Colombia's black market in foreign exchange

0305-750x/92 $5 .oo + 0.00 0 I992 Pcrgamon Press Ltd

World fkve/o/mw/. Vol. 20, No. 8. pp. 115%1207. 1992. Printed in Great Britain.

Colombia’s

Black Market in Foreign ROBERT

Exchange

GROSSE”

University of Miami, Coral Gables,

Florida

~ Decree Law 444 in 1967 hca\ily restricted ~XCSSto foreign exchange by Colombian residents. This policy spurred the dcvelopmcnt of II large hl:rck market. which had yawn to ahout US$7.5 billion of transaction WILK per year hy 19X8. Soon after the rcstrictivc pollcle\ were imposed. marijuana and then cocaine smuggling became large industries in C‘olomhia: this

Summary.

narcotics boom has appurcntly driven down the black market cxchangc rate and largely eliminated the exchange rate premium that is usual I’or huch markcth. While Colomhia’~ gowrnmcnt ha\ opcwd up the official foreign exchange market in the early IOOO~. the black market is cxpccted to continw to thrive. due to the smugglers’ need to rcbort to underground means lor transacting their huainess.

1. INTRODUCTION Parallel markets in foreign exchange have arisen in many countries throughout history in response to government controls on access to foreign exchange.’ Typically, the controls are imposed to protect a government’s limited stock of foreign exchange rcscrves. The need for this protection, in turn. is caused by trade deficits and/or capital flight that result in net demand for foreign exchange at the central bank. Once a government imposes the limitations on holding foreign exchange or on transferring it overseas, demand for an alternative source of that currency arises. In response to these conditions, a parallel market in foreign currency develops. When the market operates outside the legally permitted system of financial transactions in a country, it is called a black market. (This phenomenon also is called variously the “grey market,” the “street market,” the “unofficial market.” and other names in different countries.)’ Foreign exchange parallel markets imply a range of important issues for the countries where they operate. First, the existence of a parallel market typically results from a current or past international payments imbalance, so it is usually associated with an excess demand for foreign exchange in the local market. This disequilibrium is a target of government policy. both in terms of controls on access to foreign exchange (that produced the parallel market) and of domestic monetary policy (that often produces inflation. which in turn adds pressure for devaluation of the currency). The underlying imbalance. thus, is the

crucial issue that policy makers confront. with the parallel market as one of its symptoms. This imbalance mav be the result of several factors, separately or in combination. An excessive rate of monetary growth can lead to inflntion. If that inflation exceeds price increases in other countries. it will lead to demand for foreign currencies, both to buy foreign goods and to hold stable currencies. A lesswealth in more developed country may find that investors confidence in the economic prospects of the country is weak, and that they conscquentlv look to place their investments in other count&s (nnd therefore in other currencies); this again produces excess demand for foreign exchange and contributes to the capital flight that such countries want to avoid. Even without the previous two factors. a country may find that goods produced locally are perceived as inferior in quality to similar goods produced abroad; and thus demand arises for foreign exchange to buy those higher quality foreign products. A second variety of reasons for the develop-

ment of parallel, and particularly black. markets is the existence of an underground (or informal. in the country (Pitt, or grey, etc.) economy 1984). When some participants in the local economy

choose

to enter

into

illegal

business

*I would like to thank Martha Osorio and Rodrigo Seda for their excellent research assistance on this project and Evan Tanner. Lynn Summers, and two anonymous referees I’or their helpful comments on earlier March

drafts of this papa. 2. 1992.

Final

revision

accepted:

1104

WORLD

DEVELOPMENT

activities - such as sale of contraband products - then a need arises for financial services that circumvent the legal financial system. The underground economy can be as “harmless” as the street vendor phenomenon that exists widely throughout less developed countries. Street vendors obtain contraband products ranging from toothbrushes to televisions and sell them from unregistered places of business (usually from sidewalk tables or literally in the street). Another type of underground economy that has become quite significant in a few countries is the trafficking of narcotics. This business involves products such as marijuana, heroin and cocaine, which are illegal to produce and sell, thus requiring the use of nonconventional means of production, transport, sale and financing. The black market in foreign exchange provides both of these kinds of business access to the foreign currency needed to purchase the contraband goods - and in the case of narcotics, the access to local currency that can be bought in exchange for foreign currency earned in the drug trade. Most studies of foreign exchange parallel markets have focused on two types of economic issues: the impact of the market on the local economy, and explanation of the premium on the parallel market price of foreign exchange in comparison with the official exchange rate. In the former category. Greenwood and Kimbrough (1987) demonstrated that the operation of a parallel market in foreign exchange mitigates the effect of exchange controls on domestic reiative prices and the level of consumption of imports. They also showed that, if the government sets exchange controls to maintain policy goals ignoring leakage of funds into the parallel market, then the parallel market may reduce economic welfare. Pitt (1984) examined the relationship of smuggling and capital flight to the parallel market, demonstrating that government controls on access to official foreign exchange (imposed, for example, due to a trade imbalance or shortage of foreign exchange in the central bank) are not necessary for the existence of a currency parallel market. As discussed below, Pitt showed that existence of a contraband market whose participants could not use the official foreign exchange market is a sufficient condition for development of a black market? Nowak (1984) developed a general equilibrium model for examining the impact of exchange controls on economic welfare when a parallel market exists. He demonstrated that the parallel market mitigates the inflationary impact of official exchange rate devaluation; and that exchange controls that produce a currency parallel

market lead to a parallel-market devaluation that has similar inflationary influence as would an official exchange rate devaluation. Additional studies have examined the relationship between the parallel market rate and the official market exchange rate. Michaely (1954) studied black markets in commodities. showing that the black market price (exchange rate) would be an upper bound of the free-market price of the commodity (or currency) in question. Other writers following this first step have argued that the premium on the parallel market price of foreign exchange is an adjustment to the official exchange rate; this premium then can be evaluated to judge whether purchasing power parity is reflected in the rate. These authors generally have found parallel market rates to be strongly related to purchasing power parity, that is, to reflect the difference between domestic inflation and inflation in the foreign country (typically the United States) whose currency is being compared. Culbertson (1975) found strong correlations between the inflation differential and the parallel market exchange rate for three less-developed countries during 1952~71. Koveos and Siefert (1985) found similar strong correlations for several Latin American countries during 1973-83. Both Koveas and Siefert (1983, Aggarwal (1990) and others have found that parallel markets in lessdeveloped countries tend to be efficient markets in their tendency to achieve purchasing power parity (that is, the parallel markets generally reflect publically available information about inflation in each country at the time of the exchange transactions). This paper examines the foreign exchange black market in Colombia, focusing on the reasons for its existence, the way in which the and the implications for market functions, Colombia. The next section presents an overview of such markets. followed by a section analyzing the Colombian market’s structure and functioning. The fourth section examines the determination of the exchange rate in Colombia’s black market during the past two decades. The final section uses the empirical findings to draw some conclusions. 2. CHARACTERISTICS EXCHANGE BLACK

