Abstract
As a part of the proceedings of the Eleventh Annual Meeting of the Board of Governors of the International Monetary Fund, an Informal Session on “Recent Developments in Monetary Analysis” was held on September 25, 1956. The three papers which were presented at that Session by Dr. M. W. Holtrop, President of De Nederlandsche Bank, Dr. Paolo Baffi, Economic Adviser to Banca d’Italia, and Dr. Ralph A. Young, Director of the Division of Research and Statistics, Board of Governors of the Federal Reserve System, are reproduced below, together with the background paper, “Monetary Analyses,” prepared by the Statistics Division of the Research and Statistics Department of the International Monetary Fund.
The government of peru, largely on the recommendation of an economic and financial mission which a U.S. firm of consultants1 had been invited to send to Peru, adopted in November 1949 a system of dual fluctuating exchange rates: a certificate rate applying to most trade transactions and to some approved nontrade items and a draft rate for all other transactions. Up to October 1954, when the certificate rate was pegged by the Central Reserve Bank at 19.00 soles per U.S. dollar, both rates were permitted to fluctuate (Chart 1). The reform of November 1949 was followed by an immediate strengthening of the sol in both the certificate and the draft markets, although the trend in export prices was unfavorable up to the outbreak of the Korean war. From the outbreak of the war to the end of 1952, when export prices were very favorable, the exchange rates were remarkably stable, despite considerable internal monetary expansion. In 1953, however, the certificate rate depreciated by 28 per cent, because world market prices of Peruvian exports fell sharply while domestic costs and prices continued to rise on account of internal inflation. Exchange rate stability was restored in 1954 as the decline in export prices ended and some internal stabilization measures became effective in the second half of the year. Although it was known that, from time to time, the monetary authorities intervened by buying or selling, in order to smooth out short-term fluctuations in the certificate rate, it was claimed that no attempt was made during this period to prevent long-term adjustments. Thus, for nearly five years, Peru provided an interesting example of a fluctuating exchange rate system.
Exchange Certificate Rate and Draft Rate of Peru, Monthly, 1949–54
(In soles per U.S. dollar; logarithmic vertical scale)
Citation: IMF Staff Papers 1957, 001; 10.5089/9781451960211.024.A003
Peruvian Exchange Practices Before November 1949
A system of severe exchange restrictions was first imposed in Peru in January 1945. It was modified from time to time but, under the impact of recurrent inflationary pressures, there was constant difficulty in attempting to maintain balance of payments equilibrium. A new exchange system was accordingly adopted in December 1948. All exporters, except producers of gold, were required to take 55 per cent of their export proceeds in the form of exchange certificates from the Central Reserve Bank and to sell the remaining 45 per cent to the Central Reserve Bank at the official buying rate of 6.485 soles per U.S. dollar. Gold producers received exchange certificates for 100 per cent of their output, and similar treatment was extended in August 1949 to exporters of minerals and exporters to the sterling area. The maintenance for nonmineral exports (which normally constitute well over 60 per cent of Peru’s total exports) of the requirement to sell 45 per cent of exchange proceeds to the Central Reserve Bank at the official rate implied a discriminatory export tax, from which exports of minerals were free.2
Under the system adopted in December 1948, exchange was provided for imports of essential goods, i.e., basic foods and medicines (constituting about 20 per cent of total imports), at the official selling rate of 6.5 soles per dollar. For all other imports that were not prohibited, exchange had to be acquired by purchasing exchange certificates in the certificate market. This implied a discriminatory subsidy on imports of essential goods, the avowed purpose of which was to prevent encroachment of the demand for nonessential imports upon the foreign exchange available for essential imports, so that an “adequate” supply of essential imports could be assured for domestic consumers.
