透過您的圖書館登入
IP:3.236.240.48
  • 學位論文

美國費城半導體股市、台灣股市、台積電股價的關聯性之實證研究

An Empirical study of the Relationship Among Philadelphia Semiconductor Index in American, Taiwan Stock Index and TSMC Stock Price

指導教授 : 吳祥華
共同指導教授 : 鍾麗英
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


過去有關美國股市與台灣股市關聯性之研究,所得到的檢定結果多是道瓊工業指數最具影響力,其次才是NASDAQ指數或費城半導體指數,且均為單向關係居多。 但隨著產業的變遷,特別是台灣已經成為世界電子科技業代工重鎮,電子業亦成為台灣主流產業,尤其是半導體業者,在晶圓代工領域,台灣更是獨佔鰲頭,台積電龍頭地位自然成為眾所矚目之投資標的,這幾年來費城半導體指數日趨重要,且在19家重要半導體廠成份股中,特別納入唯一亞洲的台積電。 為探討美國費城半導體股市、台灣股市、台積電股價的關聯性(領先落後關係)分析與之間的動態、交互關係,並尋找可作為投資決策上的參考意見。研究過程首先將報酬率數列進行單根檢定,確認為定態後,再以Granger 因果關係檢定,另一端則以股價進行共整合分析、誤差修正模型來檢測數列間彼此之長短期關係結構。 研究期間自2003.6.24至2008.2.12止,共1111筆日資料,實證研究之結論歸納如下: 1.我們發現費城半導體指數影響台積電股價(晶圓代工產業)的程度比台股來得強烈,即不論在Granger因果關係檢定、共整合檢定及誤差修正模型等均顯著,而費城半導體指數報酬率與台灣加權指報酬率在Granger因果關係檢定亦顯著。 2.在共整合檢定中,則發現費城半導體指數、台灣加權指數與台積電股價三者存在著長期均衡關係,其次在兩兩共整合檢定下,費城半導體指數與台積電股價、台灣加權指數與台積電股價間存在著長期均衡關係,投資者可採用某一數列資料來預測另一數列資料,不過費城半導體指數、台灣加權指數則沒有共整合現象,這可能因半導體產業的摩爾定律現象影響,造成費城半導體指數波動會較明顯。 3.在誤差修正模型下得知費城半導體指數與台積電股價均受到雙方前1期之影響,不僅有短期及長期之關係,且兩序列會有交互影響。然而台灣加權指數與台積電股價中;僅台積電股價前4期影響台灣加權指數外,其餘前1~4期交叉變數來解釋當期變數時均不顯著。

並列摘要


As regarding to the correlative study between Taiwan stock market and American stock market in the past, the inspected result indicated that the first one which owned the great authority was Dow Jones Industrial Average index. The second one was NASDAQ index or PHLX Semiconductor Index; and mostly the correlation between them was single correlation. However, by following the transition of industries, electronics industry turns into main industry in Taiwan, especially Taiwan becomes a place of strategic importance in OEM for worldwide electronic technology industries. In the field of wafer foundry, Taiwan occupies the most important position particularly. The leading position of TSMC naturally becomes the investing target for most people. These years, PHLX Semiconductor Index is getting more and more important. It subsumes TSMC, the only selected ASIA semiconductor factory, to the constituent stocks of nineteen great semiconductor factories. To investigate the correlation ( ahead or fall behind) among PHLX Semiconductor stock market, Taiwan stock market and TSMC, and analyze the dynamic and mutual relation for finding reference opinions of investment policy, the first thing is to carry out Unit Root test for rate of return sequence. After confirming the steady state, the study will use Granger Causal Relations to test it. Next, use the stock to carry out co-integration analysis and error-correction model to test the long-term and short-term relation structure among sequences. The study started from 2003.6.24 to 2008.2.12 with 1111 daily data. The conclusion of the experimental study is generalized as follows: 1.We find that the influence of PHLX Semiconductor Index to TSMC stock is stronger than Taiwan stock index. That is, the result is remarkable no matter in Granger Causal Relations, co-integration test and error-correction model. In addition, the relation between the rate of return for PHLX Semiconductor Index and rate of return for Taiwan Weighted Stock Index also is remarkable by using Granger Causal Relations test. 2.In co-integration test, we find that long-term equilibrium relationship existed among PHLX Semiconductor Index, Taiwan Weighted Stock Index and TSMC Stock. Next, comparing all of them with co-integration test, two by two, PHLX Semiconductor Index & TSMC Stock and Taiwan Weighted Stock Index and TSMC Stock exist long-term equilibrium relationship. Investors can use one of the sequence data to predict another one. However, there is no co-integration between PHLX Semiconductor Index and Taiwan Weighted Stock Index. It might be influence by Moore's Law in semiconductor industries and caused the significant fluctuation of PHLX Semiconductor Index. 3.From the error-correction model, we can know both PHLX Semiconductor Index and TSMC stocks are influenced by previous stage. Not only they have short-term and long-term correlations, but also both sequences will interact for each other. For Taiwan Weighted Stock Index and TSMC Stock, however, only the previous four stages of TSMC Stock would influence Taiwan Weighted Stock Index. When we use cross variable to explain the variable at that stage for the previous one to four stages, we can find that the result is not remarkable.

參考文獻


2. 吳其定 (2007),「滬、港、台、美四地股市指數與區域經濟成長關聯性及共整合之研究─以中、港CEPA 實施前後期為例」,中央大學財務金融研究所碩士論文。
1. Chan, K. C., Cheng, T. W. and Fung, K. W. (2002), “Ownership Restrictions and Stock-price Behavior in China”, The Chinese Economy, Vol. 34, pp.29-48.
2. Dickey, D. A. and Fuller, W. A. (1979), “Distribution of Estimators for Time Series Regressions with a Unit Root”, Journal of the American Statistical Association, Vol.74, pp.427-431.
3. Engle, R. F. and Granger, C. W. J. (1987), “Cointegration and Error Correction:Representation, Estimation and Testing”, Econometrica, Vol. 55, pp.251-276.
4. Eun, C. S. and Shim, S. (1989), “International Transmission of Stock Market Movements”, Journal of Financial and Quantitative Analysis, Vol. 24, pp.241-257.

被引用紀錄


周政光(2012)。情緒指標與股價報酬相關性探討 -以台灣電子類股指數為例〔碩士論文,中原大學〕。華藝線上圖書館。https://doi.org/10.6840%2fcycu201200033
王重尹(2016)。探討中國匯率政策的改變是否對東亞四國及美國具有影響力-以股票及匯率市場進行探討〔碩士論文,國立中正大學〕。華藝線上圖書館。https://www.airitilibrary.com/Article/Detail?DocID=U0033-2110201614055255

延伸閱讀