# Show whether changes in the U.S. stock price cause changes

Show whether changes in the U.S. stock price cause changes in the Chinese stock prices or the other way using the Granger causality test.

## Show whether changes in the U.S. stock price cause changes

Whether changes in the U.S. stock price cause changes in the Chinese stock prices or the other way using the Granger causality test

Software: STATA

Utilize the following regression techniques to estimate the model, and you have to use Stata software
(1) Granger Causality Tests
(2) VAR (Vector Autoregressive Model)
(3) Impulse Response Function (IRF)

More details;

### Granger causality tests

The structures of the causal relationships between variables were analyzed through the Granger causality approach. The Granger causality test is a statistical hypothesis test for determining whether one time series is useful for forecasting another. If probability value is less than any <span id="MathJax-Element-1-Frame" class="MathJax_SVG" style="box-sizing: border-box; margin: 0px; padding: 0px; display: inline-block; font-style: normal; font-weight: normal; line-height: normal; font-size: 20px; text-indent: 0px; text-align: left; text-transform: none; letter-spacing: normal; word-spacing: normal; overflow-wrap: normal; white-space: nowrap; float: none; direction: ltr; max-width: none; max-height: none; min-width: 0px; min-height: 0px; border: 0px; position: relative;" tabindex="0" role="presentation" data-mathml="α”>α level, then the hypothesis would be rejected at that level.

In Table E1, when nine lags are applied, the hypothesis that LNCM does not involve Granger causality of LNTE can be rejected at the 1% level of significance, and the hypothesis that LNPM does not involve Granger causality of LNTE can be rejected at the 5% level of significance. Thus, we found unidirectional causality running from LNCM to LNTE, and we found unidirectional causality running from LNPM to LNTE.

In Table E2, when nine lags were applied, we found unidirectional causality running from LNCX to LNTE at the 1% level of significance and unidirectional causality running from LNPX to LNTE at the 5% level of significance. In Table E3, when nine lags were applied, we found unidirectional causality running from LNCM to LNUE at the 1% level of significance and unidirectional causality running from LNPM to LNUE at the 5% level of significance. When five lags, six lags, and eight lags were applied, we found unidirectional causality running from LNUE to LNCM at the 10% level of significance.

#### In Table E4, when six lags were applied, we found unidirectional causality running from LNPX to LNUE at the 10% level of significance.

When nine lags were applied, we found unidirectional causality running from LNCX to LNUE at the 1% level of significance and unidirectional causality running from LNPX to LNUE at the 5% level of significance.

The results in Table E5 indicate the existence of a unidirectional causality running from LNCM to LNRE when one lag and two lags were applied at the 1% and the 5% levels of significance.

Unidirectional causality running from LNPM to LNRE when one lag and two lags were applied at the 1% and 5% levels of significance was revealed. The results in Table E6 indicate the existence of unidirectional causality running from LNCX to LNRE when one, two, and three lags were applied at the 1%, 5%, and 10% levels of significance. Unidirectional causality running from LNPX to LNRE when one, two, and three lags were applied at the 1%, 5%, and 10% levels of significance was also revealed.