Macroprudential Policies and Economic Performance: Some Reflections from Latin America

Transcription

1 Macroprudential Policies and Economic Performance: Some Reflections from Latin America José De Gregorio Universidad de Chile Peterson Institute for International Economics October 2015 Latin America performed exceptionally well during the financial crisis of Historically, whenever there have been international turbulences, they have been amplified in the region. Even during the Asian crisis of late 1990s, Latin America s performance was even worse than that of Asian countries. Macroeconomic policies improved significantly during the last decade, the financial systems were resilient, and, together with the terms of trade boom, were the key factors that explain the unprecedented performance. In particular, Latin America has suffered recurrent financial crises and, for this reason, financial innovation has been prudent and macroeconomic management coherent with financial stability. But there have also been some tensions coming from the international economy, in particular currency appreciation and volatile capital flows. In this paper I want to focus on some relevant issues related to policies designed to foster financial stability. I will first discuss the role of macroprudential policies, interactions with monetary policy and some experiences in Latin America. Then, I will discuss the role of exchange rate flexibility and capital account management, to close with some concluding remarks. 1. Macroprudential polices and financial stability Significant progress has been made on the ways to conduct monetary policy to achieve price stability. Inflation targets have become a reasonably good strategy to pursue price stability. Despite some improvements that can be made and new dimensions that can be explored, inflation targets have been a successful anti- inflation strategy in emerging market economies. However, the instruments to preserve financial stability as well as the interactions with price stability are blurrier. Even the goal of financial stability is difficult to define. Is it the objective to avoid crises at any cost? Or is it to smooth the financial cycle? The latter seeks Prepared for the 2015 SEACEN Policy Summit, Macroprudential Policies: Practical Implementation, Kuala Lumpur, Malaysia. Many of the issues discussed in this paper are treated with greater detail in De Gregorio (2014). 1

2 to reduce financial vulnerabilities, but it does not imply the probability of crisis will be zero, or even reduced substantially. Moreover, what is the extent and the rationale to smooth the credit cycle? Is it optimal? We should smooth suboptimal cycles. Avoiding crises is also an elusive definition and difficult to monitor. Assume a central bank decides that the probability of financial crisis should be 1 percent. Only in the long run, by the law of large numbers, will the probability of crisis be the same as the actual frequency of crisis. But, triggered by some random effect, a country seeking a low probability may have an earlier crisis. Certainly, there would be a lot of post- mortem examination of what went wrong, and certainly the public would not be satisfied with an answer that is based on the realization of a bad random draw. Contrary to the objective of monetary policy, the target for financial stability is difficult to define, but the costs are evident and could be very large. Perhaps, the first lesson in this area is caution in terms of financial deepening. It must go together with increased monitoring and appropriate risk evaluation. However, regulators cannot expect to foresee all risks, in particular since it may induce moral hazard problems. Regarding instruments, financial stability policies are the now famous macroprudential tools. Most of them are standard microprudential tools, which target the financial health of specific institutions, but the difference is that macroprudential tools are calibrated to deal with financial instability on both the time- and the cross- section dimensions. On the time dimension, the idea of macroprudential tools is to avoid the buildup of financial vulnerabilities in the upturn of the business cycle, and to have a cushion for the downturn. The goal would be to smooth the credit cycle. On the cross- section dimension the idea is to limit spillovers and for that reason the focus has been on too- big- too fail institutions. Additional capital buffers have been proposed for systemic institutions in order to make them more resilient to financial turbulences. But the definition of systemic institutions is still blurred, and given the evolution of financial innovations, what today may be a non- systemic institution could eventually become systemic. A non- systemic institution on a worldwide basis could be systemic from the point of view of particular economies. This should make policymakers not to look just for global guidelines, but to have a country specific view, since being systemic at a country level is quite different from being systemic in the global financial system. Regarding the interactions between macroprudential policies and monetary policy, the traditional view has been influenced by the Tinbergen principle, by which there should exist as many instruments as policy targets. Moreover, a desired feature would be for no interactions, so policies would be completely independent, as depicted in figure 1 panel (a). However, there are many interactions and in many directions, as shown in panel (b). This is not the place to discuss all potential interactions, but, for example, the monetary transmission channel depends on the health of the financial system. On the other hand, a stable economic environment may induce search for yield and weaken the financial system. 2

