摘要
本論文以外國人海外直接投資(Foreign Direct Investment, FDI)議題為研究基礎：首先，引用引力模型(Gravity Model)，探討海外直接投資行為之基本面因素，並利用差異性斜率隨機係數效果(Random Coefficient Effect Model)探討以中國為中心，其前十大FDI投資國之隨機效果。其次，將縱橫資料平滑轉換模型(Panel Smooth Transition Regression model; PSTR model)應用於引力模型中探討中國大陸前十大投資國的非線性特徵及其門檻效果。最後，延伸PSTR模型為縱橫資料平滑轉換自我迴歸模型探討10個OECD會員國間財政政策與貨幣政策分別對國家吸引海外直接投資的非線性門檻影響結果。 第一篇文章 - The determinants of China’s foreign direct investment inflows: an application of the random coefficient gravity model.旨在探討於中國的基本面因素對國家的吸引海外投資結果中所存在的個別國家差異。實證結果顯示：第一，影響中國吸引FDI流入因素有匯率波動、相對GDP、CPI差、利差以及進口權重距離。第二、隨機係數效果模型顯示影響中國吸引外人直接投資的因素是跨國而不同，中國若想要吸引更多的外人直接投資可以設法訂立不同的獎勵政策來吸引不同國家的投資。例如：香港、韓國及德國可以用貿易開放政策來獎勵。相對有效實質匯率可以吸引大多數的投資國進行投資。而增加日本及台灣匯兌人民幣的匯率波動可以有效的擴大日本及台灣至中國大陸的直接投資。 第二篇文章 - The Differentiated-Nexus Of FDI And Determinants In China. A panel smooth transition regression approach.旨在建立縱橫資料平滑轉換的非線性引力模型，以檢視中國大陸與其前十大外人直接投資國家間的非線性動態影響因素。本文使用縱橫資料平滑轉換模型，並加入匯率波動的落後期當作轉換變數。實證結果發現：第一、中國前十大投資國的決定FDI因素呈現非線性狀況。第二、落後四期的相對匯率波動顯著影響中國FDI的決定因素。第三、自2006年以後，FDI流入的速度是較以往都來的快速，此狀況是無法藉由線性模型所描述識別的。若中國希望促進GDP的成長及有效的增加外匯存底，擴大對目標國家的匯率波動程度將可有效地引入FDI。若區域性風險擴大了匯率波動，則中國的外匯存底將會持續增加。 第三篇文章 - Nonlinear impacts of debt ratio and term spread on inward FDI performance persistence.進一步延伸了縱橫資料平滑轉換模型(Panel Smooth Transition Regression model)為縱橫資料平滑轉換自我迴歸模型(Panel Smooth Transition Auto-regression model)探討以10個OECD會員國中國家績效(以UNDP所編之FDI流入績效指標為研究標的)的非線性持續性效果。自我回歸模型是針對特定的經濟變數進行檢定持續性的一項適合工具，然而，卻無法捕捉到FDI流入績效的非線性特色。PSTR模型是可以考慮橫斷面異質性及非線性狀況的模型。因此，本研究延伸模型而成為縱橫資料平滑轉換自我迴歸模型(PSTAR model)以解決文獻上所提出的自我迴歸模型缺失，並藉以探討FDI的績效及其持續性。實證結果顯示FDI績效持續性是具有非線性特色且跨國、跨時間而變動。國家增加負債比政策會對大部分的國家造成負向的FDI績效持續性，透過政府發行負債的財政政策會較貨幣政策更容易造成持續性短暫波動的影響結果。
並列摘要
This Ph. D. dissertation entitled “Three Essays on Foreign Direct Investment” includes the following three essays: (1) The determinants of China’s foreign direct investment inflows: an application of the random coefficient gravity model. (2) The threshold effect of exchange rate volatility on FDI-determinant nexus – A panel smooth transition regression approach. (3) Nonlinear impacts of debt ratio and term spread on inward FDI performance persistence. The first essay adopts the panel random coefficient model to estimate the fundamentals of FDI inflows on China. Empirical results can be summed up as follows. First, the fundamentals that significantly attract China’s FDI inflows include exchange rate volatility, relative GDP, CPI difference, interest rate differential, and import-weighted distance. Second, the estimated random coefficients show that the impacts of the fundamentals on China’s FDI inflows vary with FDI source countries. Evidently, making differential policy incentives for different FDI source countries is required for China to attract more foreign capital inflows. For example, trade openness policy is beneficial for China to stimulate FDI inflows, especially for Hong Kong, Korea, and German. Relative real effective exchange rate is the key variable that influences the most FDI source countries to enter China for direct investment. In addition, increasing the volatilities of JPY against RMB and TWD against RMB rates can enlarge the direct investment from Japan and Taiwan to China. In the second essay, we study the differentiated nexus of FDI determinants in China for its top ten FDI invested countries from 2000 – 2011 by adopting a panel smooth transition regression (PSTR) model with the lagged exchange rate volatility as the transition variable. Empirical results can be summed up as follows. First, the FDI inflows of China display a nonlinear path. Second, the relative exchange rate volatility four-period lagged significantly and nonlinearly affects the FDI fundamentals. Third, the speed of FDI inflow in period 2006-2010 is higher than that before 2006. However, this result cannot be identified under traditional linear models. If China tried to promote GDP growth and accumulate more foreign exchange reserves, enlarging exchange rate volatility to main target countries could attract FDI inflows effectively. In other words, if regional risks can enlarge the relative exchange rate volatility, then China’s FDI inflows and foreign reserves would increase. The third essay employs a panel data of ten OECD countries in period 1996-2010 to examine the nonlinear impacts of debt ratio and interest rate differential on inward FDI performance. Auto-regressive (AR) model is a proper method to examine the persistence behavior of a specific economic variable; however, this treatment is still unable to capture the probably nonlinear behavior of FDI inflows. PSTR model considers the heterogeneity of cross-sectional units and the nonlinearity of the variable under investigation. Thus, we revise the PSTR model as a PSTAR one to resolve the shortcomings occurred in previous studies and investigate FDI performance and its persistence. Empirical results show that inward FDI performance persistence displays a nonlinear characteristic and varies with time and across countries. Debt issuance has caused a negative inward FDI performance persistence for most sample countries, implying that a transitory disturbance on the persistence and an unstable persistence effect frequently occur. In addition, fiscal policy through debt issuance is more easy to cause a transitorily volatile effect on the persistence (i.e., a negative persistence effect) than monetary policy.
參考文獻
延伸閱讀
- 黃聖文（2007）。投資環境評估模式與吸收海外直接投資關係之研究〔碩士論文，中原大學〕。華藝線上圖書館。https://doi.org/10.6840/cycu200700575
- 古永嘉、孫明德（2011）。「外人直接投資偏誤」之衡量與決定因素－行為財務觀點。商管科技季刊，12(2)，113-144。https://doi.org/10.30083/CMQ.201106.0001
- 石素娟（2006）。海外直接投資型態之決策模型。科技管理學刊，11(2)，127-153。https://doi.org/10.6378/JTM.200606.0127
- 蔡宜臻、邱俊榮（2007）。Foreign Direct Investment and Research and Development。經濟論文，35(1)，53-82。https://doi.org/10.29628/AEP.200703.0003
- (2003). Foreign Direct Investment. In 國家發展委員會. (Ed.), Economic development, r.o.c (taiwan) (pp. 16-17). 國家發展委員會.. https://www.airitilibrary.com/Article/Detail?DocID=P20191031003-200305-201911050013-201911050013-16-17