Interbank network
An interbank network is the web of lending relationships between commercial banks, central banks, and other depository institutions that enables the clearing of payments, the redistribution of liquidity, and the propagation of financial stress across national borders. It is not merely a payment infrastructure; it is the circulatory system of the modern economy, and like any biological circulatory system, its architecture determines whether a local wound becomes a systemic hemorrhage.
Topological Structure
The topology of the interbank network is strikingly asymmetric. Empirical studies of overnight lending markets — the federal funds market in the United States, e-MID in Italy, the eonia in the euro area — reveal a core-periphery structure: a small number of large, well-connected banks form a dense core that intermediates between a much larger periphery of smaller, less connected institutions. The core banks are not merely bigger; they are structurally essential. Their removal would fragment the network into disconnected components, freezing interbank lending and triggering cascading defaults.
This core-periphery structure is not an accident of size. It is an emergent property of the network formation process. Large banks have stronger credit ratings, lower perceived default risk, and deeper relationships with counterparties, which makes them natural intermediaries. Smaller banks, unable to lend directly to each other because of information asymmetries and credit-risk concerns, route their transactions through the core. The result is a network that is efficient under normal conditions — liquidity flows smoothly through the core to wherever it is needed — but fragile under stress, when the core banks themselves become suspect and the periphery has no alternative channels.
The mathematical signature of this fragility is visible in the network's degree distribution. Unlike random graphs, where degree distributions are Poisson, interbank networks exhibit heavy tails: a few nodes have degrees orders of magnitude larger than the mean. These high-degree nodes are the super-spreaders of financial contagion. When Lehman Brothers failed in 2008, it was not the size of its balance sheet alone that made it systemic; it was its position in the interbank network — its centrality, its role as a clearing counterparty, its connections to every major dealer bank in New York and London.
Contagion Dynamics
The interbank network is the primary transmission channel of financial contagion. The mechanism is simple: when a bank defaults on its interbank obligations, its creditors suffer losses. If those losses are large enough, the creditors may become insolvent, defaulting on their own obligations, and the failure propagates. This is direct contagion — the domino effect that macroprudential regulation was designed to prevent.
But direct contagion is only half the story. Indirect contagion operates through the liquidity channel: when a bank's solvency is questioned, its counterparties stop lending to it, even if it is fundamentally sound. The refusal to lend is rational — why expose yourself to counterparty risk? — but it is also self-fulfilling. A bank that cannot roll over its overnight funding must sell assets, depressing prices, which weakens the balance sheets of other banks, which makes their counterparties reluctant to lend to them, and so on. The interbank network transmits not only default but also fear.
The Bank for International Settlements and the Basel Committee on Banking Supervision have attempted to mitigate these dynamics through capital requirements, liquidity coverage ratios, and stress testing. But these tools address the nodes, not the network. A bank that passes a stress test is still vulnerable if its counterparties fail, and the network topology that concentrates risk in the core remains unchanged. The interbank network is a system-level property that cannot be regulated one bank at a time.
The Network as a Constraint
From a systems-theoretic perspective, the interbank network is not a neutral infrastructure through which autonomous banks interact. It is a constraint topology that shapes what banks can do, what risks they can take, and how crises unfold. The network does not merely transmit shocks; it transforms them. A liquidity squeeze in one market segment becomes a solvency crisis in another not because of any change in fundamentals, but because the network's architecture channels stress along particular paths.
This insight has implications for the design of financial stability policy. Central clearing counterparties (CCPs) were introduced after 2008 to mutualize counterparty risk and reduce the complexity of the interbank network. But CCPs themselves have become new systemic nodes: if a CCP fails, the entire cleared market fails with it. The solution to network fragility — replacing a mesh of bilateral exposures with a hub-and-spoke architecture — has created new forms of fragility at the hub.
The interbank network is the skeleton of modern finance, and it is a skeleton with osteoporosis. Its core-periphery structure concentrates risk in a handful of institutions that are too connected to fail, while its opacity — the bilateral nature of most interbank exposures means the full network topology is unobservable even to regulators — makes it impossible to map the pathways of contagion in real time. We regulate banks as if they were independent entities, but they are nodes in a network whose architecture makes their fates inseparable. The next crisis will not be caused by a single bad bank. It will be caused by a network that we did not understand until it was too late.