but also a network of relationships in which individuals’ competitive and cooperative behaviors give rise to a unity that performs one or more functions.

This feature is ubiquitous in social contexts. Deepening its roots in Aristotle’s (1989, Book 8, 1045a) notion that a whole requires more than its components to create unity, social emergence became the fulcrum of theories developed by French sociologists Auguste Comte and Émile Durkheim to connect individual actions to a larger spontaneous order. Based on these insights, social systems have been explained as deterministic rather than random since any change in the current state influences its future shape and helps perpetuate its existence over time. In other words, borrowing from the concept of path dependence developed in the social sciences, it can be said that “history matters” (North Reference North1990; Urry Reference Urry2005, 239). Complex systems adapt and evolve, and their past influences their present behavior (Cilliers Reference Cilliers1998, 90).

Third and consequentially, the behavior of a complex system is irreducible; it cannot be inferred from the behavior of its constituent parts. As Comte (Reference Comte1875, 153) argued, “[a] society … can no more be decomposed into individuals than a geometric surface can be resolved into lines or a line into points.” Therefore, to understand the dynamics of a complex system, the focus must be shifted from its components to the whole. Specifically, unlike complicated systems that are studied by segmenting their structure into smaller parts to isolate different (linear) relationships, a complex system is primarily examined through functional analyses focusing on the activities performed by the system and their evolution.

Financial Complexities

The financial system displays complex properties. Comprised of various activities—most notably, banking, investment, and insurance sectors—finance is characterized by the co-existence of heterogeneous groups of social actors, including public entities, private firms, platforms, investors, and retail customers. The recursive (competitive or cooperative) interactions between different actors give rise to a self-organized structure of relationships that performs the critical societal function of deploying and transferring economic resources between social actors over time (Merton and Bodie Reference Merton, Bodie, Dwight, Kenneth, Andre, Robert and Tufano1995). In short, complexity explains why the aggregate actions of individuals, each animated by self-interest, result in market trends operating as an “invisible hand,” to borrow from Adam Smith’s (Reference Smith1902, 160) famous metaphor.

Nonlinearity describes fundamental financial processes. Through financial leverage, for instance, market participants can gain (or lose) exponentially from their initial investment, be it a mortgaged house, college savings, or a stake in a publicly listed company. Through maturity and liquidity transformation performed by banks, short-term liabilities (deposits) are converted into long-term assets (loans), supporting the real economy and, thus, inducing the creation of new deposits and loans. Through financial intermediation, an ever-expanding network of interdependent relationships connects different social actors, ranging from depositors and borrowers to corporations and governments seeking funds. Compared to other complex systems, however, finance presents additional elements of complexity, reflected in the strategic actions of economic agents fueling a constant process of self-innovation.

Market participants engage in strategic behaviors. Firms, customers, and investors are cognitive agents. Their choices are based on changes in the external world and on expectations regarding the consequences of their actions (Arthur Reference Arthur2015, 107). Empirical studies on price formation in capital markets corroborate this point. The efficient capital market hypothesis, a pillar of neoclassical economics, posits that individuals participate in market transactions with fully formed preferences.Footnote ¹¹ Since the hypothesis implies that investors are rational and unbiased, any price movement should reflect a reaction to choices taken on the information previously available. However, financial economist Robert Shiller (Reference Shiller2008, 47) demonstrated that traders disregard independently collected information and act instead on general information and investment strategies followed by others, resulting in herd-like dynamics. Informational cascades and feedback loops influence movements in stock prices that, in turn, spur an increment in trading activities, prompting further changes in prices and volatility (Shiller Reference Shiller1989, 374; Reference Shiller2008, 45–46).

Nonlinear interactions between market participants are amplified by a process of innovation, whereby the financial system evolves and expands its reach. Generally identified as both a propelling force of capitalism and a source of disruption (Schumpeter Reference Schumpeter1947, 132), innovation reflects the adaptive nature of economic systems engaged in a process of continuous self-destruction and reconstruction (Arthur Reference Arthur2015, 3). Through technological advancements, transactional structures, and business processes, financial institutions develop new techniques and products to reduce costs, generate revenues, and manage risks (Avgouleas Reference Avgouleas, Moloney, Ferran and Payne2015). Securitization, for instance, is a form of financial innovation that pools various types of assets to repackage them into different categories (or “tranches”) of interest-bearing financial instruments. Accordingly, a bank can transform illiquid long-term assets—such as loans, mortgages, or credit card debts—into liquid assets sold to various investors, including public institutions and pension funds. While freeing new capital and incentivizing banks to lend more, this technique heightens the interconnectedness among financial institutions and blurs the distinction between investment and banking activities (Snider Reference Snider2011, 126; Davis and Kim Reference Davis and Kim2015, 208). Financial innovation also fuels profound socioeconomic changes, epitomized by the process of financialization (Krippner Reference Krippner2005; Van der Zwan Reference Van der Zwan2014). Structured financial products have been fueling the growing reliance on finance as a source of economic growth, connecting social actors who are traditionally distant, such as households and fund managers. The innovations brought by integrating financial activities with digital technologies has further amplified this process.

The digitalization of financial transactions and activities has been changing the relationships between social actors, resulting in new complexities. Starting with the dematerialization of financial instruments in the 1990s, digital finance has become a bedrock of the current financial inclusion policies (Arner, Castellano, and Selga Reference Arner, Castellano and Selga2023; Castellano, Arner, and Selga Reference Castellano, Arner, Selga, Chander and Sun2023). Yet the possibility for individuals to place their orders with a tap on the phone has replaced the interpersonal component of financial intermediation, also resulting in the “gamification” of finance with coordinated herd behaviors, as evidenced by the “meme stock” hype of 2021 (Tierney Reference Tierney2022) and by the banking crises of 2023, which are discussed in more detail below. At the wholesale level, high-frequency trading exploits price oscillations through algorithms that execute large and automated transactions at the “speed of light” across multiple markets (MacKenzie Reference MacKenzie2023). Recursive interactions between machines, while supporting the expansion of the financial system, exacerbate market movements by adding new liquidity when markets grow and reducing it during economic downturns (Lin Reference Lin2013, 698).

Collectively, these dynamics explain why finance is not in a natural status of stability. The general equilibrium theory, a mainstay of neoclassical economics, posits that free and competitive markets—with perfect information—have a natural tendency toward a point of optimal equilibrium because goods and services are allocated to those who value them most (Arrow and Debreu Reference Arrow and Debreu1954; McKenzie Reference McKenzie1959). However, complexity studies in economics, sociology, and political science have long noted that markets operate in a far-from-equilibrium status precisely because market participants constantly change their actions, reacting strategically to the system-wide dynamics they mutually create (Cilliers Reference Cilliers1998, 122; Arthur Reference Arthur1999, 108; Urry Reference Urry2005, 238). Put differently, instability is inextricably interwoven with complexity.

Legal Nonlinearities

Like finance, law is a complex phenomenon that exhibits nonlinearities, self-organization, and emergent properties. Gunther Teubner (Reference Teubner1988) and Niklas Luhmann (Reference Luhmann and Albrow2014, 284), most notably, have posited that the legal system is governed by autopoiesis, indicating that, through a self-produced logic, it perpetuates, evolves, and renews itself even if challenged by the outside.Footnote ¹² The result is the emergence of a large variety of subsystems of sectoral rules, each responding to their inner, self-perpetuating logic. Narrowing the focus on finance, it is possible to note that a fragmented bundle of subject-specific rules and logical deductions governs financial relationships and institutions. These branches routinely overlap, giving rise to composite regimes, termed “commercial law intersections,” that govern transactions and corporate actions (Castellano and Tosato Reference Castellano and Tosato2021). Such intersections constitute new (narrower) systems of rules and logical deductions that enable economic actors to conduct their operations. For example, a transaction in which a bank extends a loan collateralized with the debtor’s personal property gives rise to a commercial law intersection represented by the coming together of secured transactions law, contract law, and banking regulation (Castellano and Tosato Reference Castellano and Tosato2021, 1027). As these legal intersections proliferate, law can be conceived as a complex system composed of subsystems that evolve and coalesce, creating smaller unities in fractal-like organizations.Footnote ¹³ This view explains the complexities within the law itself. Nonetheless, to understand the role of law within the financial system, a deeper analysis of how legal norms shape, and are shaped by, the recursive interactions of social actors is needed.