OF FOREIGN MARKETS

A black market in foreign exchange is one that circumvents the local legal financial system; it is the result of government intervention to restrict trade or capital flows which pushes the restricted users of foreign exchange to find alternative (nonofficial) means of carrying out their trans-

COLOMBIA’S

BLACK

actions. In most contexts, the black market operates as an illegal (but often broadly accepted) source of foreign currency. Thus, the cost of buying foreign exchange in this market is generally higher than in the official, legal market. That is, because of the excess demand for foreign exchange that spills over from the official exchange market, and because of the risk of government sanctions against participants in the black market, the buyer of foreign exchange will pay more local currency (e.g., pesos) for foreign currency than in the official market. Conversely, the suppliers of foreign exchange-often foreign tourists and local residents with holdings of foreign currency abroad - will receive more local currency for their money than in the official market.’ Examples of black markets in recent years demonstrate exchange rate premia ranging from a few percentage points up to more than 100% of the official market rate.s In Brazil under the second Cruzado Plan at the end of 1986, the official rate was about 14.9 cruzados per dollar, while the black market rate was about 31.5 cruzados per dollar. In Peru during late 1988, the official inti traded for about 700 intis per dollar, while the black market rate was approximately 2,000 intis per dollar.h In India, through most of the 1980s. the black market rate maintained a premium on foreign exchange of about 25-35x over the official rate. There are also some cases where the black market rate more closely approximates the official rate, as in Israel, where the black market sheqel has been valued at about 10% over (i.e., more sheqels per dollar) the official sheqel during most of the past decade. The Colombian example is especially striking for its absence of such a premium; in fact, it shows an occasional discount on black market foreign exchange. The appendix shows graphically the difference between official and black market exchange rates for several of these countries, demonstrating the premium on purchase of foreign exchange in the black market. In Colombia the black market under discussion here has existed since the imposition of Decree Law 444 in 1967.’ During the mid-1950s Colombia experienced an abundance of foreign exchange from its trade balance, due to a boom in coffee exports. Subsequent to that time, the country experienced notable fluctuations in the trade balance (and a consistently negative current account balance), as well as repeated pressures from international creditors to stabilize the economy so that debt servicing could be maintained. When coffee exports (by far Colombia’s largest foreign exchange generator) dropped sharply during 1965566, the trade sur-

MARKET

1195

plus evaporated. By 1966 the country had reached an impasse in negotiating with the International Monetary Fund (IMF) on appropriate stabilization policy, at which point President Lleras Restrepo responded with the development of the exchange control regime that began operation in early 1967. For various reasons since that time, Colombia chose to leave the controls in place, despite some periods of balance-of-payments favorable generally conditions.x It was only in January 1991 that the system of controls was largely dismantled and the official exchange market opened to much broader participation. In early 1991, Colombia’s government formulated a new foreign exchange policy that allowed Colombians to hold dollars and otherwise liberalized the use of foreign exchange. This policy was implemented through Law 9 of January 17, 1991 and Monetary Board Rule 4 of January 31, 1991. The discussion in this article focuses on the period before then. The black market continues to function in the new environment, although the characteristics of the market may well change importantly in the coming months and years. Decree Law 444 and implementing policies that were imposed subsequently disallowed Colombian residents from holding foreign exchange domestically or abroad. Article 4 required exporters to turn in their foreign exchange earnings to the Central Bank in exchange for pesos or Certificates of Exchange (Certificados de Cambio). Article 5 required foreign investors similarly to hold their funds in Colombia in pesos. Article 6 authorized access to foreign exchange to some classes of users: e.g., for importers of those products authorized by the government (under Article 7) to pay for both the goods and shipping costs; specifically for importers of crude petroleum; for expenses of students studying abroad; for servicing of foreign debt; and a few other uses. Access to foreign exchange was denied to importers of restricted products (products such as explosives, armaments, and a few others, that were prohibited by the importlicensing agency, INCOMEX). Access was also prohibited to Colombian investors seeking to place their investments abroad (in Articles 144 149). These restrictions made the black market attractive to a wide range of Colombian residents, in otherwise legal enterprises as well as in contraband trade. The black market functions in parallel to the official market in Colombia as elsewhere. The official market provides dollars and other foreign currency to importers for paying suppliers and to banks for satisfying their foreign debt commit-

WORLD

11%

DEVELOPMEN

men&. The official market is supplied largely by legal exporters whose foreign currency earnings are transmitted to the Central Bank, that in turn provides the exporters with local currency, namely pesos. The black market provides dollars to Colombian importers whose desired imports are restricted by the government through high tariffs, quotas, and/or red tape (e.g., luxury goods and products that are widely available in Colombia from local suppliers) and to Colombian investors who want to hold their wealth outside of the country. The black market is supplied with dollars by exporters who choose not to follow the legal process of exporting (e.g., “contrrrh(mdi.slrrs” selling Colombian gold and emeralds without paying taxes). by Colombians living abroad who choose to remit some of their wealth to relatives in Colombia. by black market vendors of all sorts of products on the frontiers with Venezuela and Ecuador, and. principally, by drug traffickers. In all. both foreign exchange markets serve similar purposes: the black market is the provider of foreign exchange service to those who are unable or unwilling to use the official market. The black market is important in the Colombian context for several reasons. First. it indicated (until January 199 1) an economic inefficiency due to the exchange controls that caused transactions to move into the black market. This inefficiency had to be considered in comparison with alternatives such as eliminating the controls and dealing with the possible outflow of foreign exchange reserves through other policies (such as import tariffs or quotas) or automatic adjustment (which would lead to a peso devaluation and, hopefully. ;I rise in exports plus a fall in imports sufficient to balance the demand for foreign exchange”). Second. it provides an indicator of the equilibrium exchange rate that would have to exist to eliminate the excess demand for foreign exchange that typically has been shunted from the official market into the black market. In Colombia in particular, the black market is especially important in the context of narcodollars. the earnings of Colombian narcotics traffickers. Casual observation shows very small differences between official and black market exchange rates over the past IS years. which could be attributed to the supply of dollars to the market from narcotraffics exports, initially marijuana and more recently cocaine. This point is considered in some detail below. 3. STRUCTURE AND FUNCTIONING THE BLACK MARKET”’ The