3 Another issue in the recent discussion is whether macroprudential policies are substitutes or complements of monetary policy. This depends on how they are implemented. For example, many emerging markets tighten the reserve requirement to stabilize the economy and to avoid raising rates in order to reduce pressures on the exchange rate. In this case they are being used as substitutes. This has happened in Latin America as will be discussed below. However, if monetary policy is tightening together with a tightening of macroprudential policies, they are being used as complements. Despite all the interactions, it is still useful to utilize monetary policy for price stability and macroprudential policies for financial stability, without disregarding the interactions, but focusing on specific instruments with specific goals. For an aggregate purpose, the interest rate has broad- based effects, while macroprudential tools have more targeted impacts and unexpected side effects could be significant. On the other hand, macroprudential policies can focus on specific financial dislocations, something that the monetary policy interest rate cannot do. However, some side effects can be significant, especially because the shadow banking system may circumvent regulations and be the recipient of all risks to financial stability. Latin America was using prudential tools long before the expression macroprudential became fashionable. The long history of financial collapses, where most of the time the central banks had the lead in the rescue teams, have made regulation much more prudent than in many advanced economies that suffered financial crises in the late 2000s. In the case of Chile, the law of 1989 that granted autonomy to the Central Bank, establishes that the objectives are price stability and the normal functioning of internal and external payment systems, the latter being the financial stability mandate, in the domestic as well as the international sphere. Within this context, the Central Bank has attributions that complement the roles of supervisors. For example, it can define limits to external investment by pension funds, in the understanding that they have systemic implications. Although most of the powers the bank has on these areas are not regularly used, they are important in case of critical situations, and also to discipline regulators. In Latin America there have been occasions where macroprudential policies have been used as substitutes of monetary policy, in order to stem the business cycle without inducing too much pressure on the exchange rate. This has been the case of bank reserves. Colombia and Peru increased them before the Great Recession, to avoid tightening monetary policy in a period of sharp appreciation of regional currencies. Later, Brazil raised reserve requirements after the crisis to avoid further appreciation in the so- called currency war episode. Figure 3 shows the evolution of reserve requirements. Results reported in Tovar et al. (2012) indicate that their effects are modest and transitory, and their effectiveness improves when used in conjunction with monetary policy. This would indicate that they should be used more as a complement than as a substitute. 3

4 In Latin America, some forms of dynamic provisions have been used, such as those used in Spain since the early 2000s. Of course an informal evaluation could be biased given the poor performance of the Spanish economy during the crisis. However, the evidence would indicate that this regulation helped, especially in the downturn, when alternative sources of funding dried up. As Peydró (2014) finds: The estimates show that countercyclical dynamic provisioning smoothes cycles in the supply of credit and, in bad times, strongly upholds firm financing and performance. In good times there is no discernible contraction of credit available to firms, since additional sources of financing are available. In Latin America, Bolivia (2009), Colombia (2007), Peru (2008) and Uruguay (2001) have implemented some forms of counter- cyclical provisioning. However, their effectiveness and in particular their countercyclicality are not well established yet (Calderón and Servén, 2013). This could be a potentially interesting set of instruments; however, too much discretion must be avoided, and implementing them in the form of rules may strengthen financial stability and makes easier to recalibrate parameters. Too much discretion may end up in inaction. One of the main problems of macroprudential tools is that they induce a shift of funding to unregulated segments of the market, shadow banking, which could undermine their effectiveness, and shift risk from banks to shadow banks. This could create the same vulnerabilities that are intended to be avoided, in particular when those alternative banks are too big or too interconnected to fail. In Latin America the shadow banking system is still small, but its growth could be induced, and accelerated, by implementing tighter policies. 2. Exchange rates: overcoming fear of floating At the center of most crises in Latin America and most emerging market economies, have been currency collapses. Currency mismatches, owing largely to the authorities promises of exchange rate stability, create severe financial distress when the currency is adjusted. Banks can be well regulated in terms of direct currency exposure, and this has been the norm in Latin America, but the imbalance is rooted in the non- financial corporate sector. For this reason, in terms of regulation, monitoring and market discipline, it is essential to focus on foreign exchange imbalances outside the banking system and evaluate how they could contaminate the financial system. The fear of financial collapse and an inflation outburst have been the main causes for widespread fear of floating in emerging markets until the turn of the century (Calvo and Reinhart, 2002). However, and perhaps surprisingly, a credible flexible exchange rate is by itself a way to overcome the fear of floating. The pass- through from the exchange rate to inflation declines and corporations reduce their currency exposure when the exchange rate floats. 1 1 For further discussion of the Chilean experience, see De Gregorio and Tokman (2007). 4

