Skip to main content

Shock Propagation in Interacting Economic Networks

Andras Borsos
Monday, July 19, 2021, 3:00 pm – 5:00 pm

Please register for this event under the link provided on the right. We will send the link to registered attendees 1 hour before the talk starts. Please note that registration closes at 2:00pm CEST on July 19, 2021.

Members of the Dissertation Committee:

Chair: Tiago Peixoto (voting)

Supervisor: János Kertész (non-voting)

Co-supervisor: Ádám Zawadowski (non-voting)

Member: Fariba Karimi (Complexity Science Hub Vienna, voting)

Reviewer: Giulia Iori (City University London, voting)

Reviewer: Doyne Farmer (Oxford Institute of New Economic Thinking, voting)

ABSTRACTThe economy can be considered a complex system, in which phenomena such as prosperity and crisis are the results of feedback-based interactions of many heterogeneous components. One of the most important factors in the emergence of economic outcomes is very often an underlying spreading process among these constituent parts, which facilitates the propagation of information, sentiments, risks, resources or losses. To be able to analyse these mechanisms in a formalized way, we need to use a wide range of modeling techniques. Among these, network science is one of the most promising approaches, which can equip economists with the necessary tools to build models capable of capturing the intricacies of complex systems. This thesis contributes to these endeavours in a threefold way: (i) by providing insights into the hidden structure and the unique traits of micro-level firm network data; (ii) by proposing a model of shock spreading in firm-level production networks; (iii) and lastly, by offering a novel way of modeling feedback channels between the financial sector and the real economy in the context of interacting economic networks.

In order to be able to investigate firm networks, I obtained access to sensitive datasets about the ownership links and the supplier connections among Hungarian firms. This way, it has become possible to construct the multi-layer representation of the Hungarian firm network, which enabled us to gain insight into its previously unobserved structure. Network analysis provided suitable techniques to explore several topological traits on micro-, meso-, and macro-scale as well, which can be conducive to contagious mechanisms via supplier links. Furthermore, it was also possible to assess the significance of economic entities regarding the extent to which they can influence and control the economy via their ownership relations.

These pieces of information also enabled the simulation of shock propagation in the production network. The granularity of the data made it possible to rectify several shortcomings of industry-level supply chain analyses. The proposed model features heterogeneous production functions at the firm-level, differentiation in the importance of input types and replaceability of defaulting suppliers. With these advancements, the model is capable of quantifying short-term damages after supply chain disruptions, assessing the systemic risk of individual firms, and testing countermeasures, which has relevance for policy making.

Lastly, I propose a computational model of contagious mechanisms in the banking system complemented with feedback channels towards the real economy. The framework incorporates the interactions between the network of banks and the network of firms which systems are linked together via loan-contracts. The model has been embedded into the liquidity stress test of the Central Bank of Hungary, and the results proved the importance of the real economy feedback channel, without which systemic risks could potentially be severely underestimated. To illustrate the versatility of this modeling framework, two further applications have been elaborated. The model can be used to identify systemically important financial institutions (SIFIs), furthermore, it is also suitable to assess the financial stability impact of shocks originated in the real economy.

BIO / András is a Ph.D. candidate at the Department of Network and Data Science with a particular interest in economic networks. He is also affiliated with the Central Bank of Hungary, where he works as an applied researcher focusing on financial stability. In his PhD thesis, András has been conducting research on modeling contagious mechanisms of complex economic systems and designing regulations to prevent instabilities in the economy.