The drivers of systemic risk of the CALL and REPO Argentinean Interbank Markets

DEBORAH NOGUERA; GABRIEL MONTES ROJAS

Bogotá

Conferencia; II Regional Conference on Payments and Financial Market Infrastructures - CEMLA; 2023

Centro de Estudios Monetarios Latinoamericanos

This paper analyzes the determinants of systemic risk in the Argentinean interbank market. As explanatory factors, we consider both the financial variables of the participating institutions and the topology of the interbak network. We calculate systemic risk using the Differential DebtRank methodology, considering different levels of initial shock. To analyze the determinants of systemic risk, we apply three machine learning models: Adaptive Boosting (AdaBoost), eXtreme Gradient Boosting (XGBoost), and Random Forest. We find that the relevance of different factors in driving systemic risk varies depending on the size of the initial shock and whether systemic impact or vulnerability is the dimension of risk under analysis. Topological characteristics are most important in predicting systemic impact, while financial variables are the primary drivers of systemic vulnerability. Finally, the interconnection structure in the interbank market is more relevant in explaining systemic risk, in both of its dimensions, in the guaranteed market compared to the unguaranteed market.