Ana səhifə

Dealing with volatile external finances at source: the rôle of preventive capital account regulations

Yüklə 248 Kb.
ölçüsü248 Kb.
1   2

a = capital controls imposed; b and c = capital control strengthened.

Source: IMF, (2004a).

Figure 12.B, however, highlights a major problem in studying the effects of controls on debt profiles. The crucial question is: what is the counterfactual? Is it a matter of controls being effective because they reduced the share of short-term debt vis-à-vis its own historical trend, or were they effective because they helped the countries that adopted capital account regulation (Chile, in this case) to avoid the trend of rising shares of short-term debt experienced by other developing countries that did not impose controls?

As it happened, vis-à-vis its own trend, until 1995 capital controls seem to have had little lasting effect in Chile, but a significant effect after the strengthening of controls in that year.25 However, if the comparison is made vis-à-vis the trend followed by countries that did not impose similar controls, such as Thailand and Brazil, then controls in Chile seem to have had quite a remarkable effect from the very beginning. In fact, in Chile the share of short-term debt was at a similar level to Thailand’s before the imposition of controls (about one quarter of the total), but by 1995 Thailand had a share twice as large as Chile’s. Furthermore, at the beginning of the Plano Real and full financial liberalization in Brazil in 1994, Chile actually had a share of short-term debt five percentage points higher than that of Brazil; however, by 1998 Brazil’s share was nearly four times higher than Chile’s.

In the case of Malaysia following its 1994 controls, a new wave of debt accumulation developed in that country, but the debt profile was kept at more prudential levels than in other Asian countries that were hit by the 1997 crisis (Kaplan and Rodrik, 2002; Palma, 2000a).

F. Effects on asset price dynamics

Finally, Figures 13 and 14 show that, in the three countries, capital account regulations had sufficient capacity to slow down and, in some cases, to pierce asset bubbles.


a = first democratic government after Pinochet; b = second democratic government; c = Mexican crisis; d = East Asian crisis; and e = Russian default.

a and b = progressive opening of the capital account; c = Mexican crisis; d = East Asian crisis; and e = Russian default.

a = East Asian crisis.
Source: DataStream.

As is clear from Figure 13.A, Chile was experiencing an asset bubble in its stock market in early 1991: in the four quarters preceding the first imposition of controls, the index had jumped more than three-fold; the 1991 and 1992 regulations stopped this trend for about seven quarters. However, as with the levels of net private inflows and the broader macroeconomic effects studied above, this effect soon ran its course and together with the huge new increase in inflows in 1994 the index jumped again, this time more than two-fold. This time the strengthening of controls in 1995 had an immediate impact on this new bubble, bringing the index down considerably; and when it began to recover again in early 1997, with the new increases in inflows, its progress was halted by the mid-1997 East Asian crisis.

Something similar, but even more pronounced, took place in real estate after 1995 (see Figure 14).


Quarterly Real State Index

a = first democratic government after Pinochet; b = second democratic government; c = Mexican crisis; d = East Asian crisis; and e = Russian default.

a = East Asian crisis.
Source: DataStream.

In this market, Chile was facing another bubble when capital controls were imposed in 1991. In this case, the (short-term) reduction in net private inflows that came with inflow-controls did not have such an immediate effect as on the stock market. However, the strengthening of regulations in May 1992 coincided with the interruption of the real estate boom, which by then had already increased close to five-fold in just six quarters. Nevertheless, as in the stock market, the respite was only temporary, and this index doubled again between the end of 1993 and the strengthening of capital controls in the third quarter of 1995 (following the renewed increase in inflows). The subsequent fall is remarkable (more so than in the stock market), even though the economy continued to grow rapidly until 1998.26

In Colombia, booming capital inflows since 1992, partly fuelled by growing capital account liberalization, led to a seven-fold increase in stock prices. This process was sharply reversed by the August 1994 controls and the accompanying monetary policy. A similar reversal took place after the controls adopted in early 1997, but here the preceding boom had not reached its peak. Although no similar indices exist for real estate prices, partial evidence in this regard (the evolution of housing rents) indicates that a similar reversal took place, coinciding with the 1994 regulations.