black

market

in foreign

exchange

OF

in

Colombia has several layers of participants. On the demand side. they range from individuals seeking to convert US$1620 equivalent of pesos, into dollars all the way to large cornpanics buying millions of dollars with their pesos. On the supply side. they range from c~ontrahrrt7di.strr.s seeking to sell their excess dollars received for sales of coffee or cattle in the underground economy all the way to narcotraffickers seeking to convert millions of dollars they have earned from sales in the United States (and elsewhere). The intermediaries in most black market transactions are umhis~a.s (foreign exchange dealers). who exist at both retail and wholesale levels.” The simplest form of black market foreign exchange dealing occurs in the street market. in which a small-scale c~ccmhistrr buys dollars from foreign tourists and sells them pesos. typically in quantities of US$lO-100. Similar retail transactions occur in the contraband markets in itll major Colombian cities. where the “Srrn Atdrc.sito.s” (contraband sellers”) sell imported electronics, clothing, and many other articles that have been brought into the country without being registered, and thus without paying taxes or tariffs. Another form of simple black market dealing. and the one that stimulated market growth in the late lY6Os and early lY7Os, is contraband trade along the borders with Venezuela and Ecuador. Since both of these neighboring countries are major oil exporters. they wcrc able to maintain overvalued currencies for many years. making imports cheaper without affecting oil exports (that arc priced in US dollars, not local currency). In contrast, Colombia depends heavily on commodity exports of coffee. sugar, bananas, and cut flowers ~ all of which respond to local currency devaluations. Frequently during the past 20 years the Colombian peso has been sufficiently low in value relative to the bolivar or hucrc such that all kinds of products were cheaper to import into Venezuela and Ecuador from Colombia than to obtain locally. Thus, contraband in many products. especially cattle. moved into these countries in exchange for bolivars. sucres. and dollars. As Venezuela and Ecuador have moved to establish more open foreign exchange regimes. this kind of contraband trade has slowed greatly. More complex transactions occur when Colombian gold or emeralds are shipped abroad and paid for in dollars. In this case. when the contraband exporter wants to sell dollars to buy pesos for use in Colombia, the foreign exchange originates abroad. The exporter then must either transport the dollars in currency or cheek to Colombia. or contract with ;I crrrnhktrr for sale of the dollars to some buyer who can pay with pesos

COLOMBIA‘S

BLACK

in Colombia. In this case the dollars are transferred to the buyer’s account in the United States and the pesos are delivered to the gold or emerald exporter in Colombia either in the form of a check or other financial instrument.‘3 Even more complex transactions occur when the seller of dollars is a narcotics trafficker. In this case, a wide variety of means are used to deliver the dollars. One mechanism is to physically carry the dollars to Colombia for direct exchange for pesos in cash. Another mechanism is to convert the dollars into money orders or checks in the United States. and then ship them to Colombia for sale in exchange for pesos. A third means of selling the dollars is to deposit them in bank accounts in the United States and to arrange for bank transfers into accounts of the purchasers who reside in Colombia. In this instance, the dollar buyer specifies a bank account to receive the dollars and pays in pesos to the c~nrhisra in Colombia. typically with a personal check.“’ Many more processes have been devised to convert the narcodollars into pesos - just as the narcotraffickers have devised numerous means for holding and moving their dollar wealth outside of Colombia. A notable feature of these transactions is that most of the situations involve delivery of dollars owsidc of Colombia, in exchange for pesos delivered in Colombia. Note that on the demand side for dollars are Colombian businesspeople. who seek to obtain dollars for their business needs and/or to hold their wealth overseas. These two motives are the basic ones in virtually all of the demand for black market dollars. The businesspeople who buy these dollars tend to deal in fairly large quantities of money (e.g., US$lO.OOO or more), and they generally want to keep the dollars outside of Colombia. The camhistus who service the large-scale black market generally operate out of offices or homes, though without advertising since the business must be carried out surreptitiously.li While the Colombian government has been unable to stop the black market from functioning, it nevertheless has penalized hundreds of participants each year during the 1980s. Penalties range from forfeiture of the funds that have been passed through the black market to imprisonment if the funds are not delivered to the Central Bank.“’ Many of the large cambistas function legally as stock market brokers. These firms then operate money exchanges on the side - however, seven of the major securities brokers were forced to leave the Bogota stock exchange in 1988, apparently due to their business in black market foreign exchange. ” In summary. these categories of participants

1197

MARKET

constitute the black market in Colombia at the end of the 1980s and into the early lY9Os: SUPPlY (a) ;;;y;raffickers (a) 0 (b) cor~truhur~distus selling coffee, gold, emeralds (10%) (c) cor2trriblrrl$i.rtrr.s (b) selling cattle, etc. to Venezuela & Ecuador (10%) (d) transfers of funds(c) by Colombians living abroad to their families (IO%) (e) return of flight (d) capital by Colombian buxincsspeople (IO%)

Demand importers who want to escape the costs and bureaucracy of buying dollars legally (30%) importers who buy contraband products (&III Andrc~sitos) (25%) wealthy Colombians who want to hold their wealth abroad

(30’%,) Colombians wanting to hold dollars in Colombia (IS’%)

The percentages noted on both sides of the market represent market shares, as measured below.

(a) Market size estimutes While no precise measure of the volume of money transacted in the foreign exchange black market can be obtained. several estimates are available. The remarkable point about these estimates is their similarity. The Central Bank has tried to estimate the size of this market, as part of its concern with monetary policy in Colombia. The black market obviously detracts from the Bank’s ability to regulate the amount of foreign exchange in the economy and also has a consequent impact on domestic money holding. A recent estimate of the dollar volume of black market exchanges during the 1980s shows that approximately US$1.775 billion per year entered the black market from narcotraffic and about US$S25 million per year from return of capital flight (including contraband sales of gold, cattle, and other products overseas). (Gomez, lY90, p. 15). This study presented estimates of annual funds flows for 1981-X8; here an annual average for the period is constructed. The drug flow data are taken from reports by the National Narcotics Intelligence Consumers Committee (NNICC). a US government organization that pools information from the law enforcement agencies that deal with narcotics trafficking. A second study using the NNICC data with several additional adjustments reflecting factors such as narcotics seizures, crop eradication.