Significance for Export and Import Rates of 1949 Exchange Reform
The reform of November 1949 simplified the exchange system as a whole, but the use of exchange certificates was still maintained. Exporters were required to surrender their exchange proceeds to the Central Reserve Bank, receiving in return exchange certificates, to the extent stipulated by the Government; for the greater part of the period concerned, this requirement covered 100 per cent of U.S. dollar, sterling, and French franc proceeds, and 10 per cent of Argentine peso proceeds.3 These certificates were freely negotiable, but were valid for only a limited period, after which they had to be sold to the Central Reserve Bank at the current certificate rate less 2 per cent.4 The surrender requirement was imposed in order to channel export proceeds into the certificate market, to meet the demand for imports and other authorized uses of foreign exchange. Receipts from invisibles were free of control and could be sold in the draft market. By abolishing the requirement that 45 per cent of export proceeds had to be sold at the unfavorable official exchange rate, the implicit discriminatory tax on the greater part of Peru’s exports was eliminated.
At the same time, the implicit subsidy on imports of so-called essential goods also disappeared. There were no longer any exchange transactions at the official exchange rate, and quantitative import controls were removed. At first there was a list of prohibited imports, but this was soon reduced, and on January 30, 1951 it was abolished; however, in November 1953, imports of automobiles were banned temporarily because of the sharp worsening of Peru’s balance of payments and the rapid depreciation of the sol. This ban was replaced by a quota restriction in August 1954. Importers could obtain exchange either by purchasing in the market exchange certificates for which the authorities provided the corresponding exchange, or by buying exchange directly in the draft market.
Payments for certain types of current invisible transactions, and even for some capital transactions, when authorized by exchange licenses, could be made in certificate exchange.5 Exchange for other nonapproved invisible transactions could be bought only in the draft market.
By allowing the certificate market to determine a single exchange rate that would maintain equilibrium in the balance of payments, and by applying that rate uniformly to all exports and imports, the exchange reform of November 1949 enabled Peru’s foreign trade to be carried on closer to the optimum level, where the marginal cost of a unit of any export is equated to the value of the imports that can be brought back in exchange for it.6
Relationship Between Certificate and Draft Markets
The establishment of two separate markets for foreign exchange was intended to prevent volatile speculative capital movements from disturbing the flexible certificate rate, which applies to current transactions, while giving the full freedom to capital movements regarded as necessary to attract an inflow of foreign investment. The limitation on the period of validity of exchange certificates was also intended to protect the certificate rate from speculative influences by discouraging the holding of certificates in anticipation of its appreciation, though speculation in the form of postponement of exports and speeding up of imports (or the reverse) could not be prevented.
The adoption of two separate markets meant that a specified percentage of export proceeds must flow into the certificate market and that the supply of foreign exchange in that market could be drawn upon only to meet import demands and for approved payments for invisibles and capital transactions. To meet other demands for foreign exchange, drawings had to be made on the uncontrolled supplies allowed to flow into the draft market. Theoretically, the exchange rates in the two separate markets were free to move divergently, but there were implicit links between the demands and the supplies in the two markets.
One obvious link was created by the fact that all the foreign exchange demands for which drawings in the certificate market were permitted could also be met by drawings upon the supplies in the draft market, although the reverse was not true. This ensured that the draft rate would not fall below the certificate rate; otherwise, all the demands in the certificate market would be shifted to the draft market, thus forcing the certificate rate down and the draft rate up. But this asymmetric link between the two markets could not prevent the draft rate from rising above the certificate rate, since the demands on the draft market could not be transferred to the certificate market to take advantage of a more favorable rate in the latter.
Nevertheless, except during the periods November 1949-November 1950 and February-June 1954, the draft rate seldom exceeded the certificate rate by more than 2 per cent (Chart 1). Since the authorities are known to have interfered from time to time to smooth out short-term fluctuations in the certificate market but to have left the draft rate entirely free to find its level, the persistent narrowness of the spread between the two rates over a long period of time in spite of fluctuations in both of them suggests that there must have been other links between the two markets. One such link might have been provided by the under-invoicing of exports which would have enabled a clandestine diversion of exchange from the certificate market to the draft market. It is unlikely, however, that such illegal measures took place to any important extent, since the spread between the two rates was seldom large enough to provide adequate incentive to take the risk and trouble involved. A more important link is probably found in the fact that, throughout this period, Peru was consistently a net importer of capital. Particularly after the boom in raw material prices following the Korean conflict, there was a surge of foreign direct investment in mining and oil. In 1950 there was a net repatriation of foreign direct investment of $8.2 million,7 but thereafter there was a net inflow, which increased from $30.2 million in 1951 to $50.3 million in 1952, and then declined to $37.4 million in 1953 and $19.7 million in 1954 (Table 1).
