The Crucial Role of Network Analysis in Mitigating Systemic Risk
One of the biggest financial crises the world ever had to recover from was the 2008 crash, which put the global economy into an 18-month recession. According to the U.S. National Bureau of Economic Research (the official arbiter of U.S. recessions), the recession began in December 2007 and ended in June 2009. This crisis may have started in the United States, but its impact was global, affecting the entire world. It is thus only logical to ask ourselves why European countries were under fire even though they were not the original nations implicated. The reason is far from simple but can be understood through the following notion: the interconnectedness of financial markets.
To make everything clear, we have to define what a financial market is. A financial market, according to the Oxford Dictionary, is "a market in which financial assets are traded." Even though this definition is accurate, we can translate it into plain English for all readers not specialized in finance. All jokes aside, a financial market is a platform where certain participants, such as individuals (like retail investors), institutions (such as mutual funds and pension funds), or even governments, meet to trade different things. For example, the actors can meet on the forex market to trade currencies (e.g., USD/EUR exchange rate). They also trade stocks that represent ownership in companies (e.g., Apple shares). Furthermore, participants can also trade bonds, which are debt instruments (e.g., U.S. Treasury bonds). Finally, commodities are also traded and include gold or oil futures. You might have guessed, a financial market is like a huge place where everything is traded under certain regulations, of course.
Now that we all have a grasp of the relationship between network analysis and financial markets, we have to see the different applications it represents. The first application of network analysis will be to the interconnectedness of banks and thus the relationships they have through interbank networks. Each node represents a bank (an entity), and you can map out the borrowing and lending relationships between banks (the edges). Thus, you can evaluate which bank is at the center of the system by considering the evaluation of the edges that represent these interactions with other banks. Furthermore, you can also see which banks are serving as intermediaries in transactions between other banks. Lastly, we can also perceive the degree of speed of connectivity between banks with the edges.
This system is really important to evaluate the balance of the connections between banks. For example, higher degree centrality indicates a bank is highly connected; high betweenness centrality means the bank plays a crucial role in the flow of funds, and banks with high closeness centrality can spread or absorb shocks rapidly. All of these deductions can be seen with network analysis.
The second application is through the cross-ownership networks that will map out the power relationships of different actors inside an enterprise. This application shows the equity of different actors in one entity, which indicates the level of distress of the entity and the impact it will have on the different actors. This can be seen through either direct contagion, which means that if a company experiences financial difficulties, its direct equity holders are immediately affected due to the decline in the value of their investments. Additionally, the network analysis will also show indirect contagion, which means that the impact can spread further if the affected equity holders, in turn, experience financial distress and affect other companies they hold stakes in. In other words, it provides information on the stability of an entity's financial state by showing how distress can spread directly through equity ties and indirectly through interconnected investments.
To effectively analyze and use network analysis in relation to cross-ownership networks, several network metrics can be used, such as eigenvector centrality. Eigenvector centrality measures the influence of a node in the network, considering not just its direct connections but also the connections of its neighbors. In other words, a node is influential if it is connected to other highly influential nodes. This is the tool you need in order to then make deductions on the importance of each node and the relationship it will have in the network. This tool will not only be a local reference marker but also an extremely important indicator of the connection with other influential entities and thus indicate the impacts one entity will have on other entities that are connected. Another metric that can be used is network density. Network density measures the proportion of potential connections in the network that are actual connections. A high-density network has many interconnections between nodes. If we go back to the financial markets, we thus understand that in a highly dense cross-ownership network, financial distress can quickly propagate due to the numerous connections, which underlines the importance of understanding the density as it helps in assessing the overall risk of contagion. A dense network indicates a high level of interconnectedness, which can lead to the rapid transmission of financial shocks.
