Table 4.
Results of MCS test.
| Forecasting models | QLIKE |
MSE |
||
|---|---|---|---|---|
| Range | SeimQ | Range | SeimQ | |
| Panel A: h = 1 | ||||
| MIDAS-RV | 0.0748 | 0.0517 | 0.8417 | 0.8892 |
| MIDAS-RV-OVX | 1.0000 | 1.0000 | 1.0000 | 1.0000 |
| MS-MIDAS-RV | 0.0748 | 0.0289 | 0.4609 | 0.6597 |
| MS-MIDAS-RV-OVX | 0.3092 | 0.3092 | 0.8417 | 0.8892 |
| Panel B: h = 5 | ||||
| MIDAS-RV | 0.0954 | 0.0449 | 0.6194 | 0.4028 |
| MIDAS-RV-OVX | 0.0954 | 0.0449 | 0.6194 | 0.4028 |
| MS-MIDAS-RV | 0.0954 | 0.0449 | 0.6194 | 0.4028 |
| MS-MIDAS-RV-OVX | 1.0000 | 1.0000 | 1.0000 | 1.0000 |
| Panel C: h = 20 | ||||
| MIDAS-RV | 0.0008 | 0.0073 | 0.0028 | 0.0184 |
| MIDAS-RV-OVX | 0.0053 | 0.0167 | 0.0094 | 0.0445 |
| MS-MIDAS-RV | 0.6794 | 0.6794 | 1.0000 | 1.0000 |
| MS-MIDAS-RV-OVX | 1.0000 | 1.0000 | 0.9449 | 0.9449 |
Notes: This table shows the results of MCS test, and the p-value is calculated according to the range and semi-quadratic (SeimQ) statistics. Two criteria, i.e., QLIKE and MSE are adapted. The p-values > 0. 25 and p-values = 1 are overstriking and underlined. Different horizons h = 1, 5 and 20 are shown in Panel A, B, C. In this table, the MCS test p-value of 1indicates a better prediction.