At a glance, law appears to be a social construct that cannot be parted from economic dynamics. Modern capitalism depends on rules, enforcement systems, and adjudication mechanisms that, in turn, originate from “spontaneously evolved customs” (Deakin et al. Reference Deakin, David Gindis, Hodgson and Pistor2017, 190). Markets are premised on the validity and enforceability of legal commitments. Crucially, as noted by Katharina Pistor (Reference Pistor2013, 317), finance is “a complex, interdependent web of contractual obligations … that link market participants to one another.” Thus, law can be conceived as the syntax constructing and executing nonlinear interactions between market participants. Relationships within the financial system are legal relationships, and the enforceability of legal commitments is a determinant factor for economic agents to calculate expected returns and limit possible losses. However, drawing from complexity theory, law appears to be more than a mere vehicle to execute financial complexities.

Critically, this article qualifies law as an agent of complexity because its interactions with social actors within the financial system generate new and unique nonlinearities. First, market participants respond to regulatory standards in more articulated ways than a binary “comply” or “non-comply” reaction (Black Reference Black2013, 416). This is because strategic behaviors and other idiosyncratic factors, including personal values, peer pressure, and compliance costs, influence the response to regulation (Baldwin and Black Reference Baldwin and Black2008, 69–70) and the decisions of regulators (Castellano and Helleringer Reference Castellano, Helleringer, Avgouleas and David2019). This is evident not only when firms adapt their choices to ensure compliance with rules (Gilad Reference Gilad2014; Krawiec Reference Krawiec2019) but also when they enact regulatory arbitrage strategies, whereby transactions and corporate actions are structured to avoid regulatory requirements while ensuring formal compliance (Fleischer Reference Fleischer2010, 230). Periods of regulatory tightening follow financial crises when political momentum is gained to initiate reforms and reinvigorate enforcement actions. Hence, while supervisory agencies adapt their strategies to address novel market practices,Footnote ¹⁴ rules are routinely changed. An example is offered by capital adequacy standards for internationally active banks, enshrined in the Basel Framework.Footnote ¹⁵

The Basel Framework aims to ensure the soundness of individual banks and the financial system’s stability. It follows a risk-weighting approach, whereby the amount of regulatory capital, consisting of the bank’s “unborrowed” funds (such as shareholders’ equity) varies depending on the level of risk borne by the bank each time a loan is extended (Armour et al. Reference Armour, Dan Awrey, Luca Enriques, Gordon and Payne2016, 290–315; Cranston et al. Reference Cranston, Emilios Avgouleas, Hare and van Sante2018, 27–40). In essence, as noted elsewhere, regulatory capital connects the extension of credit to a portion of the bank’s own funds (such as equity and retained earnings) determined depending on the level of risk attributed to each borrower and transaction.Footnote ¹⁶ The rationale is that lending to borrowers deemed risky should require more “skin in the game” than lending to less risky borrowers. Given that raising regulatory capital is more expensive than deploying borrowed money,Footnote ¹⁷ banks are incentivized to limit risk exposure to maximize their return on equity. Nevertheless, the coefficients used to determine the level of risk for different classes of exposures are legally constructed and politically determined.Footnote ¹⁸ Hence, a gap between the actual risk of a deal and its regulatory classification may emerge. To exploit this gap, banks can structure their investments to reduce capital requirements even though the risk they face is materially higher (Jones Reference Jones2000). For instance, until 2008, under unbelievably favorable regulatory treatment, securitization and other derivatives were engineered and sold mainly to reduce banks’ regulatory capital.Footnote ¹⁹ When the GFC unfolded, and significant losses materialized, it became painfully evident that capital requirements did not capture the effective riskiness of these transactions. Hence, a process of reform was initiated.

Second, rules emerge, change, and evolve due to interdependence with social actors. Organizations “are both responding to and constructing the law that regulates them,” rendering law endogenous to the areas it is intended to regulate (Edelman, Uggen, and Erlanger Reference Edelman, Uggen and Erlanger1999, 407). Therefore, financial law and regulation include dialectic processes (Georgosouli Reference Georgosouli2010) and routinized procedures involving bureaucrats, lawyers, and back-office employees (Riles Reference Riles2011, 10). Cooperation between market participants yields practices recognized and absorbed by the legal system. For instance, legal protection has been granted to standard-form contracts developed by business associations (Riles Reference Riles2011, 169). Similarly, the Basel Framework has been designed around models first developed in the industry (Tarullo Reference Tarullo2008, 178). Conflicts between market participants also engender extensive legal change. Simple dispute resolution mechanisms, involving two opposing parties and a dispute resolver, trigger a process that reshapes rules and relationships within social systems, tracing broader policy trajectories (Stone Sweet Reference Stone Sweet1999).

Third, interactions between the law and environment result in novel nonlinearities that disrupt the legal status quo and reshape the relationship between law and society. For instance, the interplay between law and technology does not simply result in new rules; it also changes how social actors interact with the legal system as compliance and decisions can be automatized and solutions based on artificial intelligence are increasingly deployed in legal processes traditionally requiring human intervention.Footnote ²⁰ Similarly, the challenges of climate change have disrupted established legal institutions, resulting in completely new legislative frameworks and regulatory regimes to gauge novel risks and attribute responsibilities (Fisher, Scotford, and Barritt Reference Fisher, Scotford and Barritt2017). These dynamics indicate that law itself is subject to an ongoing process of innovation, which is often at odds with the idea that legal systems are stable but perfectly fitting with the notion of complexity.

Endogeneity of Systemic Risk

Disasters emerge from the complex interactions between fortuitous events, human activities, and social elements. In particular, the international community indicates that disasters result in “continuously present conditions of risk” as they depend on three factors: exposure to hazards, the vulnerability of societies, and the availability of resources to limit exposure and address vulnerabilities (United Nations 2016, 14). Disaster risk is, thus, the possibility—expressed in probabilistic terms—that a given hazard materializes in human, material, and economic or environmental losses that exceed the community’s ability to cope using its own resources—that is to say, that disasters stem from the complexities of the systems they threaten.Footnote ²¹

Accordingly, there is no such thing as a “natural” disaster. Although a hazard can be natural or non-natural (United Nations 2016, 18), the materialization of a disaster always depends on human factors (Farber et al. Reference Farber, Chen and Verchick2015, 3–4). Specifically, the risk of a disaster is directly proportional to both the hazard exposure and the vulnerabilities of the affected social systems but inversely proportional to its preparedness and response capacity. Hence, events that are external to a given social system, such as a novel disease, heatwaves, or an earthquake, are triggering factors that a system’s nonlinear dynamics may (or may not) transform into a systemic failure, or into inter-systemic failures, when more social systems are affected concomitantly (Heyvaert Reference Heyvaert2022).Footnote ²² A system-wide disruption may be triggered by a relatively minor technical failure, as in the case of the Chernobyl accident in 1986, or by an external event that is not under direct human control, as in the case of the Fukushima Dai-ichi disaster in 2011 (Perrow Reference Perrow2011; Funabashi and Kitazawa Reference Funabashi and Kitazawa2012). Regardless, these events only escalate into disastrous consequences due to the nonlinear interactions within affected systems.

Similar dynamics can be observed in the financial context. The emergence and busting of price “bubbles” clearly illustrate the endogenous nature of financial systemic risk. Given that market participants tend to imitate a dominant investment strategy (Shiller Reference Shiller2008, 47), a self-reinforcing mechanism is initiated, conflating market prices through iterated trading activities (Sornette Reference Sornette2017, 52–54). A bubble forms when prices rise far above the actual value of the underlying assets and burst when a sufficient number of investors sell such assets, fearing further losses (Cunningham Reference Cunningham1993, 594–95). For instance, a fall in housing prices, such as the one that occurred in 2007, is a shock that is exogenous to the choices of individual financial institutions. However, the growth and diffusion of financial products linked to the mortgage market supported the expansion of indebtedness and the rise of prices; the subsequent downfall of the housing market, in turn, triggered disinvestment strategies, whereby financial institutions purged depreciated assets to limit their losses (Adrian and Shin Reference Adrian and Song Shin2010; Shleifer and Vishny Reference Shleifer and Vishny2011). The ensuing herd-like behavior resulted in inflated sales, further depreciating financial assets and compromising the financial position of large financial institutions, such as Bear Sterns, JP Morgan Chase, and Lehman Brothers (Judge Reference Judge2012, 145). When Lehman Brothers Holdings Incorporated (LBHI) filed for bankruptcy, investors, fearing further losses, closed out their positions en masse, depressing financial markets even further, while the US authorities decided not to bail it out.Footnote ²³ The failure of LBHI is a manifestation of broader complex dynamics. The nonlinearities of the financial system, magnified by regulatory and legal frameworks unequipped to address system-wide disruptions, transformed losses of one hundred to two hundred billion dollars in subprime mortgage products into losses in the trillions, rampant unemployment, and a global recession with a persistent socioeconomic impact.

