This item is only available as the following downloads:
STERILIZED INTERVENTION AND RESERVE ACCUMULATION IN TWO SMALL OPEN ECONOMIES BY JUAN ROBERTO ELIAS DEL HIERRO A Thesis Submitted to the division of Social Sciences New College of Florida In partial fulfillment of the requirements for the degree Bachelors of Arts Under the sponsorship of Tarron Khemraj Sarasota, FL May, 2013
ii Acknowledgements First and foremost, I would like to thank my family for always being there for me and for convincing me to never settle, and that I can alway s accomplish anything I set my mind to. I want to thank the Goat House, the Embassy, momma Long, residents of Y 4 123 and many others for their hospitality this past year. I want to thank my professors for always challenging me and providing me with uncondition al guidance and patience I want to thank New College as a whole for being an amazing journey, for molding me in ways I could have never imagined, and for making me be myself at all costs I want to thank the family I built here; those people who will alwa ys stay in my heart and with whom I have shared my most memorable experiences enjoy I want to thank those people who inspired me and opened win dows when doors were being shut: a Mexican fisherman, a Cuban taxi driver, a N icaraguan adult literacy program, an Argentine footballer and my men tor and friend Carlos Gonzalez. This has been a great story, a great adventure. I will hold on to all the memories gathered as much as I can, and I will take new college with me everywhere I go.
iii Table of Content s Introduction ................................ ................................ ................................ ......................... v Ch. 1 Literature Review ................................ ................................ ................................ ...... 4 i. Foreign Exchange Interv ention: The Need to Manipulate the exchange rate ......... 4 ii. The Case for Competitiveness ................................ ................................ .................. 4 iii. Exchange Rate and Inflation ................................ ................................ .................... 5 iv. Exchange Rate Volatility and International Trade ................................ .................. 6 v. Foreign Exchange intervention: Sterilization. ................................ ......................... 7 vi. Sterilization: Dual Anchors ................................ ................................ ..................... 8 vii. ................................ ................ 9 v iii. Stylized Facts on International Reserves ................................ ................................ ... 13 ix. A Quick Review on Currency Crisis ................................ ................................ ...... 14 x. Rethinking the Trilemma in Internat ional Finance ................................ ............... 16 Ch.2 Economic and Monetary History of Peru and Colombia ................................ ......... 21 Ch. 3 Methodology ................................ ................................ ................................ ........... 34 Coefficients and their meaning ................................ ................................ ................... 34 Data acquisition ................................ ................................ ................................ .......... 35 Results ................................ ................................ ................................ ......................... 38 Research Limitations ................................ ................................ ................................ .. 43 Ch. 4 Further Specifications on the FOREX Intervention Mechanisms in Peru and Colombia ................................ ................................ ................................ ........................... 49 Colombia ................................ ................................ ................................ ..................... 49 Peru ................................ ................................ ................................ ............................. 53
iv Conclusion ................................ ................................ ................................ ........................ 58 Further Re search ................................ ................................ ................................ ........ 59
v STER ILIZED INTERVENTION AND RESERVE ACCUMULATION IN TWO SMALL OPEN ECONOMIES Juan Roberto Elias Del Hierro New College of Florida, 2013 ABSTRACT This thesis is an inquiry into the recent trend of EMEs Internati onal Reserves accumulation. As it has been noticed by the academic community, EMEs have been accumulating massive amounts of reserves since the 90s for many different reasons. As this thesis suggests, the main two reasons for countries to int ervene in the FOREX market are cautionary and m ercantilist. The Monetary trajectory and mechanisms of two small open economies of equal size and region (Peru and Colombia) will be studied; and their respective sterilization coefficients will be found in order to quantif y the research. Both countries manifest high degrees of partial sterilization, which means the coefficients were barely lower than 1 Sterilizing most of their intervention helps the country maintain a stable monetary base, buffering against inflationary r isks. The high degree of intervention and the high degree of sterilization suggest that both cou ntries engage in dual anchors: exchange rate volatility, and the money s upply.
1 Introduction Economic stability is a main criterion for the development parad ox. Emerging economies need to follow a trend of sustainable growth and avoid economic shocks that might render their development path to ashes if they want to escape their state of underdevelopment and low standards of living. Historically, the global s outh has been a victim of international shocks that disrupted their growth and mitigated the gains achieved during stability. For Central Banks in the region, it is a main priority to foster stability; through the decades they have tried to dodge recession s and depressions by following different monetary practices. These policies might have given good results on the short run, but ultimately failed after world events they could n ot upset disrupted their economies Currently, it is generally accepted that monetary stability requires foreign asset accumulation, and EMEs have been employing this policy accumulating record amounts of international reserves through foreign exchange sterilized intervention. In this thesis, we first attempt to identify the theore tical justifications for this monetary policy strategy, and then we will compare and contrast the monetary trajectory and current intervention mechanism of two similar emerging economies (Peru and Colombia) from the same region. We will also find the steri lization coefficients for both countries in an attempt to gauge the extent to which these countries sterilize and to see how this tendency has changed during the past decade. The regression will be conducted from January 2001 through December 2012 with mon thly data, and it will be a similar regression used by other studies such as Aizenman and Glick (2008) to find the sterilization coefficient.
2 websites. According to t he research, EMEs accumulate reserves for two main reasons: cautionary and mercantilist. Cautionary in the sense of buffering external current and capital shocks, and mercantilist in the sense of maintaining a competitive exchange rate in comparison to the developed world. Sterilizing the intervention mitigates the effect that an expanding MB might have on inflationary risks since an open purchase of foreign currency floods the market with domestic currency, increasing the monetary base. Another main aspect the research suggests, countries use their reserves as a fourth variable (making it a Quatrilemma), which allows them to take advantage of all three monetary strategies to a certain ex tent. My regressions results suggest a very high sterilization of most incoming capital inflows. Colombia appears to have a slightly higher coefficient than Peru, but both nonetheless sterilize heavily. Hypothesis for the coefficients will be presented in Ch. 3. economic collapse. A dummy variable will be present for the months leading after the Lehman Bros. collapse. surpluses in Peru, and previous monetary shocks and hyperinflation. In terms of similarities: Both countries have a FIT (managed float with inflation targeting, which
3 suggests dual nominal anchors), similar GDP, geographical location, and they are both commodities export led economies.
4 Ch. 1 Literature Review i. Foreign Exchange Intervention: The Need to manipulate the exchange rate Foreign exchange intervention occurs whe n monetary authorities (central banks or c entral governments in some cases) buy or sell foreign currency with the home c urrenc y in order to affect the exchange r ate as desired (Taylor 1995) When a central bank buys foreign currency with domestic currency, it is increasing the supply of d omestic currency while lowering the supply of the foreign currency. A higher supply of domest ic currency drops its relative price in comparison to the foreign currency, hence depreciating the exchange rate 1 or requiring more units of domestic currency to buy the same units of the foreign one. This is a useful tool for central bank s who might need to alter the exchange rate since the exchange rate affects major macroeconomic indicators such as price stability, domestic financial stability, economic growth, a nd external/internal balances. ii. The Case for Competitiveness One of the main reasons for cen tral bank s to intervene in the Forex 2 market is to maintain a competitive exchange rate (undervalued). Historical narrative of policies in different regions (especially in Latin America) and empirical studies support the idea that a depreciated exchange ra te tends to foster economic growth (Frankel and Rapetti 2011) A country whose exchange rate is undervalued attracts investment since the goods they produced are, in comparison, more affordable than the goods from the other country, 1 Exchange rate is the amount of home currency needed to buy o ne unit of foreign currency. 2 Foreign Exchange/Currency Market.
5 allowing the nation to have a trade and current account surplus. Many countries have engaged in this practice since the 90s, China being one of the precursors and main currency dev aluer. iii. Exchange Rate and Inflation Another reason for exchange rate intervention r ests on the strength of the exchange rate pass through ( ERPT ). ERPT is the change in the price of imports that occurs after a change in the exchange rate (Goldberg and Knetter 1997) There exists conventional wisdom that ERPT is higher in emerging economies; however, that view by authors such as Taylor (2000), Cunningham and Haldane (1999), who argue that after the 90s most countries noticed that exchange rate fluctuations had a much lower impact on the general level of prices, hence a lower importance of the ERPT (Winkelried 2003) However, an ERPT is still theoretically plausible especially in emerging economies, for this reason monetary authorities always keep it in mind when design ing monetary policy. A country with a high ERPT is more likely to suffer inflationary volatility if the exchange rate is volatile. Countries whose inflationary levels are strongly dependent on the exchange rate (Emerging E conomies) need to avoid volatile exchange rate s since a depreciation of their currency can lead to a general increase in import prices, hence inflation. Although inflation was initially seen as an output driver back in the Keynesian years (Taylor Rule), the relationship between inflation and unemployment is only prevalent in the short run. A more detailed argument regarding low inflation will be made in a la ter section of the Lit Review.
