Avoiding Systemic Risks and Tackling Recessions: A Network Perspective

“Network theory provides a critique of standard wisdom on how to create a stable financial system.” –Professor Joseph Stiglitz of Columbia University

Network analysis has steadily found its way into the study of economics and politics. Its application helps explain how consumers, firms, and institutions are linked, and how their connections influence how each interacts with the economy. Failure to understand these connections leads to decisions that fail to give appropriate attention to the risks faced by others. The Financial Crisis of 2008 is one critical example: as financial analysts failed to take into account how risk transfers between consumers and financial institutions, the housing market became a highly risky investment, for which consumers paid the price.

Risk is contagious, and the networked structure of our financial system helps spread it. This metaphor of risk-as-disease explains the notion of systemic risk. Defined as any potential risk that threatens the stability of the entire economic, political, or social system, systemic risk must be better understood by policymakers.

Currently, our understanding of these networks is limited. In a paper aimed at understanding the nature of how such risks travel through financial systems, Acemoglu et al. (2015) model different kinds of financial networks and analyze their stability when subjected to shocks (such as a sudden mass withdrawal of deposits). They do this by treating liabilities between banks as the definition of how one bank is connected to another in the network. In so doing, the authors seek to explain how sudden shocks to individual banks or groups of banks can snowball into a financial crisis. The authors base their analysis on the assumption that no one individual has the power to create a shock large enough to destabilize the whole economy.

The authors’ findings suggest that financial regulators should be cautioned against treating financial networks as a single monolith. Depending on their structure, they can behave differently under crises and hence may need individual policy designed accordingly. For example, when a bank’s liabilities are held equally by other banks, during a crisis, the loss is shared, and the system does not collapse. On the other hand, when few banks hold all the risky investments, crises will quickly flow down to all the other small banks and result in systemic collapse.

Consequently, when small shocks occur, top-performing banks can easily rescue small ones without a great effect on the larger system. However, the authors qualify this statement by arguing that, when the frequency of such shocks increases beyond a threshold, the interconnectedness and diversified liability might itself cause the risk to spread across all connected banks. In these cases, ‘weakly connected’ financial networks—in which different subsets of banks have minimal claims on one another—are significantly less fragile.

The authors also define a novel measure, called ‘harmonic distance,’ which captures the susceptibility of each bank to failure. The level of harmonic distance depends on how close the debtors of one bank are to another and also the magnitude of risk posed by these debtors to their banks, if they should fail to honor their loans. They argue that this is a better measure of identifying ill-performing financial institutions than the measures currently available in the literature.

The need for network-based analysis is not limited to the US. Internationally, systemic risk analysis is becoming increasingly relevant to economic planners and policy analysts. The de Larosière report, for example, has established the European Systemic Risk Board (ESRB), which aims to monitor systemic risks of financial markets in the EuroZone. Additionally, the new Basel III norms put forth by the Basel Committee on Banking Supervision consist of a global and voluntary regulatory framework which explicitly attempts to avoid systemic risks and crisis scenarios, like the 2008 Financial Crisis, by calling for higher capital reserves and countercyclical buffers, as well as the reduction of counter party risks through hedging.

The global impetus towards the adoption of a network-based analysis as a means for understanding and preventing systemic economic risk should not go unnoticed. For policymakers, the existence of systemic risk implies that the effect of a policy adopted in one domain could be felt in another. There must be an effort to use tools that developments in computer science, statistics, and econometrics have provided to us to better predict and address such reaction chains.

Article Source: Acemoglu, Daron, Asuman Ozdaglar, and Alireza Tahbaz-Salehi. “Systemic Risk and Stability in Financial Networks,” American Economic Review 105, No. 2 (2015): 564–608.

Featured Photo: cc/(alessandroguerriero, photo ID: 85858311, from iStock by Getty Images)