Endogeneity explains two fundamental dynamics of systemic risk. First and most fundamentally, the same nonlinear interactions transforming a set of seemingly unrelated interactions into a self-organized whole also engender the risk of its demise. Primarily observed in the natural world and in the context of disaster risks, the connection between complexity and instability has also been explored in finance. Drawing from epidemiology and natural sciences, the fragility of the financial system has been explained as a result of its inherent complexities (Haldane and May Reference Haldane and May2011, 353). This explanation is consistent with the seminal work of Hyman Minsky (Reference Minsky2015), who noted that financial instability is a cyclical phenomenon that emerges from a series of progressive phases and is fueled by an uncontrolled accumulation of debt. Crucially, risk taking and credit growth increase in a low-risk environment, rendering the financial system more fragile (Danielsson, Valenzuela, and Zer Reference Danielsson, Valenzuela and Zer2023). Hence, actions that ordinarily fulfill physiological market functions, such as investments, lending, or disinvestments, ultimately develop into an excessive accumulation of risk, reaching a breaking point represented by a diffused inability to service debt obligations. Looking at the interactions between market participants, Steven Schwarcz (Reference Schwarcz2008, 206) persuasively noted that financial systemic risk results from a “tragedy of the commons, in which the benefits of exploiting finite capital resources accrue to individual market participants, each of whom is motivated to maximize the use of the resource.” This is to say that market participants are not naturally incentivized to maintain financial stability; the rent-seeking behaviors of firms, if left unchecked, stimulate excessive risk taking and widespread distress.

Nonlinear interactions between market participants can also cause financial instability when reacting to external events, such as deteriorating economic conditions. Similar to an epidemic outbreak where people react with “flight or hide” behaviors to minimize the risk of getting infected (Epstein et al. Reference Epstein, Parker, Cummings and Hammond2008, 1), a change in economic conditions can prompt depositors, investors, and lenders to withdraw their funds, disinvest, and withhold capital. Unlike during epidemics, when flight strategies spread disease and hiding lowers the transmission rate by limiting social contact, the indiscriminate withdrawal of funds (flight) or the hoarding of capital (hide) from the financial system generates contagion. This outcome is also a manifestation of the adaptive strategies of market participants. Profit-seeker investors and risk-averse depositors react to real (or perceived) risks. In anticipation of potential losses, investors hold on to their capital, whereas depositors “run to the bank” to withdraw their savings. Collectively, these actions reinforce asset depreciation, limit liquidity, erode confidence across the system, and, without any public intervention, may determine chain reactions, known as contagion, whereby (relatively) healthy financial institutions fail. Irrespective of the triggering event, systemic risk is a product of the complexities of the financial system.

Second, the more complex a system, the higher the chances of instability. This is a direct consequence of the observation that complex (social or natural) systems exist in a far-from-equilibrium state, as they sharply transition from stable to unstable once their complexity exceeds a critical threshold. With the proliferation of interdependencies and recursive interactions, the nonlinearities of complex systems intensify, and the possibility of small events triggering significant system-wide failures grows (May Reference May1972, 414). In epidemiology, for instance, multiple interactions between individuals exponentially increase the transmission rate of disease (Kucharski 2020, 25). In industrial organizations, high technical and organizational complexity results in a higher probability of system-wide failures, rendering industrial accidents a “normal” occurrence (Perrow Reference Perrow1984). Similarly, the complexities brought by financial innovation (Judge Reference Judge2012, 158), technological advancements (Omarova Reference Omarova2020), and the burgeoning interconnectedness of financial institutions (Haldane and May Reference Haldane and May2011, 352) create new risks and new vulnerabilities.

The 2023 banking crisis provides clear evidence of these dynamics. The failures of SVB and CS had a variety of different underlying causes, but they shared at least one commonality: they showed how social media, in combination with digital banking, created new complexities that amplified the endogenous dynamics of systemic risk. Following a series of heavy investments in long-term bonds during a period of low-interest rates and booming deposits—primarily gathered from technology-intensive start-ups and venture capital-backed companies—SVB experienced significant losses when interest rates were increased to combat inflation and the value of its assets fell (Federal Reserve Reference Reserve2023, 25). Following the voluntary liquidation of a technology-focused bank, Silvergate Bank, SVB attempted to reassure depositors and investors by communicating its intention to shore up new capital to absorb part of its losses. This decision did not yield the desired effect. On March 9, SVB’s clients withdrew forty-two billion dollars in a single day.Footnote ²⁴ Coordinated via online platforms and executed through online banking services, their actions resulted in the fastest bank run to date. Expecting an additional outflow of one hundred billion dollars and lacking the liquidity necessary to meet such rapid drainage, SVB was closed and placed into receivership the following day (Federal Reserve Reference Reserve2023, 24). But fear spread further through social media, causing a rapid flight of funds from regional mid-sized banks, resulting in the closure of Signature Bank on March 12 (New York State Department of Financial Services 2023) and the liquidation of First Republic Bank on May 1.Footnote ²⁵ Internationally, a similar plot resulted in the demise of CS, a systemically important financial institution at the global level (a global systematically important financial institution).Footnote ²⁶ After years of difficulties, speculations of its imminent collapse circulated across the Internet, resulting in an outflow of over $120 billion in the last three months of 2022. Yet, as contagion spread beyond the United States, a new wave of withdrawals started soon after SVB’s failure, requiring Swiss authorities to halt the bleeding by orchestrating the emergency acquisition of CS by UBS Financial Services on March 19.Footnote ²⁷

These cases are not anomalies. They reflect the process of innovation that reshapes the interactions among the components of the financial system and within its environment. A new path is thus traced, disrupting the existing status quo. Bank runs have occurred throughout financial history; rumors have often been their precursor, and deposit insurance their cure (Diamond and Dybvig Reference Diamond and Dybvig1983). Yet in a highly financialized and digitalized economy, the relationships between social actors have changed. As individuals have gained constant access to cash and information, rumors can avalanche into a system-wide disruption at lightning speed.

Law and Systemic Risk

Embedded in the fabric of social systems, law is part of the complexities that may give rise to systemic risk. Rules, principles, and institutional arrangements are social constructs that govern, organize, and facilitate social endeavors and interactions. The same legal constructions that support the emergence of a self-organized structure may also threaten its functioning. Legal institutions, in fact, influence the ability of societies to curb systemic risks by increasing or reducing exposures to potential threats as well as the vulnerabilities and preparedness of social systems. For instance, land law rules, urban planning, and conservation requirements, through the implementation of calibrated restrictions, are essential to reduce the probability and impact of a wide range of disasters arising from flooding and earthquakes (Farber et al. Reference Farber, Chen and Verchick2015, 29–31) as well as diseases and climate change (Heyvaert Reference Heyvaert2022, 960). However, regulatory measures often adopted to curb risks in the short term may thwart long-term prevention and generate new vulnerabilities.Footnote ²⁸ Similarly, public funds, private insurance, and risk-transfer mechanisms influence the incentive structure of social actors, thus promoting or discouraging investments in preventive measures and recovery plans (Monti Reference Monti2009, 158; Castellano Reference Castellano2010, 404–6). More broadly, the institutionalization of emerging systemic risks, such as those related to new technologies or climate change, poses significant normative challenges, disrupting established legal concepts and often requiring decisions to be taken in circumstances of scientific uncertainty (Fisher, Scotford, and Barritt Reference Fisher, Scotford and Barritt2017, 179–81). The ensuing legal nonlinearities may increase, rather than reduce, risk exposures and vulnerabilities.

Finance is no exception. While premised on legal certainty and predictability, rules may be a source of instability.Footnote ²⁹ The legal shockwaves triggered by the failure of Lehman Brothers offer a powerful illustration of this dynamic. On September 15, 2008, LBHI filed for Chapter 11 bankruptcy protection, marking the largest bankruptcy in US history (Scott Reference Scott2016, 19) and spawning an intricate nexus of insolvency proceedings across various jurisdictions (Bridge and Braithwaite Reference Bridge and Braithwaite2013; Braithwaite Reference Braithwaite2014). Following LBHI’s demise, Lehman Brothers International (Europe), the London-based regional headquarters, was placed into administration, sparking a dispute on the applicability of English property and trusts law concerning the segregation and pooling of clients’ funds.Footnote ³⁰ Prior to the commencement of insolvency proceedings in the United Kingdom, however, the German supervisory authority, the Bundesanstalt für Finanzdienstleistungsaufsicht (BaFin), issued a moratorium,Footnote ³¹ halting all payments and operations of Lehman Brothers Bankhaus AG (Hofmann Reference Hofmann2011, 468). While intended to facilitate orderly liquidation in the context of normal corporate insolvency (Singh Reference Singh2020, 174), such a moratorium caused Lehman Brothers Bankhaus AG to default on all short-term obligations before any restructuring could occur. Moreover, it further compromised the position of the group, rendering uncertain the amount that clients could recover from the liquidation of Lehman Brothers International (Europe).Footnote ³² Other than identifying the shortcomings of cross-border insolvency regimes, these events indicate that legal nonlinearities play a role in the emergence of systemic risk. This is to say that law, unwittingly or by design, may exacerbate financial instability.

