Shock therapy and the market for foreign exchange in Russia: whither currency convertibility?
In January 1992, President Boris Yeltsin introduced a collection of radical economic and political reform measures designed to initiate the conversion of the Russian economy from one based on central planning to one oriented toward market forces. One element of Russia's "shock therapy" (Voprosy ekonomiki no 2, February 1993) involved liberalizing foreign trade and finance. Previously, planners had maintained a monopoly position in all decisions over the volume and assortment of imports and exports, as well as the exchange rate. Despite numerous policies adopted to liberalize the foreign currency market by introducing multiple rates and allowing more widespread participation, the Soviet legacy of centralized control over foreign trade and finance continued for much of 1992.
Currency convertibility has been an important component of the conversion from plan to a market in other transition economies (McKinnon 1993). Failure to adopt currency convertibility prolongs the transformation by maintaining distorted prices and preventing resources from being guided to their most efficient end uses. Maintaining a non-convertible currency also acts as a barrier to foreign trade and investment. However, incentives to avoid currency convertibility, at least in the short run, were numerous in Russia. Unlike other transition economies, Russia's resource endowment and industrial infrastructure may have allowed it to forestall opening its economy to the rest of the world. Failure to adopt currency convertibility meant Russia could continue to run budget deficits and print rubles without any external discipline. Thus, while there were discussions of the importance of unifying the exchange rate and moving toward achieving ruble convertibility, schizophrenic policy-making begun during perestroika, Mikhail Gorbachev's program to restructure and revitalize the Soviet economy, was continued during the initial stages of Yeltsin's economic reform program in Russia. The net result shows up in an insignificant narrowing of the parallel market premium, the percentage difference between official and black market ruble-dollar exchange rates. Moreover, using weekly exchange rate data, we find that despite declining variability in buy-sell spreads in the official and black foreign currency markets, mean buy-sell spreads remained much higher in 1992 than in developed countries.
To analyze the impact of Russia's "shock therapy" on the foreign currency market, we first provide an overview of the economic and policy conditions characterizing the situation at the beginning of the transition process. Section 2 describes the exchange rate data utilized in this study and the methodology employed to evaluate whether the official exchange rate was appropriate for introducing currency convertibility. In Section 3, using weekly ruble-dollar official and black market exchange rates in 1992, we estimate the percentage difference between the official and black market ruble-dollar exchange rates, the parallel market premium (PMP), for Moscow, St. Petersburg, and Vladivostok. We find remarkable conformity in the PMPs across disparate regions of the former Soviet Union. Calculating buy-sell spreads within the official and black markets as an independent check on the veracity of our PMP results, we find no significant reduction in the spreads in 1992. If constraints on foreign currency transactions had been removed, both the PMP and the buy-sell spreads would have fallen significantly. Section 4 summarizes our findings and offers an assessment of whether Russia, more than five years after the initiation of transition, has established the successful preconditions for successfully introducing currency convertibility.
From Perestroika to Shock Therapy
Soviet planners viewed foreign trade as an important mechanism for acquiring materials and technology to enhance overall economic performance. Planners, under the auspices of the Ministry of Foreign Trade, maintained the exclusive right to govern the activities of hundreds of foreign trade organizations, FTOs, that specialized in the buying/selling of particular goods or materials so that each year the quantities of goods sold were sufficient to finance the requisite imports, as called for in the economy-wide material balance plan (Gregory and Stuart 1993). FTOs did not communicate among themselves, nor did they communicate with the domestic firms that supplied or used the materials or products that they traded. That is, the FTO selling Soviet oil had no contact with the FTO buying drill bit equipment, nor with the FTO selling lumber. Lumber mills were not informed about the destination or price of the lumber they delivered to the FTO, nor did they receive any of the hard currency earned from the sale of lumber they supplied. MOFT maintained its monopoly position over foreign trade activities through legal restrictions that prohibited firms and other organizations from contacting foreign companies, as well as by maintaining an inconvertible currency. MOFT established arbitrary exchange rates to settle accounts rather than to facilitate foreign trade transactions: the ruble-dollar exchange rate of .64 rubles to $1, for example, remained constant for two decades.
Persistent and pervasive shortages in the Soviet economy made hard currency a much sought after commodity since it gave the holder access to goods not readily available in domestic markets. In the planners' view, access to hard currency was an enormous carrot. Greater access to this carrot came in September 1986 when planners, under the guise of perestroika, not only allowed select industrial ministries and firms to negotiate directly with foreign counterparts,(1) but also permited firms to establish joint ventures with foreign companies (Lawrence and Vlachoutsicos 1990). Planners expected the incentive effect associated with access to hard currency to cause firms to produce a greater volume of high quality output for sale in Western markets. Export earnings could then be used to finance the renovations required to make Soviet industry competitive in global markets.
Soviet firms were slow to respond to their newly-granted freedom. First, firms previously isolated from global markets and without adequate communication equipment found it difficult to gather information about buyers, sellers, prices, relative quality and so forth. Finding a 20' or 40' container for shipping purposes often proved impossible. Second, the technological level of many firms precluded effective competition in world markets. Third, the incentive to engage in foreign trade activities was reduced by the requirement that firms surrender 60-100% of their hard currency earnings (not profits) at the state-determined rate (Jeffries 1993). Further complicating foreign trade transactions were the numerous exchange rates in place at the time -- the rate that a firm was required to use for any particular transaction depended upon the product and country involved in the transaction!(2) Moreover, export licenses were required, but were limited in number and not available for many goods or materials that could readily be sold in world markets.(3) For example, firms wanting to export aluminum scrap or used copper wire were unable to do so, even if they found buyers.(4) Nonetheless, a growing number of firms engaged in foreign sector activities by early 1990, thus MOFT's monopoly position over imports and exports diminished.(5)
Reducing restrictions on domestic firms and allowing joint ventures represent one step toward liberalizing foreign trade. Interestingly enough, however, not until February 1988 were banks given the right to conduct hard currency operations, and only then with the permission of the Bank for Foreign Economic Relations (Vneshekonombank).(6) Moreover, Soviet citizens were prohibited from legally obtaining or holding hard currency;(7) only foreigners had the right to hold hard currency, which they could spend in special tourist shops or exchange for rubles at planner-designated locations at the planner-determined rate. Thus, by any measure, the first step toward foreign trade liberalization was a small one.
The second step toward decentralizing the foreign trade sector of the economy during perestroika was facilitated by the foreign currency auction organized in November 1989 by the Bank for Foreign Economic Relations (Vneshekonombank). However, because only state-owned firms with "surplus" foreign currency were eligible sellers, the trading volume was small, representing less than 0.5% of the hard currency import bill (Hanson 1990, p. 9).
The third step toward foreign trade liberalization occured with the introduction, also in November 1989, of a dual exchange rate system. One planner-determined rate was to be applied to commercial transactions; that is, planners set a rate of 6.24 rubles per $1 (a ten-fold ruble devaluation) to establish a firm's hard currency surrender obligations.(8) This rate also was applied to the settlement of current account transactions, as well as for accounting and customs valuation purposes. A second rate was to be applied to non-commercial transactions; that is, one to be used by foreigners trading dollars for rubles at the bank and by Soviet citizens providing documentation of the right to travel abroad. This "tourist" rate not only involved a much higher ruble-dollar exchange rate, but also restricted to $200 the amount of hard currency that could be obtained.
Further steps toward liberalizing foreign trade and finance occurred in 1990. Internal currency convertibility was achieved in August when Soviet citizens were granted the legal right to hold hard currency and open bank accounts without having to explain the source of their hard currency earnings -- hard currency accounts previously were restricted to those who had worked abroad or who were in contact with foreign companies.(9) In October 1990, a presidential decree granted all Soviet firms the right to buy/sell hard currency for rubles at "market" (Vneshekonombank currency auction) prices, although individuals and foreign firms remained ineligible to participate in currency auctions.
Forward progress was reversed by a number of presidential decrees in November 1990 designed to centralize the control and management of foreign currency earnings, in part to ensure that foreign debt obligations were met.(10) Rather than establish a single exchange rate based on currency auction results, four exchange rates were put in place: a planner-determined commercial rate which served as the foundation for foreign trade transactions as well as for establishing a firm's hard currency surrender obligations, a planner-determined official rate used for statistical purposes and for calculating debt payments by developing countries; a planner-determined tourist rate, least 50% of their hard currency earnings. Countering liberalization with control, the number of imports subject to tariffs rose from 319 to 5,019 (Jeffries 1993); export licenses were required for more than 70% of the raw materials and goods sold abroad; controls also were established to prevent capital flight. As if this were not complicated enough, foreign firms, still restricted from participating in currency auctions, were arbitrarily subject to planner-determined commercial rates for some transactions and adjusted auction rates for others. To suggest that foreign trade or the exchange rate regime had been decentralized or simplified during the course of perestroika grossly misrepresents reality.
