Regime-dependent drivers of the EUR/CHF exchange rate

Swiss Journal of Economics and Statistics

Table 5 OLS regression results for the peg period (October 2011–December 2014)

	Peg period (October 2011–December 2014)
	Dependent variable: Latent EUR/CHF
	(1)	(2)	(3)	(4)	(5)
Constant	− 0.101	− 0.101	− 0.101	− 0.101	− 0.101
	(0.835)	(0.823)	(0.815)	(0.816)	(0.810)
CHFEUR1m	− 0.308	− 0.582**	− 0.571**	− 0.465	− 0.468
	(0.133)	(0.024)	(0.022)	(0.103)	(0.123)
		VIF = 1.120	VIF = 1.127	VIF = 1.179	VIF = 1.179
CHFEUR10y2y				0.480*	0.181
				(0.065)	(0.558)
				VIF = 1.084	VIF = 1.279
CHFEUR30y			− 0.162	− 0.261	− 0.252
			(0.752)	(0.588)	(0.655)
			VIF = 1.437	VIF = 1.484	VIF = 1.484
DAX		0.840	0.748*	0.708*	1.252**
		(0.118)	(0.079)	(0.083)	(0.033)
		VIF = 1.120	VIF = 1.588	VIF = 1.595	VIF = 2.238
VSMI					0.934
					(0.148)
					VIF = 1.893
Observations	38	38	38	38	38
Mallows’s Cp	2.53	1.42	3.33	4.28	4.01
Selected model		X
\(R^{2}\)	0.013	0.097	0.099	0.128	0.189
Adjusted \(R^{2}\)	− 0.015	0.045	0.020	0.022	0.062

In this table, we show results from regressing the latent EUR/CHF exchange rate from Hanke et al. (2019) onto an increasing set \(\mathcal{J}_0\) of variables that are determined by the stepwise forward selection method described in Sect. 2.1. p values (in parentheses) and corresponding significance stars are based on Newey–West adjusted standard errors with 6 lags, where the lag order was determined based on a full-sample analysis. Variance inflation factors are shown underneath each variable to indicate potential multicollinearity. The selected model in column (2) minimizes Mallows’s \(C_p\)
*\(p<0.1\); **\(p<0.05\); ***\(p<0.01\)