5 The perils of fear of floating are shown in figure 2, which plots the evolution of exchange rates in Latin American countries that floated their currencies during the global financial crisis and compares it with their behavior during the Asian crisis. In both cases I consider a two- year window and normalize the exchange rate to 100 for the period average. 2 During the global crisis (panel (a)), exchange rate tensions were acute, and except for some moderate intervention and provision of liquidity in foreign exchange, currencies floated sharply and depreciations were severe. Indeed, Brazil, Chile, Colombia and Mexico experienced depreciations of about 60 percent from their strongest levels previous to the Lehman collapse in a period of just a few months. Peru s exchange rate was more stable, depreciated about 20 percent, as it also was the one most subject to intervention among these five countries. Inflationary consequences were muted. In the past, a sharp depreciation like the ones presented in panel (a) would have caused serious financial distress. This time was different, and financial systems remained broadly strong. The movements of exchange rates were consistent with the deterioration of the global outlook, which resulted in a decline in commodity prices. This was a fundamental reason to expect a weakening of the currencies. In addition, the rapid depreciation contributed to deter speculation against Latin currencies, since allowing sharp depreciation signaled the authorities commitment to let the exchange rate adjust accordingly, reducing incentives to fight against incredible exchange rate defenses and obtain financial gains from central banks fighting market developments. During the Asian crisis there was much more reluctance to let the exchange rate adjust, providing a clear case of fear of floating. The evolution of exchange rates was more stable (panel (b)), but this accumulated tensions to speculate against currencies and made the adjustment more sluggish. The fear of inflation and financial instability led to efforts to defend the currency by hiking interest rates and using reserves. The defense of the currencies coupled with tightening of financial conditions explain part of the poor performance of Latin American countries during the Asian crisis, while expansionary macroeconomic policies in the context of a flexible exchange rate help to understand the success during the global financial crisis. Regarding currency exposure and financial stability, after exchange rates were allowed to float, the level of liability dollarization in Latin America declined significantly (see figure 3). Overall, the pattern is clear pre- and post- flexibility of the exchange rate. The private sector has the incentives to hedge against currency risk when the central bank is not doing it via exchange rate management, and that is a positive development that reduces financial vulnerability when the exchange rate floats, overcoming fear of floating. 2 The exchange rate measures the price of the foreign currency in terms of the domestic currency, so a higher value of the index denotes a depreciation. 5