6 iv. Exchange Rate Volatility and International Trade An exchange rate that is highly volatile possesses a higher risk than a more stable exchange rate trade. In theory, a higher exchange rate volatility leads to higher costs of transaction internationally and uncertainty. If firms are uncertain of the exchange rate they will tend to lower the volume of international transactions with the given country since the uncertainty hinders their capabi lity to measure profitability Traditionally, a fixed exchange rate might be a preferable regime for importers and exporters, followed by a regime that focuses on maintaining exchange rate volatility to a minimum ; however after the collapse of the Bre t t on Woods system, it was clear that fixed rate regimes hinder economic efficiency and made countries vulnerable to speculative shocks when capital mobility was present Adding more to it, fixed exchange regimes prevent the exercise of monetary policy by the central bank since it requires a currency board A floating system; on the other hand, allows the central bank to retain autonomy and exercise monetary policy. Most empirical research has found it difficult to assert the assumption that volatility in the exchange rate hinders international trade (Ozturk 2006) As Cote (1994) points ou trade, this review reveals that the effects of volatility are ambiguous (Cote 1994) v. Intervention and the Dutch disease Both economies studied in this thesis have a severe risk of Dutch disease because of their commodities export nature. A Dutch disease occurs when the demand for a money demand appreciating
7 their currency and their exchange rate. Both Peru and Colom bia have an inherent necessity to intervene and sterilize because of this reason. If t sterilize at a time when thei r economies are expanding due to the world demand for their commodities, there is a chance that their manufacturing sec tor can be negatively affected by the appreciation of their currencies. adversely in both ways: deindustrialization and dependency in a very volatile industry ( primary exports ). The appreciation of the domestic currency m akes the manufacturing and services sector lose competitiveness. A central bank faced with a high risk of Dutch disease will need to intervene in currency. If they fail to d o so, the economy might end up losing competitiveness in manufacturing, and jobs will shift to the primary export industry. Colombia, to a much larger extent, considers the Dutch disease risk in their policy making. Their central bank has advocated for mor e aggressive sterilization in order to cope with the rising prices of commodities. Another approach, such as building the competitiveness of the manufacturing and services sector can also be used to cope with the Dutch disease, but that is a fiscal policy issue, not monetary. vi. Foreign Exchange intervention: Sterilization. There are two different types of Foreign exchange intervention: Sterilized and non sterilized intervention (Taylor 1995). Sterilized intervention occurs when the monetary authorities make sure that the monetary base (MB), which is a main
8 determinant of the money supply, is not affected by the foreign exchange intervention (Dominguez and Frankel 1990) Non Sterilized intervention aims at affecting the exchange rate by a change in suppl y of home currency and the demand for foreign currency, affecting both the exchange rate and the money supply. Sterilized intervention is a process of two transactions: one is needed to affect the exchange rate and one to ect the intervention had on the money supply (Obsfeld 1996) vii. Sterilization: Dual Anchors If a central bank intervenes plenty in the foreign exchange market ( de facto pegged exchange rate ), and the n s terilizes all said interventions, it can be establi shed that the bank h as dual anchors: One being the exchange rate, and the other one being the
9 monetary b ase. The MB is an important target for central bank s because it is an important determinant of inflation. This is an im portant attribute of a central ba nk implement an inflation targeting regime with adjusting exchange rate s. As previously expressed, the M B is a main determinant of the money supply, and in turn, the money supply is a main d eterminant of the price level (i nflation). According to the Quantity Theory of Money, the MS has a direct relationship with the Price level (Fisher n.d.) Consider the equation of exchange: M*V = P*Q, where M is the total amount of money in circulation, V is the velocity of money, P is the price level, and Q the real value of final expenditures. If M or V is increased, either P or Q has to increase to maintain the identity. If the v elocity of money is constant, and the real value of all final expenditures remains constant as well, and increase in M will ulti mately lead to an increase in the price level P. viii. Sterilization and the Central Bank As previously stated, central bank s st erilize to mitigate the effect f orex intervention has on the money supply. In order to sterilize, most central bank s utilize the domestic assets (Domestic Credit) and foreign assets (International Reserves) they posses to make sure they offse t the effect the previous OMO (o pen market oper ation) had on the money supply (Feenstra and Taylor 2009) C entral B ank Simpl ified Balance Sheet
10 Assets Liabilities Foreign Assets (Reserves) Money Supply (Currency in circulation) Domestic Assets (Credits/Bonds) imagine a given central bank wants to maintain a pe g to a given currency, or simply manage the exchange ra te they want to make sure th e MB stays the same because of i nflationary cautions. If the exchange rate depreciates, which means less domestic currency is needed to buy foreign currency; it poses a th reat to countries that want to maintain a competiti ve (undervalued) home currency (Frankel and Rapetti 2011) In this instance, the central bank will need to intervene in the f orex market by buying foreign currency in order to appreciate the exchange rate back to competitiveness. The central bank will engage in two transactions: 1. An open market purchase of foreign currency, which will end up increasing f oreign assets (foreign reserves) and increase the supply of domestic currency in c irculation (Liabilit ies of the central bank ). This change in the balance of Foreign /Domestic currency will appreciate the exchange rate back to competitiveness because the price for domestic currency will be lower in comparison to the price of foreign currency. Let s assume t he central bank purchased 1000 dollars of foreign reserves: Keep in mind that a Balance Sheet must always maintain the alance Sheet, the identity is FA + DA = MB. The previous transactions do not violate the identity; it is assumed that the M increased by the same amount than the FA
11 Balance Sheet Movements Assets Liabilities Foreign Assets 1000 Monetary Base (1000) Domestic Assets 2. The second transaction is needed in order to tak e the monetary b ase back to it s origi nal amount, leaving the money s upply untouched. The central bank will sell domestic assets in order to ase. Assets Liabilities Foreign Assets 1000 Monetary Base 1000 1000 Domestic Assets ( 1000 ) The net resul t of both transactions will be: Same amount of monetary base (liabilities) Lower amount of domestic assets. Higher amount of international reserves. An appreciated exchange rate hopefully back to competitiveness. It is important to mention that ste rilization is not limited to only offsetting effects on the MB after an intervention aimed at manipulating the exchange rate It can also be used to off set changes in the MB after an intervention to finance government borrowing:
12 In that case, the central bank will print money and buy domestic assets (Bonds). This purchase increases the MB, which was used to buy the domestic bonds. In order to bring the MB back down, the central bank will sell FA (International Reserves) to buy back the cashed issue to fin ance the first transaction. The net effect of the transaction is: Higher Domestic assets Lower Foreign Assets Same amount of MB Appreciation of the Home currency. There exist some cri tiques on the effectiveness of f orex intervention and on the effective ness of sterilization. Taylor argues that there was consensus amongst economist that sterilized intervention had only a very small and transitory effect on the exchange rate ; however, much of the research has not been able to establish the inefficacy of ex change rate intervention, hence the heavy and regular intervention in the major exchange rate markets since the late 80s (Taylor 1995). An important observation made by Kenen (1982) specifies the need for imperfect asset substitution between domestic ass ets and foreign assets. Were DA and FA perfect substitutes, right after the open market sell of DA (2nd transaction) private agents will sell Foreign Bonds until the value of the Foreign Bonds is equal to the value of the Domestic Bonds, inundating the eco nomy with foreign currency equal to the amount of the 1st transaction, mitigating the effect on the exchange rate to 0. Domestic and Foreign bonds need to be less than perfect substitutes in order for the intervention to have an effect on the exchange rate The exchange rate determination Model that assumes imperfect substitutability between foreign and domestic assets is the Portfolio Balance
13 Model, and it is distinguished from other exchange rate models mainly by the fore mentioned assumption. As Mundell (1968) explained, sterilized intervention has been inefficient in advanced countries, and has had more effectiveness in developing ones. One reason for this being that bonds and equity from EMEs are not good substitutes for international financial assets s ince th ey pose a higher risk premium. ix. Stylized Facts on International Reserves The two economies studied in this case have expanded their international reserves exponentially since the 90s. This is not a unique occurrence since several EMEs (Emerging Ma rket Economies) have also expanded their reserves exponentially after almost depleting them during the crisis years of the 80s. There is an inherent need in having reserves for monetary purposes: a country who wants to hold a peg needs to have reserves in order to intervene in the exchange rate market and maintain the pe g. In the same way, EMEs have been hoarding reserves not only for the aforementioned, but also in order to prevent current and capital account shocks to their economies. During the 80s, it w as recommended by the IMF to hold reserves equal to three months of imports (current account); however, the Mexican crisis demonstrated that reserves also need to be sufficient to pay off all short term debt (capital account). This rule became known as the Greenspan Guidotti rule. The rule is important because insurance for current account shocks requires fewer reserves than insurance against capital account shocks because capital account shocks revolve around a continuous fligh t (Ghosh, Ostry and Tsangarides 2012)
