Should gold be stored in chaotic eras?

Figures & data

Figure 1. The trends of G.P.R. and G.P.

Notes: There are thousands of geopolitical events from 1979:M1–2018:M12 in the world, due to space limitations, we only label the events with high G.P.R., which have large intensity and their influences are huge. Source: Authors’ Calculations.

Table 1. Descriptive statistics for G.P.R. and G.P.

	GPR	GP
Observations	480	480
Mean	96.505	632.386
Median	83.602	405.970
Maximum	455.022	1771.850
Minimum	23.314	227.270
Standard Deviation	55.137	418.448
Skewness	2.077	1.163
Kurtosis	10.046	2.922
Jarque-Bera	1337.810***	108.299***

Note: ***Denotes significance at the 1% level.

Source: Authors’ Calculations.

Table 2. Full-sample Granger causality tests.

Tests	H0: GPR does not Granger cause GP		H0: GP does not Granger cause GPR
	Statistics	p-value	Statistics	p-value
Bootstrap LR test	1.700	0.439	0.921	0.613

Notes: We calculate p-value using 10,000 bootstrap repetitions.

Source: Authors’ Calculations.

Table 3. The results of parameter stability test.

Tests	GPR		GP		VAR system
	Statistics	p-value	Statistics	p-value	Statistics	p-value
Sup-F	39.697***	0.000	13.004*	0.071	26.774***	0.004
Ave-F	12.391***	0.000	5.116	0.101	13.133***	0.008
Exp-F	15.029***	0.000	4.177*	0.055	9.673***	0.005
Lc					3.285*	0.059

Notes: We calculate p-value using 10,000 bootstrap repetitions. *** and * denote significance at the 1% and 10% levels, respectively.

Source: Authors’ Calculations.

Figure 2. Bootstrap p-values of rolling test statistic testing the null hypothesis that G.P.R. does not Granger cause G.P. Source: Authors’ Calculations.

Figure 3. Bootstrap estimates of the sum of the rolling-window coefficients for the impact of G.P.R. on G.P. Source: Authors’ Calculations.

Figure 4. Bootstrap p-values of rolling test statistic testing the null hypothesis that G.P. does not Granger cause G.P.R. Source: Authors’ Calculations.

Figure 5. Bootstrap estimates of the sum of the rolling-window coefficients for the impact of G.P. on G.P.R. Source: Authors’ Calculations.