Perestroika ended with the coup attempt in August 1991. In the chaos that followed, augmented by the dissolution of the Soviet Union in December 1991, the question of who would be responsible for Soviet foreign debt and other foreign economic commitments had dramatic consequences for efforts to liberalize foreign trade and finance. Representatives from the former viet republics met in Moscow, agreeing to share debt responsibilities and undertake efforts to reschedule foreign debts. Within months, this agreement was no longer valid, and Vneshekonombank found itself 30-50% short of the hard currency requirements needed to meet G7 stipulations. In December 1991, to obtain sufficient foreign currency reserves to make short-term debt payments, hard currency accounts were "frozen" by Vneshekonombank -- firms and individuals were told that at some point in the future they would receive in rubles an amount in some way related to the amount of hard currency in their account.(11) Confidence in the banking system, already diminished by the 50- and 100-ruble banknote confiscation in January 1991,(12) plummeted.
The failure of perestroika to improve the operation and performance of the Soviet economy paved the way for more radical reforms. The Russian version of "shock therapy" to initiate the transition from plan to market was introduced in January 1992. Of interest here are those policies affecting foreign trade and finance. The first, On liberalizing foreign economic activities, announced in late 1991, purported to scrap the existing multiple exchange rate regime, including the compulsory surrender of 40-50% of a firm's hard currency earnings. Registration restrictions on Russian firms engaging in foreign trade activities also were lifted. At the same time, Yeltsin signed a decree to take effect within 6 months that would ban all settlements in foreign currency on Russian territory and close all non-state-owned trading outlets selling goods for hard currency.(13) By the end of the year, however, in an effort to generate hard currency reserves to help pay off foreign debt and finance imports, Yeltsin issued another decree requiring all firms to sell at least 10% of the foreign currency earnings at the market (auction) price. Some firms, exporters of gas, oil, timber and precious metals, for example, were to sell an additional 40% of their hard currency earnings to the Central Bank at the planner-determined commercial rate. That is, firms that supplied about 80% of Russia's exports by volume were required to surrender 40% of their hard currency earnings at about half the market rate (Goldberg 1993); an additional 10% was to be supplied at the market rate.(14) After meeting the mandatory surrender obligations, firms could retain the balance of their foreign exchange earnings in special bank accounts, or sell it at weekly currency auctions (MICE). In sum, the liberalizing aspect of the initial reform program was limited to the elimination of the differentiated foreign exchange coefficients.
Similar schizophrenic behavior, decentralize yet maintain control, also is evident in the commitment to making the ruble internally convertible. On the one hand, greater reliance was placed on the role of banks in facilitating access to foreign currency: permanent residents were granted the right to establish hard currency accounts in designated banks, importers were granted the right to buy hard currency at designated banks at the market (auction) rate of exchange, and households maintained their right to establish hard currency accounts and to acquire limited amounts of hard currency for travel purposes (if visa and other required documentation presented). On the other hand, planners continued to set the tourist rate and the limits on hard currency acquisition,(15) as well as to maintain at least 25 separate exchange rates for imports, ranging from 1.7 rubles per dollar to 70 rubles per dollar.(16) Despite rampant inflation and a nascent commerical banking system, Russians faced severe restrictions on establishing accounts in foreign banks.
The voltage on the "shock therapy" increased in March. To protect their newly-granted membership in the IMF and to enhance the possibility of receiving Most Favored Nation status, Russian leaders announced a number of policies in March 1992, two of which had implications for liberalizing foreign trade and finance. One aim was to establish a "crawling peg" between 40 and 50 rubles per $1, although no action was taken at the time to implement this proposal. A second was to introduce by July 1992 a temporary dual exchange rate system, replacing the multiple exchange rate system that had been adopted in 1991. Under the dual exchange rate regime, one rate would apply to capital flows into Russia that would be set in such a way as to make it more costly for foreigners to buy up domestic assets; a second rate would apply to all other transactions. In addition to these policies, IMF appeasement required that the Central Bank stabilize the ruble, which, despite the hyper-inflationary situation in the first half of 1992, it was able to do for two months by selling its reserves. Indeed, the Central Bank supplied about 80% of the foreign exchange sold in the MICE in 1992. Efforts to stablize the ruble over a longer term were relatively ineffective, however, given the monetary and fiscal policies pursued at the time. One example suffices to illustrate the situation: within days of the June decree on bankruptcy, sufficient additional credit was made available to forestall the closure of all firms qualifying as bankrupt.
The schizophrenic behavior continued as the second stage of the reform program was put in place in July 1992. According to announcements at the time, a single (unified) exchange rate was to be established and the ruble was to become internally convertible with respect to current account transactions by the end of the month.(17) The exchange rate utilized for foreign capital evaluations was to be based on the MICE rate, although neither the unified rate, nor the foreign capital rate was to be equal to the MICE rate. That is, on the same day the unified rate was set at 125.26 rubles per $1 (1 July 1992), based on an average rate over the previous month, the MICE rate was 144 rubles per $1.(18) Like the earlier proposed exchange rate regime, the ruble was to float for a month to find a defendable level. At this time, however, it was imagined that the rate would be 80 rubles per $1! Based in principle on twice-weekly auctions, this new unified rate was to be applied to all transactions,(19) thus establishing current account convertibility for the first time in Russia since 1917 (Lipton and Sachs 1992).
Offsetting the announced simplification of the exchange rate regime, a number of other policies which undermined the liberalization process were announced and/or adopted in conjunction with the reform program. For example, in order to maintain budget revenues, it was announced that export taxes would be doubled, to an average of 40% for raw materials exporters.(20) Perhaps it was intended that the export tax would substitute for compulsory sales of foreign currency earnings. In reality, exports continued to be subject to mandatory currency sales to the Central Bank: of the 50% required surrender, 30% had to be sold to the Central Bank and 20% to commercial banks, who in turn were required to sell some fraction in the currency auction. Moreover, an import tax with an average rate of 5% (0% on medicine and food, 10% on cars, for example) was introduced in July; by September 1992 it had increased to an average rate of 15%. Limits on taking money out of the country were maintained at $500.
Russia's shock therapy did eliminate planner-determined exchange rates set independently from currency market outcomes, and permitted more participants in the foreign currency auction. In addition, a limited degree of internal and current account convertibility was achieved. Shock therapy did not eliminate the black market in foreign currency.
In general, a black or parallel market for foreign exchange arises when excess demand for foreign currency persists because its availability is restricted by exchange controls, taxes, or other government prohibitions. Black markets in foreign currency are well-documented characteristics of (former) socialist economies (Culbertson and Amacher 1978, Goldberg and Karimov 1993, Meyendorff 1994) as well as other developing countries (Dornbusch, et. al. 1983, Quirk, et. al. 1987, Agenor 1992). In Russia's transition economy, with prices rising an average of 50% per month as a result of the price liberalization policies introduced in January 1992, and the foreign exchange surrender requirements taxing export earnings at a rate of approximately 30% (Goldberg 1993), the incentive remained strong for both households and firms to evade the official foreign currency sector.
Foreign currency acted as a hedge against ruble inflation, at a time when no other assets were available. Moreover, foreign currency represented a means to increase income through speculation. Firms, in particular, found it advantageous to borrow money from the bank, repaying the loan within the stipulated three-month period using gains realized from currency trading. Thus, the demand for foreign currency did not diminish with the introduction of shock therapy, and the supply of foreign currency and the exchange rate, as described above, remained constrained. Consequently, the black market flourished. Exporters gained by underreporting foreign currency earnings, engaging in international barter transactions, and retaining illegal accounts outside Russia. Importers gained by underinvoicing and smuggling. Both importers and exporters gained by buying and selling foreign exchange "on the street." One measure of the magnitude of the gain is the parallel market premium. This paper examines the impact of shock therapy on the percentage gap between the black market and official exchange rates in the first year of Russia's transition process.(21)
Data and Methodology
How did Russia's "shock therapy" in 1992 affect the relationship between the official and black market exchange rates? For the purposes of our analysis, the "official" exchange rate will be the rates banks charge for rubles and dollars, as reported by Commersant(22) (see Table 1). The five banks included in this analysis, Menatap AKIB, Sberbank, Kredobank, NGS Bank, and Technobank, all located in Moscow, were selected on the basis of the frequency of reported exchange rate data. All five banks participated in the Moscow Interbank Currency Exchange (MICE) auctions. In addition, four of the five banks included in this study are listed in the "top 50" Russian banks, as measured by volume and value of financial transactions (Ekonomicheskaya gazeta 1992).