Figure 13 also shows the remarkable jump in stock prices at the time of the surge in inflows in Malaysia in 1993. Before the imposition of controls, this index jumped more than two-fold in just four quarters. During the three quarters that these controls lasted in full, this index fell by 30%; it then began to recover somewhat erratically, almost reaching its previous peak again in the last quarter of 1996.27

A rapid bubble in real estate prices in Malaysia also took off in the four quarters before the imposition of controls in 1994. As in Chile, the piercing of this bubble was not as immediate as the one in the stock exchange. However, in contrast to Chile, the return of inflows in 1995 pushed this index back up with a vengeance. One major difference between the two countries was the level of interest rates. As Table 1 indicates, interest rates remained relatively high in Chile two years after the July 1995 controls. The return of inflows, low deposit rates and little life in the stock exchange (by pre-crisis standards), together with low mortgage rates, set in motion another real estate bubble: in just four quarters the index jumped again more than two-fold.


Overall, the experiences of Chilean, Colombian and Malaysian regulations on capital inflows indicate that they served as useful instruments for improving debt profiles and the macroeconomic trade-offs faced by the authorities, and for restraining asset bubbles. However, the macroeconomic effects, including those on asset prices, depended on the strength of the regulation and tended to be temporary -- operating more as ‘speed bumps’ than as permanent speed restrictions. The basic advantage of the price-based instrument used by Chile and Colombia is its non-discretionary character, whereas quantity-based controls in Malaysia proved to be stronger in terms of short-term macroeconomic effects. Thus, when immediate and drastic action is needed, quantitative controls are more effective.

The dynamics of capital flows must be taken into account when analysing the overall and relative virtues of the different types of regulations. Interestingly, given the links between the dynamics of capital flows and current account deficits in Latin America vis-à-vis East Asia discussed in part II of the paper, the policy prescription should perhaps have been the opposite of what actually happened: quantitative controls for Latin America (where ‘exogenous’ inflow-surges dominate) and price controls for East Asia (a region in which inflow-surges have tended to be mostly ‘endogenous’).

In any case, it must be emphasized that these systems were designed for countries that had initially chosen to be fully integrated into international capital markets, but later decided to ‘fine-tune’ this integration -- at least temporarily. Traditional exchange controls and capital-account regulations -- when applied effectively and transparently -- may thus be superior if the policy objective is significantly to reduce domestic macroeconomic sensitivity to volatile and unregulated international capital flows, as the experiences of China, India and Taiwan indicate.

APPENDIX: Results of the ‘Granger-predictability’ test between net private capital inflows and current account.

In all Latin American countries the test spans the period from 1950 to 2002, and the source of the data is ECLAC’s Statistical Division. Due to lack of data, for the East Asian countries it was only possible to study the period 1975-2002; the source of these data was IMF (2004a). Following the ‘Perron-sequential procedure’, unit root tests indicate that all series of net private capital inflows and current account in both regions have a unit root; furthermore, with the (significant) exceptions of Chile and Malaysia, all series cointegrate. Therefore, the Granger-test for Argentina, Brazil, Colombia, Mexico, Korea and Thailand was done in levels, while for Chile and Malaysia it was done in first differences. The specification of the regressions (i.e., the number of lags) within which the null hypothesis (of no ‘predictability’) was tested both ways was determined by the rule of choosing the minimum number of lags that would produce a residual that was not serially-correlated.

‘p’ of T1

‘p’ of T2

Lags T1

lags T2

‘p’ of Q T1

‘p’ of Q T2











































‘p’ of T1

‘p’ of T2

Lags T1

lags T2

‘p’ of Q T1

‘p’ of Q T2















T1 = test of the null hypothesis that net private capital inflows do not ‘Granger-predict’ the current account. T2 = test of the null hypothesis that the current account does not ‘Granger-predict’ net private capital inflows. ‘p’ of Q = is the level of significance at which the null hypothesis of no autocorrelation up to order 3 is rejected using the ‘Ljung-Box Q-statistics’.

