Too Central to Fail: Financial Networks, Risks and Policy Responses - E233
The GFC of 2008 prompted calls for a better understanding of financial market vulnerabilities, and the inherent risks within the system. The complexity of relationships between financial entities may be illustrated using the concept of a network. Furthermore, the multiple linkage points within a network may be thought of as channels through which shocks may be transmitted to various parts of the system. Accordingly, a network approach represents a useful framework for studying the transmission of systemic risk.
This research applies these tools, in concert with concepts and practices from financial economics and statistics, to model the network of relations between the banks, the broader Australian economy, and the international financial markets. Moreover, the study sets out a mapping of the network of perceived financial dependencies between these respective entities.
Publicly available information, share prices, and indices of the respective entities are used to construct networks of partial correlations between their returns. The networks do not reveal what caused investors to adjust their expectations regarding particular entities. However, stock price performance quantifies the changes in market perceptions of the entities’ current present values. In this way, the networks become useful tools in helping to gain a better understanding of the operation of the financial markets, and the transmission of shocks. Accordingly, they may prove a useful addition to existing mechanisms for policy and regulatory analysis.
In the context of investment diversification, the use of correlation statistics in portfolio management is well established. However, to gain an insight into the structure of the market for the purposes of financial stability analysis or policy design, partial correlation analysis comes into play.
Partial correlations can identify direct co-movements in the relative value adjustments between particular pairs of entities, while controlling for all other variables. Construction of a network of partial correlations facilitates the observation of the likely path of a shock as it spreads through the network. From the perspective of a regulator, this facilitates the identification of key relations and, in turn, provides a point of focus for regulatory oversight.
Understanding and measuring spill-over effects by means of a partial correlation matrix represents an important component of any policy aimed at reducing systemic risk. In this regard, network theory uses various measures of centrality to measure the relative importance of particular points of risk transmission.
The study makes the following findings in terms of the linkages between the various entities and sectors within the Australian economy:
• There are strong partial correlations between the big four banks and the broader financial and real estate sectors;
• There are strong direct links between the regional banks, Bank of Queensland, and Bendigo and Adelaide Bank;
• Australian economic activity presently centres around the industrials and consumer services sectors. More broadly, the study shows a decrease in the central economic role of the mining sector post 2008, with a corresponding increase in the importance of the industrial sector;
• The basic materials, oil and gas, and consumer services sectors appear to be the primary conduit through which Asian influences impact the Australian economy.
This research has been presented at:
CIFR WORKSHOP: Financial System Inquiry
Sydney (May 2014)
1st Conference on Recent Developments in Financial Econometrics and Applications
Deakin University, Melbourne
CIFR WORKSHOP: Financial System Inquiry III
Sydney (March 2015)
21st International Conference on Computing in Economics and Finance
Taipei, Taiwan (June 2015)
2015 Australian Conference of Economics
Brisbane (July 2015)