Table 1: Weekly Official Exchange Rates
SBERBANK RF KREDO (1) (2) (3) (4) (5) Buy-Sell Week Dollar Dollar Spread Dollar Dollar Ending Buy Rate Sell Rate (Percent) Buy Rate Sell Rate 1992 1/5 111 121.1 9.1 94.5 N/A 1/19 111 122.1 10.0 110 N/A 1/26 111 122.1 10.0 110 N/A 3/8 70 140 100.0 110 110 3/15 70 140 100.0 N/A N/A 3/29 90 140 55.6 110 N/A 4/5 120 132 10.0 115 N/A 4/12 120 132 10.0 115 126 4/19 120 132 10.0 115 N/A 4/26 120 132 10.0 115 N/A 5/3 120 132 10.00 115 N/A 5/8 120 132 10.0 115 N/A 5/10 120 132 10.0 115 N/A 5/14 120 132 10.0 5/15 115 N/A 5/17 120 132 10.0 115 N/A 5/21 120 132 10.0 120 132 5/29 120 132 10.0 115 126 6/11 120 132 10.0 6/15 122 132 6/21 120 132 10.0 124 136 7/3 133 146.3 10.0 131 144 7/8 133 146.3 10.0 7/10 131 144 7/17 141 154 7/18 135 148.5 10.0 7/19 135 148.5 10.0 141 154 7/23 135 148.5 10.0 151 166 7/26 135 148.5 10.0 151 166 7/30 7/31 150 165 10.0 8/2 150 165 10.0 161 177 8/23 165 181.5 10.0 170 187 8/24 165 151.5 10.0 8/27 165 181.5 10.0 8/28 8/30 165 181.5 10.0 206 235 8/31 206 235 9/3 215 236.5 10.0 221 243.1 9/18 215 236.5 10.0 231 233 9/24 215 236.5 10.0 221 243 10/2 280 317 10/9 301 331 10/11 310 324 4.5 301 331 10/18 N/A N/A N/A N/A 10/23 330 366 10.9 335 368 10/25 N/A N/A N/A N/A 10/30 375 412.5 10.0 360 396 11/6 375 412.5 10.0 396 425 11/13 410 451 10.0 395 430 11/20 410 451 10.0 440 450 11/27 445 489.5 10.0 450 N/A 12/4 410 451 10.0 398 460 12/13 416.5 467.5 12.2 450 480 12/20 425 467.5 10.0 430 470 1993 1/17 460 506 10.0 1/24 480.2 539 12.2 KREDO MENATEP NGS (6) (7) (8) (9) (10) Buy-Sell Buy-Sell Week Spread Dollar Dollar Spread Dollar Ending (Percent) Buy Rate Sell Rate (Percent) Buy Rate 1992 1/5 96.2 106.5 10.7 110 1/19 100 110 10.0 120 1/26 105 115.5 10.0 120 3/8 10.0 100 N/A 95 3/15 100 115 15.0 100 3/29 110 N/A 122 4/5 110 N/A 122 4/12 9.6 110 N/A 122 4/19 110 N/A 120 4/26 110 N/A N/A 5/3 110 121 10.0 126 5/8 110 N/A 120 5/10 110 N/A 120 5/14 110 121 10.0 115 5/15 5/17 110 121 10.0 115 5/21 10.0 110 121 10.0 95 5/29 9.6 110 121 10.0 115 6/11 110 121 10.0 6/15 8.2 115 6/21 9.7 100 121 10.0 125 7/3 9.9 123 135 9.8 130 7/8 7/10 9.9 130 7/17 9.2 7/18 123 135 9.8 130 7/19 9.2 123 135 9.8 130 7/23 9.9 138 147 6.5 140 7/26 9.9 138 147 6.5 140 7/30 138 147 6.5 150 7/31 8/2 9.9 138 147 6.5 150 8/23 10.0 162 178 9.9 171 8/24 8/27 8/28 211.5 8/30 14.1 168 182.4 8.6 211.5 8/31 14.1 9/3 10.0 205 225 9.8 211.5 9/18 0.9 207 227 9.7 211.5 9/24 9.9 10/2 13.2 272 299 9.9 285 10/9 9.9 317 349 10.1 316 10/11 9.9 317 349 10.1 316 10/18 N/A N/A N/A 10/23 322 355 10.3 10/25 N/A N/A N/A 10/30 10.0 322 355 10.3 11/6 7.3 N/A N/A 11/13 8.9 353 388 9.9 409 11/20 2.3 385 405 5.2 425.6 11/27 451.2 12/4 15.6 413.2 12/13 6.7 445 472 6.1 405.6 12/20 9.3 427.5 1993 1/17 446 N/A 440.8 1/24 509 560 10.0 503.5 NGS TEKHNO (11) (12) (13) (14) (15) Buy-Sell Buy-Sell Week Dollar Spread Dollar Dollar Spread Ending Buy Rate (Percent) Buy Rate Sell Rate (Percent) 1992 1/5 121 10.0 100 110 10.0 1/19 132 10.0 105 115.5 10.0 1/26 130 8.3 105 116.6 11.1 3/8 105 10.5 N/A N/A 3/15 110 10.0 80 N/A 3/29 135 10.7 100 110 10.0 4/5 N/A 115 126.5 10.0 4/12 N/A 115 126.5 10.0 4/19 132 10.0 115 126.5 10.0 4/26 N/A 110 121 10.0 5/3 140 11.1 105 115 9.5 5/8 132 10.0 110 N/A 5/10 132 10.0 N/A N/A 5/14 127 10.4 5/15 5/17 127 10.4 N/A N/A 5/21 105 10.5 100 110 10.0 5/29 127 10.4 100 110 10.0 6/11 6/15 127 10.4 6/21 138 10.4 110 121 10.0 7/3 143 10.0 115 126.5 10.0 7/8 7/10 N/A 120 132 10.0 7/17 7/18 N/A 120 132 10.0 7/19 N/A 120 132 10.0 7/23 154 10.0 120 132 10.0 7/26 154 120 132 7/30 N/A 7/31 130 143 10.0 8/2 N/A 130 143 10.0 8/23 N/A 160 176 10.0 8/24 8/27 8/28 N/A 8/30 N/A 170 187 10.0 8/31 9/3 N/A 9/18 235 11.1 9/24 200 220 10.0 10/2 N/A 240 264 10.0 10/9 N/A 280 308 10.0 10/11 N/A 280 308 10.0 10/18 N/A 280 N/A 10/23 10/25 N/A 300 330 10.0 10/30 11/6 11/13 410 0.2 11/20 470.4 10.5 395 434.5 10.0 11/27 499.9 10.8 415 446 7.5 12/4 456.8 10.5 12/13 450.5 11.1 12/20 493.6 15.5 1993 1/17 N/A 450 495 10.0 1/24 577.5 14.7 500 550 10.0
As seen in Table 1, over the course of 1992, dollar buy rates rose from 95-111R/$ in January to 425-430R/$ at year end. A similar pattern emerges across banks for dollar sell rates. Both Sberbank and Technobank appear to follow a fixed rule in establishing buy/sell rates: percentage changes in the buy/sell spread (columns 6 and 9) occurred only rarely.
Black market exchange rates, reported weekly in 1992 by Commersant (see Table 2), were unaccompanied by discussion as to how these data were collected. They do correspond to the authors' experiences, as well as to rates reported by colleagues in Moscow during the course of the year. Perhaps in this regard they represent an improvement over typical data reporting.