14 x. A Quick Review on Currency Crisis A currency crisis occurs when a currency depreciates rapidly and violently, making it impossible for those who hold said currency to maintain the value of their wealth. Those who ha ve liabilities denominated in foreign currency, but assets on domestic currency tend to lose money on a currency crisis since their debt increases as the domestic currency (assets) depreciates in comparison to the foreign currency (liabilities). An economy with a fixed exchange rate has to act quickly during a currency cris is: they can either defend the pe g by fueling the market with th e I nternational Reserves they have, increase domestic interest rates, or leave the fixed exchange rate regime and begin to float. Currency crisis are especially detrimental to small emerging economies since they rarely possess sophisticated Banking systems or enough reserves to counteract the depreciation (Glick and Hutchinson 2011) Several countries suffered recessions as a product of currency crisis during the 90s, the most notable were Mexico (1994), Argentina (1999 2002), Russia (1998), and capital account shock they endured during 1994 In this crisis, the failure of the Mexican banking system to provide liquidity had a spillover effect on the balance of payments as investors took their money out and got rid of all Mexican currency they possessed, this led to an exponential depreciation of the Mexican currency who at the time was trying to maintain a pe g on the dollar. To appreci ate their currency back to the pe g, the Mexican central bank engaged in massive purchases of Mexican Pesos, getting rid of most of the international reserves the y had accumulated in the previous years. The new administration of the Mexican central bank decided to let the exchange rate float since
15 pegged at 3). The USA interven ed to help the Mexican economy and the American business interests in Mexico in two ways: by buying pesos, and with 50 billion USD in loan guarantees. The dollar stayed at 6 pesos per USD. A main implication of the Mexican crisis was the need for higher s ums of reserves; reserves not only sufficient to cover for a current account shock, but a capital account shock as well. Countries have learned from the experience of Mexico and other EMEs, and have, since the 90s, accumulating international reserves for p recautionary motives. Another reason countries have been accumulating massive amounts of reserves is a result of devaluing their currency for competitive reasons, or as Dooly (2003) called it: Modern Mercantilism (Dooley, Folkerts Landau and Garber 2003) Countries engage in sterilized intervention in order to depreciate their home currency, making their home goods cheaper to foreign markets. This is an export led strategy that many EMEs have used to create current account surpluses, expand production, an d lower unemployment. Following the Mexican crisis, the Peso was undervalued at 6 per USD, and by 1996 and 1997 the economy was growing rapidly at healthy rates of 5.1 and 6.8 respectively. Devaluation of the Mexican currency was one of the main players in allowing Mexico to regain its competitiveness internationally (Lederman, et al. 2000) This shows how it can Most countries engage in s terilized intervention to maintain the exchange rate competitive. Ste product is accumulation of foreign currency. Notably, Ghosh and Kim (2009) explain how the monetary authorities can see the
16 deprec iation of the currency through s terilized intervention as an export subsidy in an economic system wh ere productivity spillovers are positive, leading to an expansion in international reserves (Ghosh, Ostry and Tsangarides 2010) This expansion occurs because s terilized intervention buys foreign currency in the open market in order to lower its supply. Th is phenomenon can help understand why countries have increased rapidly their International Reserves. Based on the theory, it is appropriate to think that countries are expanding their international reserves because of both precautionary and mercantilist mo tives, however, the empirical literature to date has not been successful at establishing a strong link with the data available (Gosh; Ostry; and Tsangarides 2012.) xi. Rethinking the Trilemma in International Finance The international financial architecture has changed many times, especially after a time of global crises. After most nations adopted the fiat currency, different monetary and financial arrangements had to be made while keeping in mind the Trilemma. The Mundell Fleming extension of the IS LM mod el for open economies was a Neo Keynesian model which put forth t he theory of the unholy trinity which had vast relevance in the crisis years leading to the 1990s. The Trilemma became an important foundation for open economy macroeconomics during the 8 0s (Obstfeld, Shambaugh and Taylor 2004) The Trilemma states that an economy has to choose two monetary policy goals out of the three: Fixed exchange rate m onetary independence, and financial integration. A mix variety of empirical work exist on the is sue of Policy goals, leading to the idea that each policy choice can have varying effects, multiplying those effects depending
17 on the other goal the central bank decides to follow. Authors such as Aizenman and Ito (2008, 2010, and 2011 ) analyzed the trilem ma the way it has been applied in practice: a generalized adaptation of all three sides of the triangle. As it is explained, countries means, countries tend to choose betw een an index of financial integration, a measure of volatility in the exchange rate and the degree of monetary independence. To further expand the idea above, a central bank can simply decide to employ all three policy goals, but it would have to do it in small adaptations. It will be impractical and risky for a central bank to have a fixed exchange rate full financial integration, and complete monetary autonomy; however, said central bank can have limited financial integration, a managed float to reduc e volatility, and still maintain some sort of monetary autonomy. This circumvention of the Trilemma has important implications for EMEs success since the benefit of optimal policy choice is of extreme significance for their economies. Regression analyses, done by Aizenman and Ito (2012), have shown the trend of EMEs to converge in an intermediate level of all three policies (Aizenman and Ito 2012) This means that most EMEs are engaging in intermediate levels of financial integration, exchange rate s contro l, and monetary autonomy. The analysis also shows that those countries that converged on the three policy goals tend to experience lower indices of output volatility. It is important to note that while EMEs have been defying the Trilemma, they have also be en increasing their International Reserves, providing with a hypothesis: IR are being accumulated in order to cushion against the risk that rises from the circumvention of the Trilemma. EMEs that are not pursuing the three intermediat e
18 policy goals tend to have higher output volatility, as those economies usually pursue the policy goals of exchange rate stability (Aizenman and Ito 2012) For EMEs, accumulation of IR has been a byproduct of their policy goal of financial openness, financial stability, and economic stability (reduced volatility of exchange rate ). Aizenman and Marion (2003) explained that insurance motives might be behind the accumulation of IR by Asian countries, since the accumulation of reserves escalated in that region after the Asian cri ses. This observation, coupled with empirical research on the subject of Trilemma choices and IR accumulation, has led to the idea that monetary policy goal (Aizenman and Ito 2012). Furthermore, empirical analysis shows that a country who is pursuing a combination of policies with high divergence, and has low levels of IR, tends to suffer greater output volatility. It can be hypothesized that accumulating good amounts of r eserves allows EMEs to have more freedom and options when choosing the policy combinations they want to establish. The cost of maintaining monetary stability and at the same time hoard international reserves has proliferated the need for more aggressive st erilization. Authors such as Aizenman and Glick (2008) have found that as IR were building up, so were the sterilization coefficients, which means countries kept increasingly sterilizing the increments of the MB which were byproducts of open market purchas e of foreign currency (Aizenman and Glick 2008) The following graph illustrates the trends of the 4 variables: Monetary Independence, Financial Integration, exchange rate stability, and International Reserves. We ca n observe how, although not the exten t of the Emerging Asian Economies, Emerging Latin America has been increasing the accumulation of International Reserves
19 since the 80s. It can also be observed that as IR increased in EAEs, Monetary Independence decreased (Aizenman and Ito 2012 ) Figure 1.1: Illustrating the Quatrilemma in Emerging Latin America Source: Aizeman, et al (2010) Figure 1.2: Illustrating the Quatrilemma in Emerging Asia
21 Ch.2 Economic and Monetary History of Peru and Colombia The economies observed in this thesis have many fundamental factors in common such as: GDP, Population, geographical location, Monetary Regime (FIT), and they are both commodities export led economies. This chapter outlines the development of th eir monetary and economic system in the context of the Latin American economic also affected these economies (For example; Las FARC and drug cartels in Colombia had a main impact in their economy) but their study is beyond the scope of this thesis. Peru In p ost WWII Peru military juntas and dictatorships controlled the economy and dr ove it to economic turmoil, currency crises, and hyperinflation During the 1950s, a surplus in government revenue from the export of natural resources allowed the populist government to engage in social reconstruction programs. Th e policies of the next 30 years which revolved around import substi tution industrialization (ISIs) proved to have no effect on the size of the Peruvian economy. Later in the decade Peru signed an agreement with the IMF and became the first Latin American country to do so. Thi s agreement consisted in negotiating exchange rate s policies with the IMF in exchange of external finan ce since at this point in time only the IMF was a substantial medium for (Frenkel and Rapetti 2010).