Balance of Payments of Peru, 1949–541
(In millions of U.S. dollars)
Based on data from International Monetary Fund, Balance of Payments Yearbook. No sign indicates credit; minus sign indicates debit.
Balance of Payments of Peru, 1949–541
(In millions of U.S. dollars)
| 1949 | 1950 | 1951 | 1952 | 1953 | 1954 | |||
|---|---|---|---|---|---|---|---|---|
| A. Goods, Services, and Donations | ||||||||
| 1. Exports, f.o.b. | 166.5 | 194.1 | 254.3 | 242.4 | 226.4 | 253.7 | ||
| 2. Imports, f.o.b. | –140.9 | –143.6 | –215.0 | –253.0 | –251.4 | –216.8 | ||
| 25.6 | 50.5 | 39.3 | –10.6 | –25.0 | 36.9 | |||
| 3. Nonmonetary Gold | 4.6 | 4.4 | 4.2 | 5.1 | 3.9 | 2.7 | ||
| 4. Transportation and Insurance | –22.6 | –31.9 | –46.4 | –35.9 | –33.9 | –28.5 | ||
| 5. Investment Income | –11.3 | –9.6 | –21.5 | –20.6 | –20.4 | –21.8 | ||
| 6. Other Services | 1.9 | –20.7 | –13.9 | 6.0 | 4.5 | –1.7 | ||
| 7. Donations (including technical assistance grants) | 4.4 | 3.9 | 4.0 | 8.7 | 6.7 | 4.8 | ||
| Total | 2.6 | –3.4 | –34.3 | –47.3 | –64.2 | –6.8 | ||
| B. Long-Term Capital Transactions | ||||||||
| 8. Foreign direct investment | 2.5 | –8.2 | 30.2 | 50.3 | 37.4 | 19.7 | ||
| 9. Drawings on Export-Import Bank loans | — | — | — | 6.5 | 11.4 | 1.6 | ||
| 10. Official Amortization | –3.2 | –0.9 | –4.3 | –3.3 | –3.5 | –2.1 | ||
| 11. Long-term assets of commercial banks | — | 0.1 | –0.1 | 0.1 | –0.1 | –0.1 | ||
| Total | –0.7 | –9.0 | 25.8 | 53.6 | 45.2 | 19.1 | ||
| C. Short-Term Capital Transactions | ||||||||
| 12. Private foreign assets held through Peruvian banks (net increase -) | – | –10.6 | 11.0 | –0.2 | –2.0 | –1.6 | ||
| 13. Private dollar balances in U.S. (net increase -) | — | 3.3 | –0.7 | –13.9 | –3.3 | –9.5 | ||
| 14. Import collections (net increase +) | 4.9 | 16.3 | 6.5 | 7.9 | 3.3 | 4.2 | ||
| 15. Commercial bank assets, including monetary gold (net increase -) | — | 1.8 | –2.8 | 2.8 | 1.1 | 0.1 | ||
| Total | 4.9 | 10.8 | 14.0 | –3.4 | –0.9 | –6.8 | ||
| D. Net Errors and Omissions | –1.0 | 1.6 | 4.0 | –2.9 | 1.5 | –1.6 | ||
| E. Total (A through D) | 5.8 | — | 9.5 | — | –17.9 | 6.6 | ||
| F. Compensatory Official Financing | ||||||||
| 16. Liabilities to IMF | — | — | — | –3.1 | — | — | ||
| 17. Government and Central Reserve Bank foreign exchange (increase -) | 2.2 | 3.0 | 5.2 | 3.1 | –1.5 | –8.0 | ||
| 18. Central Reserve Bank monetary gold (increase -) | –8.0 | –3.0 | –14.7 | – | 9.4 | 1.4 | ||
| 19. International Petroleum Company loan | — | — | — | — | 10.0 | — | ||
| Total | –5.8 | — | –9.5 | — | 17.9 | –6.6 | ||
Based on data from International Monetary Fund, Balance of Payments Yearbook. No sign indicates credit; minus sign indicates debit.