TABLE 2: Weekly Black Market Currency Rates
Moscow St. Petersburg (1) (2) (3) (4) (5) Buy-Sell Week Dollar Dollar Spread Dollar Dollar Ending Buy Rate Sell Rate (Percent) Buy Rate Sell Rate 1991 10/20 44 46 4.6 45 48 11/3 57 65 14.0 57 62 11/17 70 80 14.3 76 85 12/1 82 86 4.9 80 84 12/15 95 117 23.2 100 120 1992 1/5 115 140 21.7 115 130 1/19 135 140 3.7 140 150 3/8 110 115 4.6 85 93 3/15 115 120 4.4 130 135 3/29 126 135 7.1 130 135 4/5 135 140 3.7 133 136 4/12 125 130 4.0 125 130 4/19 120 130 8.3 122 132 4/26 120 125 4/29 120 127 5.8 5/3 125 130 4.0 129 135 5/10 125 127 1.6 127 132 5/17 122 125 2.5 120 125 5/24 122 125 2.5 121 127 5/31 122 127 4.1 120 123 6/7 125 129 3.2 122 127 6/14 127 130 2.4 123 127 6/21 130 135 3.8 128 135 7/5 140 143 2.1 137 140 7/12 142 147 3.5 140 145 7/19 147 150 2.0 143 148 7/26 150 156 4.0 143 150 8/2 165 169 2.4 160 165 8/9 165 173 4.8 150 107 8/16 180 185 2.8 178 183 8/23 186 195 4.8 180 195 8/30 200 215 7.5 200 210 9/6 230 240 4.4 230 245 9/20 220 227 3.2 220 230 9/27 265 280 5.7 265 288 10/4 300 320 6.7 280 300 10/11 305 320 4.9 300 310 10/18 315 325 3.2 310 320 10/25 335 350 4.5 340 350 11/1 370 380 2.7 375 380 11/8 405 415 2.5 395 412 11/15 407 420 3.2 410 419 11/22 445 460 3.4 453 460 11/29 470 480 2.1 485 495 12/6 425 440 3.5 410 440 12/13 460 485 5.4 460 475 12/20 465 485 4.3 460 480 1993 1/17 495 505 2.0 500 520 1/24 540 560 3.7 525 540 St. Petersburg Vladivostok Average (6) (7) (8) (9) (10) Buy-Sell Buy-Sell Week Spread Dollar Dollar Spread Dollar Ending (Percent) Buy Rate Sell Rate (Percent) Buy Rate 1991 10/20 6.7 39 42 7.7 42.7 11/3 9.8 51 53 4.0 55.0 11/17 11.8 52 55 5.8 66.0 12/1 5.0 75 90 20.0 79.0 12/15 20.0 100 150 50.0 98.3 1992 1/5 13.0 110 150 36.4 113.3 1/19 7.1 120 170 41.7 136.3 3/8 9.4 100 110 10.0 96.3 3/15 3.9 110 115 4.6 112.5 3/29 3.9 N/A N/A 128.7 4/5 2.3 110 125 13.6 129.5 4/12 4.0 110 125 13.6 123.0 4/19 8.2 110 125 13.6 118.0 4/26 4.2 120 135 12.5 120.0 4/29 5/3 4.7 120 135 12.5 126.0 5/10 3.9 120 150 25.0 125.5 5/17 4.2 120 140 16.7 123.0 5/24 5.0 125 130 4.0 124.5 5/31 2.5 120 130 8.3 122.7 6/7 4.1 120 130 8.3 123.4 6/14 3.3 130 135 3.9 127.4 6/21 5.5 130 135 3.9 130.0 7/5 2.2 125 135 8.0 133.4 7/12 3.6 135 145 7.4 140.4 7/19 3.5 142 156 9.9 144.3 7/26 4.9 142 156 9.9 145.0 8/2 3.1 160 170 6.3 158.8 8/9 13.3 180 185 2.8 170.0 8/16 2.8 180 190 5.6 177.6 8/23 8.3 185 200 8.1 185.8 8/30 5.0 185 200 8.1 193.8 9/6 6.5 240 260 8.3 224.0 9/20 4.6 N/A N/A 221.5 9/27 8.7 N/A N/A 261.3 10/4 7.1 295 N/A 289.4 10/11 3.3 340 360 5.9 309.6 10/18 3.2 340 350 2.9 313.2 10/25 2.9 340 350 2.9 338.2 11/1 1.3 380 410 7.9 380.8 11/8 4.3 N/A N/A 11/15 2.2 410 430 4.9 11/22 1.6 420 462 10.0 454.7 11/29 2.1 N/A N/A 468.0 12/6 7.3 N/A N/A 420.0 12/13 3.3 N/A N/A 455.9 12/20 4.4 N/A N/A 458.9 1993 1/17 4.0 480 520 8.3 164.4 1/24 2.9 534.0 Average (11) (12) Buy-Sell Week Dollar Spread Ending Sell Rate (Percent) 1991 10/20 45.3 6.3 11/3 60.0 9.1 11/17 73.3 11.1 12/1 86.7 9.7 12/15 129.0 31.2 1992 1/5 140.0 23.5 1/19 157.5 15.6 3/8 102.5 6.5 3/15 118.5 5.3 3/29 136.7 6.2 4/5 136.5 5.4 4/12 130.0 5.7 4/19 130.0 10.2 4/26 130.5 8.8 4/29 5/3 135.0 7.1 5/10 137.3 9.4 5/17 131.3 6.7 5/24 129.0 3.6 5/31 128.8 5.0 6/7 129.2 4.7 6/14 131.4 3.1 6/21 136.3 4.8 7/5 139.2 4.4 7/12 146.4 4.3 7/19 151.0 4.0 7/26 153.0 5.5 8/2 166.0 4.6 8/9 179.5 5.6 8/16 185.6 4.5 8/23 198.3 6.7 8/30 208.8 7.7 9/6 236.0 5.4 9/20 229.3 3.5 9/27 275.0 5.3 10/4 309.3 6.9 10/11 319.4 3.2 10/18 330.2 5.4 10/25 362.3 7.1 11/1 398.3 4.6 11/8 11/15 11/22 474.5 4.4 11/29 488.5 4.4 12/6 444.3 5.8 12/13 480.5 5.4 12/20 483.2 5.3 1993 1/17 228.3 4.6 1/24 559.5 4.8
The chaos associated with the post-coup environment is evident in the blackmarket buy/sell rates reported in Table 2. Black market buy rates rose from 44R/$ (late October 1991) to 115R/$ in early January 1992, a 161% increase in both Moscow and St. Petersburg. During this same time, sell rates went from 46 to 140R/$ (a 120% increase). With the exception of August 1992, monthly changes in the black market ruble-dollar buy and sell rates are within two percentage points of each other (see column 3). In all but April and May, both rates increased each month, usually by more than 10%, indicating a strong depreciation of the ruble against the dollar.
To analyze the magnitude and direction of the distortion in the ruble-dollar exchange rate, we first calculate the parallel market premium (PMP) for three regions: Moscow, St. Petersburg, and Vladivostok. A positive PMP indicates that the bank rate overvalues the ruble relative to the dollar. We expect to find a positive PMP at the beginning of the transition process because of prior exchange rate controls. If reforms fully decentralize the foreign currency market and lead to increasing convertibility of the ruble, the PMP should diminish over time, although since current account rather than full convertibility was the policy goal at the time, the premia need not disappear.(23)
Second, we calculate spreads between the reported buy and sell rates in the "official" (bank) and the black markets. The buy rate is used for exchanging rubles for dollars from agents (most relevant to exporters); the sell rate is used for exchange dollars for rubles (most relevant to importers). We expect the black market buy/sell spread initially to exceed the bank spread because of existing exchange controls, with the spread diminishing over time as the currency market is decontrolled. Yet, an increase in the spread would occur in response to political uncertainty or increasing macroeconomic instability, both of which tend to characterize transition economies. For example, we would expect both the PMP and the buy/sell spread in the black market to increase in the third quarter of 1992 when the government caved in to demands by various industry groups by increasing credit and subsidy allocations.(24) Printing money to finance enterprise debt spurred inflation, reversing earlier gains by the Central Bank in stabilizing the value of the ruble.
Using only the weeks during 1992 in which both the black market rates and the bank rates were available, our data set consists of 43 matched weeks between January 1992 and January 1993. To calculate the parallel market premium:
[PMP.sub.t] = [([BMR.sub.t]/[OR.sub.t]) - 1] x 100
where BMR is the black market rate, OR is the bank (or "official") rate, and t refers to time, we use black market exchange rates reported for Moscow, St. Petersburg, and Vladivostok. Buy and sell rates for each of the five banks are used, thereby generating thirty separate weekly measures of the parallel market premium. Missing observations preclude complete analysis of all series, but for the three largest banks: Menatep, Sberbank, and Kredobank, sufficient data are available for our purposes. For ease and clarity in reporting our results, the "official" or bank rates are defined by Sberbank. Sberbank, in 1992, was one of the largest banks in Russia. Formerly the division of the state bank responsible for household deposits, Sberbank had more than 40,000 branches (more than 90% of the total bank branches in Russia), and acted as the main lender in the interbank market (Lamdany 1993). Only when results are significantly different for the other four banks will they be reported.
Because exchange rate series are usually non-stationary in levels but are frequently stationary in differences, we used the Augmented Dickey-Fuller test to evaluate whether our PMP series were trend stationary. Ten of the eighteen PMP series we examined are trend stationary, thus regression coefficients based on these series require not special interpretation. We use a simple regression of the PMP against a constant, a time trend, and a dummy variable for July, when the foreign exchange market was supposedly unified and current account convertibility was achieved. Such a specification enables us to determine whether there are significant differences in the behavior of the PMP across cities. Our analysis also focuses on whether the PMP varies by time or in response to specific policies. Finally, we evaluate whether the announced foreign exchange market liberalization reforms in July represented a structural break in the behavior of the PMP. In the regression analysis, we correct for serial correlation in the residuals by running Cochrane-Orcutt regressions with an AR(1) variable.