22 In 1975, General Bermudez replaced general Velasco who at the time was major industries, and the faile d land reform implemented by him (Sheahan 2001) During the 70s, Oil prod integrating the financial markets, inundating them with liquidity, and dropping interest rates worldwide. This source of cheap loans, liberalization and progressive de regulation of dome stic financial systems convinced Latin America to tap into the Eurodollar market. Peru was one of the first countries to engage in international credit to finance for government expenditures in infrastructure and balance of payments deficits. This severe i ndebtedness and bo rrowing in foreign currency which most Latin American countries engaged in will be a recipe for disaster leading to the Latin America Debt Crises. democratic e lections brought an end to the military rule in 1980. Belaunde was elected president in the midst of inflation stress, economic difficulties, and the rise of the Shining Path guerrilla. The Latin American debt crisis was a huge problem for the liberalizati on that president Belaunde was trying to bring to Peru. The crisis was a product of the oil embargo in 73 in which oil prices went up abruptly driving economies into recessions and cutting the supply of loanable funds to Latin America. As per capita inco me declined, Argentina, Brazil, and Mexico, were hit by the debt crises also. During these crises foreign lenders ceased the supply of liquidity to public and private instit utions in Latin America because no more Oil dollars were coming in As world interest rates rose, countries were unable to pay for the debt and this triggered a financial crises. As the FX
23 markets deteriorated appreciating the value of the dollar in comp arison to other Latin American currencies countries were not able to pay some of the debt back because it was dollar denominated. During the remainder of the 80s, most countries began refinancing their debt with the help and regulations of the IMF (Frenk el and Rapetti 2010.) Alan Garcia Perez became president of Peru in 1985. His approach to the economic problem Peru was going through was placing limits on external debt payments and expansion of public expenditure, angering the IMF and other financial i ntermediaries who had provided funds for Peru initially. His policies led to the Peruvian Crisis of 1986 (Parodi 2000) When President Garcia took office, Peru was in very unfavorable economic and social conditions. The country was going through severe te rrorist attacks from the Maoist spending than through the printing money (the inflation tax). Domestic credit doubled administration instituted a fixed exchange rate ; however, the Peg could not survive the exponential expansion of domestic credit, fiscal deficits, and the drain of international reserves. Reserves fell from 2,000 Million USD in 1986 to 500 Million USD in 1988. In April 1988, the Peruvian monetary authorities decided to give up on the fix exchange rate and the Peruvian currency (the sol) be gan to float, depreciating rapidly. The exchange rate went from 250,000 Soles per USD in September to 1,200,000 in March 1989. Inflation reached 7,649% in 1990, a period known as hyperinflation (Taylor and Feenstra
24 inflation accumulated a total of 2,200,200%. By mid 1985 3 which was replaced 20%, the average income per capita dr opped to 720 USD (well below the average income in 1960), poverty increased by more than 10%, and Peru was in bad terms with the IMF (Pastor 2012) 4 Figure 2.1: Peru GDP Source: ww w.tradingeconomics.com/theworldbankgroup The 90s was characterized by high liquidity and low world interest rates. As developed countries embraced the Neo Liberal agenda, Latin America had no other who were the ones who provided debt relief and other instruments for their recovery. The 3 indigenous peoples of the Andes, and was the official language of the Incan Empire. 4 Studies done by the National Institute of Statistics and Informatics and the United Nations Development Programme.
25 demand for developing world financial assets, fuelin g capital from the first world into EMEs (Calvo, Leiderman and Reinhart 1993) Massive amounts of capital inflows occurred during the 90s. Peru bega n the decade of the 90s with hyperinflation, a stagnant economy, and social problems that originated in th e 80s. The lack of inflation adjusted instruments led the public to seek foreign cu rrencies especially the dollar as an important medium of wealth preservation leading Peru to become the most dollarized country in the region (Rossini, Quispe and Rodri guez, Capital Flows, Monetary Policy, and Forex Intervention in Peru 2011) The monetary policy followed during the early 90s helped curve inflation down significantly even though they did not engage in a fix exchange rate regime. After the Peruvian centra l bank gave up on the nominal Peg in 1988, they began targeting monetary targets and managing the float. Peru stopped the indiscriminate printing of money and this led to a Real exchange rate appreciation because the rate of inflation was significantly hig her than the rate of depreciation of the nominal exchange rate This in time helped mitigate inflation even further (Dacourt 1999) The Banco Central de Reserva del Peru ( BCR ) maintained this regime until 2002, relying on sterilized FX interventions to con trol the monetary aggregates and the nominal exchange rate This approach to Monetary Policy proved to be very efficient at accomplishing two things: Reducing the inflationary trends, and avoiding recessions. The economy expanded moderately from 1994 to 20 02. Mexico, Argentina, and Brazil went through some really bad crises d uring the 90s, but Peru did not. H owever, Peru did suffer from the Asian and Russian crises regardless of the flexibility of the exchange rate the nominal exchange
26 rate depreciated and capital flew the country once again, leading to a financial crisis in 1999. After 2002, Peru engaged in pure floating and inflation targeting regime (FIT). At first, the central bank still targeted monetary aggregates mainly, then in 2003 switched to in terest rate and sterilized intervention. Although Peru had adapted to FIT, they did not fully allowed the exchange rate to float freely (Chang 2008) Not only Peru, but most FIT countries in Latin America have declared the need to intervene in the FX marke t in order to manipulate the exchange rate Peru has become the most dollarized country in the region, and monetary policy is aimed at mitigating the fluctuations of the nominal exchange rate which can affect the behavior of the financial system and inves tment. Because of the heavy dollarization of Peru, the reaction of financial agents to sudden movements in foreign liquidity and volatil it y in the exchange rate is magnified (Rossini, Quispe and Rodriguez 2011) The interventions have accumulated reserves enough for the central bank to be a lender of last resort in the case of bank runs (Dancourt, 2009 and Armas and Grippa, 2006). The volatility of the Peruvian exchange rate is the lowest of the region since 2009 (BCRP.) Most of the monetary emissions that the central bank creates is sterilized through the placement of Deposit Certificates (CDBCRP) and with treasury deposits from the fiscal sector into the central bank It is im portant to note that credit to the private sector did not shrink during the 2008 crises, on the contrary, it expanded by 13% since September of 2008 until the end of 2009, and still to this date credit keeps expanding. This credit access was the highest ra te of the Latin American region. The BCRP acted
27 promptly in providing liquidity to the private sector in both domestic and foreign currency, preventing the domestic sector from being hit by the credit restriction from the international markets. The liquidi ty injected to the financial sector by the central bank was 9% of GDP. An important aspect of the Peruvian economy is the fiscal surplus. On August 2012, Peru recorded a fiscal surplus of about 7% of GDP. The fact that Peru can depend on the fiscal secto r for liquidity is a central aspect for this paper: Because Peru has a fiscal surplus, they can depend on the fiscal sector for cheap sterilization bonds. This allows the central bank to engage in sterilization more efficiently, without increasing its debt Figure 2.2 anco C entral de R eserva del P eru )
28 Figure 2.3 : Source: www.tradingeconomics.com/institutonacionaldeestadistica *A s we can see in the graph above, the price volatility in Peru was pretty high during the 90s. During the 2000s, the v olatility ceased. Colombia production of coffee. Back i however, by the 1920s it made up about 75%. Several inflows of capital had public and private destinies thanks to the strong performance of the commodity (Steiner and Vallejo 2010) as an export. Coffee had an important social impact as well: Coffee was cultivated in small owned farms, unlike mining and banana plantations. This aspect of the coffee industry created a class of small successful business owners, which in time will create a need for people to settle in different are a s of Colombia equally, decentralizing population
29 barely affected by the Great Depression of the 30s. After WWII, Colombia began constructing a more int egrated economy thanks to the profits that the coffee industry brought. Since the 1950s, the Colombian model for development was ISI (Import Substitution Industrialization), prioritizing domestic consumption of imports. Just like Peru, Colombia signed an agreement with the IMF that dealt with exchange rate and international borrowing. During the 1960s, Colombia managed to control their balance of payments issues with the use of a crawling peg, which was a very innovative move at the time, and only used in Argentina, Chile, and Brazil. During the implementation of the crawling peg, monetary authorities adjusted the peg an average of once a week. The crawling peg was used not only to provide Real exchange rate stability, but competitiveness in exports as well Colombia experienced high economic growth after the implementation of the crawling peg regime. After 1976, primacy was given to the promotion of exports, especially in the manufacturing sector and coff ee derivatives (Steiner and Vallejo 2010). During t he 70s, the price of coffee significantly. The high demand and price for coffee doubled its production in one decade. The expansion of exports and revenues helped Co lombia build international reserves, while the central bank did all it could to stop inflation from rising. During this time of economic expansion, inflation was averaging 20% per year (Hernandez Gamarra 2001) The central bank then decided to begin imple menting a entral B ank but as an intermediary as well.