Balance of Payments of Peru, 1949–541
(In millions of U.S. dollars)
| 1949 | 1950 | 1951 | 1952 | 1953 | 1954 | |||
|---|---|---|---|---|---|---|---|---|
| A. Goods, Services, and Donations | ||||||||
| 1. Exports, f.o.b. | 166.5 | 194.1 | 254.3 | 242.4 | 226.4 | 253.7 | ||
| 2. Imports, f.o.b. | –140.9 | –143.6 | –215.0 | –253.0 | –251.4 | –216.8 | ||
| 25.6 | 50.5 | 39.3 | –10.6 | –25.0 | 36.9 | |||
| 3. Nonmonetary Gold | 4.6 | 4.4 | 4.2 | 5.1 | 3.9 | 2.7 | ||
| 4. Transportation and Insurance | –22.6 | –31.9 | –46.4 | –35.9 | –33.9 | –28.5 | ||
| 5. Investment Income | –11.3 | –9.6 | –21.5 | –20.6 | –20.4 | –21.8 | ||
| 6. Other Services | 1.9 | –20.7 | –13.9 | 6.0 | 4.5 | –1.7 | ||
| 7. Donations (including technical assistance grants) | 4.4 | 3.9 | 4.0 | 8.7 | 6.7 | 4.8 | ||
| Total | 2.6 | –3.4 | –34.3 | –47.3 | –64.2 | –6.8 | ||
| B. Long-Term Capital Transactions | ||||||||
| 8. Foreign direct investment | 2.5 | –8.2 | 30.2 | 50.3 | 37.4 | 19.7 | ||
| 9. Drawings on Export-Import Bank loans | — | — | — | 6.5 | 11.4 | 1.6 | ||
| 10. Official Amortization | –3.2 | –0.9 | –4.3 | –3.3 | –3.5 | –2.1 | ||
| 11. Long-term assets of commercial banks | — | 0.1 | –0.1 | 0.1 | –0.1 | –0.1 | ||
| Total | –0.7 | –9.0 | 25.8 | 53.6 | 45.2 | 19.1 | ||
| C. Short-Term Capital Transactions | ||||||||
| 12. Private foreign assets held through Peruvian banks (net increase -) | – | –10.6 | 11.0 | –0.2 | –2.0 | –1.6 | ||
| 13. Private dollar balances in U.S. (net increase -) | — | 3.3 | –0.7 | –13.9 | –3.3 | –9.5 | ||
| 14. Import collections (net increase +) | 4.9 | 16.3 | 6.5 | 7.9 | 3.3 | 4.2 | ||
| 15. Commercial bank assets, including monetary gold (net increase -) | — | 1.8 | –2.8 | 2.8 | 1.1 | 0.1 | ||
| Total | 4.9 | 10.8 | 14.0 | –3.4 | –0.9 | –6.8 | ||
| D. Net Errors and Omissions | –1.0 | 1.6 | 4.0 | –2.9 | 1.5 | –1.6 | ||
| E. Total (A through D) | 5.8 | — | 9.5 | — | –17.9 | 6.6 | ||
| F. Compensatory Official Financing | ||||||||
| 16. Liabilities to IMF | — | — | — | –3.1 | — | — | ||
| 17. Government and Central Reserve Bank foreign exchange (increase -) | 2.2 | 3.0 | 5.2 | 3.1 | –1.5 | –8.0 | ||
| 18. Central Reserve Bank monetary gold (increase -) | –8.0 | –3.0 | –14.7 | – | 9.4 | 1.4 | ||
| 19. International Petroleum Company loan | — | — | — | — | 10.0 | — | ||
| Total | –5.8 | — | –9.5 | — | 17.9 | –6.6 | ||
Based on data from International Monetary Fund, Balance of Payments Yearbook. No sign indicates credit; minus sign indicates debit.
Although most of these direct investments probably took the form of imports of equipment and materials by foreign firms, financed with their capital resources abroad, it is conceivable that, whenever the spread between the draft rate and the certificate rate tended to widen, it may have become worth while for some of these investing firms to sell part of their foreign exchange funds on the draft market and use the sol proceeds to buy certificates for their imports of equipment and materials rather than use their foreign exchange resources directly on such imports. Such arbitrage operations would at once increase both the supply of foreign exchange to the draft market and the demand for foreign exchange on the certificate market, and hence tend to narrow the gap between the two rates. Probably this possibility of arbitrage on the part of foreign investors was primarily responsible for maintaining the remarkably narrow spread between the two rates.
