Networking the Euro crisis

I just wanted to flag a great little graphical analysis, published in the New York Times, of the links between different economies and the Eurozone. Each country is represented in a circle proportional to the size of its outstanding debt obligations with arrows between different countries representing external claims (the arrows are proportional to the size of the claims as well). Mouse over each of the spheres to get a sense of the numbers.

The application of network theory to finance has gained steam after the contagion that followed the collapse of Lehman Brothers in September 2008. The central idea is the analysis of the networks formed between different financial institutions both within and across jurisdictions to better understand systemic risk – or aggregate risk that affects the entire financial sector as a whole.

As the little graph in the NYT shows, a basic sketch of the networks (here illustrated through contractual relationships) between different economies gives us a powerful image of the extent of contagion that could ensue if one (or more) of the peripheral European Union economies fails to make good on its debt. A more substantial network analysis would require the links between individual financial institutions so that central banks and financial regulators can identify the banks and other institutions that are most at risk from the bankruptcy of any one economy.

Of course, contractual relationships aren’t the only determinants of networks between financial entities: interlinkages may exist in the absence of a financial contract as well. For instance, the negligible financial relationship between Greece and Spain will not particularly help the Spanish economy in the event of the Greek economy declaring bankruptcy. The perceived links between the two economies (particularly via the common currency) may prove to be just as strong.

Nonetheless, central banks are turning to network analysis to arm themselves with the information to spot a potential crisis before it arrives. A crucial question, though, is what does a financial regulator do if it identifies a concentration of risk between a number of financial institutions? Is the regulator to formulate a set of rules according to which a set of networks may be deemed too complicated and can the regulator act to dismantle these links? Given that network analysis is typically based on information supplied by banks themselves, and at a bit of a lag, it will likely not be useful when contagion has actually occurred and the financial sector is distressed.

In any case, a full-fledged network analysis will allow the researcher to simulate a crisis and then plot the development of the crisis, as well as the build-up of contagion. By improving the quality of crisis simulation exercises, regulators can become more alert to the scale and extent of contagion and the potential for risks to increase across the financial system.

And, of course, it makes for pretty pictures.