Chaos and Synchronization in Financial Leverages Dynamics: Modeling Systemic Risk with Coupled Unimodal Maps
Marco Ioffredi, Stefano Marmi, Matteo Tanzi · 2026
Read Chaos and Synchronization in Financial Leverages Dynamics: Modeling Systemic Risk with Coupled Unimodal Maps by Marco Ioffredi, Stefano Marmi, Matteo Tanzi (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-01505.txt.
Introduction
Marco Ioffredi, Stefano Marmi, and Matteo Tanzi model how bank leverage can destabilize a whole financial system. Banks target a leverage ratio, assets over equity, under value-at-risk rules. When asset prices move, every bank adjusts the same way at the same time, and those adjustments move prices again. That feedback loop is procyclical: it amplifies booms and busts.
The authors encode the loop in coupled unimodal maps, simple dynamical systems famous for chaos. A single bank's leverage dynamics already shows logistic-like behavior with a rich pattern of bifurcations as parameters shift. Coupling several banks lets contagion and shared shocks enter. The framework shows how systemic risk can grow out of behavior that looks rational bank by bank. For regulators, the lesson is that leverage rules written for safety at one bank can, through synchronization, manufacture danger for the system as a whole.
Catalog record
| CREATOR | Marco Ioffredi, Stefano Marmi, Matteo Tanzi |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 37 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-01505
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Chaos and Synchronization in Financial Leverages Dynamics: Modeling Systemic Risk with Coupled Unimodal Maps Marco Ioffredi, Stefano Marmi, Matteo Tanzi Abstract Systemic financial risk refers to the simultaneous failure or destabilization of multiple financial institutions, often triggered by contagion mechanisms or common exposures to shocks. In this paper, we present a dynamical model of bank leverage — the ratio of asset holdings to equity — a quantity that both reflects and drives risk dynamics. We model how banks, constrained by Value-at-Risk (VaR) regulations, adjust their leverage in response to changes in the price of a single asset, assumed to be held in fixed proportion across banks. This leverage-targeting behavior introduces a procyclical feedback loop between asset prices and leverage. In the dynamics, this can manifest as logistic-like behavior with a rich bifurcation structure across model parameters. By analyzing these coupled dynamics in both isolated and interconnected bank models, we outline a framework for understanding how systemic risk can emerge from seemingly rational micro-level behavior. Financial systemic risk refers to the event in which several institutions fail simultaneously, possibly leading to large financial crisis. To model the evolution of the level of risk associated with a system, it turns out to be crucial to look at the evolution of banks’ leverage ratios, quantities that can amplify both gains and losses. Building on the works 16 18 10 , we propose a simple analytical dynamical model for the evolution of coupled leverages and explore how this can lead to complex behavior within the banking system. In particular, we investigate how the heterogeneity in bank size and strategies of single banks may impact the stability of the whole system. These insights contribute to a deeper understanding of the mechanisms that can precipitate systemic crises and inform strategies for enhancing financial resilience. I Introduction Several mechanisms may be responsible for the emergence of systemic risk in a system of financial institutions (or banks for simplicity), such as informational contagion leading to bank runs and liquidity shortages, direct contagion in which banks fail to pay back loans to other banks, etc. In a broad sense, systemic risk may also be considered to include the risk associated with a system collapsing simultaneously as a result of a shock, either endogenous or exogenous ( 11 ). For instance, in the case in which multiple banks hold positions in the same asset, a severe loss of its value would cause all banks to fail together. This paper will primarily focus on this broader definition of systemic risk. As briefly explained below, the dynamics of the returns of the assets (and thus the evolution of the level of systemic risk in the “broad” sense) are intimately related to the evolution of the leverages of the banks. The leverage \lambda of a market agent, such as a bank, is defined as the ratio between the value of the assets held A , i.e. the economic resources the bank owns, and the value of its equity E , i.e. what remains of the asset value after subtracting the liabilities to which the bank is subject to. As the name itself suggests, a large leverage means an amplification of resources invested, which is made possible by the bank borrowing money to invest (a “pumping” of money from the creditors allows the bank to amplify its invested resources) and which translates into an amplification of gains, but also losses. The relationship between leverage and risk has indeed already been suggested by many ( 1 ; 2 ; 3 ; 12 ; 14 ; 19 ), and depends on the following mechanism. In order to maximize gain, banks want to be as leveraged as possible, as a larger leverage will in general allow for more opportunities to make profit by buying on margin. On the other hand, banks have to face limitations on the maximum attainable leverage imposed by financial regulations in order to make the financial system more robust and resilient. In particular, regulators require banks to define a risk measure, the Value at Risk (VaR) \Lambda_{\text{VaR}} , which is associated with a certain probability of loss P_{\text{VaR}} and is implicitly defined by \int_{-\infty}^{-\Lambda_{\text{VaR}}}f(r^{p})dr^{p}=P_{\text{VaR}} r^{p} being the portfolio return and f its probability density function. In other words, banks are required to loose more than \Lambda_{\text{VaR}} with a probability smaller than P_{\text{VaR}} . Having defined \Lambda_{\text{VaR}} , it is required that the maximum loss, \Lambda_{\text{VaR}} , be smaller than the equity of the bank. This translates in a maximum leverage that banks are allowed to set, i.e. \lambda\leq\frac{1}{\Lambda_{\text{VaR}}} . Notice that \Lambda_{\text{VaR}} depends on the variance of r^{p} , i.e. on the volatility of asset prices. For example, for Gaussian returns with variance \sigma^{2} , a probability of loss P_{\text{VaR}}=5\% would give \Lambda_{\text{VaR}}\approx 1.64\sigma . Banks will thus make an estimate of this variance and set a target leverage \lambda=\frac{1}{1.64\sigma} . This leverage-targeting strategy banks may adopt can result, due to the finite liquidity of the market, in a positive feedback between leverages and asset prices ( 2 ), possibly leading to instabilities very much like what happens in physical systems with positive feedback. Being concrete, what happens is that if things go well, in order to maintain a certain leverage level, banks are led to borrow more; while if things go bad banks will start selling assets. For example, consider a bank having a target leverage \lambda=10 , which it realizes having an asset value equal to 100 and an equity equal to 10 . If there is a 1\% portfolio return, at the next time the bank will have an asset value of 101 and an equity equal to 11 , with a resulting leverage \frac{101}{11}\approx 9.2 , which is less than the target leverage. Thus, to increase its leverage, the bank will borrow some money and increase its asset value, determining as a consequence an increase in its price. This is also known as procyclicality of the Value-at-Risk constraint and has the effect of amplifying prices movements, which is particularly relevant during a falling period of a financial crisis. On top of the above described feedback mechanism, another positive feedback effect may play a role. Indeed, many estimates of risk are based on observation of recent price movements. However, the choice of the time window in the past to consider to perform such an estimate is far from trivial, since there is a trade off between choosing a long temporal window in order to improve statistical precision and choosing a short window in order to capture a more timely measure of risk. This choice, determining the future trading strategies, endogenously moves the asset prices, resulting in a second feedback effect. The models developed here will take both of these feedback effects into account. This being said, it should be clear that by looking at the evolution of the leverages one could get insights in the level of systemic risk of the system. In the case in which the banks are “similar” in the sense that they have comparable sizes and risk estimation strategies, one can reduce to study the evolution of a “representative” leverage. This is what has been done for instance in a paper by Lillo et al. 16