Table 6

Campbell–thompson (C-T) test results for in-sample and out-of-sample forecasts: ADL-MIDAS with green innovation versus AR benchmark

ADL-MIDAS (PCR) vs ARADL-MIDAS (TCR) vs AR
In-sampleOut-of-SampleIn-sampleOut-of-sample
h = 4h = 8h = 12h = 4h = 8h = 12
Communication service0.66780.71450.73370.82480.62240.67350.69520.7995
Consumer discretionary0.52320.52050.46930.49020.56990.55120.49260.5125
Consumer staple0.62910.60290.57900.64160.63180.58930.52500.5956
Energy0.03660.04020.14770.11280.08810.08900.19000.1569
Financial0.08520.07730.10910.12160.09320.09840.10500.1176
Health care0.73360.74190.75370.75900.74870.74620.75690.7621
Industrial0.50610.49240.38740.38290.55840.53850.43270.4285
Information technology0.77590.77540.75040.79010.77490.77560.75890.7972
Material0.42170.41850.40200.52880.47730.46640.42910.5501
Real estate0.99030.99110.99200.99240.99010.99020.99130.9918
Utility−2.6369−2.1915−2.2121−2.1011−1.7606−1.4225−1.6233−1.3344

Note(s): The C–T test results are based on the forecast performance comparison of the unrestricted ADL-MIDAS-based predictive model as against the AR process. Hypothetically, a positive C-T statistic or value implies that the unrestricted ADL-MIDAS based predictive model outperforms the AR process and the reverse holds if the statistic is negative

or Create an Account

Close Modal
Close Modal