Serial correlation in the residuals in some of the OLS regressions (evident in the Durbin-Watson statistics), caused us to respecify those equations and use what amounts to be generalized least squares -- the original equation is transformed using a serial correlation coefficient, [Rho], and an interative Cochrane-Orcutt procedure, so that the model we estimate is:
[Y.sub.t] = [Rho] [Y.sub.t-] + ([X.sub.t] - [Rho] [X.sub.t-1])B + [e.sub.t]
The coefficient estimates in this equation are efficient. Thus, the estimates of p are provided in Table 4 as an indication of positive serial correlation in those equations where it appears.
The regressions we estimate using OLS or GLS are:
PMP = a + bTREND + e,
PMP = a + bTREND + AR(1) + e,
PMP = a + bJULY + e,
PMP = a + bJULY + AR(,) + e, where Ho: a = b = 0.(25)
Following our analysis of the parallel market premium, we examine the buy/sell spreads in both the "official" (bank) and the black markets. To determine whether these spreads narrowed or widened significantly during 1992, we calculated the mean, variance, and correlation of the spreads across time and over the three regions. The buy/sell spreads in both the "official" (bank) and black markets were stationary during 1992.
Impact of Shock Therapy on the Parallel Market Premium and Buy/ Sell Spreads
If Russia's "shock therapy" program liberalized the foreign currency market, the parallel market premium between official and black market rates should diminish. As seen in Figure 1, where Sberbank is used to typify results of the other banks, there is no obvious downward trend in the PMP during 1992. Rather, there is continued oscillation of the premia around zero.(26) The strongest clustering of the positive spikes occurs in weeks 2429, mid-August through late-September 1992, when the Central Bank dramatically increased credits to enterprises to preclude bankruptcy. Such an increase in the money supply caused the ruble to depreciate against the dollar (see Tables 1 and 2 for the official and black market rates). The countervailing effects of the exchange rate unification measures and the looser credit policies, combined with increases in import and export taxes, caused the parallel market premium to rise.
[Figure 1 ILLUSTRATION OMITTED]
The average parallel market premium by city and by bank in 1992 is reported in Table 3. Several results are interesting. First, in Moscow and St. Petersburg, the average PMP (column 1) calculated from dollar buy rates was very similar, ranging from 4.2% to 16.8%. The ranges for the dollar sell rates are smaller, and in some cases, the average PMP over the first year of Russia's transition is negative.(27) In Vladivostok, the average premia was smaller for all of the banks. Overall, these PMP are quite large by industrialized country standards, but not large when compared to developing countries such as Ecuador, Nigeria, Morocco, and Malaysia. While no similar study has been conducted for other transition economies, anecdotal evidence culled from The Economist (7 January 1989, p. 44, 18 February 1989, p.83) indicates that in Poland when each dollar bought 500 zlotys in the official market (bank), on the street each dollar was worth seven times more. Anecdotal reports put the black market exchange rate in Hungary at 15% greater than the official ruble-forint rate; in Romania, Bulgaria, and Slovakia, estimates of the black market rates are 3-5% higher than the official rates.
Table 3: Parallel Market Premia Summary Statistics
A. PMP Buy Rates Moscow Mean S.D. Max Min C of V Menatep 9.73 7.52 35.00 -6.94 .773 Kredobank 4.20 6.00 22.73 -5.84 1.43 Sberbank 8.98 13.64 60.71 -6.25 1.52 NGS 5.67 6.05 27.37 -6.65 1.07 Technobank 16.82 7.71 40.63 4.35 .458 St. Petersburg Mean S.D. Max Min C of V Menatep 9.24 9.36 40.00 -15.00 1.01 Kredobank 3.84 8.19 27.27 -15.00 2.13 Sberbank 8.44 15.05 85.71 -4.69 1.78 NGS 4.77 7.93 30.00 -10.53 1.66 Technobank 16.38 11.29 62.50 -8.11 .689 Vladivostok Mean S.D. Max Min C of V Menatep 8.65 5.85 20.00 0.00 .676 Kredobank 2.13 5.20 12.13 -10.20 2.44 Sberbank 6.31 13.13 57.14 -8.33 2.08 NGS 3.09 8.63 28.95 -12.53 2.79 Technobank 14.70 9.26 37.50 -4.35 .630 B. PMP Sell Rates Moscow Mean S.D. Max Min C of V Menatep 6.17 8.07 31.46 -9.74 1.31 Kredobank -1.60 4.74 14.20 -9.57 -2.96 Sberbank .298 7.58 17.34 -17.86 25.44 NGS 1.05 6.17 19.05 -8.21 5.88 Technobank 11.06 7.47 28.34 -1.37 .675 St. Petersburg Mean S.D. Max Min C of V Menatep 5.55 9.19 36.36 -19.48 1.66 Kredobank -3.38 5.89 15.84 -15.45 -1.74 Sberbank -.309 8.23 22.85 -33.57 -26.6 NGS .404 7.41 22.73 -11.43 18.34 Technobank 10.60 9.09 29.87 -20.24 .858 Vladivostok Mean S.D. Max Min C of V Menatep 9.86 10.76 50.00 -4.76 1.09 Kredobank -.848 5.27 8.76 -17.02 -6.21 Sberbank 1.95 10.63 35.14 -21.43 5.45 NGS 2.90 8.96 25.00 -11.1 13.09 Technobank 12.92 10.81 42.86 -5.66 .837
PMP standard deviations and coefficients of variations are reported in columns 2 and 5 of Table 3 (panel A). For the dollar buy rates, the lower coefficients of variation for the Moscow PMP indicate somewhat more stability than the PMP in either Vladivostok or St. Petersburg, although all estimates are all fairly close together. Given communication and transportation barriers between Vladivostok in the Far East and Moscow/St. Petersburg, the similarity of the estimates is remarkable. Similar results emerge using the dollar sell rates (see panel B).
The correlation and covariance between the parallel market premia in different cities is relatively high for the banks. Specifically, all PMP correlations are positive, many above 70%. The correlations between Moscow and St. Petersburg PMP are consistently higher than those between either of these cities and Vladivostok. This is true for both the dollar buy and sell rates. All in all, the premia appear to have behaved quite similarly across cities during 1992.
The presence of negative PMP indicate instances in which agents actually and her explanation is that it is due to the use of nominal exchange rate series rather than properly constructed effective exchange rates which reflect the impact of taxes relevant to the particular transaction under consideration. Given the aggregated nature of our data (we do not have access to transaction-specific tax information nor prices to construct real effective exchange rate series), we cannot evaluate whether this explains all of the negative premia we find, but we suspect that more may be going on than this explanation can address. For example, while many of the negative premia occur during the same weeks across many of the banks, not all of the negative premia match in this way, indicating that perhaps other bank- or market-specific forces may be at work.(28)
We regressed the stationary PMP series against a constant and a time trend, separately estimating a dummy variable, JULY, to capture the impact of the foreign exchange market reforms announced by the government in July 1992. Our objective is to discern how much of the trend in the PMP can be explained by policy changes, and how much is attributable to movement of the premia over time. We hypothesize that the intercept will be positive and significant in both sets of regressions--this coincides with our observation of the PMP data. We also hypothesize that the time trend and the dummy variable, JULY, will have negative coefficients, indicating a decline in the PMP over the sample period. Such a result would show a positive role of shock therapy on liberalizing the currency market.
We estimate ten regressions, one for each of the stationary PMP calculated. In the trend regressions (Table 4A), we find evidence of both a significant downward trend and a significant positive intercept in five of the ten regressions.(29) In addition, the intercept was significant and positive in two of the other five regressions in which the trend coefficient was not significant. In only one case, the PMP between Kredobank and St. Petersburg, did we get the strange result indicating a significant negative trend in the PMP over time.
Parallel Market Premia: Regression Results
4A. Trend Regressions Dependent Var. Constant Trend AR(1) [R.sup.2] APM2 11.89(*) -.29(*) .309(*) .38 (2.80) (.107) (.136) APP1 16.81(*) -.344(*) .21 (2.61) (.103) APP2 14.28(*) -.397(*) .29 (2.42) (.096) APV2 16.87(*) -.383(*) .16 (3.32) (.246) KRM1 10.76(*) -.292(*) .29 (1.78) (.074) KRM2 -2.67 .044 .01 (2.47) (.093) KRP2 -7.98(*) .187 .01 (2.93) (.130) KRV1 2.78 -.012 .00 (2.04) (.097) MP1 18.4(*) -.417 .404(*) .25 (7.50) (.290) (.140) MM1 20.26(*) -.440 .667(*) .53 (10.2) (.372) (.118) Dependent Var. F statistic #Obs. APM2 12.13 42 APP1 11.1 42 APP2 17.1 42 APV2 6.09 34 KRM1 15.71 41 KRM2 .221 28 KRP2 2.87 28 KRV1 .02 33 MP1 6.63 42 MM1 1.7 42
(*) (**) denotes significant 5% (10%) level; standard errors in the parentheses.
