Abstract
Central bank swaps have flourished since the 2008 financial crisis and the intertwining swap lines have formed a typically complex network, in which bilateral swap relationships have been deeply embedded. This paper constructs a global swap network of central banks and then utilises network indices to measure the network positions and resources of swap central banks. From the perspective of the global swap network, it employs the Spatial Durbin Model to empirically estimate both direct and indirect effects of swap positions on foreign reserves and then further explores the applicability of these effects. The findings confirm a negative impact of network positions on foreign reserves. Besides, swaps can significantly substitute for foreign reserves of relevant banks by back-up rescue and demonstration mechanisms. The fact that spatial spill-over effects are much larger than direct effects justifies the necessity to assess swap performance from a spatial perspective. Applicability tests show that the substitution effect is larger in less developed countries and those with lower foreign exchange stability and financial openness. The conclusions have important policy implications for the construction of a global financial safety net and the coordination of counter-measures to cope with crises.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 Information on central bank swaps disclosed on the websites mainly covers the swap partners, quota, currency and tenor. The withdrawal amount is only available on the Fed’s website. Since 2013, earlier temporary swap lines between the Fed and five developed central banks have been turned into standing lines with an unlimited amount. Thus, the network cannot be constructed based on contracted amount or withdrawal amount. The undirected binary network based on the existence of swap links in this paper may not be accurate enough to reveal the full picture of central bank swaps; however, analytical indices based on the network are able to depict central banks’ capabilities to control resources.
2 See the news release by the PBOC on 12 December 2008, available at: http://www.pbc.gov.cn/goutongjiaoliu/113456/113472/2894599/index.html.
3 The different signs of the Moran’s I index and do not mean that they are contradictive, since the Moran’s I index just testifies the spatial relationship between central banks’ foreign reserves without excluding the effects of other variables and thus provides a rough estimation on the spatial auto-correlation.
4 Statistical analysis shows that in the sample, the average ratio of foreign reserves to the GDP in developed countries is 26%, higher than that of 19% in developing and emerging economies.
5 The index of institutional affinity between RMB swap recipients and China is calculated based on government governance indicators from the World Bank’s Global Governance Database. The first step is to calculate institutional distance in which and represent the m-th indicator for countries i and j in year t, respectively, and is the variance of the m-th indicator. Then, the value of institutional affinity equals the reciprocal of institutional distance LD.
6 Referring to Calvo, Izquierdo, and Mejia (Citation2004), the paper defines a phenomenon as a crisis if the payment reduction exceeds one standard deviation below the mean change or the negative deviation of payments from the average payment over the previous five years is larger than one standard error.
7 The paper empirically confirms the promoting effects of RMB swap lines on bilateral trade, which are measured by the ratio of bilateral trade to the total trade amount of a recipient and the ratio of the recipient’s imports from China to its total imports.