30 Source: www.tradingeconomics.com/theworldbankgroup Colombia was an exception to the Latin Am erican Debt crises of the 80s. Because of the coffee boom during the 60s and 70s, Colombia did not tap the International Market for Loanable funds to the same extend as other Latin American countries who suffered from the crises. However, the price of coff ee did go down, taking Colombia to a current account deficit. After the central bank begins taping on its foreign reserves because of the balance of payments problem, they decided to ease the restrictions on foreign borrowing, and filled their economy with liquidity. The government of Betancur implemented a plan of stabilization, which tried to correct the fiscal and external imbalances (Hernandez Gamarra 2008). Colombia was the only major Latin American country that had a positive GDP growth rate through t financial needs to overcome the crises, the aid that was given to it by the IMF and World Bank, the stable exchan ge rate and the inflation suffered after the devaluations during the early 80s. Towards the end of the 80s, Colombia went through major economic changes,
31 most of them aimed at liberalizing certain accounts, trade, and foreign exchange (Steiner and Vallejo 2010). During the 1990s, Colombia faced tremendous amounts of social problems, most notably, drug cartels and the FARC. The economic changes the country made were mostly aimed at fiscal decentralization and progress towards social reconciliation and pea ce. The new constitution of 1991 made the central bank independent, establishing the rule for a low and stable rate of inflation (Hernandez Gamarra 2001). The favorable conditions of the international environment worked well with the new economic and socia however, in 1996 these two policies became inconsistent (Steiner and Vallejo 2010.) For most of the first half of the 90s, Colombia enjoyed significant economic success asi de from the social chaos and violence that haunted the country. Leading to the second part of the 90s, higher spending led to deficits, which in turn, led to public debt, making the country vulnerable to negative international shocks. During the Russia and Asian crisis, Colombia had their first recession since the 40s. The central bank had to devalue the exchange rate because it came under international pressure. A large contraction in aggregate demand occurred as a byproduct of the adjustment to the curren t account because of the lack of international lending. The contraction in demand r esulted in a contraction in GDP, which was accompanied by a real state bubble. All of the sudden, a banking crises erupted. The government tried to help the banking sector b ut that required more public spending. Some banks had to close and some were taken over by the government. In 1999, the government and central bank made the decision to allow the exchange rate to float freely, without any kind of government intervention. H owever,
32 to the point of default (just like the case of Mexico). Colombia sign ed a pact of extended fund facility for three years. With this pact, confidence in the peso was restored and that confidence allowed the central bank to focus on inflation control and fiscal spending sustainability. With the pact, the government increased taxation, reformed the pension system, and privatized the banks it had previously natio nalized. Colombia began its recovery by early 2000s with the help of the competitive exchange rate which drove exports. The Colombian economy moving into the new millennia saw the growing importance of remittances, which increased from 745 million USD in 1 996 to 3 billion industry. The Colombian peso kept going through some devaluation throughout the 2000s because of the floatation of the exchange rate but in 2004 and 2007 it appreciated. This volatility obligated the export sector of the Colombian economy to limit the volume of business they were providing. The business sector then put pressure on the policy makers to mitigate the uncertainty of future exchange rate s. This pressure demonstrated that the exporters were not ready for a fully floating economy, since they were not using financial coverage or exchange rate insurance. Net foreign reserves in terms of debt payments increased substantially, increasing the credibili ty and lowering the chance of a balance of payments crisis Today, Monetary Policy in Colombia has the objective to reach and maintain a low and stable inflation rate. The Banco de la Republica defines the quantitative inflati on targets. The Departamento Administrativo Nacional de
33 Estadisticas (DANE) is responsible for formulating the measurements of the CPI. The main monetary tool used by the central bank is the manipulation of interest rates, which dictate the liquidity in the economy. In the same way, the central bank has objectives regarding the exchange rate The central bank must maintain a good number of reserves, which should be enough to fight current and capital account shocks. The exchange rate should also not be to o volatile, and should be moderate in appreciation and devaluation because the nominal exchange rate can affect future inflation targets (Banco de la Republica de Colombia n.d.) Figure 2.5 Source: www. tradingeconomics.com/DANE
34 Ch. 3 Methodology To quantify our study on the Sterilization patterns of both Peru and Colombia we proceed to estimate the extent to which they Sterilize. In order to estimate t he extent of sterilization, we must find the st erilization coefficients of both Peru and Colombia, and for this we need to run two separate simple regressions. There are different ways to run regressions to find the sterilization coefficient in this case, we shall use the simple regression utilized by Aizenman and Glick (2008), with two major, but fundamentally insignificant changes: Aizenman and Glick use quarterly data when running their regression, we shall use monthly data. Also, Aizenman lags the MB 12 months (4 lag the MB beca use we will assume that the central banks do not wait for the MB expansion to affect inflation, they simply sterilize right after they purchase FA s I will also include a dummy variable to control for the 2008 financial crisis. Coefficients and their mea ning To find the coefficients we will regress the central bank change of the The change w ill be measured monthly We will use the Ordinary Least Squares method (OLS) to run our regression. The equ ation is as follows: FA / MB + Dummy
35 The sterilization coefficient is the term. The values of should range between 0 and 1. A coefficient of 1 implies full sterilization of any incoming unit of FA, a coefficient of 0 means no sterilization at all. If the value found is between 1 and 0, then partial sterilization is assumed. A sterilization coefficient tells us on average how much each country sterilizes. The advantage of utilizing monthly data is that we can find coef ficients for short periods such as two years. It can be implied that the sterilization coefficient from 2001 2002 might not be the same than the coefficient from 2008 2009. Another important observation will be the event that < 1, if this happens it means that not only are the monetary authorities sterilizing for changes in the MB as product of FA purchases, but also mitigating any natural expansion of the MB as a product of an increase of money demand, in one word: contract ionary policy. In the same way, a sterilization coefficient greater than 0 implies expansionary monetary policy, because it is allowing the full expansion of the MB from the purchases, and at the same time is supplying more liquidity by increasing the MB e ven further. Data acquisition I obtained the data for both countries from their respective central bank websites. Both central bank s are required by law to publicize all the information they can regarding their actions, income statements, open purchases, and balance sheets. In this case, the information required for the regression is found on their Balance Sheets. The Balance sheets for central bank s are not as simple as it was explained in the Literature Review central bank and liabilities that need to be ignored.
