原油價格波動是影響經濟與股市的重要因素,台灣是以出口為導向的國家,然而原油大多只能仰賴進口,原油價格上漲會使生產成本增加,對股市與經濟產生影響。 本研究期間為2006年1月1日至2015年12月31日的日資料,研究資料包括布蘭特原油價格、波羅的海指數、上市航運公司股價及台灣加權股價指數。研究採用路徑分析、仲介效果、單根檢定、ARMA、ARCH模型以及GARCH模型來探討原油價格對上市航運公司股價、波羅的海指數與台灣加權股價指數之相關分析。 實證結果發現以台灣加權股價指數為仲介變項下,由於估計參數縮小可得知,在航運公司股價為依變數與波羅的海指數、台灣加權股價指數為自變數的情況下呈現部分仲介效果,再由路徑分析可得知原油價格對航運公司股價之直接影響效果為成本面,而台灣加權股價指數為預測經濟之指標,所以由此為仲介變項之間接影響為經濟面,結果顯示萬海與原油為負向相關,裕民、陽明、新興、萬海、長榮皆受原油成本面之影響大於經濟面之影響,波羅的海指數對裕民、陽明、長榮、新興皆為正響影響關係但波羅的海指數與萬海則為負向影響關係。利用建構ARMA模型初步判斷模型之落後期數,並且利用AIC值最小的方式來選取最適當的模型,由於選取之ARMA模型均拒絕H_0之假設故存在異質變異。接著利用ARCH(1)的診斷發現除了裕民與新興之外皆無法拒絕H_0之假設故不存在自我相關與異質變異,而裕民與新興持續透過GARCH模型診斷,最後利用AIC選取最小的方式選取最適當之GARCH模型,裕民最適當模型為GARCH(1.2),新興最適當之模型為GARCH(2.2)。
Taiwan is an export-oriented country, However, most of the crude oil can only rely on imports. Rising crude oil prices will increase production costs, the stock market and economic have a significant impact. This research period from the first day of 2006 through the end of 2015.The data used in the analysis include information on the Brant crude oil price, Listed shipping company stock price, Baltic Dry Index and Taiwan Accumulate Stock Indexes. This paper adopts some of path-analysis, mediator, unit root test, ARMA, ARCH and GARCH to study for the analysis of the relationship between crude oil and Listed shipping company stock price, Baltic Dry Index and Taiwan Accumulate Stock Indexes. The empirical results show that in Taiwan stock index as the mediated variables, Due to the estimated parameter reduction is that, in the shipping company's share price is according to the variables and Baltic Dry Index and Taiwan Stock Index as the independent variable has a partial mediation effect. From the path analysis, it can be found that the direct effect of crude oil on the shipping company's stock price is the cost side, The Taiwan stock index is the index to predict the economy, So it is between the mediated variables affect for the economy, The results show the WANHAI and crude oil negatively correlated. EVER, U-MING, YANG MING,WAN HAI and SINCERE impact of the cost side is greater than the economy. Baltic Dry Index and EVER, U-MING, YANG MING, WAN HAI positive correlated but the WANHAI and Baltic Dry Index negatively correlated. Using the construction of ARMA model to estimate the lag period of the mode, And using the least AIC value to select the most appropriate model, Due to the selection of the ARMA model, the H0 hypothesis is rejected, so existedheterogeneity variation and the ARCH(1) model received H0hypothesisthen no autocorrelation and heterogeneity variation, along with U-MING and SINCERE. But UMING and SINCERE continuously employ the GARCH model. Finally, UMING and SINCERE use AIC to select the appropriate model of the GARCH, which are GARCH(1.2) and GARCH(2.2).