APM = Akib-Moscow PMP; APP = Akib-St. Petersburg PMP; APV = Akib-Vladivostok PMP; KRM = Kredobank-Moscow PMP; KRP = Kredobank-St. Petersburg PMP; KRV = Kredobank-Vladivostok PMP; MP = Sberbank- St. Petersburg PMP; MM = Sberbank-Moscow PMP. 1 refers to the ruble-dollar buy rate, 2 refers to the ruble-dollar sell rate.
4B. July Regression
Dependent Var. Constant JULY AR(1) APM2 8.04(*) -4.74 .420 (2.6) (3.1) (.126) APP2 10.26(*) -7.79(*) (2.05) (2.63) APP1 13.46(*) -6.97(*) (2.14) (2.75) APV2 12.94(*) -6.17(**) (2.54) (3.59) KRM1 8.07(*) -5.89(*) (1.43) (1.87) KRM2 -2.28 .916 (1.82) (2.10) KRP2 -5.93(*) 3.41 (2.19) (2.53) KRV1 2.77(**) -.416 (1.44) (2.01) MP1 13.12(*) -6.88 .440(*) (5.98) (7.30) (.140) MM1 13.53(*) -5.94 .698(*) (7.30) (7.90) (.114) Dependent Var. [R.sup.2] F statistic # Obs. APM2 .34 10.0 43 APP2 .18 3.5 43 APP1 .14 6.4 43 APV2 .08 3.0 34 KRM1 .20 9.92 41 KRM2 .007 .19 28 KRP2 .07 1.81 28 KRV1 .001 .043 33 MP1 .23 5.96 42 MM1 .52 21 42
(*) (**) denotes significant at 5% (10%) level; standard errors in the parentheses.
APM = Akib-Moscow PMP; APP = Akib-St. Petersburg PMP; APV = Akib-Vladivostok PMP; KRM = Kredobank-Moscow PMP; KRP = Kredobank-St. Petersburg PMP; KRV = Kredobank-Vladivostok PMP; MP = Sberbank-St. Petersburg PMP; MM = Sberbank-Moscow PMP. 1 refers to the ruble-dollar buy rate, 2 refers to the ruble-dollar sell rate.
The JULY dummy variable is negative and significant in only four of the ten cases, and these regressions all have lower it-squared statistics than do the trend regressions. When both the time trend and the JULY dummy variable appear in the same regression, only the time trend remains significant and the it-squared statistic is almost unaffected.
Our results offer some evidence of a downward trend in the PMP for these banks across cities, although given the non-stationarity of two-thirds of the PMP series, we could not assert this with much confidence. In addition, there is scant evidence to support the hypothesis that a structural break occurred when exchange rate unification was announced in July. We conclude that despite the official announcements of impending policy changes to make the ruble more stable and convertible, the black market continued to play an important role. These results suggest both firms and individuals were willing to pay a premium to use it.(30)
Buy/sell spreads in the bank and black markets, in effect, represent the transactions costs of obtaining foreign currency. Our calculation:
% Buy-Sell Spread = [(Sell-Buy)/(Sell+Buy)/2]x100
generates near-perfect correlation between the buy rates in both the bank and black markets (but not between the markets), and also between the sell rates within each market (see Table 5). While we expected the spreads to be closely correlated, we found this not to be the case. In fact, the correlation coefficient for the Moscow and St. Petersburg black market spread was only .459, and the average spread was about 5%.(32)
Table 5: Percentage Buy-Sell Spreads Summary Statistics (Full Sample)
A. Black Market Spreads City Mean Std. Dev. Max. Min. Coef. of Variation Moscow 5.05 2.75 19.61 2.35 .54 St. Petersburg 4.94 2.71 14.43 1.32 .55 B. Bank Spreads Mean Std. Dev. Max. Min. Coef. of Variation Bank Sberbank 12.83 13.15 66.67 4.42 1.02 Menatep (Akib) 8.39 2.33 14.39 4.44 .28
For the bank exchange rates, the means were higher, ranging from 8.4% to 12.8%, and the correlation coefficients were generally positive; two-thirds of them were larger than .20. We find little correlation between the spreads across cities. That is, the correlation coefficients between the Moscow black market buy-sell spread and the bank spreads ranged from .051 to .109(1) The correlation coefficients between the banks located in Moscow, and the black market spreads in St. Petersburg also were low.
Splitting the sample allowed us to examine how the spreads differed in their behavior and size before and after implementation of the July reforms (see Table 6). We found that the coefficient of variation was lower for seven of the eight spreads examined, due mostly to the fall in the standard deviations of the spreads. The ranges also narrowed between the two sub-periods. The correlation coefficients increased slightly between the bank and black market spreads from the period prior to the reforms to the period after, although the coefficients stayed fairly small with one exception. The Moscow black market spread and the Menatep Bank spread had a correlation coefficient of .514 in July 1992-January 1993 period; this is in sharp contrast to the coefficient = -.0094 in the pre-July period.
Table 6: Percentage Buy-Sell Spreads Summary Statistics (Split Sample)
A1. Black Market Spreads (1/5/92 - 7/19/92) Mean Std. Dev. Max. Min. Coef. of Variation Moscow 5.03- 3.70 19.61 2.35 .54 St. Petersburg 5.39 2.96 14.43 2.23 .55 A2. Black Market Spreads (7/26/92 - 1/24/93) City Mean Std. Dev. Max. Min. Coef. of Variation Moscow 5.07 1.63 8.24 2.67 .32 St. Petersburg 4.56 2.48 10.52 1.32 .54 B1. Bank Spreads (1/5/92 - 7/19/92) Bank Mean Std. Dev. Max. Min. Coef. of Variation Sberbank 16.89 18.63 66.67 8.70 1.10 Menatep (Akib) 8.71 2.9 14.39 4.44 .33 B2. Bank Spreads (7/26/92 - 1/24/93) Bank Mean Std. Dev. Max. Min. Coef. of Variation Sberbank 9.29 1.38 11.54 4.42 .15 Menatep (Akib) 8.10 1.71 9.91 5.06 .21
Our results suggest that the buy-sell spreads in both the official (bank) and black markets did narrow slightly and became less variable as a consequence of shock therapy. However, we found a rather small degree of correlation between the spreads, even within the same city. This indicates continued rigidity in the foreign exchange market, despite reform efforts, which maintained a relatively high transactions costs of buying and selling currency.
Summary and Conclusions
Analysis of black market and official (bank) exchange rates in Russia during the first year of transition suggests that attempts to liberalize the foreign exchange market did result in a slight narrowing of the parallel market premium, though the degree of volatility in either the black market or the bank rates changed little. We find that the buy-sell spreads in both of the markets also narrowed somewhat but they remained fairly high during 1992 and were not highly correlated across markets or cities. Despite declining variability of the buy-sell spreads, the mean spreads, ranging from 5% to 7% in the black market and 10% in the official market, remained much higher than in developed countries.(33)
Our analysis indicates that the black market for foreign exchange in Russia remained a viable, independent market throughout the first year of transition, and the persistence of large buy-sell spreads in both black and official markets indicates significant inflexibility. While we did find some evidence of a negative trend in the PMP during 1992, the size of the coefficients was small (generally in the range of -0.29 to -0.39) and they were only significant in four of ten stationary PMP series out of the thirty series we calculated. There is similarly weak evidence that a structural break occurred with the July exchange rate unification reforms.
Has the foreign exchange market continued to evolve towards one with the black market playing a less significant role, a more stable ruble, and increasing convertibility? There are some positive signs. Russia's rapid privatization program resulted in 62% of GDP produced by the non-state sector in 1994,(34) compared to less than 15% in 1992. Currently, more than 95% of Russia's retail shops are privately owned. Equally important, many of the basic institutions required for a market economy to function have been put into place in Russia: in January 1995, Russia's first contract law was implemented, as well as the basic legislation necessary to protect property rights and bankruptcy law. The banking and financial sector is also developing with over 2500 licensed commercial banks in operation in 1995, a functioning stock market, the availability of over 600 investment funds, and foreign currency exchanges in all major cities.(35)
At the same time, however, many firms remain largely state-controlled and many sectors remain monopolized which impedes the restructuring required to improve the efficiency of many operations. The fact that, officially, Russian enterprises and individuals hold $24 billion in foreign accounts (and probably that amount again illegally) suggests that appropriate incentives are not yet fully in place.