The impact of legal non-linearities on the emergence of system-wide disruptions can also be observed at the macro level when regulatory changes are examined. While financial regulation is not solely crisis driven and reflects a larger process of policy innovation (Conti-Brown and Feinstein Reference Conti-Brown and Feinstein2021), the alternation between regulatory relaxation and regulatory tightening is a common dynamic. Owing to the convergence of ideological and political interests (Harvey Reference Harvey2007), regulatory capture (Baker Reference Baker2010) and, more broadly, liberalization and deregulatory initiatives (Born Reference Born2011; Stiglitz Reference Stiglitz2015, 23–24) are often indicated as the root of recent financial crises. The flaws exposed in the aftermath of a crisis, in turn, trigger the implementation of more stringent regulatory regimes that are diluted during periods of relative stability (Snider Reference Snider2011). Yet the perspective adopted in this article indicates that changes introduced in the aftermath of a crisis or during a period of perceived stability may equally contribute to financial instability.

Consistent with the notion that law is an agent of complexity introduced earlier, law may contribute to financial instability during regulatory easing and tightening periods. Regarding periods of regulatory easing, the deregulatory initiatives that preceded the 2023 banking crisis in the United States exemplify this dynamic. Enacted as a response to the failure of LBHI, section 165 of the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010 (Dodd-Frank Act) contains measures to address the “too-big-to-fail” problem by establishing a standardized set of enhanced prudential standards for banks that are classified as systemically important.Footnote ³³ In 2018, however, the Economic Growth, Regulatory Relief, and Consumer Protection Act (EGRRCPA) amended section 165 in two ways: it raised the minimum asset threshold for applying enhanced prudential standards from $50 billion to $250 billion, and it mandated the Federal Reserve to enact a tailored approach for banks with total assets between $100 billion and $250 billion.Footnote ³⁴

In a general political climate where supervisory enforcement was actively discouraged (Snider Reference Snider2020), the Federal Reserve enacted a lengthy and elaborate matrix of bank-specific parameters to effect the EGRRCPA provisions.Footnote ³⁵ As a result, mid-sized banks, such as SVB and Signature Bank, were not treated as systemically important and were not subjected to the standardized and enhanced requirements but, instead, to a tailored approach. Ironically, five years later, facing an exceptionally fast bank run driven by unsecured depositors, the Department of the Treasury, in coordination with the Federal Reserve and the Federal Deposit Insurance Corporation,Footnote ³⁶ invoked the risk of a systemic failure in order to resolve SVB and Signature Bank in a manner that would protect all depositors (insured and uninsured). The Federal Reserve (Reference Reserve2023, 10; emphasis added) later indicated that the “tailoring approach in response to EGRRCPA and a shift in the stance of supervisory policy impeded effective supervision by reducing standards, increasing complexity, and promoting a less assertive supervisory approach.” This is to say that a reform enacted to reduce regulatory pressure generated new nonlinearities that required exceptional public measures to avert a system-wide disruption.

Regulatory regimes implemented to foster financial stability may have similar effects. For instance, risk-based capital requirements tend to incentivize investments during good economic times as the risk of failure on individual investments is perceived as being lower when the general economic outlook is positive. However, during an economic downturn, capital requirements incentivize financial institutions to adopt a more conservative approach by limiting lending and, possibly, liquidating their position (Black Reference Black2012, 1046; Awrey and Judge Reference Awrey and Judge2020, 2334). Similarly, mark-to-market accounting rules, whereby the value of an asset should reflect current market prices, push financial entities to mark down their assets to fire sale prices, forcing otherwise solvent firms to liquidate depreciated assets (Schwarcz Reference Schwarcz2009, 232; Scott Reference Scott2010, 674). Another paradoxical result emerges: rules designed to promote the soundness of the financial system may amplify the cyclical movements of markets.

New rules have been implemented to address the critical flaws unveiled by the GFC. For instance, a new set of international standards has been adopted to ensure the orderly resolution and liquidation of troubled financial institutions. Specifically, according to the Key Attributes of Effective Resolution Regimes for Financial Institutions, distressed financial institutions should continue critical economic functions, and early termination of rights should be avoided (Financial Stability Board 2014).Footnote ³⁷ Moreover, starting from 2011 with a progressive implementation to be completed by 2028, the Basel Framework has been amended.Footnote ³⁸ While the core structure and risk-based rationale have been maintained, specific measures have been introduced to curb the cyclical movements of markets. Most notably, financial institutions that are deemed systemically important have been required to implement a counter-cyclical buffer, consisting of a cushion of capital that banks must maintain when their balance sheet is expanding (during good times) and release to absorb losses when a contraction occurs (during bad times).Footnote ³⁹

Similarly, a novel leverage ratio has been implemented, requiring banks to maintain a minimum amount of regulatory capital relative to their total assets. This new ratio is not riskbased: it is not intended to reflect the level of risk associated with any specific exposure; instead, it has been designed to counteract the accumulation of excessive leverage in the banking system and subsequent deleveraging during an economic downturn.Footnote ⁴⁰ The newly introduced counter-cyclical buffers helped to mitigate the adverse economic impact of the COVID-19 pandemic; by releasing the buffer, banks were encouraged to lend even if the economy halted. Other measures, however, heightened systemic distress. For instance, to keep the statutory ratio between total assets and regulatory capital, banks were disincentivized to purchase US treasuries when, at the outset of the COVID-19 crisis, investors in need of cash wanted to sell them. To avoid a systemic disruption and inject liquidity, the Federal Reserve had to purchase an unprecedented sum of treasuries while temporarily excluding treasuries and cash reserves from the calculation of the leverage ratio (Kress and Zhang, forthcoming, 52).

The observation that easing or tightening regulatory requirements may equally contribute to the emergence of systemic risk does not imply that nothing can be done. Financial crises like disasters are inevitable, but their likelihood and impact can be minimized once system-wide dynamics are understood and addressed. Yet, notwithstanding the monumental reforms implemented in the aftermath of the GFC, the normative premises for financial regulation to address systemic risks have remained unaltered. Specifically, as further illustrated in the next section, existing regulatory policies neglect complexity and conceive both systemic risk and law as exogenous to market dynamics: the former is an anomaly; the latter is its cure.

Financial Regulation and the Exogeneity of Systemic Shocks

In 2009, as the GFC unfolded, the International Monetary Fund, the Bank for International Settlements, and the Financial Stability Board (2009, 2) defined systemic risk as “[a] risk of disruption to financial services that is (i) caused by an impairment of all or parts of the financial system and (ii) has the potential to have serious negative consequences for the real economy.” While seemingly reflecting the endogenous nature of systemic risk, closer inspection reveals that this definition is incomplete and does not support a genuinely systemic view. Constructed as a tautology—where a potential “disruption” results from an “impairment”—the definition neglects the relationship between systemic risk and financial dynamics. Unlike disaster risk governance, financial regulation falls short of indicating that finance exists under a constant condition of risk as potential damages to the financial system are a function of its vulnerabilities and preparedness for reoccurring disturbances. This is not to say that existing policy responses are not geared to address vulnerabilities altogether. For instance, building upon the above definition of financial systemic risks, the International Monetary Fund, the Bank for International Settlements, and the Financial Stability Board (2016, 4) have indicated that macro-prudential regulation should limit systemic risk by addressing vulnerabilities related to both “the build-up of risks over time” and “the interconnectedness and the associated distribution of risk within the financial system.” However, this understanding of vulnerability reveals fatal ambiguities once juxtaposed with the one adopted to mitigate disaster risks. Disaster risk governance defines vulnerability as a scalar variable that gauges the idiosyncratic ability of actors and systems to absorb or suffer damages resulting from identified hazards.Footnote ⁴³ Therefore, the build-up of risk and the interconnectedness amongst its components are not vulnerabilities per se; they are a manifestation of complex dynamics that may contribute to the emergence of a systemic failure if (and to the extent to which) the affected system is unable to absorb losses.