36 Since the identity MB = FA + DA have to be maintained or approximated, I simply found the total amount of the MB of both Colombia and Peru, and then subtracted the am ount of net international reserves each country possessed. The amount left is the Domestic Assets. To be more specific, the MB is the amount of currency in circulation central bank Net Foreign Assets are the international reserves accrued by the central bank minus the long term international liabilities that the central bank has. Domestic assets are basically debt that is owed to the central bank by domestic agents such as the financial sector, the private sector, and the government. In both Colombia and Peru, DAs are negative, and this can be explained by the necessity of the central bank s to borrow from domestic agents in order to sterilize. A negative DA means the central bank owes money to d omestic agents. Authors such as Lavoie and Wang (2011) explain a principle known as the compensation thesis, which states that a movement in the central bank movement in another element of the balance sheet in the opposite dir ection (Lavoie and Wang 2009 ) As Lavoie (2001) pointed out earlier; the movements are a normal response from the financial system to the disequilibrium created by the increase in FA, which in time will affect liquidity (Lavoie 2001) In this case, the mov ement in the opposite direction occurs in the DA side, in an attempt to sterilize the intervention. All data was collected in the home currency of each respective country. Adjustments Before I ran the regression I made sure to adjust for seasonal facto rs. I used the hich identifies any monthly determined pattern and
37 ary base growth, which peaked every December. There were other minor adjustments to the variables but not as significant. Peru showed to have much less significant adjustments in comparison to Colombia.
38 The coefficient estimation was run with adj usted variables, so the foreign assets, monetary base, and the domestic assets were all adjusted for both countries. If the regression had been run with unadjusted variables the sterilization coefficient s would have had minor changes een 84, and Colombia .91 both numbers greater than adjusted results. I also added a dummy variable to control for the 2008 financial crisis. The dummy variable begins in September 2008 because of the Lehman Bros. collapse, and extends until august 20 09. Results
39 The sterilization coefficients found for both countries and their respective p values are shown in the following table. Country Sterilization coefficient ( ) ( no dummy v. ) P Value Peru 0.80 0.00 Colombia 0.84 0.00 Monthly from February 2001 to Dece mber 2012. Legend DA: Domestic Assets FA: Foreign Assets MB: Monetary Base P ER : Peru COL: Colombia _SA: Seasonal Adjustment Dependent Variable: D(DAPER_SA)/MBPER_SA Method: Least Squares Sample (adjusted): 2001M02 2012M12 Included observations: 143 a fter adjustments HAC standard errors & covariance (Bartlett kernel, Newey West fixed bandwidth = 5.0000) Variable Coefficient Std. Error t Statistic Prob. C 0.010031 0.004545 2.207093 0.0289 D(FAPER_SA)/MBPER_SA 0.83 9504 0.056327 14.194 0 DUMMY4 0.052735 0.013096 4.026837 0.0001
40 R squared 0.647489 Mean dependent var 0.01436 Adjusted R squared 0.644989 S.D. dependent var 0.078809 S.E. of regression 0.046957 Akaike info criterion 3.26529 Sum squared resid 0.310897 Schwarz criterion 3.22385 Log likelihood 235.4682 Hannan Quinn criter. 3.24845 F statistic 258.9872 Durbin Watson stat 1.788091 Prob(F statistic) 0 As we can see, the for Peru is 0. 83 which implies partial s terilization; h owever, it is a very high sterilization coefficient. We also measured the recursive coefficients for Peru. A recursive coefficient graph shows how the coefficient varies through time. Figure 3.1: Recursive Coefficient of Peru
41 As we can see in the above graph, the sterilization coefficient grows steadily from 2001 through the end of 2012. The graph suggests that as time passed Peru increased their extent of sterilization continuously. The resul t s for Colombia were as follows (after the adjustment): Dependent Variable: D(DACOL_SA)/MBCOL_SA Method: Least Squares Sample (adjusted): 2001M02 2012M12 Included observations: 143 after adjustments HAC standard errors & covariance (Bartlett kernel, Newey West fixed bandwidth = 5.0000) Variable Coefficient Std. Error t Statistic Prob. C 0.008988 0.002791 3.219967 0.0016 D(FACOL_SA)/MBCOL_SA 0.837923 0.046225 18.12725 0 DUMMY4 0.002652 0.006760 0.392233 0.6955 R squared 0 .699743 Mean dependent var 0.00311 Adjusted R squared 0.697613 S.D. dependent var 0.058941 S.E. of regression 0.032412 Akaike info criterion 4.00671 Sum squared resid 0.148123 Schwarz criterion 3.96527 Log likelihood 288.4795 Hannan Quinn crite r. 3.98987 F statistic 328.5972 Durbin Watson stat 2.553372 Prob(F statistic) 0 As we can appreciate, the sterilization coefficient for Colombia is 0. 83 which means a high degree of partial sterilized intervention and insignificantly higher than
42 Figure 3.2: Recursive Coefficient of Colombia As we can see, the sterilization coefficient for Colombia remains mostly constant from 2001 stays relatively the same through the 2 000s. The dummy variable had no effect whatsoever in the regression for Colombia; however, it decreased the sterilization coefficient ( .79 without the dummy variable) to .83 for Peru. On e hypothesis that can explain this is how both countries approached the crisis. Peru, on one hand, might have increased the sterilization during the crisis because of capital flight into Peru. Colombia, on the other hand, might have rather continuing with the same level of sterilization, allowing the financial sector to internalize the risk. It is also statistically important to note the P value for the Dummy variable for Colombia is higher than .005.
43 Research Limitations One limitation I encountered whi for an increase in the demand for money. In Aize n man and Glick (2008), they use nominal GDP growth in order to control for natural expansions of domestic assets which occur in order to meet increasing deman d for money. Monthly data for nominal GDP is not issued by either country. O ther regressions on sterilization coefficients that I encountered did not mention anything about controlling for DA expansion for money demand increases. Another problem encounte red was an accounting problem. As we know, FA are kept in foreign currency in the vaults. As exchange rate s fluctuate, the value of the FA also fluctuates causing monetary expansions. In Clavijo and Varela (2003) explain how accounting for the changes in t he N et International Reserves can be complex because credit variations are measured in local currency. In times of appreciating exchange rate s, the FA has a lower accounting value, although the monetary value has already been used. In my regression, I do n because the exchange rate s are somewhat stable throughout the decade. The information was gathered from two different Central Banks who collect their data their own different ways and with the ir own different means. Although I would have rather use data from the IMF instead, the IMF does not have the variables I w as looking for. Figure 3.3: Colo
44 The graphs above illustrate the changes in FAs in Colombia and Pe ru. As we can s increased from 20,000,000 million Pesos in 2001, to almost
45 70,000,000 million Pesos (38,335 million USD) in 2012 For Peru, the FA increased even further: from 25,000 million soles to 160,000 million (61,776 million USD) It is evident that Peru has increased their FA eight fold, while Colombia only by a factor of 3.5. Today, Peru has 23,441 million dollars more than Colombia in foreign assets. Figure 3.4: Monetary Base of Both Peru and Colombia
46 The Monetary Base s for both Peru and Colombia have increased in the past in December and then co ntract back; however, it grows overall. It appears as if Colombia intervene s in their MB more often than Peru. The differences can also be a product of the different instruments and bylaws of each monetary system.
47 Figure 3.5: Domestic Assets Peru and Colombia
48 Lastly, the domestic Assets for Colombia changed a lot this past decade; however, it is almost the same amount it was in 2001 9,889,463 billion of Colombian Pesos or 5,414 million USD Peru, on the other hand, has decreased their DA by a f actor of 8 from almost 20,000 million Nuevos Soles (7,722 million USD) to almost 90,000 million Soles (34,749 million USD) Another important observation is the big decline in DA lained by the financial crisis in which both central banks had to provide liquidity to the domestic sector.