119x

WORLD DEVELOPMENT

money confiscations, and others, produced similar estimates of narcotics money flows on the order of US$2.5-3 billion per year during the 1980s (Sarmiento. 1991). This study also discussed but did not estimate the sources of black market dollars coming from underinvoicing of exports (in which case exporters leave some portion of their receipts overseas, and some of those funds may enter the black market) and from remittances of Colombians living abroad. Sarmiento’s estimate was that about $US900 million to US$I .3 billion of narcodollars annually enter the black market and the “ventunilla siniestra.” Unfortunately, he did not estimate the size of the total black market. A third study presented at a recent conference on the underground economy in Colombia produced an estimate of US$2.5 billion in total dollars generated by the underground economy, of which approximately US$1.5 billion were from narcotics trafficking (Borrero. 1989.) According to this study, the rest of the foreign exchange entering Colombia through this market resulted from contraband exports of coffee. cattle, cement, and other products. A fourth study was undertaken by Urrutia (19YO). lie obtained an estimate of US$230-250 million for the value of narcodollars entering Colombia in 1988. His estimate was based on the value of dollars entering the Colombian Central Bank’s “ventanilla siniestru” (see below), which is not part of the black market. though it probably does receive large quantities of narcodollars, as explained by Urrutia. That is, in addition to the narcodollars entering Colombia through the black market, there are more funds that enter through the ventanilla siniestra, but which are part of the official market in Colombia. Urrutia explicitly did not consider the dollars that are delivered outside of Colombia to Colombians who buy them in the black market, since his concern was to examine the impact of the drug traffic and funds flows in Colombia.” A fifth estimate of the size of the black market was obtained from a large-scale carnhista in Bogoti. In an interview in May of 1990, he estimated a daily volume of about US$25 million. or about US$6 billion for the year 1990. This is about twice the previous estimates, which covered mainly the mid-1980s. This carnbtsta estimated that about 60% of the supply of black market dollars comes from narcotraffic - and virtually all of that (SS%) is dollars delivered outside of Colombia. The remainder of foreign exchange supply was estimated as 20% from return of flight capital, 10% as contraband sales on the Venezuelan and Ecuadorean borders. and 10% as transfers from Colombians living over-

seas to their local relatives. This carnhkta also provided estimates of the distribution of demand for foreign exchange in the black market: about 30% of the foreign exchange was estimated to be purchased by each of three categories of market user - general importers, by San Andresitos. and by flight capitalists-while about IS’% of the demand was by Colombians wanting to hold dollars domestically in Colombia. Kalmanovitz (1990) presented an estimate that Colombian narcotraffickers earned approximately US$4.0 billion per year during the 198Os, of which US$3.5 billion annually were taken into Colombia. This estimate, if coupled with the previous estimates of black market volume, would show virtually all of the black market dollars to be coming from this one source, a result that clearly is overstated due to the known supply of dollars from the other sources mentioned. Kalmanovitz assumes that almost all of the traffickers’ incomes are remitted to Colombia. which is at odds with all other discussions of this phenomenon. He does estimate that a total of US$4.6 billion per year were used by various classes of black market purchasers of foreign exchange, including the ventanil/a siniestra. This would place the narcotraffic as providing 76% of the total black market dollars. In sum, it appears that more than US$2.5 billion entered the black market annually by 19%). and perhaps significantly more according to the estimation of the large camhista. About twothirds of this money came from narcotics sales, principally cocaine sold in the United States.‘”

(b) The ventunih

siniestra

A foreign exchange phenomenon in Colombia that does not belong directly to the black market, but which does similarly take place outside of the commercial banking system is the Central Bank’s practice of receiving foreign currency from individuals without demanding evidence of its source. Similar to an amnesty program for failure to comply with taxes or other government regulations. this program operates through a special window at the Central Bank where anyone can turn in foreign exchange and receive pesos without incurring any penalty. Foreign exchange is received from various types of holders, such as tourists, professionals who provide personal services abroad, donations from abroad to Colombian families, and remittances of dividends and interest from abroad. While the program has operated since before the narcotics boom. all estimates show that it has become a

COLOMBIA’S

BLACK

major recipient of narcodollars during the 1980s. This means of exchanging foreign currency for pesos has become known as the vrntanilla siniestra (the sinister window), because of the suspected narcotics connection.2” Annual flows of funds through this account are listed in Table 1. Note that the value of services and remittances by 1989 had grown to over US$l billion, indicating that a very large quantity of narcodollars probably were being “laundered” through this vehicle. These funds do not enter the black market, since they are used by the Central Bank for official purposes such as financing legal imports and paying foreign debt - these dollars are sent physically to the US Federal Reserve Bank branch in Houston for crediting the Colombian Central Bank’s account.” 4. DETERMINATION OF THE BLACK MARKET EXCHANGE RATE The exchange rate in a black market may be expected to reflect several important relationships. First, it will usually offer an exchange rate worse than the official rate to sellers of local currency, both due to the excess demand for foreign exchange spilling over from the official market and since they pay a premium over the legal rate to compensate sellers of the foreign Table 1. Flows of fundr through the Ventmillu

Sinirstro

Year

Millions of US dollars

1970 1971 1972 1973 I974 1975 1976 1977 197x 1979 1980 1981 1982 1983 1984 19x5 1986 1987 19X8 1989

129.5 130.1 147.2 232.7 253.8 465.4 872.6 922.6 960.1 I .452.6 1,2x1.2 1,00x.2 720.0 489.9 403.8 s70.7 945.3 1J142.6 1,149.0 1,269.3

Source: series on otros”.

Banco de la Republica “servicios laborales y

MARKET

1199

exchange for the risk of this illegal activity (Michaely, 1954; Pitt, 1984). Second, it may show a relation between foreign and local currency that more closely reflects parity conditions between the two than the official exchange rate, since there are no restrictions on the black market other than its illegality (Koveos and Siefert, 1985). That is, participants on both supply and demand sides of the market are free to buy and sell at mutually agreed exchange rates, resulting in a basically free market (biased only by the transactions cost of paying an intermediary for the risk of undertaking the illegal transaction). Therefore, the black market rate can be expected to equilibrate supply and demand for the currency, and thus to reflect an equilibrium in the market. These two points are not inconsistent, since the government’s intervention in the official market leads to a nonparity condition, which is relieved in the freely determined black market. In virtually all cases, the black market exchange rate has a higher value of local currency per unit of foreign exchange than the official market, because of the foreign exchange scarcity that caused the market to develop. The Colombian case is quite striking for its lack of such a relationship. During the past 20 years the black market rate has closely tracked the official rate, sometimes even offering a discount under the official market! This section presents a model of the Colombian black market and seeks to explain the exchange rate there. The purpose of this section is to present a simple economic fundamentals model of determination of the black market exchange rate in Colombia during 1970-89. The main factors that can be expected to influence the exchange rate between pesos and US dollars are monetary and balance-of-payments influences (Connolly and Taylor, 1984; Dornbusch, 1983; Culbertson, 1989), plus the narcotraffic. For most of the relevant variables, monthly data were available for this period. End-ofmonth exchange rates for both official and black markets were obtained from the Central Bank.“’ The narcotraffic was only measured annually in the black market, so a moving-average time series interpolation was employed. In addition, another estimate was based on the flow of funds through the ventanilla siniestra, which is expected to demonstrate similar movement as the flow of narcodollars into the black market, though not at the same level as the black market flows. The model tested was: Black market exchange rate = a + hl (price level Colombia/price level United States)

1700

WORLD

- h2 (interest United States)

rate differential,

Colombia

DEVELOPMEN’I

vs.