This hypothesis is borne out by the relative movements of the two rates over the five-year period. During the three years 1951-53, when, as pointed out above, the inflow of direct investment was particularly heavy and when, in consequence, the private capital account of the balance of payments (including both long-term and short-term capital transactions) was always in substantial surplus, the spread between the two rates was invariably very small. Even in 1953, when the sol was under great pressure and depreciated considerably in both markets, the spread remained very narrow.
The exceptionally wide spread between the two rates in the first half of 1950, and the unusually long period that elapsed before it narrowed to any extent, can also be explained by the small inflow, and possibly net withdrawal, of foreign direct investment during 1950. Direct investment was rather slight for 1950 as a whole, and such net inflow as did occur was probably concentrated in the second half of the year. There was, therefore, little possibility of transferring capital from the direct financing of imports to the draft market to narrow the gap between the two rates.
The gap widened again in February-June 1954, probably because of the sharply reduced rate of inflow of direct investment and the equally sharply increased rate of outflow of private short-term capital in 1954.8
Initial Strengthening of the Certificate Rate
The exchange reform of November 1949 was followed by an immediate strengthening of the sol in the certificate market. In October 1949, the month preceding the reform, the monthly average certificate rate had been 17.93 soles per U.S. dollar; in November, it was 15.68 and in December, 14.81. This was, indeed, a continuation of a trend that had started in August; in July the average certificate rate was 20.17 soles, but in August it dropped to 19.13 and in September to 19.04.
There was a strengthening of the sol prior to, as well as immediately after, the November reform; in both instances, the strengthening appears to have been due, to a large extent, to the curtailment of imports brought about by monetary contraction. After the reform, the abolition of the subsidy rate for so-called essential imports seems to have contributed significantly to the curtailment of imports. As shown in Chart 2, the general upward trend of imports (in soles as well as in dollars) from the beginning of 1949 to August of that year is obviously parallel to the steady increase in the money supply. In July 1949, the authorities introduced restrictions on credit expansion by private banks. The expansion of the money supply was immediately arrested, and from September onward imports declined. The reversal of the trend in imports and the decline from August 1949 to February 1950 evidently correspond to the contraction of the money supply. The resumption of the upswing of imports after February 1950 corresponds to the resumption of monetary expansion.9 There is no doubt, therefore, that the credit restriction after July 1949 was a major contributing factor to the strengthening of the sol after August.
Money Supply and Imports of Peru, Monthly, 1949–50
Citation: IMF Staff Papers 1957, 001; 10.5089/9781451960211.024.A003
It is also noticeable that, although the average money supply for the first half of 1950 (S/. 1,932 million) was greater than that for the first half of 1949 (S/. 1,896 million), imports for the first half of 1950 were smaller both in soles and in dollars (S/. 965 million and $68.6 million) than the corresponding figures for the first half of 1949 (S/. 1,263 million and $78.3 million). This decline in imports cannot be accounted for by the difference in the average certificate rates of these two periods; the average certificate rate for the first half of 1950 (S/. 14.67 per dollar) was actually considerably lower than that for the same period of 1949 (S/. 17.67 per dollar). It seems likely, therefore, that the abolition of the subsidy rate for the so-called essential imports (implied in the abolition of the official rate) was also an important factor contributing to the reduction of imports and hence to the strengthening of the certificate rate in the months immediately after the exchange revision of November 1949.
The stimulus given to exports by permitting 100 per cent of the proceeds of all exports to be sold in the certificate market and thus abolishing the discriminatory tax on nonmineral exports appears to have been outweighed by the unfavorable trends in the dollar prices of Peru’s exports up to the outbreak of the Korean war. Prices in world markets of Peru’s five most important export products, cotton, sugar, lead, copper, and petroleum, were lower in the first five months of 1950 than in the corresponding period of 1949. As a result, Peru’s dollar export proceeds fell from $57.0 million in the first five months of 1949 to $54.8 million in the corresponding months of 1950. This, of course, should not be interpreted as meaning that the abolition of the implicit export tax had a negative effect on Peru’s export proceeds.
