Table 4.

‘Heat wave’ and the ‘meteor shower’ effect from futures to spot markets

Explained variables
Gold spot		Silver spot		Platinum spot		Palladium spot
Panel A. Estimated regime switching parameters
α_1_gold(AR)	0.655***	α_1_silver(AR)	0.527***	α_1_platinum(AR)	0.653***	α_1_palladium(AR)	0.560***
α_2_gold(AR)	0.914***	α_2_silver(AR)	0.877***	α_2_platinum(AR)	0.774***	α_2_palladium(AR)	0.841***
β_1_gold_F	0.340***	β_1_silver_F	0.531***	β_1_platinum_F	0.244***	β_1_palladium_F	0.457***
β_2_gold_F	0.052***	β_2_silver_F	0.112***	β_2_platinum_F	0.227***	β_2_palladium_F	0.068***
σ21	−4.63***	σ21	−2.04***	σ21	−3.06***	σ21	−3.40***
σ22	−2.52***	σ22	−1.96***	σ22	−1.05***	σ22	−1.12***
Panel B. Regime properties
P11	0.46	P11	0.81	P11	0.75	P11	0.35
P22	0.81	P22	0.95	P22	0.95	P22	0.75
ED 11	1.8	ED 11	5.4	ED 11	4.0	ED 11	1.5
ED 22	5.3	ED 22	18.3	ED 22	18.5	ED 22	3.9

P11 and P22 are regime switching probabilities, while acronym ED denotes expected duration of each regime. ${\sigma }_1^2\mathrm{and}{\sigma }_2^2$ are regime-specific error variances ***, **, * represent statistical significance at the 1%, 5% and 10% level, respectively