Opinion 150 Overseas Development March 2011 Institute

Capital controls in a global economy: in search of a coordinated truce

n the aftermath of the global financial try to the risk of sharp reversals in crisis, capital flows into emerging and flows in the event of a crisis. developing economies have bounced back quickly from their slump in 2008. This has Recent developments Ibeen triggered by prospects of strong output The recent wave of capital inflows to emerging growth, investors regaining their appetite for and developing countries has caused curren- risk, and, in particular, by ‘carry trade’ prac- cies to appreciate sharply. The Brazilian real, for tices1 favoured by the exceptionally low interest example, has appreciated 38% against the US rates in developed countries. dollar since 2009. In response, an increasing Isabella Massa This surge in foreign capital has led to a number of emerging and developing countries renewed focus on capital controls, a policy have imposed or strengthened different forms option to manage large inflows additional to of capital controls (Table 1 overleaf). ‘In a globalised policy, , fiscal In 2010, increased tax on foreign world, decisions policy, intervention investment in fixed-income bonds by 2%, while taken at country level and domestic prudential regulation. The debate Thailand introduced a 15% tax on interest income have a global impact on the causes and effects of global imbalances and capital gains earned by foreign investors. In and associated capital flows lies at the heart of 2008, Colombia increased the unremunerated ... Coordination current G-20 deliberations. reserve requirements on portfolio inflows from between developed 40% to 50%. Quantitative limits or minimum and developing The theory of the ‘six fears’ stay requirements have been deployed by According to the literature (Magud and Reinhart, Indonesia that limited short-term external bor- countries is the only 2006; Ocampo and Palma, 2008; Epstein, rowing to 30% of capital in 2010 and introduced way forward’ 2005), there are six fears that drive countries to a one-month minimum holding period for central adopt capital controls: bank money market certificates. Other countries 1. fear of appreciation: massive and rapid cap- such as Russia are considering the introduction ital inflows may generate upward pressure of capital control measures to prevent further on the real exchange rate, damaging export currency appreciation. competitiveness 2. fear of ‘’: short-term capital inflows Do capital controls work? may cause destabilising distortions and After decades of criticism, capital controls have increase exposure to capital flow reversal regained legitimacy. The International Monetary 3. fear of large inflows that can disrupt the Fund recognises them as a ‘legitimate part of financial system, even if they are not all ‘hot the toolkit to manage capital inflows’ but only money’ as temporary measures and under specific cir- 4. fear of loss of monetary autonomy: the so- cumstances: the economy should be running called ‘trilemma’ of international macroeco- near its potential, the level of reserves should nomics means that a country cannot achieve be adequate, and the exchange rate should simultaneously perfect capital mobility, not be undervalued (Ostry et al., 2010). But do monetary policy autonomy and exchange they work? Governments considering capital Overseas Development Institute rate stability. So, to avoid exchange rate controls should learn from past experience and ODI is the UK’s leading independent appreciation and sustain an independent recognise the following issues. think tank on international develop- monetary policy, a country should give up Time horizon: empirical evidence shows ment and humanitarian issues. full capital mobility that the effectiveness of capital controls tends ODI Opinions are signed pieces by 5. fear of asset bubbles: large inflows of -for to diminish over time, with markets likely to out- ODI researchers on current develop- eign capital may feed unsustainable asset smart any type of control in the long-run (Carvalho ment and humanitarian topics. price bubbles and Garcia, 2008). So governments should resort This and other ODI Opinions are avail- 6. fear of capital flight: herding behaviour by to capital control measures only temporarily. able from www.odi.org.uk international investors may expose a coun- Tool selection: the type of capital controls

ODI at 50: advancing knowledge, shaping policy, inspiring practice • www.odi.org.uk/50years Opinion

Table 1: Capital control measures in emerging and developing Coordination is the answer countries, 2008-2010 Under certain conditions, capital control measures may be a legitimate and effective tool for developing Country Tax measures Quantitative Time Unremunerated reserve and emerging economies to avoid the drawbacks limits requirements requirements linked to sudden and massive surges in capital

