Contagion measure important in a crisis

Contagion is an important problem in finance. It does not happen often, but if you are not prepared for it, the cost can be high. Foremost, returns can swing quickly during a contagion event, but perhaps as important there can be a change in correlation between markets. The classic view is that correlations will move to one during a crisis. Risk will, of course, also increase with this higher correlation. Finally, the direction of causality will change as the contagion spreads. Causality will move from a primary driver to secondary markets. The spread can be measured which will help with nay portfolio decisions. 

A recent paper focuses on measuring contagion in the currency markets, see "A new way of measuring effects of financial crisis on contagion in currency markets", but more importantly, the authors look at the causal relationship that are generated during a crisis. The linking of markets during a contagion crisis is a risk management problem, but determining how the contagion spreads allows for better portfolio responses and the opportunity to generate profits. 

Contagions can be mapped as a network with the strength of the causal relationship measured by estimating the autoregressive properties of an asset with the autoregression betas of cross-assets, the network contagion part. The network contagion factor (NECOF) is the volatility explained by the cross-asset divided by the autoregressive volatility, the network contagion volatility and the error volatility.