The results of the Granger tests (in levels) indicate that in Argentina, Brazil, Colombia and Mexico net private capital inflows are a good predictor of the current account, while in Korea and Thailand there is evidence of predictability the other way around. In Chile (tested in first differences), there seems to be a two-way predictability phenomenon; in Malaysia, meanwhile, (also in first differences and up to a 10% level of significance) the null hypothesis of no predictability cannot be rejected either way; however, if the level of significance is increased to 10.7%, the test indicates that Malaysia seems to follow the same pattern as Korea and Thailand.


Agosin, Manuel (1998), ‘Capital Inflow and Investment Performance: Chile in the 1990s’, Capital Inflows and Investment Performance: Lessons from Latin America, Ricardo Ffrench-Davis and Helmut Reisen (eds.), Paris and Santiago, OECD Development Centre/ECLAC.

Agosin, Manuel and Ricardo Ffrench-Davis (2001), ‘Managing capital inflows in Chile’, Short–term Capital Flows and Economic Crises, Stephany Griffith–Jones, Manuel F. Montes and Anwar Nasution (eds.), New York, Oxford University Press/United Nations University (UNU)/World Institute for Development Economics Research (WIDER).

Ariyoshi, Akira, Karl Habermeier, Bernard Laurens, Inci Ötker-Robe, Jorge Iván Canales-Kriljenko, and Andrei Kirilenko (2000), Capital Controls: Country Experiences with Their Use and Liberalization, Washington, D.C., International Monetary Fund, Occasional Paper 190.

Cárdenas, Mauricio and Felipe Barrera (1997), ‘On the Effectiveness of Capital Controls: The Experience of Colombia During the 1990s’, Journal of Development Economics, vol. 54, N 1, October.

--------------------- and Roberto Steiner (2000), ‘Private Capital Flows in Colombia’, in Felipe Larraín (ed.), Capital Flows, Capital Controls, and Currency Crises: Latin America in the 1990s, The University of Michigan Press, Ann Arbor.

De Gregorio, José, Sebastián Edwards and Rodrigo Valdés (2000), ‘Controls on Capital Inflows: Do They Work?’, Journal of Development Economics, vol. 63, No. 1.

ECLAC (2004), Balance Preliminar de las Economías de América Latina y el Caribe, 2004, Santiago, ECLAC.

Epstein, Gerald, Ilene Grabel and K.S. Jomo (2003), ‘Capital Management Techniques in Developing Countries’, in Ariel Buira (ed.), Challenges to the World Bank and the IMF: Developing Country Perspectives, London: Anthem Press, chapter 6.

IMF (2004a), World Economic Outlook database.

IMF (2004b), International Financial Statistics database.

Kaplan, Ethan and Dani Rodrik (2002), ‘Did the Malaysian Capital Controls Work?’, NBER Working Paper Series, No. 8142, Cambridge, Ma., February.

Kindleberger, Charles (1984), ‘The 1929 World Depression in Latin America – from the outside’, in R. Thorp (1984).

Larraín, Felipe, Raúl Labán and Rómulo Chumacero (2000), ‘What Determines Capital Inflows? An Empirical Analysis for Chile’, in Felipe Larraín (ed.), Capital Flows, Capital Controls, and Currency Crises: Latin America in the 1990s, The University of Michigan Press, Ann Arbor.

Laurens, Bernard (2000), ‘Chile’s Experience with Controls on Capital Inflows in the 1990s’, in Ariyoshi et al., (2000).

Le Fort, Guillermo and Sergio Lehmann (2000), ‘El encaje, los flujos de capitales y el gasto: una evaluación empírica’, Documento de Trabajo Nº 64, Central Bank of Chile, February.

Nagara Bank, (2003), Annual Report, Kuala Lumpur, Nagara Bank.

Ocampo, José Antonio (2003a), ‘Capital account and counter-cyclical prudential regulations in developing countries’, in Ricardo Ffrench-Davis and Stephany Griffith-Jones (eds.), From Capital Surges to Drought: Seeking Stability for Emerging Markets, London, WIDER/ECLAC/Palgrave.

