CoRisk: Quantile Regressions in Practice

Jorge A Chan-Lau

The previous chapter analysed techniques traditionally used to assess the transmission of negative shocks across different institutions or sovereign countries. In particular, risk measures built on extreme value theory, as described in the previous chapter, are suitable for capturing the impact of large negative shocks. One common problem associated with these measures, however, is that by focusing only on tail (or extreme) realisations, they neglect to use the information contained in the whole data sample. This problem becomes more acute the shorter the length of the data sample. In addition, extreme value theory is not helpful for examining how linkages between institutions differ between normal and crisis periods.

As an alternative to extreme value theory, this chapter introduces the concept of CoRisk, which measures how the default risk of an individual institution changes whenever other institutions are subject to large negative shocks after controlling for the effect of other common factors, either fundamental or technical, that drive default risk. CoRisk, or default risk co-dependence, therefore captures contagion or risk spill-overs across institutions as implied from