This specification’” hypothesizes ;I positive rclntionship between the ratio of the aggregate price level in Colombia to that in the United States and the exchange rate defined as pesos per dollar. The greater the difference between Colombian and US prices. i.e.. the relative rate of Colombian

inflation.

the greater

the devaluation

of the peso

(i.e.. more pesos per dollar). This is one measure of the purchasing parity argument. in addition, the relative rate of return available to Colombian investors between instruments in pesos and in dollars is expected to itffcct the black market exchange rate. The preater the premium on peso-denomin~rted lnvcstments (such as hank deposits). the lower the expected dem~~nd for dollars and hence the lower the exchange rate (pesos per dollar). Thus. there should be ;I negative relationship between the black market rate and the interest differential (defit:cd as the peso interest rate on three-month certificates of deposit (CDs) minus actual devaluation for the period. minus the US dollar interest rate for three-month C’Ds). A third factor that COLII~ be expected to intlucncc the exchang,e rate is the trade balance. which is a proxy tor Colombia’s ability to generate forcipn exchange to supply the (official) market. In fact. a positive trade balance could he expected both to generate foreign exchange that put\ dowjnward pressure on the peso/dollar rate (i.e.. for the peso to revalue upward compared to the dollar) ~17d to generate confidence in Calombia’s ability to service its foreign debt both of which lead to an expected negative rrlationship between the trade balmce and the black market exchange rate. An alternative \pecific:ition of this relationship could ubc the chnpc in Colombia’s official foreign excl~anpc reserves to measure the pressure on the exchange rate. Since the

official

imbalances

reserves

balance

records

both

trade

c777dcapital account imbalances. it is really the bottom line in the balance of payments that would be expected to affect the exchang,e rate. As official reserves decline, the peso 1s expected to devalue relative to the dollar. because the reserve outflow results in a greater scarcity of foreign exchange. The fourth factor that is expected to influence the rate is the supply of narcodollars into the market. Clearly, if about 60% of the market is supplied by narcotics proceeds, then the flows of funds in that traffic will have an important impact on the supply side and hence on the price of

foreign exchange in the black market. The greater the supply of narcodollars, the lower the value of the pcsoidollar exchange rate expected in the black market. The estimation of this equation using monthly data from 197&%J resulted in the model shown in Table 2. Five separate specifications of the model appear in the table. each listed vertically in one column. The basic model (column one) produced highly significant coefficients for all four independent variables, as well as explaining almost ~111 of the variation in the black market exchange rate during the IO-year period of the 19XOs. The price level ratio between Colombia and the United States explained most of the variation in the black market exchange rate. as is gcnerall~ found 111 analyses of exchange rates In Jc\sdeveloped countries. That is. the m:tin factor contributing to the dcvaluatmn of Colombia’s peso relative to the US dollar wa5 the greater rate of growth c;f Colombia’s aggrcgite price level. The devaluation trend of the black market peso followed the price level diffcrcntial closely: in the model the relative price variable explained about 99% of the variation in the black market rate. The interest differential between short-term bank deposits in pesos and in (offshore) dollar\ by itself could explain about 7’?<) of the peso’s devaluation in the black market. Thi\ factor u’;I~ significant at the .01 level and correctly signed. Its relatively low ability to explain the exchange rate change\ in comparison to the other independent variables may rcflcct the fact that interest differentials are less of ;I consideration than purchasing power and other factor\ in (‘olombians decisions to obtain dollars. Narcotics traffic (as measured by Gomez) cxpl:~inrd an additional part of the variation in the black market exchange rate. This variable was highly significant (at the .OI level), and by itself explained over 70% of the variation in the black market exchange rate. Because this sariable was estimated as a time-series interpolation of annual data. and was only available for the IOXOs. another measure is u\ed below to retest the model. The trade balance coefficient is incorrectly signed but significant (at the .05 Icvel). By itself. this variable explains some 42% of the variation in the black market exchange rate. This outcome may be due to misinterpretation of the direction of causation between the trade balance and the exchange rate; it is possible that the devaluations of the peso led to improved trade balances during this period of time. This outcome alternatively may bc due to the fact that Colombia suffered through the same debt crisis as other Latin American countries throughout the Ic)XOs, and ;I

COLOMBIA’S

‘l‘ablc Dependent lY70

variable:

through

BLACK

2. Drtcrrnirurrion of ~‘olombiu’s

Rluck

black market exchange rate (monthly

December

Model

I

(OLS) -2.5.78 (.036)

term

Measure

of inflationary

prcssurelColombian price index/ US price

index

Measure of dollar availability$ trade balance

Measures flow5

Markc/

cwhcwgr

data supplied by the Banco de In Reptiblica

Parameter estimates (significance level) Model 1’ (GLS)

214.73

Model 1 (GLS)

-7.83 (.IXl)

ps.71 (.007)

NY. 76

2OY.30

(.OOO)

( .OOO)

0.02 (.04X)

0.01

(.OOO)

Model 2 (OLS)

Model

2

(GLS)

-8.74 (.OOO)

-0.x2 (.OOh)

I X7.YX (.OOO)

1x3.74

-0.17

-0.004 (.72X) -

(.OOO)

of narcotics

(.OOO)

sirrirslru 2. Cocaine mcasure

F-ratio

for January

(.X3)

I Vrrllcltlrllr

Adjusted

rrrtr. IB70-89

19X9)

Independent variables

Constant

1201

MARKET

flow

O.YY3

R’ of regression

Durbin-Watson

-0.06 (.Ol I)

value

2.442.03 (.OOO)

-0.12 (.037)

-0.0x

(.052) 0.941

0.960 327.28

756.26

(.OOO)

0.5X

(.OOO)

0.6X6 240.81

(.OOO)

O.Y7

‘This model presents only the significant variables that remained in I-The relative growth of the money supply in Colombia vs the specification. that produced similar results with a slightly wcakcr $It was not meaningful to run regressions using official rcscrve Ilows. the relative monetary growth &able

positive trade balance was virtually forced on the country in order to try to make payments on the sizable foreign debt.” Therefore, a positive trade balance may have been an indicator of the external debt problem more than an indicator of Colombia’s actual availability of foreign exchange. A better measure of foreign exchange availability would be total official reserves flows, which are discussed below. Finally, the result obtained in the present specification was foot robust to other specifications: so the trade balance may not be a key factor in determination of the black market rate, despite this first model’s outcome. Because the model showed a high degree of autocorrelation among the error terms. a respecification was carried out using generalized least squares (GLS). This specification is presented in column 2 of Table 2. Note that except for the price differential and narcotics flow variables. the

(.OOO)

