以往對於外人直接投資之研究,多以少數國家或地區為樣本為研究對象,探討經濟因素的影響,但對於直接投資為何在某些投資與被投資國組合中較為集中的現象與成因,則尚未加以探討。本研究參考行為財務之熟悉度偏誤觀點,建構「外人直接投資偏誤指標」並納入政治因素、資訊成本及文化距離等因素,以1993-2003年間占全球投資金額81%前16大投資國與占全球73%之31大被投資國數據,利用縱橫資料迴歸模式透過跨國比較加以實證。研究結果發現,直接投資偏誤除了傳統經濟因素如勞動成本及貿易等因素影響外,亦受到代表資訊成本之地理距離及語言,以及文化距離等因素之影響,而政治上之自由貿易協定因素則並不顯著。其次,受到不同變數納入的影響,也使各投資國之固定效果產生差異。本研究成果除可對現有直接投資研究成果加以補充外,由於本研究之樣本範圍遠大於以往相關研究,也有效提高本項實證結果的可信度和參考價值。
Prior research that analyzed determinants of Foreign Direct Investment (FDI) mostly used economic factors as explanatory variables, but FDI which was centralized or dispersed couldn’t be explained by these factors. Furthermore, samples were from few countries or regions. This research designed the “Foreign Direct Investment Bias (FDIB)” index by the Behavioral Finance view “Home Bias”, used the Panel Data Regression Model, employed FTA, geographical distances, language and culture distance as the explanatory variables, and expanded sample countries/regions with data from the top 16 investing countries responsible for 81% of global investing and the top 31 host countries receiving investing 73% of total global investments during the period 1993-2003. Adopting the perspective of international comparison, firstly, this research found that countries’ FDIB mainly considered the information cost of geographical distances, language, and culture distance variables rather than FTA in the full model. The fixed effect of FDIB was also mixed in different models. The results support the direct investment theory and can influence future studies. Secondly, the impacts of the small or large models are different cause the role of destination countries. By the same reason, the FDI impact factor considerations are not the same as portfolio investment.