Table 4:

Percentage of Accurately Classified Time-Varying and Time-Invariant Parameters Based on Smoothed Estimates

	Time Series Length
	T = 70						T = 200						T = 500
	Simulation Condition						Simulation Condition						Simulation Condition
	1			2			1			2			1			2
Parameter	A	B	C	A	B	C	A	B	C	A	B	C	A	B	C	A	B	C
${\text{Φ}}_{12}$	100a	100a	100a	-	85	87	100a	100a	100a	-	93	94	100a	100a	100a	-	99	99
${\text{Φ}}_{21}$	100a	100a	100a	-	80	83	100a	100a	100a	-	93	96	100a	100a	100a	-	99	100
${\text{Φ}}_{11}$	-	80	78	-	-	87	-	86	87	-	-	96	-	88	93	-	-	99
${\text{Φ}}_{22}$	-	85	87	-	-	85		81	86	-	-	94	-	80	92	-	-	99
${\text{Γ}}_1$	-	-	17	100a	100a	100a	-	-	15	100a	100a	100a	-	-	15	100a	100a	100a
${\text{Γ}}_2$	-	-	19	100a	100a	100a	-	-	20	100a	100a	100a	-	-	16	100a	100a	100a

^aParameter generated as time-varying. All other parameters were generated as time-invariant.

Note. Cells containing a single - were generated and estimated as time-invariant and for this reason misclassification was not possible.