0.966 3.225.97

the generalized Icat squares specification. United States was used in an alternative fit. since these Ilows wcrc YY’%, correlated with

other independent variables became insignificant in this specification. The revised model still explains 96% of the variation in the black market rate and is highly significant. A third specification is presented in column three to show only the significant independent variables using the generalized least squares technique. Next, the model was respecified using an alternative indicator of the narcotics money flows. namely, the flow of money through the ~rntnnilla siniestrrr. The results of this respecification also appear in Table 2. These versions weakly confirm the model of black market exchange rate determination based on relative price levels and narcotics trafficking. The ordinary least squares (OLS) model explains 97% of the exchange rate variability and has both variables significant at more than the .OOl level. Correcting for autocorrelation of the error terms using generalized least squares causes the ~cnta-

1202

WORLD

DEVELOPMENT

nilla siniesfra measure to become insignificant, though correctly signed. The mode1 now explains only 69% of the variation in exchange rates.‘s The measure of narcotics trafficking taken from US government estimates for 1981-88 as shown in the first model (columns l-3 in Table 2) was highly significant and correctly signed. supporting the inference that narcodollar inflows have a strengthening effect on the peso’s value. The measure of narcotics-related supply of dollars into the black market proxied by flows through the ventanilla siniestra provided similar support, though the mode1 was less significant. Note that the model using the direct estimates of narcotics trafficking only include data for the 1980s so this model does not cover the previous decade.2h In sum, the Colombian black market, peso/ dollar exchange rate is highly positively correlated with the rate of inflation (Colombia vs. United States) and highly negatively correlated with the inflow of narcotics dollars into the market. The former relationship supports the hypothesis that purchasing power parity tends to be the most significant driving force behind exchange rate adjustment in Latin American currency black markets. The latter relationship supports the hypothesis that the currency black market is a significant recipient of narcoticsrelated dollars, and thus serves to facilitate the drug trafficking business. It is likely that the reason for the very small divergence of the black market exchange rate from the official rate during the past two decades is the very large supply of narcodollars into the black market, which push the price of dollars down in that market.

5. IMPLICATIONS OF THE BLACK MARKET TO COLOMBIA It should be noted before any discussion of the specific situation in Colombia that the possibility for eliminating a black market is quite small, unless the government becomes both willing to permit a fully free foreign exchange market as in most industrial countries today and able to eliminate the underground economy. Black markets in foreign exchange continue to flourish in less-developed countries into the 199Os, despite efforts to restrict them or make them unnecessary by opening up the official markets. Based on conditions around the world since the mid-1980s. economic openness has become widely acceptable, so that government willingness to operate a free foreign exchange market may remain in Colombia. Examination of the

balance of payments of Colombia during the past decade shows that the trade balance has regularly generated foreign exchange-and the amount of capita1 flight has been much less than in many other Latin American countries. Imports in 1988 were about US$4.5 billion (vs. exports of US$5.3 billion); this compares with an estimated black market volume of US$2.5 billion. Because of the large size of the black market relative to import spending, any change in the availability of dollars in the black market could conceivably have an important impact on the stability of the official exchange market. Despite this last caveat, it appears that a free foreign exchange market could be sustained in Colombia without excessive economic disorder - and this expectation has been borne out by the evidence through 199 1. The black market operating during the 1980s had several important macroeconomic impacts on the Colombian economy. First, it provided an outlet for excess demand for foreign exchange on the part of importers and flight capitalists. Second, it enabled the sellers of Colombian contraband products (including narcotics, gold, emeralds, etc.) to find buyers for the foreign exchange that they chose to remit to Colombia. These two influences apparently were of approximately equal size. such that the black market exchange rate generally fell close to the official rate. Thus, one could not conclude that the black market pressured the official exchange rate to be over- or undervalued. On the other hand, the supply of narcodollars apparently had a great influence on the black market exchange rate; without this influence the rate would have been much higher (pesos per dollar) and would have put pressure on the official peso to devalue and make Colombian exports cheaper abroad. The income from narcotraffic, therefore, allowed Colombia to maintain an exchange rate that was less favorable to exports than would have obtained without this factor. A third macroeconomic impact of the black market on the Colombian economy results from any added consumption and investment that were enabled through this mechanism. All of the suppliers of foreign exchange through the black market were able to obtain pesos, which presumably were used in purchases and investments within Colombia. This activity provided a stimulus to the economy. The flight capital that left the country through the black market weakened local investment. since this part of Colombian savings was shifted abroad. In addition, the incremental imports that were paid through the black market may have competed with Colombian products, again depressing the local economy. On balance, the income effect of the

COLOMBIA’S

BLACK

black market is ambiguous though some studies have found that the net impact of the narcotraffic is negative on the Colombian economy (Sarmiento, 1991). As in other instances in Latin America and elsewhere, the foreign exchange black market is used for tax evasion. The Colombian black market is also used in the operation of the underground economy, particularly in contraband exports of gold, emeralds, coffee, and cattle. Unless the tax regime and other regulatory policies are altered to make the black market less attractive to these users, it will still be used. In addition, the narcotraffickers must resort to some form of black market, since their entire business is illegal and their revenues are primarily received in foreign exchange. Thus, to avoid taxes and/or narcotics violations, those black market participants involved in contraband trade will continue to require a black market, regardless of the balance-of-payments situation in Colombia. It makes sense in this context to eliminate the restrictions on access to foreign exchange for importing and for investment overseas by Col-

1203

MARKET

ombians. By legalizing these activities, the Colombian government may be able to isolate the narcotics traffic and unreported exports of coffee, gold, and emeralds in the black market that remains. Thus, if the opening of the legal foreign exchange market to accommodate Colombian importers without rationing and to permit “capital flight” when investors so desire does not destabilize the market with overwhelming demand for dollars, then such a policy can continue to be used. The black market clearly has deleterious effects in the domestic Colombian economy. By facilitating the operations of the narcotraffickers, it contributes to a huge social and political problem with international ramifications.” By rewarding the purchasers of foreign currencies with lower cost access to foreign exchange than the official market (after taxes are considered), it again contributes to the subversion of the domestic system. These costs must weigh heavily in comparison with the inflow of dollars and other foreign currencies that Colombia’s black market permits.

NOTES

I. Among the numerous studies of this phenomenon are several recent analyses of Latin American countries, including Dornbusch et al. (19X3), Canto (1985), and

Grosse

(1991).

