Economics Letters 112 (2011) 60–62 Contents lists available at ScienceDirect Economics Letters journal homepage: www.elsevier.com/locate/ecolet De facto financial openness and capital mobility Javed Younas ∗ Department of Economics, American University of Sharjah, P.O. Box 26666, Sharjah, United Arab Emirates article info Article history: Received 16 July 2010 Received in revised form 12 March 2011 Accepted 23 March 2011 Available online 8 April 2011 JEL classification: F30 F02 Keywords: Feldstein–Horioka puzzle De facto financial openness Dynamic panel model abstract This paper shows that de facto financial openness does (does not) increase capital mobility in developing (developed) countries and that capital is (is not) freely mobile in the most financially open developing (developed) countries. © 2011 Elsevier B.V. All rights reserved. 1. Introduction Feldstein and Horioka (1980), hereafter FH, obtained a corre- lation of domestic savings and domestic investment close to one in their analysis of sixteen OECD countries which they interpreted as evidence of capital immobility. 1 This finding led to the contro- versial conclusion of the existence of significant home bias in the allocation of domestic savings. Subsequent analysis also finds this correlation to be high for OECD countries, but less than what was observed in the 1960s and 1970s (Feldstein and Bacchetta, 1991; Tesar, 1991). In light of the increased integration of world capital markets, evidence of capital immobility among developed coun- tries came to be known as the FH puzzle (e.g., Obstfeld and Rogoff, 2000). It is not difficult to concede that there will inevitably be some home bias due to barriers to entry, information cost, and attitude toward risks in investing in international portfolios. Contrarily, financial reforms and capital account openness in a country would weaken this correlation as domestic savings flow to countries with the highest marginal returns. A few recent studies have captured this phenomenon by employing either de jure ‘‘perceived’’ measures of financial openness or a time trend in ∗ Tel.: +971 6 515 2612; fax: +971 6 558 5065. E-mail address: jyounas@aus.edu. 1 In the FH sense, a higher estimated value of the coefficient of savings rate indicates low capital mobility, while its lower value reflects high capital mobility in a country. an augmented FH equation (e.g., Isaksson, 2001; Georgopoulos and Hejazi, 2005; Younas and Chakraborty, 2010; Younas and Nandwa, 2010). These studies attribute their findings of increased capital mobility to the financial markets’ liberalization in the world. I argue that both de jure measures as well as a time trend are weak indicators of capital account openness and financial reforms in a country. The former are indices based on the extent of restrictions on capital account transactions across borders. Their obvious drawback is that they can at best be described as the ‘‘intention’’ to restrict capital flows which, by itself, does not accurately reflect an economy’s actual level of integration in the world financial market. The problem with the latter is that this may also capture a host of other factors such as business cycle effects and changes in policy regimes in a country. Consequently, their use in the FH equation would result in an upward bias in the savings retention coefficient, thus implying lower capital mobility than is actually the case. 2 This paper contributes to the empirical FH literature in a significant way by incorporating two major improvements over the existing research: First, this is the first study that uses data on de facto financial openness to measure its impact on the degree of capital mobility. Since the purpose of this study is to examine the impact on capital mobility of actual restrictions on domestic 2 The coefficient on the savings rate is also called the saving retention coefficient. It shows the extent to which an increase in domestic savings is used to finance domestic investment or the fraction of one dollar of savings that is invested domestically (Georgopoulos and Hejazi, 2005). 0165-1765/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.econlet.2011.03.028