Capital account liberalization and economic performance: Evidence from Korea

AbstractThis study examines whether capital account liberalization is beneficial for inducing foreign investment and positively affects allocative efficiency and economic performance in Korea. Using panel data on Korean industries and firms from 1991 to 2019, we confirm that capital account liberalization is positively associated with foreign investment. This effect is more pronounced for industries with high external financing dependence. Additionally, we conduct a firm-level analysis and find that capital account liberalization improves capital allocative efficiency, indicating that it helps firms with high growth opportunities to make more investment. Furthermore, changes in allocative efficiency attributable to capital account liberalization positively affect firms’ future financial performance. Overall, our study highlights the importance of capital account liberalization in attracting foreign capital, improving allocative efficiency, and enhancing the economic performance of firms in a transition economy.KEYWORDS: Capital account liberalizationforeign investmentallocative efficiencyeconomic performanceKoreaJEL Classifications: F32F38G15O16 AcknowledgementsWe wish to thank the editor (Lex Zhao) and a reviewer for their guidance and constructive comments. Any remaining errors are ours.Disclosure statementNo potential conflict of interest was reported by the author(s).Notes1 The “flight to quality” refers to the investors’ herding behavior to shift out of risky securities in bear markets (that suffer from financial recessions) and preference for relatively more safe assets such as bonds. It is the synonym of the “flight to safety.”2 The Chinn and Ito (Citation2006) capital account openness index is available on the research page for Menzie Chinn and Hiro Ito: http://web.pdx.edu/~ito/Chinn-Ito_website.htm.3 We use all firms in both KSE and KOSDAQ, including currently delisted firms.4 We use the KSIC code to identify financial firms and remove them from our sample. The KSIC is the abbreviation of “Korean Standard Industry Classification” and is based on the International Standard Industrial Classification (ISIC) adopted by the United Nations. According to the classification, the KSIC code K indicates financial firms.5 Table A1 shows the standard classification of the Korean industries. Among 33 industries, after excluding “Finance and insurance (i.e., KSIC code K),” we use the remaining 32 industries for our firm-level analyses.6 For more information, see Chinn and Ito (2008).7 The detailed definition of FDI in Korea is explained in the Korea Legislation Research Institute (KLRI). See https://elaw.klri.re.kr/eng_service/lawView.do?hseq=44628&lang=ENG.8 We use the annual sales as the output level of firms. Capital stock is proxied by the value of tangible fixed assets deflated using the GDP deflator, and labor input is measured by the number of employees.9 There are several methodologies to estimate production functions using firm-level panel data. First, Olley and Pakes (Citation1996) developed a two-step estimation method and showed how capital and investment could be used as a proxy variable for unobserved time-varying productivity. Ackerberg, Caves, and Frazer (Citation2006) indicated that variable inputs such as labor are also determined by unobserved productivity, and the coefficient on inputs may not be non-parametrically identified in the first-stage estimation. To mitigate the potential problem with the identification of parameters, Wooldridge (Citation2009) developed the generalized method of moments (GMM) setup that allows the first-stage estimation in previous methodologies to include identifying information for parameters on variable inputs.10 The value ranges from zero to one, becoming higher when the industry is closer to a monopolistic market.11 It should be noted that our results are intact when we use the other industry-level financial stability measure instead of the debt ratio, such as the fixed liability ratio (i.e., current liability divided by equity capital) and the borrowing dependency (i.e., borrowings divided by total equity).12 According to the definition, Performance¯i,t=(Performancei,t+1+Performancei,t+2+Perfor−mancei,t+3)/3, where Performance is one of the following performance measures: TFP, AssetTurnover, Salesgrowth, and ROA.13 We also use the normalized Chinn and Ito (Citation2006) index scaled from zero to one instead of the original index. Although coefficient variables are quantitatively different, statistical significances are exactly the same in all columns.