Premised on a limited understanding of the complex dynamics, regulatory regimes addressing systemic risk suffer from critical flaws. First, although the prevention of systemic risk ranks now as one of the primary objectives for financial regulators, its incomplete understanding results in vague policy mandates. In the United States, for instance, the Financial Stability Oversight Council (FSOC), comprised of the heads of the major regulatory agencies chaired by the secretary of the Treasury, has been entrusted to “identify risks” and “respond to emerging threats to the stability of the United States financial system.”Footnote ⁴⁴ In practice, the FSOC acts as a coordination body and is only entrusted with the powers of designating systemically important financial institutions (therefore, subjecting them to the Federal Reserve’s enhanced supervision) and issuing nonbinding recommendations to regulatory agencies (Kress, McCoy, and Schwarcz Reference Kress, McCoy and Schwarcz2019, 1473). Crucially, systemic risk is not institutionalized as a function of system-wide dynamics but, rather, as a disruption arising from different events that regulatory agencies must identify.

Second, systemic risk is construed as a phenomenon exogenous to financial dynamics as it rests outside the control of individual financial institutions. This understanding is embedded in the risk models developed by the industry and further reflected in capital requirements adopting the value-at-risk (VaR) methods. VaR models are intended to estimate the maximum loss that an investment portfolio might suffer over a given period and with a certain level of probability (or “confidence interval”).Footnote ⁴⁵ Their construction requires a vast amount of data typically gathered from the observations of small and frequent gains and losses routinely incurred by financial institutions (Danielsson Reference Danielsson2008, 326). In these models, system-wide disruptions are not gauged as they are deemed independent from the choices of individual financial institutions (Danielsson and Shin Reference Danielsson, Song Shin and Field2003, 87). Losses generated by system-wide disruptions, such as a financial crisis, are reflected indirectly in historical observations of smaller events, provided that such occurrences fall within the observed period.

Hence, following traditional economic theory, the activities of financial institutions in response to a shock are considered not to affect the formation or amplification of systemic risk because, in perfectly competitive markets, every economic agent is considered a price taker (that is, it cannot solely influence a change in prices). Although it is recognized that VaR models are not designed to measure systemic risk (Danielsson Reference Danielsson2008, 327), their use is ubiquitous and embedded in financial regulation. In particular, a few years after their deployment in the financial industry, VaR models were included in the Basel Framework and, thus, elevated to a global regulatory standard.Footnote ⁴⁶ In ensuring a consistent homogenous application of international capital requirements, a paradoxical result may ensue: every firm in the financial system considers systemic risk to be a variable that is unrelated to its actions. This understanding is also reflected in the methodology adopted to conduct stress testing—a technique often hailed as an example of the macro-prudential ethos of post-GFC reforms. Stress tests are simulations (based on hypothetical but possible scenarios) run by regulators to identify criticalities affecting markets, transactions, and firms (Armour et al. Reference Armour, Dan Awrey, Luca Enriques, Gordon and Payne2016, 428). Depending on the outcome, additional capital requirements may be imposed. While intended to remedy the shortcomings of capital requirements calculated at the individual level, the Federal Reserve deploys an equilibrium model to analyze how each financial institution, discretely considered, would react to an exogenous shock (Enriques, Romano, and Wetzer Reference Enriques, Romano and Wetzer2019, 370). Second-order consequences and feedback loops resulting from the interaction between market participants are excluded.

All in all, the institutionalization of systemic risk as a phenomenon external to individual financial institutions’ decisions is paired with the notion that markets lean toward a natural state of equilibrium. As discussed below, this understanding reflects a broader, reductionist view that discounts complexity to explain core financial dynamics.

The Stability of Reductionism and the Exogeneity of Law

Implicit in financial regulation is the notion that finance tends toward stability, represented by an optimal equilibrium. From a normative standpoint, this assumption has two critical implications. First, markets are considered to have the almost thaumaturgical ability to self-heal. Because shocks are assumed to be an external factor that affects prices, market participants are deemed to adapt to price changes, thus reaching a new equilibrium. The confidence in the ability of the financial system to correct itself caused regulators to refrain from intervening in the early signs of financial distress in 2007. In October 2008, during a historic hearing in the US Congress, Alan Greenspan, former chairman of the US Federal Reserve, admitted that the decisions of the Federal Reserve in the early days of the GFC were taken under the assumption that the market would correct itself.Footnote ⁴⁷ Greenspan further added that such an assumption—referred to as an “ideology” or a “view of the world”—was flawed.Footnote ⁴⁸ Yet, post-crisis reforms have not provided an alternative view of the world as systemic risk is still considered an exogenous anomaly causing a departure from the natural state of equilibrium.

Second, the equilibrium assumption is a benchmark to determine when regulation is justified. Accordingly, any form of regulatory intervention is needed only when financial markets cannot reach an optimum due to imperfections that limit the ability of individuals to engage in private negotiations (Baldwin, Cave, and Lodge Reference Baldwin, Cave and Lodge2012, 15). Such imperfections could result from market failures such as monopolies and asymmetries of information (Bator Reference Bator1958),Footnote ⁴⁹ in addition to transaction costs, representing the costs incurred when participating in economic exchanges (Coase Reference Coase1937). According to this view, if transaction costs and market failures were hypothetically absent, legal institutions would be unnecessary or even irrelevant, given that the parties to any economic transaction would be free to negotiate their rights and obligations to maximize their idiosyncratic preferences.Footnote ⁵⁰ However, since market imperfections exist in the real world, this logic indicates that law should be crafted to facilitate private negotiations, which, in turn, are expected to restore the natural equilibrium.

Such an understanding implies that legal and regulatory norms perform a corrective function and, thus, are external to the dynamics of markets.Footnote ⁵¹ Accordingly, legal and regulatory norms are considered to facilitate or stifle market development by reducing or increasing transaction costs and market failures. However, they are not considered intertwined with economic dynamics, nor is it deemed essential for the functioning of the markets. While market failure rationale may be a simple but effective diagnostic tool for identifying and addressing specific issues, such as the asymmetries of information in individual transactions, its applicability in other contexts is problematic.Footnote ⁵² Critically, as further indicated below, it is unequipped to grasp system-wide, complex dynamics.

The Market Failure Fallacy

In finance, systemic risk falls under the rubric of market failures. Specifically, systemic risk has been qualified as a negative externality, which is a cost not accounted for in market prices, like pollution (Buchanan and Stubblebine Reference Buchanan and Craig Stubblebine1962; Dahlman Reference Dahlman1979).Footnote ⁵³ Therefore, a systemic failure disrupts market dynamics because it imposes costs for which individuals neither accept nor receive retribution. Consequently, the optimal allocation of resources ensured by the market’s invisible hand cannot be attained, and public intervention is justified. To restore the market’s dynamics, neoclassical economics requires bringing such non-negotiated costs within a transactional framework. A tax on firms or activities generating negative externalities is a typical solution to ensure that the cost of externalities is borne by those causing them. Such a fiscal levy, known as the “Pigouvian tax” from British economist Arthur Cecil Pigou (Reference Pigou2013), is determined by calculating the expected overall costs of a given negative externality and the marginal contribution of each firm to the expected losses, net of any social benefits.

Although the original theory has been heavily criticized, also by Ronald Coase (Reference Coase1960),Footnote ⁵⁴ it gained some traction in different policy circles, as the carbon tax exemplifies. Since pollution is perceived as a negative externality, putting a price on greenhouse gas emissions has been considered an effective tool to reduce carbon emissions, as the cost of pollution is transferred to firms and, ultimately, to consumers. Drawing from the recognition that carbon taxes contribute to reducing greenhouse emissions (Haites Reference Haites2018), proponents of a Pigouvian tax to redress financial systemic risks suggest that large financial institutions should be charged a finely geared levy that is designed to cover pro rata the losses of their potential failure (Acharya et al. Reference Acharya, Pedersen, Philippon, Richardson, Fouque and Joseph2013, 229; Masciandaro and Passarelli Reference Masciandaro and Passarelli2013; Poledna et al. Reference Poledna, Rovenskaya, Dieckmann, Hochrainer-Stigler and Linkov2020). However, this approach presents some fatal issues.

A financial crisis, like a disaster, cannot be reduced through an ex ante compensation of its potential damages. Leaving aside any ethical considerations, a Pigouvian tax is effective only if it accounts for all the possible losses that a systemic failure brings with it. But, for financial systemic risk, the total damages can be larger than the capitalization of the financial system since the broad economy may suffer. Furthermore, regardless of its feasibility, even a hypothetical, perfectly calibrated tax would not limit the likelihood or the impact of system-wide disruptions for it does not entail a reduction of existing vulnerabilities or increased preparedness. Environmental policies offer conclusive evidence of this point. As carbon taxes are intended to increase the cost of socially undesirable activities, they are effective only when they can shift consumers’ preferences toward more environmentally sustainable activities (Metcalf Reference Metcalf2009). However, since alternatives to the traditional financial industry (if available) are offered through unregulated entities, a Pigouvian tax would steer customers toward unregulated markets, thus creating new nonlinearities.

