Table 4.
Panel GMM estimation results for changes on stock market VIX in case of Europe vs. Asia and whole sample.
| Dependent Variable |
VIX changes |
|||||
|---|---|---|---|---|---|---|
| Independent Variables | Exp. Sign | System GMM | Independent Variables | Exp. Sign | System GMM | |
| Model A |
All countries |
Model B |
j = EUR |
j = Asia |
||
| VIX changes(t-1) | +/− | −0.2279 | VIX changes(t-1) | +/− | −0.4002 | −0.1250 |
| (0.00)*** | (0.00)*** | (0.32) | ||||
| C | +/− | 0.0056 | C | +/− | 0.0024 | 0.0060 |
| (0.00)*** | (0.58) | (0.04)** | ||||
| GTR changes(t) | + | 0.0531 | Dj x GTR changes(t) | + | 0.1355 | 0.1648 |
| (0.01)** | (0.04)** | (0.30) | ||||
| (1-Dj) x GTR changes(t) | + | 0.1693 | 0.0605 | |||
| (0.17) | (0.04)** | |||||
| Stock changes(t) | – | −1.6939 | Dj x Stock changes(t) | – | −0.988 | −2.5684 |
| (0.00)*** | (0.09)* | (0.00)*** | ||||
| (1-Dj) x Stock changes(t) | – | −1.5651 | −1.6834 | |||
| (0.04)** | (0.00)*** | |||||
| Stock changes(t) x GTR changes(t-1) | – | −5.8253 | Dj x Stock changes(t) x GTR changes(t-1) | – | −3.1651 | −2.4563 |
| (0.09)* | (0.04)** | (0.47) | ||||
| (1-Dj) x Stock changes(t) x GTR changes(t-1) | – | −2.48617 | −6.6170 | |||
| (0.32) |
(0.04)** |
|||||
| Number of Observations N =(ixT) | 767 | Number of Observations N =(ixT) | 767 | 767 | ||
| Arellano-Bond test AR(1) (p-value) | (0.04)** | Arellano-Bond test AR(1) (p-value) | (0.01)** | (0.03)** | ||
| Arellano-Bond test AR(2) (p-value) | (0.55) | Arellano-Bond test AR(2) (p-value) | (0.93) | (0.24) | ||
| Sargan test (p-value) | (0.99) | Sargan test (p-value) | (0.99) | (0.99) | ||
Notes: The left hand side of the table (Model A) presents the Panel GMM estimation results for VIX changes regressed on stock market changes, google trends indicator (GTR) contemporaneously (t), and on cross terms of stock changes with GTRt-1 measurement for the period from 02 January 2020 to 09 April 2020 across all countries studied. The column Exp. Sign refers to expected signs of the coefficient of the relevant explanatory variable. The right hand side of the table (Model B) reports the Panel GMM estimation results for changes on stock market VIX in Europe vs. Asia, regressed on cross terms of the dummy variables Dj and (1-Dj) with i) google trends indicator (GTR) contemporaneously (t), ii) stock changes contemporaneously (t) and finally iii) stock changes and google trends indicator (GTR) with a lag (t-1). By using dummy variables Dj, (1-Dj) we decompose the sample between j geographic area and the rest. The j takes the value of either EUR for European markets studied or Asia for Asian markets. The lower part of the table presents results for Arrelano-Bond tests concerning the hypothesis that the error term is not serially correlated and the results for Sargan test examining the overall validity of the instruments. Finally, p-values are in parentheses. *,** and ***indicate statistical significance at the 10%, 5%, and 1% level respectively.