6 Since early 2013, the currencies of emerging markets and commodity exporters have weakened significantly. The decline in commodity prices, the slowdown of emerging market economies, and the perspectives of a tightening of financial conditions in the US have triggered massive depreciations. Until September 2014, the Chilean peso, the Colombian peso and the Brazilian real had depreciated with respect to the US dollar by 40, 70 and 100 percent. Twenty years ago this could have been a cause for a full financial collapse and inflation explosion. This has not happened and it owes a lot to the exchange rate regime. 3. On capital account management and reserves accumulation One issue that has been dominant in international finance in small open economies has been the role and potential destabilizing effects of capital flows. Indeed, one of the main focuses currently of macroprudential policies in emerging markets is capital account management. Periods of rapid capital inflows may end up in very costly sudden stops. Reserves accumulation and capital controls are the main tools policymakers have used to manage the capital account. However, I want to raise some issues on the interpretation and policy implications of capital flows. The first one refers to the distinction of gross versus net flows and their implications regarding macroeconomic and financial stability. Net capital inflows are the counterpart of current account deficits, when there is no reserves accumulation. Excessive net inflows may be an indication that the economy is running an unsustainable current account deficit. What is the driving force: the current account imbalance or capital inflows? This is an unsettled issue, and of course causality should go both ways, but this requires further analysis. If the cause of large net inflows is the current account, one has to think of macroeconomic policies to contain aggregate expenditure. In contrast, capital account management seems more appropriate when the driver is financial flows. As a first approximation, the current account or net inflows is what matters for exchange rates, in particular for the real exchange rate, which is the relative price that gives the signal for resource allocation and demand patterns consistent with saving and investment decisions. Gross inflows, in turn, are the response to portfolio allocation and are central to financial stability. The form and volume that gross flows take have a direct impact on the vulnerability of the financial system. It has long been argued, and rightly so, that foreign investment and equity flows are more stable, while banking flows are more likely to be subject to sharp reversals. Therefore, it is important to be clear about the objectives of different policy actions. Whether they are geared to affect the volume of net flows or the composition of gross 6

7 flows is relevant regarding the choice of instruments. In particular, we would expect that concerns regarding the exchange rate have to do with net flows, while financial stability with gross flows. For net flows, any policy must be broad based, while discrimination about the type of flows is more reasonable to deal with financial vulnerabilities stemming from different types of gross flows. A second issue relevant to capital flows is how we interpret them in the presence of reserves accumulation. In this case net capital inflows (deficit in the financial accounts) equals the current account deficit plus reserves accumulation. In other words: 3 CA =FA+ ΔR, where CA and FA are the current and financial accounts balances, respectively, and ΔR is the change in international reserves. Therefore, net capital inflows may be the result of either or both a current account deficit (- CA) and international reserves accumulation (ΔR). Or, with causality running the other way around, accumulation of reserves may be the result of large capital inflows. In the case that reserve accumulation has no effects on the exchange rate, and therefore no effect on the current account, as in a frictionless world, FA will change one to one with ΔR. Therefore, for reserves accumulation to be effective in reducing capital flows it is necessary for it to have effects on the exchange rate and the current account. The evidence is unsettled, but it is interesting to contrast the evidence of surges of capital flows during the 1990s and the 2000s to illustrate that they may be caused by different phenomena (De Gregorio, 2014, chapter 5). Table 1 shows episodes of surges of net capital inflows since 1990 in a large sample of countries. The surges are those identified in Gosh et al. (2012) and I add the current account balance and the increase in reserves in both decades for three groups of countries. In the 1990s, capital inflows to Asia and Latin America were much more significant than in the 2000s. In particular, in Asia the average surge was 7.6 percent of GDP, while the current account deficit was 4.8 percent of GDP. This is consistent with the fact that some Asian countries where having increasing current account imbalances before the crisis. In contrast, during the 2000s only Korea had a couple of surge episodes, but associated with a large accumulation of reserves with current account surpluses. Current accounts were in surplus, while net inflows were the result of reserves accumulation. In Latin America, net capital inflows during the 2000s were slightly less than in the 1990s, although the composition between current account deficit and accumulation of reserves was quite different. During the latter period there was much 3 This is the convention following the IMF Balance of Payment Manual 6 of 2013, which equalizes the current account with foreign assets acquisitions, and hence a surplus in the current account equals a surplus in the financial account. 7

