1 Systemic risk can be defined as “the external effects of banking problems on the rest of the financial system or the real economy” (Laeven et al., 2016). 2 Benoit et al., (2017) and Moch (2018) provide a more detailed review of the theoretical literature on sources of systemic risk. And the contribution of bank i to the systemic risk of banking sector j in equation (3) has the following form:
8) (9) where Xti and Xtsystem/i are respectively the returns of bank i and the returns of system j. Furthermore, we add local variables deemed necessary in the context of the GCC countries. Furthermore, Caccioli et al., (2012) and Lu and Hu (2014) theoretically approved that size is one of the most important factors for systemic risk.
The table shows the correlation coefficient matrix among the banking indices of each of the GCC stock markets. The magnitude of the tail losses from average levels is replicated for all GCC-SIBs. Summary Statistics for the Systemically Important Banks in the GCC Region The table reports summary statistics for the logarithmic returns of the GCC SIBs.
Summary Statistics for the Market and Bank Indices The table reports summary statistics for the logarithmic returns of the GCC SIBs.
RESULTS
Scatter plots of the GCC-SIB's VaR and ∆CoVaR for all GCC Banking Indices (continued). Scatter plots of the GCC-SIB's VaR and CoVaR for all GCC Banking Indices (continued). To understand these differences, we provide the cross-sectional correlations for each of the GCC countries' market and banking indices in panels A and B of Table 5 , respectively.
It shows that, on average, the cross-sectional correlations of VaRi,t and CoVaRi,t for market and banking indices are around 0.5. When it comes to the cross-sectional correlation between VaRi,t and ∆CoVaRi,t, the cross-sectional correlation for market indices (0.063) is weaker than for banking indices (0.228). We also note that the slightly higher cross-sectional correlation for banking indices is driven by the larger simultaneous movement of VaRi,t and ∆CoVaRi,t for Bahrain, Kuwait and Oman.
Our findings are consistent with evidence from empirical studies showing that larger banks are the most important determinant of US systemic importance. In Table 6, we provide the results for QNB of Qatar and NBK of Kuwait in panels A and B, respectively. The one-day CoVaR shows that QNB and NBK contribute significantly to the VaR of both markets and banking indices in the GCC countries.
∆CoVaR indicates that as the QNB tail risk moves from the normal state to the disturbed state, tail bursts can further increase by an amount of 13 bps for OMI, at a minimum, to 167 bps DBI, at a maximum. The table shows the correlation estimate of GCC-SIB' VaR, CoVaR and ∆CoVaR. The table reports summary statistics for CoVaRsystem|i and ΔCoVaRsystem|i, which is defined as the change in VaR of financial system j attributable to the stress of SIB i relative to its average condition.
The column called CoVaR shows that Saudi banks contribute large parts of the VaR of the GCC market and banking indices. When it comes to ∆CoVaR, the delta CoVaR contributions associated with NCBs, the largest increase in tail losses is of magnitude 63 bps and the smallest increase is 6 bps. Here, we note that the largest increase in tail risk of all Saudi SIBs is for the Qatar market index.
The smallest loss contribution is for KBI when linked to distressed states NCB, ARB, SFG and RB. For the market/financial system j 13 GCC index, shown in the first column of the table.
CONCLUDING REMARKS
This effect is further supported by the weak cross-sectional correlation of the isolated, idiosyncratic risk of a SIB, i.e. VaR, and the SIB's contribution to systemic risk, measured as the difference between the CoVaR in the distressed state and its state average, i.e. ΔCoVaR. While cross-sectional correlations between SIB VaR and system CoVaR, i.e. market/banking sector VAR conditional on other SIB VaRs show a strong correlation. Especially when our results show that almost all systemic risks of SIBs carry significant spillovers not only to the domestic market in which they operate, but also to at least one other GCC market and banking sector.
Our results show that, except for QNB and NCB, the systemic risk of the SIBs to the tail risk of the rest of the markets is loosely related. In these spillovers, we note that QNB's tail risk explains a large part of the systemic risk of the rest of the GCC markets – both broad-based market index and the banking sector. Our findings are important for regulators, financial risk managers and portfolio diversifiers to understand the systemic relationships and how it spills over within and across GCC countries.
For example, spillovers exist from SIB tail risk to other domestic banks in the GCC countries, and SIBs listed in Qatar, Kuwait and the UAE affect the changes in tail risk for all GCC banking sectors except in the case of Saudi Arabia. To prevent medium and small banks from transferring risk from large banks, regulators in the region can adopt new policy reforms to regulate the banking sector effectively. Portfolio diversifiers may consider the stocks of these major banks when they intend to reduce the risk of their portfolio investments.
Since GCC countries have fixed exchange rates and limited taxes, regulators are constrained by fiscal and monetary policies. The impact of large changes in asset prices on intra-market correlations in the domestic and international markets.