Similar considerations can be advanced for other regulatory remedies designed to address financial systemic risk as a form of market failure. The cost of complying with regulatory standards limiting excessive risk taking has been considered akin to a Pigouvian tax (Bernanke Reference Bernanke2016; Levine and Macey Reference Levine and Macey2018). Assessments of domestic post-crisis reforms, such as the Dodd-Frank Act, reveal that, while specific issues have been addressed, unintended consequences have emerged (Baily, Schardin, and Swagel Reference Baily, Schardin and Swagel2016). The increased amount of regulatory pressures on large financial institutions, particularly banks, have spurred the development of a more opaque, less regulated sector, such as the “shadow banking” sector, wherein a new network of complexities arose amongst financial institutions that are substantially similar to traditional banks but fall outside the regulatory purview (Awrey and Judge Reference Awrey and Judge2020, 2304).

The preceding analysis indicates that the current institutionalization of systemic risk within financial regulation is incompatible with the notion that finance is a complex social system. As long as finance is understood as a linear system leaning toward a natural state of stability and law is conceived as an exogenous remedy, regulatory actions remain confined within narrow boundaries. Attempts to curb systemic risk, at best, address specific facets, but their broader effects on the stability of the financial system will remain uncertain and potentially detrimental. New solutions require a different conceptual framework, one that does not exclude complexity a priori.

The Limits of Simplifying Complexity

Ambitions to reduce complexity in finance face fundamental limitations. First, complexity is an inherent feature of finance and, therefore, is not necessarily undesirable. Hence, regulatory simplification should target, at least theoretically, only those nonlinearities that are deemed to add to the instability of the financial system. In practice, however, this task is unattainable. Consistent with the irreducibility property, since the behaviors of complex systems cannot be inferred from the behaviors of their components, it is impossible to connect a specific dynamic to a given outcome. This point is also directly observable in the context of regulatory attempts to govern financial innovation.

As a propelling force stimulating the emergence of the financial system and a potential source of instability, the impact of any given financial innovation cannot be assessed a priori. In some instances, innovation may result in a simplified user experience while concealing significant organizational complexities. Mobile banking services, for instance, have simplified user experience and promoted financial inclusion while relying on articulated data governance frameworks (Arner, Castellano, and Selga Reference Arner, Castellano and Selga2023, 253–72). Nonetheless, they have invertedly resulted in new fragilities, as epitomized by the 2023 digital bank runs. In other circumstances, the complexities brought about by innovations reveal their positive or negative social effects only over time. For instance, the emergence and diffusion of new digital assets deploying distributed ledger technologies (or “crypto assets”) has been indicated as a potential source of systemic risk (Financial Stability Board 2023). Yet their diffusion is prompting radical changes in the supply of money, possibly reducing the risk of liquidity crises. Jurisdictions are currently exploring the possibility of issuing digital currencies that are directly linked to, and guaranteed by, the issuing central bank. As a result, commercial bank deposits and the fractional reserve system may become a thing of the past, granting individuals direct access to a perfectly liquid monetary instrument. This new form of digital money has evolved from crypto assets (Dell’Erba Reference Dell’Erba2019; Buckley et al. Reference Buckley, Arner, Zetzsche, Didenko and Van Romburg2021; Jackson Reference Jackson2022) to deploy a technology that, ironically, was popularized by bitcoin with the promise of subverting traditional monetary and payment systems.Footnote ⁵⁵ This indicates that whether the complexity stemming from these (or other) financial innovations yields any socio-economic benefit cannot be determined in advance.

A second fundamental problem that may compromise regulatory approaches aimed at reducing complexity relates to legal nonlinearities. As an agent of complexity endogenous to finance, law contributes to the emergence of the financial system and systemic risk. Hence, consistent with the “butterfly effect,” noted earlier, even a simple set of rules can bring forth momentous consequences as market participants react strategically to any change in the system. The evolution of international capital adequacy standards enshrined in the Basel Framework illustrates this dynamic. For half a century, the Basel Framework has undergone profound changes reflecting a tension between simplicity and granularity (Tarullo Reference Tarullo2008). To limit strategic behaviors, policy makers have overhauled the initial approach to capital regulation, which was praised for being straightforward but not sufficiently precise, in favor of a more granular approach intended to reflect more accurately the level of risk taken by banks. However, such an increased risk sensitivity also significantly increased complexity without necessarily disincentivizing regulatory arbitrage practices (Romano Reference Romano2014; Gerding Reference Gerding2016). Not surprisingly, the Basel Framework is still primarily criticized for its intricacy, which is deemed to compromise its effectiveness (Aikman et al. Reference Aikman, Galesic, Gigerenzer, Kapadia, Kothiyal, Murphy and Neumann2021, 328). Nevertheless, a closer look at the mechanics and implementation of the Basel Framework indicates that cries for simplification may harbor deeper problems.

A popular argument for simplification purports that the intricacy of the Basel Framework results in detrimental effects, especially on small banks. The assumption is that only large and sophisticated entities have sufficient resources to fully implement articulated regulatory standards (Gurrea-Martínez and Remolina Reference Gurrea-Martínez and Remolina2019). While more rigorous empirical studies to support such claims are needed, it must be noted immediately that the Basel Framework is grounded on a proportionality principle.Footnote ⁵⁶ Accordingly, domestic authorities, while implementing the central tenets of the Basel Framework, may not implement the whole. A recent survey indicates that out of ninety jurisdictions, 85 percent apply the principle of proportionality, which consists of a simplified and standardized framework for calculating capital requirements (World Bank and Basel Committee on Banking Supervision 2021, 3). As a result, most banks (globally) are not required to comply with the most sophisticated parts of the Basel Framework, which was first introduced in 2004. Instead, capital requirements are commonly calculated following a classification of credit risk exposures introduced with the First Basel Accord (Basel I) of 1988 and often praised for its simplicity.Footnote ⁵⁷ Critically, this version was very close to the one applied by SVB (and all other mid-sized banks), reflecting a deregulatory stance made in the name of simplicity.

The pursuit of simplification through law is, indeed, a complex matter. Conceptual and normative limitations affecting the understanding and governance of nonlinear dynamics may thwart financial stability objectives even when (relatively) simple legislations are implemented. Moreover, absent clear parameters, once simplification is elevated to a primary goal, regulatory policies will rely on subjective determinations of the level of complexity deemed acceptable. This is why pleas for simplification of regulatory standards mask, consciously or unwittingly, a deregulatory agenda.

A Risk Governance Perspective

Risk governance refers to a composite approach deployed to deal with risks that escape linear cause-and-effect relationships (Power Reference Power2007, 19; van Asselt and Renn Reference Van Asselt and Renn2011, 432). It embodies an organized structure where legal institutions, administrative discretion, and private and public actors are coherently coordinated to limit exposures, increase preparedness, and reduce the vulnerabilities of the social groups exposed to complex risks (Renn Reference Renn2008, 5). Unlike financial regulation, which typically deploys different government-oriented models oscillating between command and control and self-regulation (Castellano, Jeunemaître, and Lange Reference Castellano, Jeunemaître and Lange2012), a governance approach entails a combination of ordered rules (top-down) and collective actions (bottom-up) to achieve the desired goal (Stoker Reference Stoker1998, 18; Schweizer Reference Schweizer2021, 83). By and large, risk governance encompasses a variety of measures intended to steer the decision-making processes of a network of public and private actors in consideration of broader behavioral, social, economic, and legal variables. Its core tenets are distilled in a set of principles issued by the International Risk Governance Council (2017), an independent organization based in Switzerland.

Central to risk governance is the recognition that societal factors add to the magnitude and complexity of disaster; therefore, adequate measures should be adopted. Unlike mainstream economic theory, which influences much of the regulatory policies for financial systemic risk, disaster risk governance is not anchored to the assumption that people react rationally to risks. Instead, it is acknowledged that risk is a subjective matter influenced by psychological, social, cultural, and political factors that determine idiosyncratic levels of tolerance.Footnote ⁵⁸ If a risk is categorized as “intolerable,” a prohibition is established to forbid any activity contributing to the risk under consideration. In contrast, if a risk is deemed “tolerable,” appropriate measures to mitigate its negative impacts are implemented, whereas when a risk is considered “acceptable,” risk mitigations are not necessary (International Risk Governance Council 2017, 20).Footnote ⁵⁹ For disasters stemming from natural events that cannot be controlled, such as earthquakes and infectious diseases, tolerability and acceptability relate to the preventive measures that the population must deploy to reduce exposure to the risks arising from such hazards (Renn Reference Renn2008, 144).