---------------------- (2003b), ‘Developing countries’ anti-cyclical policies in a globalized world’, in Amitava Dutt and Jaime Ros (eds.) Development Economics and Structuralist Macroeconomics: Essays in Honor of Lance Taylor, Aldershot, UK, Edward Elgar.

---------------------- (2002), ‘Developing Countries’ Anti-Cyclical Policies in a Globalized World’, in Amitava Dutt and Jaime Ros (eds.) Development Economics and Structuralist Macroeconomics: Essays in Honour of Lance Taylor, Aldershot, UK, Edward Elgar.

Ocampo, José Antonio and Camilo Tovar (2003), ‘Managing the Capital Account: Colombia’s Experience with Price-based Controls on Capital Inflows’, CEPAL Review, No. 81, December.

--------------------------------------------- (1998), ‘Capital Flows, Savings and Investment in Colombia, 1990-96’, in Ricardo Ffrench-Davis and Helmut Reisen (eds.), Capital Flows and Investment Performance: Lessons from Latin America, Paris and Santiago, OECD Development Centre/ECLAC.

Ötker-Robe, Inci (2000), ‘Malaysia’s Experience with the Use of Capital Controls’, in Ariyoshi et al. (2000).

Palma, Gabriel (2000a), ‘The Three Routes to Financial Crises: The Need for Capital Controls’, in CEPA Working papers N.17; a revised version of this paper was published in John Eatwell and Lance Taylor (eds.), International Capital Markets - Systems in Transition, New York, Oxford University Press, 2002.

---------------- (2000b), ‘The magical realism of Brazilian economics: how to create a financial crisis by trying to avoid one’, in CEPA Working Paper N. 16, 2000; a revised version of this paper was published in J Eatwell and L Taylor, International Capital Markets - Systems in Transition, Oxford University Press, 2002.

---------------- (2003), ‘The ‘three routes’ to financial crises: Chile, Mexico, and Argentina [1]; Brazil [2]; and Korea, Malaysia and Thailand [3]’, in H-J Chang, Rethinking Development Economics, Anthem Press, 2003. Reprinted in R. E. Allen (ed.) The International Library of Writings on the New Global Economy: The Political Economy of Financial Crises, Aldershot, UK, Edward Elgar, 2004.

Park, Yung Chul and Chi-Young Song (1997), ‘Managing Foreign Capital Flows: The Experience of the Republic of Korea, Thailand, Malaysia and Indonesia’, in UNCTAD, International Monetary and Financial Issues for the 1990s, Volume VIII.

Rodrik, Dani and Andrés Velasco (2000), ‘Short-Term Capital Flows’, Annual World Bank Conference on Development Economics 1999, Washington, D.C., The World Bank.

Stiglitz, Joseph E. (2003), ‘Whither reform? Toward a new agenda for Latin America’, CEPAL Review, No. 80, August.

Thorp, R., (ed.) (1984), Latin America in the 1930s: the role of the periphery in world crisis, London, Macmillan.

Valdés-Prieto, Salvador and Marcelo Soto (1998), ‘The Effectiveness of Capital Controls: Theory and Evidence from Chile’, Empirica, Nº 25, The Netherlands, Kluwer Academic Publishers.

Villar, Leonardo and Hernán Rincón (2002), ‘The Colombian Economy in the Nineties: Capital Flows and Foreign Exchange Regimes’, in Albert Berry (ed.), Critical Issues in Financial Reform: A View from the South, New Brunswick, N.J., Transaction Publishers.

Williamson, John (2000), ‘Exchange Rate Regimes for Emerging Markets: Reviving the Intermediate Option’, Policy Analyses in International Economics 60, Washington, D.C., Institute for International Economics, September.

World Bank (2004), Global Development Finance.

1 According to ECLAC’s database, in US$ of 2000 value -- from now on US$(2000) --, between 1973 and 1981 net private inflows into Latin American reached US$(2000) 440 billion; in the 1990s this figure jumped to US$(2000) 580 billion (see Figure 2 below). In turn, net transfer of resources (net inflows minus net factor payments; for this calculation net inflows includes ‘errors and omissions’, but excludes the use of IMF credit, IMF loans and exceptional financing) became negative to the tune of US$(2000) 650 billion between 1982 and 1990; and again, between 1999 and 2004, Latin America had to deal with negative transfers of a further US$(2000) 209 billion (see Figure 4).

