Performance analysis results
| Forecast error (%) | |||
|---|---|---|---|
| Model 1 | Model 2 | Model 3 | |
| Probit models incorporating indirect sentiment measure | |||
| 50% threshold | |||
| Type Aa | 15.25 | 14.28 | 13.86 |
| Type Bb | 09.81 | 09.53 | 08.22 |
| 25% threshold | |||
| Type Aa | 12.45 | 11.85 | 10.63 |
| Type Bb | 06.86 | 06.04 | 07.34 |
| Probit models incorporating direct sentiment measure | |||
| 50% threshold | |||
| Type Aa | 11.48 | 10.75 | 10.45 |
| Type Bb | 03.25 | 03.13 | 04.27 |
| 25% threshold | |||
| Type Aa | 07.45 | 07.18 | 05.84 |
| Type Bb | 03.81 | 04.63 | 04.96 |
| Forecast error (%) | |||
|---|---|---|---|
| Model 1 | Model 2 | Model 3 | |
| Probit models incorporating indirect sentiment measure | |||
| 50% threshold | |||
| Type A | 15.25 | 14.28 | 13.86 |
| Type B | 09.81 | 09.53 | 08.22 |
| 25% threshold | |||
| Type A | 12.45 | 11.85 | 10.63 |
| Type B | 06.86 | 06.04 | 07.34 |
| Probit models incorporating direct sentiment measure | |||
| 50% threshold | |||
| Type A | 11.48 | 10.75 | 10.45 |
| Type B | 03.25 | 03.13 | 04.27 |
| 25% threshold | |||
| Type A | 07.45 | 07.18 | 05.84 |
| Type B | 03.81 | 04.63 | 04.96 |
Note(s):
Probability of having a crisis without any signal emitted
Number of false signals among all signals
Sharing content requires targeting cookies to be enabled. Please update your cookie preferences to use this feature.