You might have seen that those two networks indicate patterns, and those patterns are a source for deductions about the state of an entity and the impact it will have on the actors that own equity in the entity or even the other entities that are linked with it. Moreover, network analysis is also used to plan out the payment and settlement systems. First, the network analysis will map the flow of payments between financial institutions, showing which institutions send money to whom. To understand the importance of different institutions in the system, we also need to know which institutions are at the center of validating all those transactions. It’s obvious that the transactions are not as simple as giving a $100 bill. These transactions are regulated with clearing that verifies the transactions and then settlement that transfers the money. By analyzing the network and identifying which institutions handle the most exchanges, we can see which institutions have the most of these procedures and then adapt to prevent risks. By analyzing the network of clearing and settlement systems operations, we can detect potential points of failure and thus prevent the collapse of the system. Furthermore, to ensure the robustness of the system, we can put it under various stress scenarios and see what impacts some institutions have if they are put under certain stress. Then we can balance the entire network by reducing the weight of each institution on the system to make the network robust.
There are many ways to reduce the weight of certain institutions on the health of the network. The first one is diversifying the risk by not putting all the eggs in the same basket. The second is to enhance capital requirements to ensure the feasibility of the entities when making those transactions. This can be done by requiring institutions to hold capital proportional to the riskiness of their assets. This method comes with enhancements of the regulations and oversight on the transactions and the entities involved. The institutions that control and oversee can implement transaction caps to limit risks, enhance the transparency of the actors involved to better evaluate the risks of the transactions, and promote decentralization of entities to prevent the domino effect. Furthermore, on the entity level, they can increase their technological protection to avoid being victims of cyberattacks. Finally, another important way of making the network robust is to reduce the “Too Big To Fail (TBTF)” risk by breaking up large institutions that might become Diogenes' lantern.
At this point you might have understood that mapping out the networks, whether it is inside an entity or between banks, the method is the same. We analyze to understand who are the biggest actors and thus who will impact the others the most when something bad happens to them. This can be used in market trading networks where we try to understand who are the biggest firms that have the biggest flow of liquidity in the market. With this, we can detect clusters of highly interconnected traders or firms that could be vulnerable to simultaneous failure.
In order to help the process, many technologies have been developed over the years and have thus helped the analysis of the networks. The first technological advancement is enhanced data processing with advancements that have revolutionized the handling and processing of "big data." "Big data" refers to extremely large datasets that can be analyzed to reveal patterns, trends, and associations, especially relating to human behavior and interactions.
Furthermore, AI also improved the processes by integrating machine learning into the analysis of the network. Once again, this technology will be able to process large amounts of data and identify the patterns, correlations, and anomalies we were talking about before in order to then draw conclusions and act accordingly. For example, machine learning algorithms can detect hidden relationships between entities, predict potential risks, and simulate various scenarios to assess network resilience. Lastly, the power of AI and technology in general enables us to have a continual analysis in real time. This real-time capability allows for swift responses and informed decision-making in volatile market conditions.
In conclusion, understanding the interconnectedness of financial markets through network analysis is crucial for preventing future financial crises and ensuring global economic stability. The 2008 crash underscored the far-reaching impact of systemic risks, showing how the collapse of financial institutions in one country can trigger a domino effect worldwide. By mapping out the relationships between banks and other financial entities, we can identify key players, potential points of failure, and the overall health of the financial network. Network analysis helps us see beyond the surface, revealing the intricate web of connections that define our financial systems. By evaluating metrics like degree centrality, betweenness centrality, and network density, we can better understand the flow of funds, the role of intermediaries, and the speed of connectivity. These insights allow us to take proactive measures, such as diversifying risks, enhancing capital requirements, and promoting decentralization to mitigate the risk of systemic collapse. Moreover, technological advancements in big data and AI have significantly enhanced our ability to process and analyze vast amounts of financial data in real-time. This capability not only improves our understanding of current market conditions but also allows us to predict potential risks and respond swiftly to emerging threats. Ultimately, the integration of network analysis into financial systems is not just a tool for risk management but a necessary step towards building a more resilient and stable global economy. By continually refining our analytical methods and leveraging cutting-edge technology, we can better anticipate and mitigate the risks that threaten financial stability, ensuring a more secure future for all.
Comentários