Measuring systemic risk in the Nordic countries - An application of CoVaR

Abstract: Spillover effects and systemic risk contribution of institutions, as measured by their CoVaR and delta-CoVaR respectively, is one way of assessing risk both for an institution in isolation, as well as for regulators and the economy as a whole. CoVaR is the q%-VaR of an institution conditional on another institution already being at its q%-VaR level, whereas delta-CoVaR measures each institution’s marginal risk contribution. This essay applies the CoVaR methodology proposed by Adrian and Brunnermeier (2011) on the Nordic stock market (OMX Nordic 40 Index) in order to measure systemic risk contribution of 36 firms on this market, during the period January 2002 to March 2014. Publicly available stock market data is used to estimate abovementioned measures by applying quantile regression. The results, which are aggregated at sector level, suggest that systemic risk contribution is higher during times of financial distress and sectors generally show a similar pattern in how risky they are over time. VaR is further not positively correlated with CoVaR, i.e. even if a sector is considered risky in isolation as measured by its VaR, it is not necessarily the case that it spills over this risk to other sectors/institutions. However, there are some sectors that contribute more to systemic risk than they are risky in isolation, as measured by their delta-CoVaR and VaR. Sectors contributing the most to Nordic systemic risk are Forestry and Construction, as well as the European stock market as measured by the EuroStoxx50 Index. The banks included in the OMX Nordic 40 Index are also examined in a separate case study, finding Swedbank the most risky and Nordea the least risky in isolation, but the other way around when measuring risk contribution (delta-CoVaR) of these two banks, to other banks.