مقاله انگلیسی همکاری درون سازمانی شبکه را در معرض ریسک سیستمیک قرار میدهد
Abstract
Keywords
1. Introduction
2. Literature review
3. Data
4. Methodology
4.1. Network metrics
4.1.1. Centrality
4.2. DebtRank
4.3. Beyond inter-bank exposures
4.4. Reconstruction of firm-to-firm network
4.4.1. Binary adjacency matrix via the fitness model
4.4.2. Assigning weights via the RAS algorithm
4.5. Building a network of effective exposures of banks towards firms
5. Results
6. Conclusions and further work
Declaration of Competing Interest
Appendix A. Additional results
References
Abstract
We propose to use a systemic risk metric for an extended network which includes the inter-bank network, the banks-firms bipartite network, and the intrafirm exposures network in Uruguay. This is the first work, to the best of our knowledge, in which the intra-firm exposures network is estimated with good accuracy by using information from a firm survey. Given that the survey only includes the three most relevant debtors and creditors, we complete the full intra-firm exposures matrix by resorting to Maximum Entropy, Minimum Density and a new method which takes into account the known entries of the matrix obtained from the survey. We show that ignoring intra-firm exposures results in an important underestimation of systemic risk. Moreover, if the marginal liabilities are used as an indicator of the systemic relevance of firms, important network effects are ignored. To conclude, the paper contributes with a precise estimation of the impact of intra-firm exposures to overall systemic risk.
1. Introduction
The increasingly complex and interrelated connections in the financial system are considered one of the main sources of risk amplification and propagation of shocks. This was made evident in the worst possible way during the Global Financial Crisis (GFC) after the fall of Lehman Brothers. Since then, macroprudential policies and the interconnections in the financial system have taken central stage in the financial stability agenda.
These interconnections among financial entities have been modelled by resorting to network theory and models. Subsequently, researchers have modeled financial entities and their relationships by financial networks. Extensive literature exists on the structure of these networks and the effects of these structures on the propagation of both microeconomic and macroeconomic shocks (see Battiston, Martinez-Jaramillo, 2018, Martinez-Jaramillo, Carmona, Kenett, 2019) for an introduction.
Nevertheless, contagion through commercial indebtedness among firms or economic sectors has received less attention (Acemoglu et al., 2016), mainly due to the lack of information. However, this situation has recently changed, and now it is possible to find works that include the real sector of the economy and its relationship with the banking system: Poledna et al. (2018) and Silva et al. (2018).
This work aims to contribute to filling this gap as well by building a commercial and financial debt network for Uruguay. We resort to a firm-level survey that included questions on the main debtors and creditors for each firm that participated in the survey (see Baron et al., 2020). Additionally, the links between the firms and the banking system are also known from the credit registry. Finally, even though the inter-bank market is small and we find low contagion through this channel, the exposures network among banks is also considered.