With respect to the soundness of Russia's macroeconomic policies, there is much work to be done. Inflation continues to be a major problem in 1996 as it was in 1992, caused primarily by the printing of rubles to cover the large government budget deficit. In 1992, the deficit grew as a consequence of subsidies to state-owned loss-making firms, social service expenditures, defense conversion, and aid to the former Soviet republics. In 1993, privatization eliminated many state-owned firms and thus, many of the subsidies, and the end of the ruble zone eliminated the need for Russia to make large financial transfers to the CIS. At the beginning of 1995, a new law was adopted making it illegal for the government to finance its deficits by printing rubles, but more than half of the expenditures identified in the 1995 budget was earmarked for subsidies to loss-making firms. The 1995 budget also necessarily included financing the war in Chechnya.
Finally, regarding the level of international reserves and an appropriate exchange rate, there have been many articles published in prominent Russian journals decrying the lack of foreign exchange reserves, citing the need for and/or the consequences of implementing additional controls over foreign currency movements, and proposing the introduction of new trade barriers, distortionary taxes, and the criminalization of various foreign currency transactions. In 1994, the ruble/dollar rate exceeded 3000; in May 1995, policymakers celebrated the strengthening of the ruble from 5,088 to 5,056 rubles per dollar in a single day. Russia's consistency and credibility in policy-making and enforcement are still seriously questioned, as is its commitment to market reform. In short, it appears that even at the end of 1996, full ruble convertibility and a complete evolution of the foreign exchange market in Russia are goals that are still far from being met.
(1.) Designated ministries and firms were granted the right to start direct foreign trade operations with non-socialist countries; all firms were granted the right to deal directly were firms in socialist countries. Firms received the right to apply for hard currency credits to finance their foreign trade activities. Twenty ministries and departments (such as Gosagroprom, Ministry of Chemical Industry, Gossnab) and 76 major individual enterprises (VAZ plant, URALMASH) and amalgamated enterprises (obedineniie) were granted right of direct participation on their own behalf in export-import transactions in world market, and to retain a designated proportion of their foreign exchange earnings. By 1990, about 10% of the firms (14,000 firms) were granted the right to engage in transactions with foreign organizations.
(2.) "Differentiated foreign trade coefficients" were established by planners in 1987 in an attempt to more closely link domestic and world market prices. In fact, these coefficients initially reflected the same commodity-specific tax/subsidy pattern previously used by planners. The coefficients applied to the planner-determined exchange rate ranged from 0.1 to 15.9, end numbered 1,600 in 1987; by 1988, there were 10,000 coefficients (IMF et al 1991). In effect, each transaction utilized a different exchange rate. The fact that firms pressured planners to raise the coefficients applicable to the goods they exported suggests that firms wanted greater hard currency earnings and thought they could sell their products at higher prices.
(3.) Planners' loss of control over the composition and volume of trade, as well as the balance between exports and imports was offset by numerous trade regulations. More than 70% of the exports were subject to license, where the licenses were explained by the need to protect domestic supplies. Just under 10% of the imports required license; in this case the license was explained by the need for planners to obtain sufficient hard currency to meet economy-wide needs. Export quotas also were established to protec domestic supplies: licenses were given only to firms that had "quota" designation, and these firms were able to export only above-plan production (that is, only if they fulfilled all domestic contracts and state orders). Quotas were revoked even if these two conditions were met if planners viewed the product in short supply (defitsitny).
(4.) For nickel, copper wire, and other "strategic" materials, export licenses were granted to fewer than 5 firms in the entire country. For electronic or avaiation scrap, no export licenses were granted.
(5.) The foreign trade situation changed dramatically in July 1990 with the decision to abandon the `transferable' ruble used by CMEA partners and begin in January 1991 to settle all accounts and make all payments either in national currencies or hard currency.
(6.) In 1988, the Central Bank (Gosbank) created 5 distinct banks, each specializing in particular activities; Vneshekonombank specialized in international trade and finance activities. Not until beginning of 1989 was legislation adopted to permit the formation of commercial and joint stock banks (non state-owned banks). Even these five newly-created banks did not operate independently from the Central Bank, but rather simply distributed Central Bank credits to local firms or authorities.
(7.) Two exceptions to this rule were allowed: individuals employed by foreign companies or joint ventures, and farmers. That is, in August 1989, planners promised farmers hard currency in exchange for above-plan production. However, the amount of hard currency involved was minimal. For farmers living in rural communities where the infrastructure was seriously underdeveloped, access to and perhaps interest in tourist shops was limited, as was access to foreign travel.
(8.) At this same time, firms were bidding between 13 and 15 rubles per $1 in the foreign exchange auction (Financial Times, p. 1, 6 November 1989). This is twenty-two times the planner-determined commercial rate. Thus the impact of the auction was limited in the first year, in part due to a low trading volume, and in part due to the planners ignoring the results.
(9.) That is, previously, individuals employed by joint ventures, for example, or those hired by foreign firms to provide consulting services, could legally receive payment in foreign currency as long as the payment was made directly to a special bank account established for this purpose. Planners exercised control over the currency by requiring a "declaration of origin" every time foreign currency was deposited.
(10.) One result of the decentralization of foreign trade decision making wasserious delay in payments by Soviet firms to western suppliers (Economist, March 1991), perhaps reaching as high as 10% of trade payments to the West. Other estimates range from $3$5 billion (Peel, Financial Times, May 1990, October 1990, and International Herald Times, September 1991).
(11.) The popular press describes the hard currency confiscation with speculations that ten cents on a dollar would be payable in rubles at the planner-determined rate, within 3 or 4 years. Actually, Yeltsin stated that banks must begin repaying those whose accounts had been frozen by Vneshekonombank by July 1993. The total amount involved was announced as $300 million. Interestingly enough, given the privatization process initiated in June 1992, many of the organization whose accounts were frozen in December 1991 did not exist in July 1993 -- that is, state-owned firms established the account from which currency was confiscated, but that same state-owned firm did not exist after privatization (step one in the privatization process requires the firm to establish/register itself as a joint stock company), and thus had no real legal claim to the confiscated money.
(12.) People had 30 days to turn in large denomination ruble banknotes (50, 100); plus they faced limits on the maximum amount they could trade in for smaller denomination notes. According to the Soviet press, this policy was designed by planners to curtail the black market for goods. Since black marketeers dealt in dollars rather than rubles, they would not be affected by the removal of large denomination banknotes. In fact, given the liquidity in the black market, they were likely to benefit from the program given its rationing effects. Planners next suggested this policy was designed to foil a conspiracy by foreign banks against the Soviet economy (Wall Street Journal, 20 January 1991, p. A1).
(13.) Persons convicted of violating Article 88 of the Russian Criminal Code, that is, convicted of buying, selling, or exchanging foreign currency or hard currency assets on Russian territory or of using these as a form of payment without the appropriate authorization from the Central Bank can receive the maximum prison sentence.
(14.) For every $1 earnings, these firms had to hand over 50 cents for conversion into rubles at a rate of 62 rubles to $1. At the time, the auction rate was 150 rubles per $1 (Ekonomika i zhizn' 1992, Sutela 1993). It is not surprising to hear exporting firms describing barter deals where they take as payment goods desired by employees (for use or for sale). Barter enables exporters to avoid foreign currency surrender requirements.
(15.) Planners devalued the tourist rate in November 1991 from 32 rubles per $1 to 47 rubles per $1. At this time, one could easily get well over 100 rubles per dollar on the street in Moscow. Planners raised the limit that Soviet citizens with visas and tickets to travel abroad could obtain from $200 to $500 -- this figure had nothing whatsoever to do with an individual's bank account, either in rubles or dollars.
(16.) When these rates were in effect, on what amounted to nearly 60% of the goods imported, the auction rate was 150 rubles per dollar.
(17.) The proposal called for a fixed rate against the dollar, with adjustments of 7.5% on either side of par. This proposal was based in large part on the idea that a ruble stabilization fund would be provided by the international community. The stabilization fund did not materialize in 1992, and the ruble fell dramatically against the dollar.
(18.) At the time, there were only 12 participants in the auction market, thus the volume of trading involved was small relative to the currency earnings of the country.
(19.) Without an adequate bank clearing system, internal ruble convertibility was only applicable for cash transactions in July 1992 (Fischer 1992).
(20.) Previously, these firms paid an export tax of 20% and had to surrender 40% of their hard currency earnings at the planner-determined rate (typically half the auction rate). Thus, the new program left their foreign currency position nearly unchanged.
(21.) Agenor (1992) shows, in his graphs of PMP for several developing countries, that there is a great deal of difference in the size of these premia across countries. For example, PMP on the order of 100% or more have been recorded in Ecuador and Nigeria, while for Morocco and Malaysia, the premia ranged from -4% to as high as 20%. Typically, parallel market premia display large fluctuations overtime, particularly during periods of uncertainty and political instability.