2 The remarkable size of net inflows into East Asia before the 1997 crisis -- according to the IMF (2004a), ‘Developing Asia’ received more than US$(2000) 500 billion between 1989 and 1996 -- makes Alan Greenspan's oft-quoted post-East Asian crisis remarks something of an understatement: ‘In retrospect, it is clear that more investment monies flowed into these economies than could be profitably employed at modest risks’ (1997, p. 1).

3 According to Kindleberger, international financial markets can do one thing that is more damaging for developing countries than ‘over-lending’: to halt that lending abruptly! (see Kindleberger, 1984).

4 See also the papers by Frenkel, and by Epstein, Grabel and Jomo in this volume.

5 The latter figure is from ECLAC (2004).

6 Only time will tell whether the unprecedented current account surpluses of 2003 and 2004 (US$ 8 billion and US$ 22 billion, respectively -- the first current account surpluses of the region since the end of the Korean War) will represent a break in this historical feature of Latin America. So far, these surpluses seem to be the result of a particular combination of factors, such as high commodity prices, slow regional growth between 1997 and 2003, low international interest rates and low country-risk premiums for many countries in the region; however, Latin America’s economic history seems to indicate that unfortunately these events do not tend to last very long.

7 The exceptions are Chile and Malaysia, where capital controls seem to have ‘meddled’ sufficiently (in a statistical sense) with the dynamics between these two components of the balance of payments (see Appendix). In time series forecasting, the Granger statistics is an F statistics testing whether the lags of one variable have a useful predictive content on another variable (above and beyond other repressors). The claim that a variable has no predictive power corresponds to the null hypothesis that the coefficients on all lags of that variable are zero. As is well-known, ‘Granger-causality’ has little to do with causality itself; it simply means that if one variable ‘Granger-causes’ another, then the former is a useful predictor of the latter, given the other variables in the regression (for this reason, a more appropriate name for this test would have been ‘Granger-predictability’ test).

8 The D-Ram price per megabyte, for example, fell from US$26 in 1995 to US$10 in 1996, US$4 in 1997 and less than US$1 in 1998. Memory-chips were one of Korea's main export-items. The collapse of the price of the D-Ram memory in 1995 was triggered by massive Taiwanese new investment in memory-chips of more advanced technology, which came on stream at a time when markets were already saturated. As is well known, in micro-electronic markets competitiveness only exists at the cutting-edge of technology; so, Korean corporations had little choice but to invest in the new technology even though collapsing prices had drastically cut short-terms returns.

9 For the different route followed by Brazil to its 1999 financial crisis, see Palma (2000b).

10 See the contribution of Epstein, Grabel and Jomo to this volume.

11 The previous regulations, dating from 1991 (which, in turn, were reformed versions of the 1967 foreign exchange regulations) had established a minimum maturity of one year for foreign loans and maintained the traditional regulations on the final use of funds from such loans; i.e., these could only be used for trade or investment financing. In February 1992 for the first time firms were allowed to contract foreign credits abroad for short-term working capital. In September 1993, when the URR were introduced, the traditional system of regulations based on final use was replaced by a system based on maturity. Additionally, domestic financial intermediaries were authorized to lend in foreign currency to domestic firms and residents regardless of the final use of the credit, to lend to foreigners in international currencies and to invest liquid assets abroad. The fact that the 1993 regulations were an effective liberalization of the capital account is emphasised by Ocampo and Tovar (1997 and 2003), who show that they actually increased the sensitivity of capital flows to interest-rate differentials (arbitrage incentives).

12 This took place after a short experiment with an explicit (Tobin) tax on all capital inflows, which was declared unconstitutional by the Constitutional Court two months after it was decreed.

