Table 5 Out-of-sample forecast...

Table 5

Out-of-sample forecast evaluation [h = 120]

Clark and West test
	Full sample		Pre-COVID		COVID^#
	Model 1 vs Model 2	Model 1 vs Model 3	Model 1 vs Model 2	Model 1 vs Model 3	Model 1 vs Model 2	Model 1 vs Model 3
Consumer goods	0.0892^a [4.23]	0.2764^a [5.04]	0.1007^a [4.50]	0.3975^a [5.16]	2.4765^a [4.04]	5.7999^a [4.95]
Consumer Services	0.0341^a [6.67]	0.5057^a [19.87]	0.0444^a [8.32]	0.5015^a [16.79]	2.5197^a [6.69]	9.3796^a [9.78]
Financials	0.0376^a [7.13]	0.3604^a [21.56]	0.0427^a [6.73]	0.3505^a [16.97]	3.8129^a [6.49]	10.6388^a [11.82]
HealthCare	0.0546^a [2.66]	0.1028^c [1.63]	0.0593^a [2.46]	0.1537^a [2.08]	1.4538^a [2.73]	3.8291^a [5.48]
Industrials	0.0305^a [5.27]	0.3098^a [9.45]	0.0393^a [5.99]	0.3625^a [9.73]	7.3485^a [2.03]	12.7912^a [2.40]
Materials	0.0551^a [5.89]	0.0966^a [2.17]	0.0707^a [6.27]	0.0860^c [1.57]	7.8840^a [6.17]	15.2597^a [7.12]
Technology	0.1042^a [3.22]	0.1720^a [3.32]	0.1405^a [3.57]	0.2089^a [2.73]	1.8555^b [1.80]	4.2805^a [2.12]
Telecoms	0.0537^a [3.64]	0.4602^a [3.32]	0.0397^a [3.14]	0.3429^b [1.97]	3.9582^a [3.73]	9.2347^a [4.84]

Clark and West test
	Full sample		Pre-COVID		COVID^#
	Model 1 vs Model 2	Model 1 vs Model 3	Model 1 vs Model 2	Model 1 vs Model 3	Model 1 vs Model 2	Model 1 vs Model 3
Consumer goods	0.0892^a [4.23]	0.2764^a [5.04]	0.1007^a [4.50]	0.3975^a [5.16]	2.4765^a [4.04]	5.7999^a [4.95]
Consumer Services	0.0341^a [6.67]	0.5057^a [19.87]	0.0444^a [8.32]	0.5015^a [16.79]	2.5197^a [6.69]	9.3796^a [9.78]
Financials	0.0376^a [7.13]	0.3604^a [21.56]	0.0427^a [6.73]	0.3505^a [16.97]	3.8129^a [6.49]	10.6388^a [11.82]
HealthCare	0.0546^a [2.66]	0.1028^c [1.63]	0.0593^a [2.46]	0.1537^a [2.08]	1.4538^a [2.73]	3.8291^a [5.48]
Industrials	0.0305^a [5.27]	0.3098^a [9.45]	0.0393^a [5.99]	0.3625^a [9.73]	7.3485^a [2.03]	12.7912^a [2.40]
Materials	0.0551^a [5.89]	0.0966^a [2.17]	0.0707^a [6.27]	0.0860^c [1.57]	7.8840^a [6.17]	15.2597^a [7.12]
Technology	0.1042^a [3.22]	0.1720^a [3.32]	0.1405^a [3.57]	0.2089^a [2.73]	1.8555^b [1.80]	4.2805^a [2.12]
Telecoms	0.0537^a [3.64]	0.4602^a [3.32]	0.0397^a [3.14]	0.3429^b [1.97]	3.9582^a [3.73]	9.2347^a [4.84]

Note(s): Model 1 is the Historical Average model; Model 2 is the model without control; Model 3 is the model with control. The Clark and West test measures the significance of the difference the forecast errors of two competing models. The null hypothesis of a zero coefficient is rejected if this statistic is greater than +1.282 (for a one sided 0.10 test), +1.645 (for a one sided 0.05 test) and +2.00 for 0.01 test (for a one sided 0.01 test) (see Clark and West, 2007). Values in square brackets – [] are for t-statistics. a, b and c indicate statistical significance at 1%, 5 and 10% levels respectively. #Due to the data scope for COVID, we use 10 and 20 days for out-of-sample forecast evaluation

[ViewLarge]

Sharing Unavailable