(22.) Commersant is a weekly Russian business publication which was first published in 1908 but ceased publication in 1917 "for reasons beyond its control." It resumed publication in 1990 as part of a joint venture with the Chicago-based Refco Group, Ltd., a publishing and risk-management group. This is, to the best of our knowledge, the only publicly-available data on black market and bank exchange rates during this period, and the data has been used in other empirical work (e.g., Goldberg (1993)).
(23.) That is, as long as capital account convertibility is not attained, the distortions in the foreign exchange market due to capital controls and other restrictions will likely ensure that the black market for foreign currency stays viable.
(24.) Claims that credit was tight is contrasted against the Central Bank's allocation of 457 billion rubles in finance credits to commercial banks, an amount that approximately equaled the stock of high-powered money et the beginning of 1992 (Lipton and Sachs 1992).
(25.) It suffices here to summarize the results. Detailed results from each regression are available from the authors upon request.
(26.) The positive jumps in the data are strongly correlated within banks across cities, but to a lesser extent across banks within cities.
(27.) This may be caused by limitations in our data series -- using only 42 weeks out of a possible 56 weeks.
(28.) For example, Kredobank, Sberbank, and NGS Bank have much higher proportions of negative premia than the other banks when calculations are based on sell rates. Using dollar buy rates, the incidence of negative premia is similar across all of the banks. Lack of information about bank policies and the volume of foreign currency transactions make it impossible to deter mine what bank-specific factors, if any, explain the different incidence of negative premia across banks.
(29.) For these regressions, we also find that the F-statistics are above the critical value, indicating that we can reject the joint hypothesis: a = b = 0. The banks included in this category are Menatep, Sberbank, and Kredobank.
(30.) Despite the incidence of negative premia in many of the PMP series we calculated (for all five banks), in only one regression (Kredobank and the St. Petersburg black market dollar sell PMP) is the intercept negative and significant. The fact remains that of the significant intercept coefficients, seven of the eight were positive, indicating that in general, the PMP were positive during the transition year across all three cities.
(31.) Insufficient matched pairs of buy and sell rates for Vladivostok precluded us from calculating these spreads.
(32.) Banks were legally constrained to keep their buy-sell spreads at or below a maximum of 10%.
(33.) According to Rivera-Batiz (1994), p. 21, developed country buy-sell spreads are close to .1%.
(34.) This percent-of-GDP estimate is considered a lower bound. Due to the burdensome tax situation in Russia where income, not profits, is taxed, there is tremendous incentive to underreport private sector earnings. Moreover, there is also a thriving underground economy in which perhaps as much as 20% of the working population engage in private trade without reporting any income.
(35.) The stock market is said to be valuing the best of what Russia has to offer (oil, gas, other natural resource-based companies) at one-fifth of what a similar firm would be valued in the US.
Agenor, Pierre-Richard (1992), "Parallel Currency Markets in Developing Countries: Theory, Evidence, and Policy Implications," Princeton University Essays in International Finance no. 188, November.
Ash, Timothy (1992), "Problems of Ruble Convertibility," RFE/RI Research Report 1(29), pp. 26-32.
Berg, Andrew and Jeffrey Sachs (1992), "Structural Adjustment and International Trade in Eastern Europe: The Case of Poland," Economic Policy no. 14, pp. 117-74.
Commersant: The Russian Business Weekly, various issues (1991-1993).
Culbertson, William and Ryan Amacher (1978), "Inflation in the Planned Economies: Some Estimates for Eastern Europe," Southern Economic Journal October, pp. 380-93.
Dornbusch, Rudiger et. al. (1983), "The Black Market for Dollars in Brazil," Quarterly Journal of Economics February, pp. 25-40.
Ekonomika I zhizn', "Instructions on Procedures for Mandatory Sale of Hard Currency," no. 28 (July 1992), translated in FBIS-USR-92-110 (September 1992).
Fischer, Stanley (1992), "Stabilization and Economic Reform in Russia," Brookings Papers on Economic Activity 1, pp. 77-126.
Fry, Maxwell and D. Nuti (1992), "Monetary and Exchange Rate Policies During Eastern Europe's Transition: Some Lessons From Further East," Oxford Review of Economic Policy no. 1, pp. 27-43.
Gilman, Martin (1990), "Heading for Currency Convertibility," Finance and Development 27 (Washington, DC: IMF and World Bank), pp. 32-34.
Gold, Joseph (1971), "The Fund's Concept of Convertibility," IMF Pamphlet Series, no. 14 (Washington, DC: International Monetary Fund).
Goldberg, Linda (1993), "Foreign Exchange Markets in Russia: Understanding the Reforms," IMF Staff Papers 40(4).
-- (1994), "Policy Reforms, Currency Auctions, and Production Choices in Emerging Market Economies," Journal of Comparative Economics forthcoming 1995.
Greene, Joshua and Peter Isard (1991), "Currency Convertibility and the Transformation of Centrally Planned Economies," IMF Paper no. 81 (Washington, DC: IMF).
Gregory, Paul R and Robert C. Stuart (1994). Soviet and Post-Soviet Economic Structure and Performance (New York: Harper Collins).
Hanson, Philip (1992). From Stagnation to Catastroika (New York: Praeger/ Washington Papers).
Haberler, Gottfried (1994), "Currency Convertibility," National Economic Problems Series, no. 451 (New York: American Enterprise Association).
Hewett, Ed (1988). Reforming the Soviet Economy: Equality Versus Equity (Washington, DC: The Brookings Institution).
Holzman, Franklyn (1991), "Moving Toward Ruble Convertibility," Comparative Economic Studies 33(3), pp. 1-23.
Hunya, Gabor (1992), "Foreign Direct Investment and Privatization in Central and Eastern Europe," Communist Economies and Economic Transformation 4(4), pp. 501-12.
IMF, et. al. (1991), Study of the Soviet Economy vol. 1 (Washington, DC: IMF).
IMF (1992a), The Economy of the Former USSR in 1991 (Washington, DC: IMF).
IMF (1992b), Economic Review: Russian Federation (Washington, DC: IMF).
"Transitional Arrangements for Trade and Payments Among the CMEA Countries," IMF Staff Papers 38 (June), pp. 235-67.
Kennett, David and Marc Lieberman, eds. (1992). The Road to Capitalism: Economic Transformation in Eastern Europe and the Former Soviet Union (New York: Dryden Press.)
Lamdeny, Ruben (1993), "Financial Reform in Russia: The Options," Transition 4(5), pp. 9-10.
Lipton, David and Jeffrey Sachs (1992), "Prospects for Russia's Economic Reforms," Brookings Papers on Economic Activity 2, pp. 213-82.
McKinnon, Ronald (1993). The Order of Economic Liberalization: Financial Control in the Transition to a Market Economy (Baltimore: Johns Hopkins University Press).
Meyendorff, Anna (1994), "The Black Market for Foreign Exchange in the Former Soviet Union," Comparative Economic Studies 36(4), pp. 156-71.
Polak, Jacques (1991), "Convertibility: An Indispensable Element in the Transition Process in Eastern Europe," in John Williamson, ed. Currency Convertibility in Eastern Europe (Washington, DC: Institute for International Economics).
Quirk, Peter et. al. (1987), "Floating Exchange Rates in Developing Countries: Experiences with Auction and Interbank Markets," IMF Occasional Paper no. 53 (Washington, DC: IMF).
Rivera-Batiz, Francisco and Luis (1994). International Finance and Open Economy Macroeconomics 2nd Edition (New York: Macmillan Publishing Company).
Sutela, Pekka, ed. (1993). The Russian Economy in Crisis and Transition (Helsinki: Bank of Finland).
Tarr, David (1992), "Problems in the Transition from the CMEA: Implications for Eastern Europe," Communist Economies and Economic Transformation 4(1), pp. 23-44.
Van Brabant, Jozef (1993), "Lessons from the Wholesale Transformations in the East," Comparative Economic Studies 35(4), pp. 73-102.
Voprosy ekonomiki (1993), no. 2; special issue on the fate of economic reforms in Russia.
Whitlock, Erik (1992), "Russia's Progress Toward An Open Economy," RFE/ RI Research Report 1(47), pp. 35-40.
|Printer friendly Cite/link Email Feedback|
|Author:||Krupp, Corinne; Linz, Susan J.|
|Publication:||Comparative Economic Studies|
|Date:||Mar 22, 1997|
|Previous Article:||Australian government policies and the balance of trade performance of the transportation equipment industry: a comparative perspective.|
|Next Article:||The purchasing power parity and the Russian ruble.|