The essence of the financial system itself and its relation to the real economy can be naturally captured as a network [Figure 2.4], which can be sub-divided into smaller ones.
There are many types of connection12 between financial entities13 depending on which aspect we want to study. For example, banking systems can be modeled by agents that are connected directly through mutual exposures in the interbank market and indirectly through holding similar portfolios or sharing the same mass of depositors. The network view provides implications that can hardly be devised from the traditional view of financial systems.
12e.g. liability-asset connection – where liabilities of an institution are assets of others; information connection – where institutions share certain information and behave in a coordinated way (e.g.
invest in the same assets, have the same owner); regulation connection – where institutions are under the same operation restrictions; etc.
13 Financial entities are used as a broader term than financial institutions, which include all agents that participate in financial transactions.
Global financial crisis
tighter credit conditions /credit crunch slower economic growth of foreign markets
funds more expensive funds less available tightening of bank lending standards
reduced lending to households and corporates
and/or
reduced demand for loans under certain economic prospect and higher borrowing costss reduced consumption
and investment spending
borrower debt service problems/
declining asset quality
domestic bank problems further tightening of lending
standards economic slowdown in home
country
Feedback loop between bank problems and
the real economy
reduced export to foreign countries
When looking under the lens of network theory, the externalities that caused systemic risks can be described as different kinds of contagion in financial networks. According to Cont (2012), these contagions can be of following types:
Correlation. Homogeneity behavior of FIs creates large exposures to common risk factors, which in turn lead to simultaneous losses across FIs.
Counterparty Risk/Balance sheet contagion. The default of an FI may lead to the write- down of assets held by its counterparties and consequently, leads to its insolvency.
Spirals of illiquidity. Market moves and credit events may lead to margin calls, which lead to the default of institutions that lack sufficient short-term funds.
Procyclical feedback effects. Fire sales of assets due to deleveraging can further depreciate asset prices and lead to losses in other portfolios, generating endogenous instability.
These contagions are underlying factors that facilitate cascading failures14 in financial network.
14In network theory, cascading failures usually begin when one part of the system fails. When this happens, nearby nodes must then take up the slack for the failed component. This in turn overloads these nodes, causing them to fail as well, prompting additional nodes to fail one after another in a vicious circle.
Figure 2.4: Macro-financial linkages between real and nominal (i.e. financial) economy.
Source: Galati and Moessner (2011)
Haldane (2009) noticed the similarity of global financial system – and the global crisis happened within – with various social and natural phenomenons. Using network theory, he explained the emergence of complexity and homogeneity as two primary characteristics of modern financial networks. This intuitive work of Haldane has laid the ground for many macroprudential and systemic risk studies later on.
Acemoglu et al. (2013) established a stylized network model [Appendix 5] that confirmed main ideas of Haldane (2009). In which, for small negative shocks, a more densely connected financial network (corresponding to a more diversified pattern of interbank liabilities) enhances financial stability. Beyond a certain threshold, larger shocks are amplified by such densely connected network leading to more fragile system. Therefore, factors that improve resilience under certain conditions can be a source of systemic risks under others.
Galati and Moessner (2011) restated these findings by referring to the research of Gai and Kapadia (2008) and Nier et al. (2008). They constructed artificial homogeneous networks of banks and analyzed the effect of an idiosyncratic shock on the resilience of the network.
Both found non-linear effects of net worth (i.e. size of the bank) and network connectivity (the probability that one bank has lent to another bank) on contagion. These results
illustrate that the financial system is likely to have a robust-yet-fragile tendency – while the likelihood of contagion may be reduced by greater connectivity; the potential impact of a shock has a much larger scale.
Along the network interpretation of financial systems, the notion of too-big-to-fail may not fully describe institutions that have high network importance. The current term to describe those institutions is systemic important financial institutions (SIFI). The new term incorporates too-interconnected-to-fail financial institutions as well as too-intercorrelated- to-fail financial structures. In which, too-interconnected-to-fail refers to FIs with large/complex bilateral exposures and too-intercorrelated-to-fail financial structure refers to the incentive of individual FIs to participate in risky behavior coordinately15 when the government have to insure the safety of the whole sector (e.g. banking, stock market) [Overview A]. Lastly, it is worth mentioning too non-traditional to fail principle, where non-interest activities may lower the risk for individual banks yet increase the systemic risk level (More & Zhou, 2014).