Stability of Rate to End of 1952
After the initial strengthening, the certificate rate was kept stable at 14.95 soles per dollar from October 1950 to August 1951. Subsequently, it was allowed to fluctuate. While the underlying trend was indeed upward, the rate rose so gently that by the end of 1952 it had not exceeded 15.60 soles. Of course, this remarkable stability was not achieved entirely without government intervention. However, except in 1951 when the Central Bank absorbed in reserves some $9.5 million of foreign exchange, in order to stabilize the rate, the authorities appear not to have intervened persistently in either direction. Neither in 1950 nor in 1952 was there any appreciable net compensatory financing of the balance of payments deficit or surplus by the monetary authorities (Table 1).
However, unlike the initial strengthening, this fairly long period of exchange rate stability cannot be attributed to internal monetary stability. The restrictions on credit expansion by private banks introduced in July 1949 were removed in 1950, and the way was thus opened for credit expansion in the following years. The data in Table 2 indicate the continuing tendency of Peru’s economy to be inflationary throughout the period under examination.
Indicators of Inflationary Tendencies in Peru, 1949–541
Based on data from International Financial Statistics (Washington).
Data as of end of year.
Indicators of Inflationary Tendencies in Peru, 1949–541
| Money Supply2 (million soles) |
National Income (million soles) |
Indices of Wholesale Prices | Index of Cost of Living (1950 = 100) |
||
|---|---|---|---|---|---|
| All goods | Home goods | ||||
| (1950 = 100) | |||||
| 1949 | 1,883 | 9,390 | 87 | 81 | 90 |
| 1950 | 2,198 | 11,984 | 100 | 100 | 100 |
| 1951 | 2,705 | 14,506 | 118 | 105 | 110 |
| 1952 | 3,151 | 16,238 | 124 | 107 | 118 |
| 1953 | 3,478 | 17,205 | 125 | 111 | 128 |
| 1954 | 3,699 | — | 138 | 122 | 135 |
Based on data from International Financial Statistics (Washington).
Data as of end of year.
Indicators of Inflationary Tendencies in Peru, 1949–541
| Money Supply2 (million soles) |
National Income (million soles) |
Indices of Wholesale Prices | Index of Cost of Living (1950 = 100) |
||
|---|---|---|---|---|---|
| All goods | Home goods | ||||
| (1950 = 100) | |||||
| 1949 | 1,883 | 9,390 | 87 | 81 | 90 |
| 1950 | 2,198 | 11,984 | 100 | 100 | 100 |
| 1951 | 2,705 | 14,506 | 118 | 105 | 110 |
| 1952 | 3,151 | 16,238 | 124 | 107 | 118 |
| 1953 | 3,478 | 17,205 | 125 | 111 | 128 |
| 1954 | 3,699 | — | 138 | 122 | 135 |
Based on data from International Financial Statistics (Washington).
Data as of end of year.
In the main, the stability of the certificate rate from 1950 to the end of 1952 was the result of the fortuitous increase in world demand for Peru’s main exports generated by the Korean war. According to data published in International Financial Statistics,10 the value of Peru’s chief exports in terms of dollars rose substantially from 1949 to 1952 (Table 3). Although export prices declined somewhat from 1951 to 1952, the increase from 1950 to 1952 was sufficient to offset the inflation in the domestic cost of production, and, in conjunction with the slight rise in the certificate rate, to stimulate a 14 per cent expansion in the volume of exports, thus providing the exchange proceeds for the inflated demand for imports.
Indices of Unit Value and Volume of Exports of Peru, 1949–541
(1950 = 100)
Based on data from International Financial Statistics (Washington).
Indices of Unit Value and Volume of Exports of Peru, 1949–541
(1950 = 100)
| 1949 | 1950 | 1951 | 1952 | 1953 | 1954 | |
|---|---|---|---|---|---|---|
| Unit value in U.S. dollars | 90 | 100 | 139 | 117 | 93 | 93 |
| Volume | 90 | 100 | 100 | 114 | 123 | 127 |
Based on data from International Financial Statistics (Washington).