49 Ch. 4 Further Specifications on the FOREX Intervention Mechanisms in Peru and Colombia describe how mechanisms work in practice. The central bank the first chapter, for example, is not a real Balance sheet in the sense that it is overly simplified (Lavoie & Wang 2012). A real balance sheet contains many other differ ent instruments which the banks use to conduct monetary policy. Different central bank s conduct monetary policy in different ways, and the instruments they use will most likely be different as well. In this chapter, we will try to examine the intervention mechanisms that both the Peruvian and Colombian central bank s utilize to conduct monetary policy, including the instruments and sterilization processes. Colombia fulfill t an informative pamphlet released for this purpose. In 2013, the latest of these pamphlets was aimed at outlining the FX intervention policy the Colombian central bank maintains, a nd to illustrate the topics of discussion that have had major importance when determining policy at the central bank The authors, Hernando Vargas, Andres Gonzalez, and Diego Rodriguez, reached the hypothesis that the effectiveness of Sterilized interventi on by the central bank of Colombia is not useful at coping with external factors such as: Quantitative easing by developed economies (especially the USA), lower risk
50 premiums on emerging economies (especially domestic assets risk), and the higher commodity prices. Furthermore, there is a tradeoff between the FX intervention and Inflation targeting since even when it is sterilized FX intervention has an expansionary effect on the credit supply and aggregate demand. The current FX Intervention of the Co lombian central bank is aimed at the following official objectives: Maintaining an adequate level of international reserves, fixing short term exchange rate misalignments, and mitigating excessive volatility of the exchange rate (Vargas, Gonzalez and Rodri guez 2013) As previously stated; a small open economy must maintain an adequate level of International reserves in order to fend determined by the size of GDP and the financial sector, but also on the exchange rate regime, price determination, and how regulated their financial system is (Edison 2003) In the case of Colombia, the credibility of the inflation target, the low level of ERPT and the sound regulation of the financial market create an environment where high exchange rate intervention was the smallest in the Latin American region. Authors such as Vargas (2011) explain how the v olatility from the floating exchange rate contributes to the previous conditions, and that the currency risk is internalized by the private sector. A recent source of potential external shocks has the Colombian monetary authorities debating on the quest ion of optimal international reserves. The source is the expansion of Colombian banks to other countries. Two approaches are being considered by the Bank: holding more international reserves to hedge against the risk of shock, or strengthening liquidity re gulation. Holding more reserves will pose a higher opportunity
51 cost of the central bank; so internalizing the risk by tougher regulation seems to be promising. 67% of the banks that operate outside of Colombia are located in Central America, a region that on average lacks regulation or a Lender of Last Resort mechanism. Were problems erupt from overseas; being loaned outside of Colombia as well. Another aspect to consider when vel of International drawbacks; how ever they are useful to prevent or a lleviate the effects of external shocks (Calvo, Izquierdo and Loo Kung 2012) The drawbacks of the mode l are important because it renders the models unusable in practice. For example, the models do not account for saving behavior of the economy, or the openness of the financial account. This is why a strong assumption must be always made on the probability of liquidity shocks. The most important drawback is the assumption that foreign liabilities as a given. This assumption leads to the misconception that the FA position of the country increases in the same amount as the purch ase of International Reserves increases. Sterilized intervention in an economy with high capital mobility creates new capital inflow, reducing the most liquid assets of the domestic private sector. In the rare event of perfect capital mobility, FX interven tion will not change the net asset l model. Because of the practical issues signal ed above, the Colombian central bank follows different indicators of international reserves in practice. Purchases are aimed at
52 keeping these indicators stable. The indicators are the following: ratio of reserves to broad money, short term debt payment, th e current account balance, level of imports, and GDP. Sterilization in Colombia is utilized in order to maintain short term interest rates at the rate announced by the central bank The main tools used fo r sterilization at the central bank have been gov ernment deposits. These deposits have allowed the central bank to remain a net lender to the financial system. In the event that Government deposits are not enough to sterilize the intervention, the central bank holds government securities which they can s ell in the open market. The Bank has in the past opened short term deposits (remunerated) for financial institutions with the mission of sucking up excess liquidity from FX interventions. In 2011, the Government began issuing Monetary Regulation Government Bonds (MRGB), and all the proceeds would go to the central bank These bonds allowed the central bank to work jointly with the government on sterilization policy. As of now, the central bank intervenes in the FX market through announced daily purchases of dollars in fixed amounts. The fixation of the amount of intervention is
53 important because it diminishes the recognition of a specific exchange rate target which strengthens the credibility of the inflation target. Empirical data from Echavarria, Lopez, and Misas (2009) have determined the effects of FX intervention on the exchange rate to be short term only. These results coupled with other econometrics findings has led the central bank to believe sterilized intervention to be ineffective at dealing wit h long term changes such as Quantitative easing, reduced risk in domestic assets, and high commodity prices. In practice, the central bank has limited itself to using Sterilized FX intervention only to correct short term misalignments. For longer term misa lignments, Capital controls are being considered. Peru The most important aspect to consider when understanding the Monetary Policy mechanisms in Peru is the fact that the economy is heavily dollarized. When monetary authorities determine the policy ins truments to use, they need to make sure they avoid transmission mechanism because the policy rate can not affect these liquidity flows nor their interest rate; a sharp depreciation can lead to a shock on the supply of credit because of the deterioration of the quality of the assets denominated in different currencies. Because of the phenome na previously described, the BCR has added several instruments central bank has to regulate the liquidity in foreign currency, they have higher reserve requirements on shor t term FX liabilities. At the same time, the BCR intervenes in the FX
54 market sterilizing the increments in the MB with instruments that only the domestic sector can buy. It is important to mention, that the BCR tries to do both: reduce the volatility of th e dollar, and accumulate IR; all this without giving a strong commitment to a given exact exchange rate (Quispe 2000). Although highly complicated, this policy mechanism has allowed the central bank prevent any substantial disruptions from the 2008 global crisis. The BCR decided to implement an inflation targeting regime in 2002, with an initial anchor of 2.5% and with a tolerance o f 1%. As previously mentioned, d ollarization of the financial intermediaries increases the risk of credit busts These busts m ight result from violent movements in foreign exchange liquidity inflows or exchange rate volatility. This problem creates solvency risks in the event that the asset side of a bank is denominated in domestic currency and the liabilities in foreign currency In order to cope with this problem, the BCR embraced a strategy consisting of three levels of liquidity: hoarding of international reserves to buffer external shocks, high liquidity requirements (especially for those holding foreign liabilities), and a conservative public financial position (fiscal policy). The quantitative instruments, such as the reserve requirement and the positioning of the BCR balance sheet, have gained notoriety because these innovative instruments helped the BCR avoid the collaps e of credit. These measures acted directly over the flows Financial intermediaries become insensitive to changes in interest rates during times of crisis because the interest rates fail at signa ling the stance of monetary policy. The instruments are part of a
55 Monetary Risk Management approach. It includes preventive and corrective measures designed to avoid credit shocks in order to maintain the transmission mechanisms. The BCR has two ways of d ealing with the problem of signaling a specific exchange rate : First, they can avoid the predictability of the market by using, what they call, a rule type intervention. This kind of intervention involves in intervening with pre announced amounts of purcha ses. However, depending on the state of the economy, the bank might be forced to abandon this option an opt for the more discretionary intervention which is without pre announced amount purchases. The main way of intervention is with direct operations with commercial banks at the current exchange rate Auctions are also used by the BCR to make temporary swaps of currency. However, auctions are only used when the futures market for foreign currency in pushing the exchange rate position of local banks. Two important factor of the Peruvian economy served valuable conditions for the Sterilization process in Peru: the Fiscal position, and the increasing demand for money. The fiscal contribution to the BCR makes up for 35% of the size of international reserves. The importance of fiscal deposits rests on the fact that it is free of interest, which relates central b ank has to pay at one point in time. Authors such as Rossini, Renzo; Quispe, Zenon; & Rodriguez Donita (2011) computed a ratio indicating how the volatility of the interest rate compared to the volatility of the exchange rate in different economies. The re sults showed that Peru had the lowest ratio, meaning interventions in Peru affect the interbank interest rate the least in comparison to other countries in the same region. This advantage can translate for a
56 lower need for Peru to focus on the interest rat e when intervening on the FX market for exchange rate reasons. In the same way, the cost of liabilities the BCR holds is exceeded by the return on their foreign assets. Episodes such as the Lehman collapse and QE2 led to a surge of capital inflows in Peru, which has been reflected on the increasing demand for BCR instruments from foreign agents, these instruments are denominated in local currency and can pose a problem for sterilization. In order to deal with the inflows, the BCR has imposed a 4% fee for the purchases and sells of BCR instruments by foreign agents, doubled the reserve requirement for deposits denominated with local currency by foreign agents, and substituted the sterilization deposit certificates 5 The treasury has also done its part to deal with inflows and the global economic turmoil: They have cut limits for net foreign positions of banks, limited the volume of FX operations by pensions, and passed a law that limits the amount of derivatives banks can
57 hold. The treasury is also tax ing capital gains by 30%. Another unconventional move was the release of bonds denominated in domestic currency but paid in foreign currency with the aim of increasing the demand for foreign currency. The reserve requirement has been used plenty by the BCR to discourage the purchase of domesti c instruments by foreign agents, a byproduct of this is the accumulation of reserves which serve as buffers in the event of crisis. Peru experienced a crisis triggered by the Russian crisis of 1998. In this episode, the surge of capital inflows expanded the banking credit, and then a sudden stop of capital inflows ended the credit expansion, leading to a credit crunch (Rossini and Quispe 2010) After this credit crunch, the BCR became vigilant of credit expansions as a result of capital inflows.