These general principles of risk governance are operationalized in the action plan developed by the international community to cope with disaster risk—the Sendai Framework. The Sendai Framework puts forward an “all hazards” approach premised on the following four priorities: furthering risk knowledge; reinforcing the institutional mechanisms to manage emergencies; stimulating investments to support disaster risk reduction efforts; and enhancing preparedness, recovery, and reconstruction vis-à-vis any form of danger.Footnote ⁶⁰ While public institutions have the primary responsibility to meet these priorities, it is recognized that disaster risk reduction entails a whole-of-society approach, whereby various social actors contribute directly or indirectly to increase preparedness and resilience.

Upon these premises, disaster risk governance develops along four cyclical phases where only two states of reality exist: a disaster has already occurred or will occur. Prior to a disaster, once potential threats are identified, preventive measures are enacted to mitigate potential losses (mitigation and prevention phase), reduce vulnerabilities, and increase the preparedness of communities under threat (preparedness phase). Once a disaster occurs, actions are taken to contain the emergency (emergency response phase), facilitate reconstruction of damaged infrastructures, and promote the rehabilitation of the socio-economic order (recovery phase). The recovery phase ushers into the preparedness phase, thus setting the foundation for all risk governance activities. This is to say that reconstruction ought to occur to limit exposure to disaster risk by addressing existing vulnerabilities, as reflected in the “Build Back Better” principle, according to which mitigation and preparedness measures are implemented during the recovery phase.Footnote ⁶¹

The Sendai Framework provides a general structure to develop more specific frameworks. While covering a diverse range of policy domains—from public health to hydrogeology—each risk governance framework shares a set of standard features underscored by the acknowledgment that social systems and systemic risks are complex phenomena. For instance, the Pandemic Influenza Risk Management guidance document entails a whole-of-society framework to ensure coordination between three key constituencies of stakeholders: public institutions, private businesses, and civil society (World Health Organization 2017). Each social group, in turn, is expected to contribute locally and internationally to the continuity of the activities in seven areas considered essential for the normal functioning of societies and identified as “health, defense, law and order, finance, transport, telecommunications, energy, food and water” (55). Finally, reflecting the cyclical occurrence of infectious diseases, governance actions are established depending on whether a pandemic has occurred or ought to occur.

Acknowledging that disaster risks are prone to subjective evaluations does not imply denying their measurability. On the contrary, it highlights the relevance of public perception for disaster risk reduction strategies. The predisposition to fear certain types of risks often pushes the public to take action and demand public interventions toward specific threats (Sunstein Reference Sunstein2005, 100; Vogel Reference Vogel2012) while disregarding others (Michel-Kerjan and Slovic Reference Michel-Kerjan and Slovic2010). In this context, scientific evidence (where available) intersects with irrational images, as shown by the social stigma attached to the Black Death epidemic of the fourteenth century and the ongoing acquired immunodeficiency syndrome (AIDS) epidemic.Footnote ⁶² More recently, during the first wave of the COVID-19 pandemic, face masks—like other non-pharmaceutical precautionary measures—were implemented spontaneously by the populations of several Asian jurisdictions, such as Hong Kong and Taiwan (Chang and Chun-Yuan Reference Chang, Chun-Yuan and Victor2020, 49). Concomitantly, in the United States, the Centers for Disease Control and Prevention discouraged the use of face masks, and public authorities indicated that only those showing symptoms should wear masks in public.Footnote ⁶³ As a result, at the outset of a global health emergency, the same preventative measure was concomitantly adopted by some and considered a sign of illness carrying a stigma by others. A governance framework aware of these dynamics should emphasize the need to implement effective communication strategies.Footnote ⁶⁴

The approach deployed to govern financial systemic risk is remarkedly different. At the most fundamental level, it can be noted that there is no equivalent of a Sendai Framework for finance. Consequently, there is no general blueprint to guide actions and define responsibilities before, during, and after a financial crisis. Instead, a plethora of regimes have been enacted under the overarching umbrella of maintaining financial stability. For analytical purposes, these regimes can be roughly grouped into policies implemented before or after the failure of individual financial institutions (Cranston et al. Reference Cranston, Emilios Avgouleas, Hare and van Sante2018, 33). Accordingly, it can be noted that capital requirements and resolution plans are intended to perform a preventative function, ensuring that financial institutions are sufficiently capitalized to withstand a shock and have identified strategies for their orderly wind-up if needed; similarly, special insolvency and resolution regimes can be classified as emergency management frameworks that grant public authorities special stabilization powers to resolve troubled financial institutions while limiting contagion (Davis et al. Reference Davis, Stephan Madaus, Irit Mevorach, Romaine, Sarra and Tirado2023, 131). Arguably, it may also be noted that stress tests may perform a preventative role for financial institutions, whereas emergency liquidity lending and deposit insurance schemes can be intended to limit the impact of unfolding crises. These considerations, however, only reflect classificatory efforts. These instruments are not part of a cohesive governance framework and are not benchmarked against clearly enunciated prevention, preparedness, responsiveness, and recovery priorities. As a result, the management of system-wide threats largely depends on exceptional measures subject to discretionary choices and broader political swings. The CS case offers a vivid illustration of this point.

Confronted with the distress of CS and fearing material consequences for the stability of the financial system, the Swiss authorities departed from international and domestic rules to resolve systemically important financial institutions. Established in the aftermath of the GFC to address the “too-big-to-fail” problem, resolution regimes entail mechanisms to ensure the orderly resolution of troubled financial institutions while protecting depositors, ensuring that shareholders bear the losses first and avoiding the use of public funds (Financial Stability Board 2014). As the crisis precipitated, the Swiss government passed an emergency ordinance to expand the discretionary power of the Swiss regulator and resolution authority, allocate additional funds (to absorb potential losses and guarantee a private bailout), and change the hierarchy of priority rights under insolvency.Footnote ⁶⁵ As a result, holders of convertible contingent bonds (CoCos)Footnote ⁶⁶ were wiped out to absorb a loss of over eighteen billion dollars, shareholders received UBS shares worth $3.25 billion, and UBS received an additional liquidity line (equal to more than $110 billion) from the central bank, backed by a default guarantee to which the Swiss government also contributed (Paz Valbuena and Eidenmüller Reference Paz Valbuena and Eidenmüller2023, 3). These actions, albeit considered necessary, disattended established resolution plans and subverted the core principles for effective resolution of troubled financial institutions, two mainstays of the post-GFC regulatory reforms. Thus, it is evident that there is no rule-based plan to follow in times of crisis, let alone a governance structure organized in phases. The result is an uneven application of legal rules across time, sectors, and financial institutions.Footnote ⁶⁷

The lack of a structured framework is particularly evident in the context of post-crisis actions when recovery and reconstruction should occur. While disaster risk governance emphasizes the need to rebuild in preparation for the next disaster, regulatory reforms often attempt to restore a pre-crisis alleged normalcy. Generally invoked as the guiding principle of post-crisis policies, the Build Back Better principle is not a formalized concept to direct financial reforms. As a result, financial regulation is locked in a path-dependent dynamic that reiterates the same flawed approach to systemic risk. In the aftermath of financial distress, public discontent triggers a political reaction to reform extant laws. However, absent a general framework that ties preparedness to the identification and reduction of structural vulnerabilities, post-crisis reforms are more susceptible to deregulatory pressures and tend to be designed to address the specificities of past crises, leaving the financial system vulnerable to new forms of distress.

Toward a Governance Framework for Financial Systemic Risk

Breaking with the past, a governance framework for financial systemic risk provides a blueprint to handle the inherent complexities that give rise to financial instability. Premised on the notion that systemic risk and law are endogenous to finance, the proposed framework entails three core elements: (1) the enactment of regulatory regimes and supervisory actions that are calibrated to address specific priorities before, during, and after any crisis; (2) the establishment of a whole-of-society approach for crisis prevention, preparedness, response, and recovery; and (3) the implementation of structured communication policies. While an exhaustive treatise of all regulatory regimes contained in the proposed governance framework is a task necessarily left to future works, an analysis of these three core elements highlights the novelty of this approach, paving the way to more detailed studies.

The identification of different priorities for each phase of a crisis implies an acknowledgment that disturbances are reoccurring and, regardless of their origins, developing through cyclical patterns. Normatively, this means that the governance framework for financial systemic risk should institutionalize the notion that vulnerabilities accrue over time due to the nonlinear interactions both within the financial system—as epitomized by financial cycles of boom and bust—and between the financial system and its environment—when, for instance, financial distress is triggered by events not directly linked to financial activities. Hence, differently from a narrow understanding of vulnerabilities,Footnote ⁶⁸ an all-hazard governance approach recognizes that financial instability comes in different forms and shifts the focus from the proximate causes of distress to the factors that impact a system’s ability to limit their impact. Hence, taking as a point of reference the classifications adopted in the context of pandemic risk (World Health Organization 2017, 22), four periods are isolated: inter-crises, which are represented by the period in between financial crises; alert, which is when economic contractions, price volatility, diffused inability to service debt obligations, unanticipated shocks, or a combination of the above indicate that a crisis might occur; crisis, which is represented by an abnormal and diffused distress that, if left uncontrolled, can spread to the real economy; and transition, which is when the risk of contagion diminishes in anticipation of a new inter-crises period.

