資本帳自由化與經濟成長—台灣的案例研究

Abstract

1980年代中起資本帳自由化風潮對開發中國家造成重大影響，依據新古典經濟成長理論，開發中國家資本帳自由化有助提升每人產出及消費，然而，實證研究結果甚為分歧。事實上，許多開發中國家在開放資本帳後發生各式各樣的金融危機。台灣在這一波全球資本帳自由化浪潮中並沒有缺席，但台灣開啟資本帳後的經濟表現卻多與理論預期相左，特別是經濟成長率呈現趨勢向下現象，問題徵點何在，仍有待推敲。 本文首先討論全球資本帳自由化的遞嬗，並論述台灣資本帳自由化的發展歷程，試圖從中了解兩者的連結。我們可以發現，台灣資本帳自由化固然與來自美國的壓力及國際組織的倡議脫離不了關係，但來自內部的力量亦不可輕忽。其次，台灣資本帳自由化採取漸進主義，亦即在經常帳持續出現巨額順差後才啟動，並循序開展。第三，台灣資本帳自由化後資本並未如所預期的流入，反而呈現流出現象，此或與之前的資本管制及長期國內超額儲蓄並存有關。 第二章建立一個跨期模型，據以探討資本移動對總體經濟的影響，並以台灣的資料作校準分析，得到三項重要的結論：首先，資本帳自由化對每人所得及消費的效果視金融市場的效率而定，這一結果呼應經濟自由化過程不應忽視自由化順序之論點。其次，資本流出時若國內利率上升幅度小於資本流出的變化幅度，將不致降低每人產出及消費。第三，產出之實體資本及人力資本彈性越大，則調整速度越慢，這一結果印證發展經濟學家所倡議之收斂理論。 第三章透過事件分析法探討「事實的」資本帳自由化之總體經濟效果，發現台灣在1980年代中解除外匯管制後，經歷相當長時間的負向產出缺口。其次，除了1990年代外，其他時段的投資成長率均下降，但波動度卻見提高；另1990年代初起開放外國人投資國內股市，卻未對股價產生推升效果；而無論是股票殖利率或Tobin q值，幾乎都指向資本帳自由化對降低資本成本並無助益。第三，台灣資本帳自由化並未使得產出和其組成項目以及各項目彼此的相關趨淡。 第四章深入探討台灣資本帳自由化後，資本的流動及對總體經濟變數的影響、經濟榮枯循環、以及台灣與主要貿易夥伴及對手國家經濟波動關係之變化。發現早在1980年代中解除外匯管制之前，台灣實際上是不缺乏投資資金的，1990年代初開放外資投資國內證券市場後，甚至可觀察到資金的淨流出。其次，台灣開放資本帳後資本自發性移動的跡象逐漸明顯，但資本帳衝擊下，除了外匯準備及貨幣供給M2對資本帳衝擊的反應符合預期外，其他變數在多數時段的反應並不明顯。第三，台灣資本帳自由化並未導致景氣榮枯循環，因其從未引發資本的大量移入。最後但卻是最重要的，過去20年台灣與美、日GDP波動相關係數呈現下降，與中國大陸GDP波動相關係數則明顯提高。

The widespread capital account liberalization (CAL henceforth) has shed a strong influence on most developing countries since mid-80s. According to neoclassical growth theory, CAL should help lift per capita output and consumption in developing countries. However, empirical studies fail to give a definite result. In fact, many developing countries have encountered various kinds of financial crisis after opening up capital account. Taiwan has not been absent in the CAL movement, the economic performance after CAL were in general at odds with what the theory tells us. Specifically, economic growth rate has followed a downward trend ever since. The reasons why it happened have yet to be scrutinized. In this dissertation I started with a discussion of the evolution of global CAL and the development of CAL in Taiwan, aiming to find a link between them. I found that in addition to the pressures from the US and the initiatives of international organizations, endogenous forces should not be ruled out in understanding the underpinning factors in Taiwan’s CAL. Secondly, Taiwan adopted a gradualism approach in opening up its capital account namely not before it had enjoyed a huge amount of current account surplus and moved forward step by step. Thirdly, Taiwan experienced capital outflow rather than inflow after CAL. This should be due to the coexistence of capital control and excess domestic savings before CAL. An intertemporal model is built in chapter two to derive the effects of CAL on macroeconomic variables. A model calibration exercise has been performed by applying Taiwan’s data. Several findings are concluded as the following: Firstly, the effect of CAL on per capita income and consumption lies heavily on the efficiency of its financial market. This result echoes the proposition that the issue of optimal sequencing should not be left aside in the process of economic liberalization. Secondly, per capita income and consumption may not decline in the case of capital outflow as long as the magnitude of increase in domestic interest rate is lower than that of capital outflow. Thirdly, it takes time for the effects of liberalization to be fully materialized. Lastly, a higher output elasticity of capital implies a lower rate of adjustment. This result lends support to the convergence theory initiated by most development economists. In chapter three I investigate the macroeconomic effects of CAL by performing event study of “de facto” liberalization. I found that Taiwan has in general experienced a long period of negative output gap ever since it lifted foreign exchange controls in mid-80s. Secondly, the rate of investment growth has plummeted except for 1990s with a higher volatility. Furthermore, the opening up of security market beginning from early 90s proved to have no effect on lifting stock prices, and either the dividend yield or Tobin’s q has pointed to the failure of CAL in lowering the cost of capital. Thirdly, CAL has no effect on brining down correlation of output with and among its components. In chapter four I probed into the effects of capital flows on macroeconomic variables, and investigate the so called boom-bust cycle and the co-movement phenomena among Taiwan and her major trading partners or competitors. It was found that Taiwan did not have any problem in the supply of investment funds prior to it lifting foreign exchange controls back in mid-80s. We can even observe capital outflow in early 1990s when it opened up its security market. However, autonomous capital flows became evident as a result of CAL. The impulse responses of macroeconomic variables to autonomous capital flows were not significant though except for foreign exchange reserves and money supply M2. Thirdly, I found that CAL did not result in boom-bust cycle because it had not brought in massive capital inflows. Last but not least, the correlation of GDP between Taiwan and the US or Japan has been declining while increasing with China in the past two decades.