58 Conclusion In this thesis we explored the trend of reserve accumulation through sterilized intervention by two small open economies It appears that both Peru and Colombia exercise sterilized intervention often. Thei r respective sterilization coefficients suggest that both countries engage in heavy sterilization. Literature in the area will suggest that there is evidence of dual anchors for Monetary Policy : Exchange Rate and Monetary Base. The central bank s target exc hange rate by intervening on the FOREX market (in order to manipulate exchange rates) and sterilize their intervention with DA in order to leave the MB untouched, mitigating effects on inflation. There is also evidence that fected the sterilization coefficient of both countries. Another central idea of the thesis is the cautionary and mercantilist reasons for accumulating reserves: Cautionary to prevent economic shocks and mercantilist to get an advantage in the terms of tr ade. The ability for countries to accumulate massive amounts of reserves allows them to circumvent the trilemma. As Aizenman explained, a 4th variable has been added to the trilemma making it a quatrilemma; that 4 th variable is international reserves. Many EMEs are hoarding reserves at an all time high, China being a precursor of this practice. There is mixed data from the literature on the efficiency of sterilized intervention. In one part, countries that are engaging in it have been able to sustain i t and, in a few cases, able to employ sterilized intervention so efficiently is the imperfect substitutability of sterilize they can have the desired effect on the MB.
59 So far the recipe for stability is clear: dual anchors and accumulation of reserves. However, we have seen many other recipes in the past fail, rendering to nothing any major progress that EMEs had accomplished. Further Research A way to control for the increase in money demand such as nominal GDP growth can be very useful in future regressions. At the same time, finding a way to value argument presented in ch.3 will also contribute to a more accurate coefficient. As the Aizenman paper stated, the cost of reserve accumulation for China is now higher than the benefit they receive from it. A study that targets the optimal reserve acc umulation for countries for different regions can stop depending on the dollar or euro? To me, it seems unfair that developing economies have to commit to buying d ollars that keep losing value because of quantitative easing.
60 Works Cited Aizenman, Joshua, and Hiro Ito. "Trilemma Policy Convergence Patterns and Output Volatility ." UCSC and the NBER Portland State University 2012. Aizenman, Joshua, and Reuven Glick "Sterilization, Monetary Policy, and Global Financial Integration." NBER Working Paper No. 13902 2008. Banco de la Republica de Colombia. n.d. www.Banrep.gov.co/exchange_policy/index.html (accessed febuary 2013). Calvo, G, A Izquierdo, and R Loo Kung. Optimal Holdings of International Reserves: Self Insurance against Sudden Stops." Working Paper No. 18219 National Bureau of Economic Research 2012. Calvo, Guillermo, Leonardo Leiderman, and Carmen Reinhart. "Capital Inflows, and Real Exchange Rate Apprec iation in Latin America: The Role of the External Factors." IMF Staff Papers 1993: 108 115. Chang, Roberto. "Inflation Targeting, Reserve Accumulation, and Exchange Rate Management in Latin America." Banco de la Republica de Colombia, Borradores de Econom ia 2008: 487. Clavijo, Sergio, and Carlos Varela. "Floating Exchange and Monetary Sterilization: The Experience of Colombia." Essays on Political Economy 2003. Cote, A. "Exchange Rate Volatility and Trade: A Survey." Working Paper (Bank of Canada), 1994: 94 5. Dacourt, Oscar. "Structural Reforms and Macroeconomic Policy in Peru, 1990 1996." Edited by Taylor Lance. After Neoliberalism, What is next in Latin America? (The University of Michigan Press), 1999.
61 Dominguez, Kathryn, and Jeffrey A Frankel. "Does Foreign Exchange Intervention Matter? Disentangling the Portfolio and Expectation Effects for the Mark." (UC Berkeley: Department of Economics, Ucentral Bank) 1990. Dooley, Michael P, David Folkerts Landau, and Peter Garber. "An Essay on the Revived Bretto n Woods System." NBER Working Paper No. 9971 2003. Edison, H. "Are Foreign Exchange Reserves in Asia too High? Three Current Policy Issues in Developing Countries." World Economic Outlook IMF 2003: 78 92. Feenstra, Robert, and Alan Taylor. International Economics. Davis: Worth Publisher, 2009. Fisher, Irving. "The Purchasing Power of Money, its Determination and Relation to Credit, Interest, and Crises." Online Library of Liberty n.d. Frankel, Roberto, and Martin Rapetti. "A Concise History of Exchange R ate Regimes in Latin America." Center for Economic and Policy Research Reports and Issue Briefs 2011. Ghosh, Atish R, Jonathan D Ostry, and Charalambos G Tsangarides. "Shifting Motives: Explaining the Buildup in Official Reserves in Emerging Markets Since the 1980s." International Monetary Fund IMF Working Paper 2012. Ghosh, Atish, Jonathan Ostry, and Charalambos Tsangarides. "Exchange Rate Regimes and the Stability of the International Monetary System." IMF Occasional Paper No. 270 2010. Glick, Reuven, and Michael Hutchinson. "Currency Crises." Working Paper Series Federal Reserve Bank of San Francisco 2011.
62 Goldberg, Pinelopi K, and Michael Knetter. "Goods Prices and Exchange Rates: What Have We Learned?" Journal of Economic Literature 1997: 1243 1272 Hernandez Gamarra, Antonio. "A Monetary History of Colombia: The Age of inflation." Bogota: Editores Villegas 2001. Kenen, Peter B. "Effects of Intervention and Sterilization in the Short Run and the Long Run." Edited by Richard N. Cooper. The Internati onal Monetary System Under Flexible Exchange Rates 1982: 58 61. Lavoie, Marc. "The Reflux Mechanism and the Open Economy." Credit, Interest rates, and the Open Economy: Essays on Horizontalism 2001: 215 242. Lavoie, Marc, and Peng Wang. "The Compensation Thesis, as Exemplified by the Case of the Chinese Central Bank." International Review of Applied Economics 2009: 287 301. Lederman, Daniel, Ana Maria Menendez, Guillermo Perry, and Joseph Stiglitz. "Mexico: Five Years after the Crisis." The World Bank Pr esentation at the ABCDE Conference 2000. Obsfeld, Maurice. "Foundations of International Finance." (Massachusetts Institute of Technology) 1996: 597 599. Obstfeld, Maurice, Jay C Shambaugh, and Alan M Taylor. "The Trilemma in History: Tradeoffs among Exch ange Rates, Monetary Policies, and Capital Mobility." NBER Working Paper No.10396 2004. Ozturk, Ilhan. "Exchange Rate Volatility and Trade: A Literature Survey." International Journal of Applied Econometrics and Quantitive Studies 3 1 (2006). Parodi, Carl os. "Peru 1960 2000." CIUP 2000: 206 208.
63 Pastor, Manuel Jr. "Latin America, the Debt Crisis, and the International Monetary Fund." JSTOR 2012. Rossini, Renzo, and Zenon Quispe. "Monetary Policy during the Global Financial Crisis of 2007 2009: The Case o f Peru." Bank of International Settlements 2010: 299 315. Rossini, Renzo, Zenon Quispe, and Donita Rodriguez. "Capital Flows, Monetary Policy, and Forex Intervention in Peru." Working Paper Series Banco Central de Reserva del Peru 2011. Sheahan, John. "L a economia peruana desde 1950." 2001: 207. Steiner, Roberto, and Hernan Vallejo. "The Economy in Colombia: A Country Study." Edited by Rex A Hudson. Library of Congress Federal Research Division 2010. Taylor, Mark P. "The Economics of Exchange Rates." Jou rnal of Economic Literature 1, no. 33 (1995): 13 47. Vargas, Hernando. "Monetary Policy and the Exchange Rate in Colombia." BIS Papers 2011: 57. Vargas, Hernando, Andres Gonzalez, and Diego Rodriguez. "Foreign Exchange Intervention in Colombia." Borradores de Economia No. 757 (Banco de la Republica), 2013. Winkelried, Diego. "Es asimetrico el pass through en el Peru?: Un analisis agregado." Revista Estudios Economicos no. 10 (2003).