Following this rationale, rules are drafted, and supervisory decisions are taken to promote prevention, preparedness, crisis response, and recovery. Rather than relying solely on the general enunciation of protecting financial stability or reducing systemic risk, regulatory regimes and supervisory actions are geared to meet the priorities that each phase of a crisis dictates (Figure 1). During an inter-crises period, priority is given to the enactment of preventative measures; as alert signals manifest and disturbances escalate, the focus shifts toward preparedness and crisis response. Once a crisis has run its course, emergency actions are progressively rolled back to prioritize recovery. An illustration of how this approach can shape a holistic, yet flexible, framework for financial regulation is offered in the context of key prudential regimes enacted to prevent, prepare, and respond to different crisis scenarios.

Figure 1.

Financial Systemic Risk Governance Framework: Core Elements, Crisis Phases, and Policy Priorities.

During periods of relative stability, the implementation of preventative measures requires the identification of system-wide vulnerabilities. Regulators typically rely on a wide array of indicators to detect the buildup of risks. Most commonly, the growth rate of equity and property prices, the gap between the level of indebtedness and economic growth (credit gap), and the difference between the actual and potential production of a given economy (output gap) signal the emergence of distress (Borio and Lowe Reference Borio and Lowe2002; Chen and Svirydzenka Reference Chen and Svirydzenka2021). Although these early warning signs identify the risks related to financial and economic cycles, primary attention should be given to the ability of the financial system to absorb different forms of losses. This means to recognize that financial cycles as well as non-financial events, such as geopolitical tensions and climate change, may expose structural vulnerabilities. Hence, stress testing can be deployed to map structural fault lines that, under plausible circumstances, can trigger system-wide disruptions (Haldane, Hall, and Pezzini Reference Haldane, Hall and Pezzini2007, 5). However, as noted earlier, most stress test methodologies, such as those adopted by the Federal Reserve, cannot be used to perform a preventative function as they are not designed to identify system-wide fragilities. Following the logic of prevention, in fact, stress tests should serve as a diagnostic tool. For this purpose, they should be designed to identify the ability of the financial system or its part to absorb the costs related to different types of events. At the most fundamental level, the assessment of system-wide vulnerabilities requires gauging the availability of short-term funding and the mechanisms that may accelerate their depletion. Such an assessment should not be limited to individual financial institutions; it should also encompass a broader analysis that considers how liquidity shortages can spread owing to nonlinear dynamics and network effects, such as the feedback loops stemming from fire sales or from the failure of financial institutions.Footnote ⁶⁹

Once vulnerabilities are identified, measures are implemented to increase preparedness and, thus, ensure adequate responses to potential threats. Through this prism, correcting the procyclical effect of existing regulatory frameworks is a priority that can be addressed in different manners. A drastic solution is to require banks to maintain more effectively liquid assets (consisting of cash and prepositioned collateral to support the central bank’s emergency lending) than liquid liabilities (comprising deposits and short-term unsecured debt). Advocated by Mervyn King (Reference King2016, 269), former governor of the Bank of England, this approach requires central banks to act as pawnbrokers, providing emergency lending based on pre-posted collateral; concomitantly, it binds banks’ abilities to expand their balance sheets to the availability of liquid assets. This solution is gaining some traction in scholarly and policy circles since it would have prevented, among others, the failures of CS and SVB (Wolf Reference Wolf2023). However, its implementation would require a drastic change in the current business models for banking activities. An alternative approach to limit procyclicality is to detach, at least partially, capital adequacy standards from standardized risk weighting, which encourages herd behaviors, and devise additional buffers that are not linked to bank size, which leads to neglecting smaller entities. Instead, capital and liquidity cushions should be calibrated to address the vulnerabilities identified through early-warning indicators and (diagnostic) stress testing. For instance, regardless of the size of the bank, additional requirements can be progressively imposed on banks that are exposed to interest rate risks or supply chain disruptions.

The second element of the proposed governance framework entails the establishment of a whole-of-society strategy. Accordingly, responsibilities should be allocated in advance proportionately among market participants depending on the crisis phase. This proposition reveals again the limitations of existing regulatory regimes and indicates how to address them. Currently, systemically important financial institutions are required to submit recovery and resolution plans, detailing how they intend to ensure a rapid and orderly resolution in the event of distress or bankruptcy.Footnote ⁷⁰ Such plans, however, were not required for SVB, which was exempted from enhanced supervisory requirements, and were disregarded in the CS case, as a private acquisition was preferred. Moreover, as they focus on the modes to wind up individual financial institutions, resolution plans neglect system-wide dynamics and remain silent on the role that entities should perform as crises unfold.

To ensure adequate crisis response, banks should have in place contingency plans similar to those deployed in the context of disasters. During disasters, private and public facilities, such as schools, universities, and corporations, are often repurposed according to predetermined criteria to limit the impact of the emergency. This technique, for instance, was adopted in the context of the COVID-19 pandemic. To avoid the collapse of the national health-care system, hospitals adopted triage schemes to prioritize urgent cases, whereas schools, clinics, and other buildings were deployed to meet the sudden surge of people in need of medical attention or to isolate contagion (Phua et al. Reference Phua, Weng, Ling, Egi, Lim, Vasishtha Divatia and Raja Shrestha2020). Similarly, the role played by market participants to avert or limit system-wide disruptions should be predetermined. Contingency plans for financial institutions, in particular, should indicate what measures will be enacted to limit the propagation of a crisis. For example, banks should indicate how they intend to communicate with clients to maintain confidence, what additional sources of liquidity can be mobilized if a crisis materializes, and what type of support they can provide to connected entities in distress, such as key counterparties. This mechanism acknowledges that an effective response to systemic risk requires concerted actions where responsibilities (and costs) to mitigate system-wide disruptions are shared proportionally among market participants.

Finally, as a direct consequence of the whole-of-society approach, systemic risk governance entails the implementation of structured communication policies addressing information gaps throughout the crisis cycle. The ability of individuals to understand key economic information and make informed decisions about financial planning, wealth accumulation, pensions, and debt has material consequences for the stability of the financial system. Empirical studies indicate that the limited understanding of basic financial concepts, such as the effects of compound interests, inflation, and diversification on savings, is directly connected both to an excessive accumulation of debt in pre-crisis periods and to the inability to face unexpected economic difficulties during the height of a financial crisis (Lusardi and Mitchell Reference Lusardi and Mitchell2014; Reference Lusardi and Mitchell2023, 9). Similarly, high levels of financial literacy amongst depositors are inversely related to the probability of a bank run (Campioni et al. Reference Campioni, Larocca, Mirra and Panaccione2017), while constituencies with low financial literacy are more likely to default on their mortgages (Gerardi, Goette, and Meier Reference Gerardi, Goette and Meier2013) or liquidate their assets once they have lost value (Bucher-Koenen and Ziegelmeyer Reference Bucher-Koenen and Ziegelmeyer2014). Given that the stability of the financial system is intrinsically linked to the decisions made at the individual level, a fundamental priority for systemic risk governance is to reduce the vulnerabilities connected to financial illiteracy. This implies going beyond current consumer protection regimes and educating the public at large about financial stability.

Recent financial crises have spurred efforts to protect people from financial exploitation due to limited knowledge. In the United States, for instance, the Dodd-Frank Act established the Consumer Financial Protection Bureau (CFPB), an independent agency housed in the Federal Reserve, with a broad range of rule-making, supervisory, and enforcement powers to protect consumers and help them to make better financial decisions by promoting financial education.Footnote ⁷¹ The CFPB’s role is in line with the proposed holistic approach as it monitors system-wide dynamics in view of the needs of different constituencies, attempting to reach any individual interacting with the financial system. Building upon this structure, the lesson from disasters indicates, in fact, that it is imperative for individuals to comprehend the impact of their actions on broader financial dynamics and possess sufficient knowledge to respond appropriately before, during, and after a period of distress. To this end, efforts to bolster financial literacy should be paired with a clear communication strategy to inform the public about each phase of a crisis—the CFPB, in coordination with the Federal Reserve, can perform this task. Clear communication indicating that an expanding economy is in fact in an inter-crisis period can be a powerful tool to enhance public awareness about potential risks and, thus, incentivize preventative actions.