預測五國航運(運輸)指數之研究 -從產業面及總體經濟面探討

Forecasting Transportation Index in Five Stock Markets -from Industrial and Macroeconomic Perspective

葉詩妍(shy-yan yeh)

指導教授 ： 胡為善

中原大學/商學院/企業管理研究所/碩士(2009年)

https://doi.org/10.6840/cycu200900910

摘要

全球化對經貿影響更加快速和廣泛，台灣是海島型國家，因原物料和天然資源的缺乏，係透過貿易來取得所需的大宗貨物，例如鐵礦砂、煤碳、穀物等，故散裝航運與我國經濟發展息息相關。當總體經濟發展時，對原物料有較大的需求，會增加整個航運的運輸量，進而影響各國航運(運輸)指數。本研究將從BDI指數、產業(全球鋼鐵指數、各國工業指數、各國航運指數)及總體經濟面(大盤指數和消費者物價指數)來探討影響台灣、中國、美國、德國及挪威航運(運輸)指數的因素並比較EGARCH模型和倒傳遞類神經網路之預測績效。經由實證結果歸納以下三點結論：
1. 美國航運指數對台灣航運指數有正向影響；挪威航運指數對美國航運指數有正向影響；台灣航運指數及挪威航運指數對德國運輸指數有正向影響；美國航運指數及台灣航運指數對於挪威航運指數報酬有正向影響。
2. 在五國的航運(運輸)指數中，倒傳遞類神經網路的RMSE、MAE和MAPE值均比EGARCH模型小，表示BPN預測的績效優於EGARCH模型。
3. 前一期的BDI指數對於當期的台灣航運指數報酬和當期的美國航運指數報酬有正向影響存在；然而其對於當期的中國運輸指數報酬、當期的德 國運輸指數報酬及當期的挪威航運指數報酬產生負向影響。

關鍵字

BPN ； EGARCH ； 波羅的海運價指數(BDI) ； 航運指數 ； 散裝航運

並列摘要

For the past two decades, there is a huge and widespread impact of globalization on Taiwan’s economy and trade. Since Taiwan is lack of natural resources, she obtains raw materials like steel, iron, aluminum, corn, etc. via international trade. Consequently, tramp shipping industry is closely linked to the economic development of Taiwan. Along with the expansion of macroeconomy, Taiwan has a high demand on raw materials increaseing the trading volume via sea shipments and affecting marine transportation indices in worldwide stock markets. This study examines the impact factors of five transportation indexes (i.e. Taiwan, China, U.S.A., Germany, and Norway) from Baltic Dry Index (BDI), various industrial perspectives (CRUspi, industrial indexes, marine transportation indices) and macroeconomic perspectives (stock market indexes and consumer price indexes in five countries). This work also attempts to forecast the transportation indexes in these five countries employing EGARCH model and Back-Propagation Neural Network (BPN) and compares the performances obtained from these two models. The empirical results are summarized below:
1. DJ US Marine Transportation Index has a positive impact on Taiwan's Shipping and Transportation Index; Norway's OSE203030 MARINE index also has a positive impact on DJ US Marine Transportation Index; Both Taiwan Shipping and Transportation Index and Norway's OSE203030 MARINE index have positive impacts on Germany's PRIME TRANSP PERF; Moreover, DJ US Marine Transportation Index and Taiwan Shipping and Transportation Index also have positive impacts on Norway's OSE203030 MARINE index.
2. Upon examining performance indexes of RMSE, MAE and MAPE via EGARCH and BPN models, this study finds that BPN outperforms EGARCH on forecasting transportation indices among five countries.
3. This work also finds that the previous BDI index has a positive impact on Taiwan Shipping and Transportation and DJ US Marine Transportation Indexes. However, the previous BDI index has negative impacts on Chinese Transportation, Germany’s PRIME TRANSP PERF and Norway’s OSE203030 MARINE indexes.

並列關鍵字

transportation index ； EGARCH ； Baltic Dry Index (BDI) ； Back-Propagation Neural Network (BPN) ； tramp shipping industry

參考文獻

江信寧 (2007)，「以總體指標作為選股策略之探討~以上市、櫃運輸類股為例」，淡江

徐毓宏 (2003)，「總體經濟指標預測股價指數報酬率之實證研究－基因及類神經網路

陳永順 (2008)，「全球散裝船海運市場分析及展望(上)」，船舶與海運通訊，第53期，

陳永順 (2008)，「全球散裝船海運市場分析及展望(下)」，船舶與海運通訊，第54期，

張瀞之、王志敏 (2008)，「面臨嚴峻挑戰的全球散裝海運市場」，船舶與海運通訊，第