2. Lindauer (1989) distinguishes among these labels based on several criteria. Parallel markets can be subdivided according to the degree of legal enforcement of the foreign exchange rules (e.g., black market implies more legal sanctions than grey market). He also uses the term parallel market to cover the broad heading of exchange markets that develop in response to government intervention in the official market that limits participation in some way(s). The parallel market concept includes both legal and illegal markets that develop in response to the restrictions on the official market; black markets in foreign exchange are illegal typically due to their involvement in effecting payment for contraband trade. 3. Implicitly here, the existence of the foreign exchange black market has a negative welfare impact on the economy, because it facilitates the operation of contraband (i.e., illegal) business. 4. Legal sanctions such as fines, funds confiscation. and imprisonment should serve to create a premium on use of the black market for both buyers and sellers of foreign exchange. Thus. the sellers typically face a better deal in the black market due to excess demand for foreign in the official market - but this deal would

have

been

even

better

if no sanctions

existed

5. The International Currency Analysis (various issues) offers a succinct and very useful commentary on currency black markets around the world. Both exchange rate quotations in black and official markets and analysis for the differences between the two markets are presented. 6. The Peruvian case offers an interesting contrast to the Colombian one, since this country is also a recipient of large quantities of narcodollars. Nevertheless, the underlying “formal” economy and balance-ofpayments situation were so weak during the 1980s, that the excess demand for dollars spilling over from the official foreign exchange market far surpassed the narcotics proceeds that were brought into the country - resulting in a very large premium on dollars in the black market. See Grosse (1991). 7. Decreto-Ley Numero 444 de 1967 (March 22) sobre regimen de cambios internacionales y de comercio international. A foreign exchange black market also existed under previous periods of exchange controls. 8. A very useful analysis of the pressures that led to Colombia’s exchange controls appears in Nelson, Schultz and Slightor (1971), chapter 7. A history of the Colombian exchange controls appears in International Currency Analysis (various years). Despite lradc sur-

WORLD

1204

DEVELOPMENT

pluses in most years since 10X-1. the country chose to maintain the exchange control regime until early IYYI. 9. At the same time. elimination of exchange controls could help to reestablish confidence in the Colombian economy, which would lad to reduced capital flight and thus II reduced demand for forcipn exchange. IO. Much of this section is based on detailed interviews with Colombian c~rrmhi.srrr,sin Colombia and in the United States. Both large-scale foreign exchange dealers (who operate in transaction5 of approximately US$lO.OOO or more) and retail dealers wt‘rc‘ interviewed. as well II\ other intermediaries involved. Additional evidence comes from published records of money laundering prosecutions of Banco tic Occidcnte and Bank of Credit and Commerce International.

11. Useful participants

descriptions 01 the market and its various appe:tr in Cuevas (15%) and Rodriguez

(1%3X). II. The .Sctr~Atldrr.si/o.\ arc named for the Colombian island of San Andre\. which lies off the coast of Nicaragua in the Carihbcan Sea. This island is H free port. into which products arc imported without taxc\ or much of the merchandise is tariffs. Subserluently. smuggled into Ihc Colombian mainland lor sale as contraband by the people from San Andres. Over time. the label Sntr Atldrc.\i!o has come to mean any contraband acller in Colombia; in the large cities these vendors arc actually gathcrcd into specific districts where they ply their wares side by Gde. Ii.

Sec. for example,

Perry

(1900).

14. This process of Iaundcring narcodollars is described in detail in the documents presented by parties in the Banco de Occidente legal GISC. Criminal cast‘ #81)-0X&A. “U.S. vs. Pablo Emilio Escohnr-Gnviria CI District of tieorgia. Laundering ol 01. ,- Northern Colombian cocaine money is ;II~o discussed in Mermelstein (IYYO) and Rodrigue7 (I%#). l5. The proliferation ol camhisko makes it certain that the regulatory authoritica are aware of this black market buaincss. Some dcgrec of enforcement of the exchange controls is indeed undertaken. In IYXY some 500 casts were brought against illegal holders of US dollars in Colombia (Esrr~tqirr, May 1YYO. p. 5.), and in IYXX half a dozen stock brokerage firms were forced to leave the Bogotli exchange, primarily due to their high-volume and high-visibility activities in black market foreign ewchangc. Set Rodriguez (IWX).

1X. This estimate would he consistent with the others, if it truly measures the dollars that physically enter Colombia. The other analyses show that most narcodollars remain outside of Colombia and are exchanged for pesos that are paid in Colombia. Following the Inrgc carnhisitr’.s estimate that 85% of the narcotics-related dollars remain outside of Colombia. Urrutia’s estimate of the dollars that do enter Colombia would be 15% of the total. which in turn would bc about US$1.533 billion for that year. This figure is quite similar to the other estimatca. IY. A final estimate that may bc useful to consider in this context coma from a study by Cahallcro (IYXX). He estimated that the cocaine trade gcncrated about USS4 billion in revenues for Colombian traffichers. of which about US$I billion would have been brought into Colombia through the black market. This estimate is lower than moat of the others presented here. primarily because of the author’s assumption 01 a far smaller return of the nnrcodollara to Colombia. 20. Discussion3 of the rsc,,~trrr~i/lrr.vioic.,lrtr appear often in the Colombian prc\\. See. for example. Cadavid (IYXX); bee i&o Gomez (IYXY). 21. Information obtnincd from the US Trcasurv. Financial Crimes Enforcement Network (FINCEN) in August. IYYO. The C‘olombian government also prohably buys proceeds of narcotraftic through its purchaacs of gold mined in the country. Since the narcotics traffickers arc known to have cntcrccl the gold huxincss. any gold that they \cII to the povcrnment would be link4 indircctty to the narcotics businas, though not ncccsa;irily to foreign exchange tr;mssction\. That i\. SO~C‘ of the gold may bc purchased by the narcotraffickcrs for dollars. and more gold i, produced by mines controlled by the narcotraffickcra. Sec. e.g.. Kalmanovitz (1000) p. 2.1. 22. ‘The black market 5crics. 01 course. i\ not “olli&I,” but the ratch have been widely publicized, appearing in several daily newspapers in Colombia during the late tYXO\. 23. An additional variable that would be cxpcctcd to affect this exchange rate is the penalty for getting caught using the black market. This penalty wa\ essentially constant durin g the period of time under study. so it did not influence variations in the black market exchange rate. ‘4.

See, for example.

Kuczynski

(10X7).

16. Thcsc pcnaltics arc discussed in Wills (19x9). WIIIS was Superintendent of Exchange Controls at the Central Bank during the Barco administration in the IYXOs. An annual statement of the prosecutions undcrway appears in the I/+rrm, ,4rllrcll of the Supcrintcndcncia de Control dc Cambios.

25. Addirional specifications lor determination of the black market exchange rate were also tried. .l’he offici;il reserves balance of payments was so highly correlated with the monetary/inflation variable (r = O.YY) that it could not bc included in the regression. The relative growth of the money supply in Colombia versus the United States provided a similar fit to that obtained by using the relative price levels, with no impact on the significance or signs of the other indepcntlcnt variables.