13 In Chile they were later abandoned altogether. Furthermore, when Chile signed a free-trade agreement with the USA in 2003 it accepted as part of the treaty a clause that would seriously restrict its capacity to implement capital controls in the future. According to this clause, now Chile could only implement capital controls at a time of ‘financial crisis’, and only for a one-year period; if controls were to continue after that year, Chile would have to compensate capital coming from the USA. However, the treaty does not specify (i) what does it take for a difficult financial situation to be acknowledged as a financial ‘crisis’; (ii) what would be the level of compensation that would need to be paid to capital coming from the USA if the capital controls continue after one year; nor (iii) what would be the mechanism to differentiate capital coming from the USA from ‘non-USA capital’ (i.e. the particularly difficult issue of ‘rule of origin’ in international financial markets).

14 There is also evidence of a strong lobby from the domestic financial sector for the government to lift the most drastic controls.

15 On the effect of these regulations, see Epstein, Grabel and Jomo (2003), Rajaraman (2002) and, particularly, Kaplan and Rodrik (2002).

16 See footnote 11.

17 Non-FDI inflows, which had reached more than US$(2000) 11 billion in 1993, turned sharply negative in 1994; FDI was the only component of capital inflows that remained unaffected by these controls.

18 The negative figures for this item in 1995 and 1996 probably reflect capital flight by Malaysian citizens (as well as peculiar accounting practices of the Malaysian Central bank, especially regarding the revaluation of their gold reserves). If this was the case, like their counterparts in Mexico before their December 1994 crisis, domestic agents may have predicted trouble with better foresight than international funds and bank managers did.

19 For documents that support the effectiveness of these regulations in Chile, see Agosin (1998), Agosin and Ffrench-Davis (2001), Larraín et al. (2000), Le Fort and Lehman (2000 and 2003) and Palma (2000a). For a more mixed view, see Ariyoshi et al. (2000), De Gregorio et al. (2000), Laurens (2000) and Valdés-Prieto and Soto (1998). Similarly, for strong views on their positive effects in Colombia, see Ocampo and Tovar (1998 and 2003) and Villar and Rincón (2002), and for a more mixed view, Cárdenas and Barrera (1997) and Cárdenas and Steiner (2000).

20 Indeed, evidence on the insensitivity of the volume of capital flows to capital-account regulations comes from econometric analysis in which URR is not included as a determinant of interest rate spreads but rather as an additional factor affecting capital flows; this may therefore be interpreted as an inadequate econometric specification.

21 Some of these mechanisms, such as the use of hedging, enable investors to cover some of the effects of these regulations, but in large part this is done by transferring risks (more specifically, the risk associated with longer-term financing) to other agents who would only be willing to assume them for an adequate reward. More generally, if there is no stable external demand for the domestic currency, hedging may be available only in limited quantities, a fact that affects the maturities and costs involved.

22 For the case of Chile, see Larraín et al., 2000; Le Fort and Lehman, 2000 and 2003; and De Gregorio et al., 2000; for Colombia, Villar and Rincón, 2002.

23 This is the very apt interpretation provided by Williamson (2000, ch. 4). Indeed, under this interpretation, the apparently conflicting evidence on the Chilean case largely disappears.

24 The low level of the URR may account for this result. Valdés-Prieto and Soto (1998) find evidence of a ‘threshold effect’, which would explain why these regulations were only effective in reducing capital flows in 1995-1996. Despite this, Agosin and Ffrench-Davis (2001) have argued that on broader grounds the macroeconomic management undertaken in the earlier part of the 1990s was more appropriate than in 1995-1996.

25 In fact, according to Chile’s Central Bank balance of payments statistics, after 1995 this share fell even further than is indicated by the IMF source used in this graph (2004b) -- from over 18% in 1994, to 16% in 1995, 12% in 1996, and less than 5% in 1997.

26 Chile’s GDP grew at 7.4% in 1996 and 6.6% in 1997.

27 The crash after the mid-1997 crisis was equally remarkable; by the third quarter of 1998 the local currency denominated index had fallen to just 38% of its early 1997 level.

1   2

Verilənlər bazası müəlliflik hüququ ilə müdafiə olunur © 2016
rəhbərliyinə müraciət