8 more reserves accumulation. During the 1990s, net capital inflows were indeed the counterpart of significant current account deficits, but not in the 2000s. The case of Emerging Europe is the opposite. Surges of capital inflows were much larger and more frequent than those of Latin America and Asia. In addition, during the 2000s they responded much more to very large current account deficits, more than double the ones experienced in the other regions. In hindsight, they really suffered a capital- inflows- unsustainable- current- account problem. This evidence shows that the interpretation of surges in capital inflows is not unique, and therefore, the policy implications are not straightforward. Beyond the traditional push and pull factors explaining capital inflows we should also add a third factor, and this is the reserves one. But, there are also other policy factors that we should look at when discussing capital inflows. Some acute periods of capital inflows have coincided with high domestic interest rates and pressures on the exchange rate. Good examples have been Chile in the 1990s and Brazil in the 2000s. In both cases, capital controls were applied to contain the appreciation of the exchange rate and limit capital inflows. However, the success, at least from the exchange rate perspective, was at most limited. 4 Interest rate differentials with the fed funds were about 10 percentage points at the peak of the inflows period in both episodes. This leads me to my final point regarding capital account management, and it is the role of capital controls. We have seen recently some interesting analytical research on the usefulness of capital controls (e.g., Brunnermeier and Sannikov, 2014, and Korinek, 2011). The economies are subject to pecuniary externalities that increase vulnerability, and hence the use of capital controls, or Pigouvian taxes, serve the purpose of macroprudential policies. It is good that new models are introducing financial frictions to understand the business cycle and its policy implications. In this regard, the issue is what the appropriate financial policies to reduce fragilities are. Indeed, more often than not, distortions have to do with excessive credit, so rather than a capital control, prudential policies regarding credit booms could be more appropriate. There is somewhat of disconnect between the analytical developments and the evidence as well as practice with capital controls. The evidence has broadly shown that capital controls have very limited effects on flows and almost none on exchange rates. 5 There are problems from translating theory into practice. When investors have incentives to take a long position in a given country and some forms of inflows are restricted, they just use other, less regulated, forms of investment. Controls may generate some short- term frictions, but unless they apply to all types of flows their effectiveness is impaired. Indeed, 4 See Cowan and De Gregorio (2007) for Chile and Chamon and Garcia (2013) for Brazil. 5 See Magud et al. (2011) and De Gregorio (2014), chapter 5. 8

9 recent research has highlighted the perils of capital controls in a very liquid global economy. Firms in emerging markets have had high incentives to borrow in global markets, as interest rates are very low or because restrictions on capital flows may create an arbitrage opportunity for firms subject to less stringent controls. This borrowing could be to finance investment or restructure liabilities, or to behave as financial intermediaries. When banks are limited in their ability to raise funds abroad, corporations may borrow and deposit the proceeds domestically. Shin and Zhao (2013) have analyzed this issue for four Asian countries: China, India, Indonesia and Korea. Non- financial firms use cash or borrow in order to invest, therefore their financial assets and liabilities move in opposite directions. In contrast, financial intermediaries borrow to lend and, hence, the correlation between financial assets and liabilities is positive. Using this insight Shin and Zhao (2013) compare the case of the Asian economies with the US. The evidence shows that while in the US the correlation is zero, in all four Asian countries it is positive, and hence, large corporations are surrogate intermediaries. Caballero et al. (2015) address this issue for Brazil and Chile for non- financial corporations during the period , and conclude that in Brazil the correlation is positive, and zero in Chile. The policy implication is that when carry trade is attractive, non- financial corporations with access to global markets could take advantage by behaving as intermediaries. If the problem is volume, and not composition, the restrictions must be broad based, something that in general is not possible to implement. Moreover, most theories refer to problems with excess credit and financial vulnerability, and regarding this concern a more efficient route would be to regulate and limit the activities of the banking industry. This has been indeed the case of emerging markets. This was done in Korea with non- core liabilities in the banking system. But in this case we could expect non- financial firms to borrow, and this measure could be good for financial stability, but it may not reduce the volume of inflows. Finally, I know of no country where one could argue that a cornerstone of successful macroeconomic management during the global financial crisis was capital controls, as is the case, for example, of exchange rate flexibility. One risk with the use of capital controls is that policymakers may not have a proper evaluation on the causes of flows and exchange rate tensions. The use of capital controls may give central bankers the perception that they can significantly raise interest rates without having exchange rate repercussions. But, if the controls have small effects on the exchange rate, severe monetary tightening will cause an appreciation of the currency and large incentives for carry trade. The cause of this appreciation would be the misperception that the currency is immune to monetary policy, and policymakers could end up blaming foreign investors for being eager to bring in capital. 9