17.

26.

See Rodrigucl

(19Xx).

chapter .5.

‘l‘hc measure

itsell

was obtained

as an annual

COLOMBIA’S

estimate. Using a simple moving-average time series to construct monthly data.

smoothin! interpolation

BLACK

technique. a was created

1205

MARKET

27. See, for example, Tokatlian and Bagley Gomcz (1990). and Sarmiento (1991).

(IYYO),

REFERENCES black marAggarwal. Raj, “The Nature of currency kets: Empirical test of weak and semistrong form efficiency,” Intm?utiona/ Trade Journul Vol. V. No. 1 (Fall 1990). pp. 1-24. and the war on drugs.” Bagley. Bruce. “Colombia Foreign Affuirs Vol. 67. No. 1 (Fall 1YXX).pp. 7&Y2. Bhandari, Jagdeep and Bernard Decaluwe. “A framework for the analysis of legal and fraudulent trade transactions in parallel exchange markets,” We/twirlschuftliches Archiv, Band 122. Heft 2 (June lY86). pp. “233-253. subtcrBorrcro. Oscar, “La finca raiz y la economia ranea.” Camacol seminario (Bogota: November X. 198’)). Caballero ArgBez, Carlos, “La Economia de la Cocaina, Algunos Estimativos para 1988,” Cuyuntura Econdmica Vol. 18, No. 3 (September 198X), pp. 179-183. Cadavid, Fernando Gaviria, “Cuanto vale el narcotr& fico,” La Reptiblicu (February 18, 19X8). pp. E7-8. Canto, Victor, “Monetary policy, dollarization, and parallel market exchange rates: The case of the Journal of Internutionul Dominican Republic,” Money and Finance Vol. 4. No. 4 (September 1985). pp. 507-521. Connolly, Michael. and Dean Taylor, “The exact timing of the collapse of an exchange rate regime and its impact on the relative price of traded goods.” Journul of Money, Credit, und Bunking Vol. 16. No. 2 (May 1984). pp. 194-207. Cucvas. Angela. Lu ofru (‘uru del Dolur (Bogotri: Tcrcer Mundo, 1986). ’ Culbertson, W. Patton, “Empirical regularities in black markets for currency.” World Drrv/opmrt7~ Vol. 17. No. 12 (December 19X9), pp. 1YO7-IYlY. Culbertson. W. Patton, “Purchasing power parity and black market exchange rates,” Sourhern Economic Journul. Vol. I3 (June lY75). pp. 2X7-296. Diaz-Alejandro, Carlos, Foreign Trude Regimes und Economic Developmenr: Colombia (New York: National Bureau of Economic Research, 1976). Dornbusch, Rudiger. et al.. “The black market for dollars in Brazil.” Quurrerly Journul of Economics Vol. 98, No. I (February 1983). pp. 25-40. “El Control dc Cambios para Out?“. Eswutegia. (May 1900). pp. 4-6. Gomez, Hernando Jest, “El Tamario del Narcotrafico v su Impact0 Econ6mico.” Economia Colombiana No. 228-7 (February-March 1990) pp. %17. Gomez, Silverio, “Mientras las Exportaciones de Bienes Crecen 11%. La Ventanilla Siniestra estj Dispaiada,” El Tiempo (April 24, 1989). pp. lB-4B. Greenwood, Jeremy. and Kent Kimbrough, “Foreign exchange controls in a black market economy,” Journal of Dn~rlopmm~ Economics Vol. 26, No. I (June 1987). pp. 129-143. Grossc. Robert, “Peru’s black market in foreign

Journul Affuirs Vol.

exchange.” World

135-164. International

of Inrerumericun Studies 33, No. 3 (Fall 1991).

und

pp.

Currency Analysis, World Currency (Brooklyn. NY: ICA, various years). del NarcotrriKalmanovitz, Salomhn. .ILa Economia fico en Colombia.‘. Economiu Co~ombiunu No. 22G 7 (February-March 1990). pp. 1X-2X. Koveos. Peter, and Bruce Siefert. “Purchasing power parity and black markets,” Finunciul Munugemenf, ~oI.~ 14. No. 3 (Autumn lYX.5) pp. [email protected]. Pedro Pablo. “The outlook for Latin (Fall 10X7). ArGerican debt,” Foreign Afluirs “Parallel. fragmented, or black? Lindauer, David. Defining market structure in developing economies,” World De~~rlopmenr Vol. 17. No. 12 (December IYXY) pp. 1x71-IXXO. Mermelstein. Max. The Mun Who Made ir Snolt. (New York: Simon and Schuster. IVY(l). Michacly. Michael, “A geometric analysis of black market behavior.” Americun Economic Review, Vol. 44, No. 4 (September 1954) pp. 627-637. Nelson. Richard, Paul Schultz. and Robert Slighton, Strucrwul Chuqc in u Developing Economy (Princeton, NJ: Princeton University Press. 1971). Nowak. Michael, “Quantitative controls and unofficial markets in foreign exchange: A theoretical framework.” IMF Srulf‘ Pupers Vol. 21. No. 3 (September lYX4) pp. 40+4x1. Perry, Guillcrmo. “Minas y Energia.” in (‘olomhiu Siglo XXI, Vol. 2 (Bogota: Edicioncs Imprcsores, 1990). pp. 129, 159-160. Pitt, Mark, “Smuggling and the black market for foreign exchange,” Journul of Infernutionul Economics. Vol. 16, No. 4 (May IYX4) pp. 243-2.57. Rodriguez, Hector Mario. Los Pirurus de IN Bolsu (BogotB: Peyre, 1YXX). Sarmiento. Eduardo, “Economia del Narcotrlifico,” in Carlos Gustav0 Arricta, Luis Javier Orjuela. Eduardo Sarmiento Palacio. and Juan Gabriel Nurcowdfico en Colombia, 2nd cd. Tokatlian. (Bogota: Universidad de 10s Andes. 1991). Shiekh. Munir, “Black market for foreign exchange, capital flows. and smuggling,” Journal of Developmetif Economics, Vol. 3. No. 1 (March lY76) pp. Y-26. Tokatlian. Juan. and Bruce Bagley (Eds.). Economicty Poiiricu de1 Nurcotrdfico. (Bogota: Universidad de 10s Andes, 1990). Urrutia. Miguel, “Anilisis Costo-Beneficio del TrBflco de Drogas para la Economia Colombiana.” C’oyuttrut-u Econdmica, (October 19YO), pp. 11.5-126. Wills, Emilio, “Regimen de Control de Cambios en Colombia - Origen. Evolution. Aspectos Generales.” Revisru Derecho Privudo (Bogota) Vol. 3, No. 5 (19X9), pp. 127p13Y. Yeurhook