Table 4.
Out-of-sample forecasting comparison criteria (1-day ahead forward prediction)
| Criteria | MAE × 102 | |||||
|---|---|---|---|---|---|---|
| Period | Pre-COVID-19 | COVID-19 | ||||
| Model | LMM | GAM | Tv-LMM | LMM | GAM | Tv-LMM |
| BPI | 0.758 | 0.834 | 0.429 | 0.981 | 0.956 | 0.613 |
| ETH | 1.560 | 1.584 | 0.891 | 0.684 | 0.657 | 0.486 |
| XRP | 1.472 | 1.409 | 0.959 | 0.859 | 0.855 | 0.731 |
| BCH | 1.232 | 1.450 | 0.673 | 1.360 | 1.361 | 1.188 |
| BNB | 1.502 | 1.415 | 1.219 | 1.459 | 1.470 | 1.385 |
| LINK | 3.995 | 3.816 | 3.442 | 1.802 | 1.796 | 1.438 |
| CRO | 1.838 | 1.802 | 1.640 | 2.899 | 2.899 | 2.778 |
| LTC | 1.133 | 1.189 | 0.690 | 1.404 | 1.398 | 1.099 |
| BSV | 3.423 | 2.740 | 1.842 | 1.106 | 1.103 | 0.765 |
| ADA | 1.327 | 1.388 | 0.817 | 1.201 | 1.201 | 1.143 |
| Average | 1.824 | 1.763 | 1.260 | 1.376 | 1.370 | 1.163 |
| Criteria | MSE × 104 | |||||
|---|---|---|---|---|---|---|
| Period | Pre-COVID-19 | COVID-19 | ||||
| Model | LMM | GAM | Tv-LMM | LMM | GAM | Tv-LMM |
| BPI | 0.983 | 1.072 | 0.344 | 1.745 | 1.627 | 0.755 |
| ETH | 4.787 | 4.626 | 1.524 | 0.730 | 0.718 | 0.411 |
| XRP | 4.490 | 4.175 | 2.698 | 1.413 | 1.410 | 1.078 |
| BCH | 2.522 | 3.221 | 0.899 | 4.074 | 3.867 | 3.123 |
| BNB | 3.971 | 3.761 | 3.161 | 3.994 | 3.965 | 3.626 |
| LINK | 25.729 | 23.759 | 18.416 | 5.934 | 5.935 | 3.407 |
| CRO | 6.436 | 6.068 | 4.981 | 19.099 | 19.099 | 17.194 |
| LTC | 2.851 | 2.828 | 1.131 | 3.862 | 3.816 | 2.049 |
| BSV | 19.631 | 14.336 | 8.556 | 2.423 | 2.421 | 1.273 |
| ADA | 2.794 | 2.830 | 1.123 | 2.679 | 2.679 | 2.364 |
| Average | 7.419 | 6.668 | 4.283 | 4.595 | 4.554 | 3.528 |
Bold values indicate the best predictability performance for cryptocurrency price in terms of the lowest MAE and MSE, respectively