10 4. Concluding remarks A natural reaction to a crisis is to think that everything is wrong and that all must be changed. Emerging markets, and in particular Latin America, performed well during the crisis, and above all during the recovery. Therefore, a first lesson must be about the factors that produced these good results, and macroeconomic management was certainly central to the rapid recovery. The financial system was resilient, which shows that the regulatory framework was appropriate. Monetary policies, organized around inflation target regimes, were very successful to provide room for expansionary policies in Latin America. Indeed, although controversy still exists, monetary policy was not the culprit of the financial collapse in advanced economies. Very low interest rates in advanced economies induced high risk- taking as financial institutions were searching for yields. Excess liquidity may sow the seeds for asset price bubbles and financial vulnerabilities, but to cause a huge financial crisis, some serious distortions in the financial system are required. Indeed, countries like Australia, Canada or Norway had very low interest rates, but their financial systems responded appropriately to the financial crisis. Despite their boom in house prices, their financial systems remained resilient (figure 5). The experience in Latin America was even more sanguine since there was no such thing as a large housing boom. Moreover, in some countries there could have been even a housing bubble, but there was not the degree of leverage as in the US, which was central in triggering the crisis. Indeed, it is useful to recall that asset price bubbles do not necessarily cause a financial crisis, as was the case with the tech bubble in the early 2000s. The bad combination is asset price bubbles with high leverage in the banking system. Regarding the interactions between macroprudential policies and monetary policy, I would like to make some final comments. Unfortunately, the distinction between financial and price stability is not simple. As I discussed before, actions of monetary policy, such as the Greenspan put, may create financial instability. The financial crisis also showed that it has effects on the business cycle and, hence, it has implications on monetary policy. Moreover, the transmission channels of monetary policy could be broken down during a crisis. Therefore, the state of the financial system should have implications on the conduct of monetary policy of economies following a flexible inflation target regime since it affects the transmission channels and the business cycle. It is more appropriate to view macroprudential tools as complements rather than substitutes of monetary policy. For example, adjusting capital requirements over the cycle or introducing dynamic provisioning may have effects on the output gap, inflation and interest rates. This is similar to the case of automatic stabilizers of fiscal policy, which of course have implications on monetary policy. Therefore, there should be no conflict between financial stability and price stability tools, as long as financial regulation is 10

