本研究利用GJR-GARCH-M模型,研究期間自2017年01月03日至2021年03月11日,共1,021筆日資料,資料來源為台灣經濟新報(TEJ)。在Phillips-Perron單根檢定中,符合研究假說一,所有研究變數之報酬皆為平穩變數。ARCH實證結果發現,新台幣升值、台股報酬及台灣上市建設公司股價指數日報酬,除R1316上曜外,資料皆符合研究假說三。GJR-GARCH-M估計結果可看出顯著不對稱效果。
This research uses the GJR-GARCH-M estimation model from January 3, 2017 to March 11, 2021, with a total of 1,021 pieces of day information attained from the Taiwan Economic Journal (TEJ). In the Phillips-Perron unit root test, in agreement with Research Hypothesis 1, the returns of all research variables are stationary variables. The ARCH empirical evidence establishes that all data gathered from the studies of the appreciation of the new Taiwan dollar, the returns of Taiwan stocks, and the daily returns of the stock price index of Taiwan's listed construction companies, except for Sun Yad Construction Company(R1316), are in line with Research Hypothesis 3. The GJR-GARCH-M estimation results also reveal the evidence of significant asymmetry effects.