Argentina √ √ inflows. However, in the current global climate, countries should understand that acting solely in Brazil √ √ their own self-interest is no longer a viable and sus- China √ tainable solution. Coordination between developed and developing countries is the only way forward. Colombia √ √ So far, the G-20 has failed to reach consensus India √ √ on coordinated capital controls. However, at the latest G-20 Summit in Paris, there was agreement Indonesia √ √ on a number of indicators to measure global imbal- Peru √ √ √ ances. While not enough, this is a step in the right direction and I hope that further progress towards South Korea √ √ greater global coordination will follow. √ A Capital Controls Charter for the G-20 might focus on: Thailand √ • improving financial regulation to discourage Turkey √ ‘carry trade’ practices in developed countries • providing trade agreements with enough flexibil- Source: Author’s elaborations on different sources. ity to allow the use of adequate, temporary and coordinated capital control measures to prevent matters. Governments should select capital con- or mitigate crises trols according to what they really intend to achieve. • encouraging emerging and developing econo- While capital controls may be more effective in mies to adopt agreed coordinated restrictions on changing the composition and maturity structure capital flows to avoid indiscriminate actions that of inflows (rather than reducing their volume), would simply redirect flows to other countries evidence suggests that tax measures work better • ensuring that capital control measures are well in easing the amount of inflows, compared to, for designed, last just long enough to counter surges example, unremunerated reserve requirements. in capital flows and can be withdrawn quickly Governments should also ensure that the chosen when they are no longer needed type of capital controls is flexible enough to adapt • promoting cooperation at the jurisdictional level Overseas Development to sudden changes in investor sentiment, as seen in between developed, emerging and developing Institute North Africa where the fall of regimes has spooked countries to avoid the circumvention of capital 111 Westminster Bridge investors, causing a contraction of net capital controls. Road, London SE1 7JD inflows to emerging markets. Tax reducing meas- Tel +44 (0)20 7922 0300 ures, for example, might be easier to adjust when Fax +44 (0)20 7922 0399 necessary than other forms of capital controls. Multilateral repercussions: in a globalised world, Email decisions taken at the individual country level are [email protected] likely to have a global impact. So, when deciding to impose capital controls, governments should Readers are encouraged to remember that this could prompt other countries to quote or reproduce material follow suit, undermining the current global recovery Written by Isabella Massa, ODI Research Officer (i.massa@ from ODI Opinions for their and exacerbating global imbalances. odi.org.uk). own publications, as long as they are not being sold commercially. As copyright Endnotes and references holder, ODI requests due acknowledgement and a 1 The practice of borrowing at low interest rates in one Magud, N. and Reinhart, C. M. (2006) Capital Controls: An country and using the borrowed funds to buy assets Evaluation. Working Paper 11973. Cambridge, MA: NBER. copy of the publication. offering higher yields in another country. Ocampo, J. A. and Palma, J. G. (2008) ‘The Role of Preventive The views presented in Regulations’ in J. A. Ocampo and J. this paper are those of Carvalho, B. S. de M. and Garcia, M. G. P. (2008) ‘Ineffective E. Stiglitz (eds), Liberalization and the author and do not Controls on Capital Inflows under Sophisticated Development. New York: Oxford University Press. necessarily represent Financial Markets: Brazil in the Nineties’ in S. Edwards Ostry, J. D.; Ghosh, A. R.; Habermeier, K.; Chamon, M.; the views of ODI. and M. Garcia (eds), Financial Markets Volatility and Qureshi, M. S. and Reinhardt, D. B. S. (2010) ‘Capital Performance in Emerging Markets. Cambridge, MA: Inflows: the Role of Controls’, IMF Staff Position Note, © Overseas Development National Bureau of Economic Research. SPN 10/04, 19 February. Washington, DC: International Institute 2011 Epstein, G. A. (2005) Capital Flight and Capital Controls in Monetary Fund. ISSN 1756-7629 Developing Countries. Cheltenham: Edward Elgar Press.