11 countercyclical. From the point of view of the inflation target, this is just part of the environment in which decisions have to be made. The advantage of using the interest rate for monetary policy is that we have relatively good knowledge of the transmission channels. Changes in monetary policy interest rates affect the cost of financing, asset prices and the availability of credit. The macroeconomic consequences of changing regulation are less well understood. Tightening restrictions on banks may create disintermediation and move credit to unregulated segments of the market. In addition, the latitude of changes in regulation is much more limited than that of interest rates. A more constructive approach may be to design rule- based countercyclical regulation with a clear purpose to minimize the risk of financial crisis. Nevertheless, we still need to learn more about the interactions between monetary and financial policies. References Brunnermeier, Mark K. and Yuliy Sannikov (2014), International Credit flows and Pecuniary externalities, mimeo, Princeton University. Caballero, Julián, Ugo Panizza and Andrew Powell (2015), The Second Wave of Global Liquidity: Why are Firms Acting like Financial Intermediaries? mimeo, IADB and the Graduate Institute, Geneva. Calderón, César and Luis Servén (2013), Macroprudential Policies over the Cycle in Latin America, in T. Didier and S. Schmukler (eds.), Emerging Issues in Financial Development: Lessons from Latin America, The World Bank. Calvo, Guillermo A., and Carmen Reinhart (2002), Fear of Floating. Quarterly Journal of Economics 177: Chamon, Marcos and Márcio Garcia (2013), Capital Controls in Brazil: Effective? Efficient? mimeo. Cowan, Kevin and José De Gregorio (2007), International Borrowing, Capital Controls and the Exchange Rate: Lessons from Chile, In S. Edwards (ed.), Capital Controls and Capital Flows in Emerging Economies: Policies, Practices and Consequences, NBER and The University of Chicago Press. De Gregorio, José (2014), How Latin America Weathered the Global Financial Crisis, Peterson Institute of International Economics, Washington. De Gregorio, José and Andrea Tokman (2007), Overcoming Fear of Floating: Exchange Rate Policies in Chile, in N. Batini (ed.), Monetary Policy in Emerging Markets and Other Developing Countries, New York: Nova Science Publishers Inc. 11

12 Ghosh, Atish R., Jun Kim, Mahvash S. Qureshi and Juan Zalduendo (2012), Surges, IMF Working Paper WP/12/22. Kamil, Herman (2012), How Do exchange Rate Regimes Affect Firms Incentives to Hedge Currency Risk? Micro Evidence from Latin America, IMF Working Paper WP/12/69. Korinek, Anton (2011), Capital Flows, Crises and Externalities, mimeo, University of Maryland. Magud, Nicolás E., Carmen M. Reinhart and Kenneth S. Rogoff (2011), Capital Controls: Myth and Reality A Portfolio Balance Approach, NBER Working Paper No Peydró, José- Luis (2014), Macroprudential policy and credit supply cycles, Banque de France, Financial Stability Review, No. 18, pp Tovar, Camilo E., Mercedes Garcia- Escribano, and Mercedes Vera Martin (2012), Credit Growth and the Effectiveness of Reserve Requirements and Other Macroprudential Instruments in Latin America, IMF Working Paper WP/12/142. Shin, Hyun Song and Laura Yi Zhao (2013), Firms as Surrogate Intermediaries: Evidence from Emerging Markets, mimeo, Princeton University. 12

13 Table 1. Surges of Capital Inflows (% of GDP) Net Capital Current Change in Inflows Account Reserves LATAM 1990s s Asia 1990s s Europe 1990s s Sources: Ghosh et al. (2012) for net capital inflows and IMF- IFS for current account and reserves accumulation. 13

14 Figure 1 (a) No interactions. Independent policies (b) Real world. Many interactions 14

15 Chile Mexico Peru Figure 2. Exchange Rates (period average=100) (a) Global Financial Crisis Brazil Colombia Chile Mexico Peru (b) Asian Crisis Brazil Colombia Source: Bloomberg. 15

16 Figure 3. Dollarization of Liabilities of the Corporate Sector in Latin America: (percentage, annual average across firms) Argentina Brazil Chile Colombia Mexico Peru Source: Kamil (2012). 16

17 Figure 4. Average and Marginal Reserve Requirements (%) (Percentage) Brazil: average Colombia: marginal Colombia: average Peru: average Peru: marginal m1 2005m1 2007m1 2009m1 2011m1 1 Simple av erage. Brazil: reserv e requirements on sight deposits and term deposits; Colombia: checking, savings, CDs and term deposits; Peru: domestic and f oreign currency deposits. Source: IMF Tovar staf et f al. estimates (2012) on teh basis of central bank data. 17

18 Figure 5. House Prices in Advanced Economies pre- Crisis (percent change from 1996 to pre- crisis peak) Australia Canada Finland Greece Japan New Zealand Norway Spain Sweden Switzerland United States Source: BIS. 18