Indices of Unit Value and Volume of Exports of Peru, 1949–541
(1950 = 100)
| 1949 | 1950 | 1951 | 1952 | 1953 | 1954 | |
|---|---|---|---|---|---|---|
| Unit value in U.S. dollars | 90 | 100 | 139 | 117 | 93 | 93 |
| Volume | 90 | 100 | 100 | 114 | 123 | 127 |
Based on data from International Financial Statistics (Washington).
On the assumption that no very great changes in the volume of exports are required to keep the balance of payments in equilibrium, the sol price of dollars—the certificate rate—could be expected to tend to adapt itself to the changes in the domestic cost of producing a dollar’s worth of exports. If, because of the lack of more appropriate indices, the cost of living index is taken to represent the index of the average domestic cost of producing exports, the cost of producing a dollar’s worth of exports appears to have risen between 1950 and 1952 by more or less the same percentage as the certificate rate.11
From the data in Table 4, it is not surprising that the certificate rate rose by only 4 per cent between 1950 and 1952 although the cost of living advanced by 18 per cent. The rise in the certificate rate of foreign exchange was modest because the cost of earning foreign exchange through exports increased only slightly.
Indices of Estimated Cost of Producing a Dollar’s Worth of Exports in Peru, 1951–541
(1950 = 100)
Based on data from International Financial Statistics (Washington).
Indices of Estimated Cost of Producing a Dollar’s Worth of Exports in Peru, 1951–541
(1950 = 100)
| 1951 | 1952 | 1953 | 1954 | ||
|---|---|---|---|---|---|
| 1. Export unit value in U.S. Dollars | 139 | 117 | 93 | 93 | |
| 2. Cost of living | 110 | 118 | 128 | 135 | |
| 3. Estimated average cost per dollar’s worth of exports (2 ÷ 1) | 79 | 101 | 138 | 145 | |
| 4. Certificate Rate | |||||
| Yearly average | 101 | 104 | 115 | 130 | |
| End of year | 102 | 104 | 133 | 127 | |
Based on data from International Financial Statistics (Washington).
Indices of Estimated Cost of Producing a Dollar’s Worth of Exports in Peru, 1951–541
(1950 = 100)
| 1951 | 1952 | 1953 | 1954 | ||
|---|---|---|---|---|---|
| 1. Export unit value in U.S. Dollars | 139 | 117 | 93 | 93 | |
| 2. Cost of living | 110 | 118 | 128 | 135 | |
| 3. Estimated average cost per dollar’s worth of exports (2 ÷ 1) | 79 | 101 | 138 | 145 | |
| 4. Certificate Rate | |||||
| Yearly average | 101 | 104 | 115 | 130 | |
| End of year | 102 | 104 | 133 | 127 | |
Based on data from International Financial Statistics (Washington).
The extraordinarily favorable prices for Peru’s exports plus the increase in the volume of exports12 stimulated by these prices and by the uniform application of the certificate rate to all exports increased Peru’s export proceeds from $166.5 million in 1949 to $194.1 million in 1950, $254.3 million in 1951, and $242.4 million in 1952. Attractive world prices for mineral exports also induced large imports of long-term capital on the part of the U.S. direct investment firms, which are the chief producers of Peru’s mineral exports. Thus the inflow of direct investment, which was practically down to zero in 1950, expanded to $30.2 million in 1951 and $50.3 million in 1952.
Imports also increased rapidly, the c.i.f. figures (for imports other than gold) as recorded by customs rising from $167 million in 1949 to $176 million in 1950, $262 million in 1951, and $287.5 million in 1952. This increase, save insofar as it was directly associated with investment by foreign mining companies, is attributed chiefly to the expansion of domestic money income.
The demand for imports can be most conveniently analyzed in terms of an expenditure function of the following type:
Here, M is the physical volume of imports, Pm the price of imports in foreign exchange, E the exchange rate expressed in terms of local currency per unit of foreign exchange, Y aggregate money income of the importing country concerned, and Pd the price of domestically produced and consumed goods.
For the purpose of analyzing the effect of changes in the exchange rate on the demand for imports in terms of foreign exchange, each term in equation (1) should be divided by E:
The movement of Peru’s imports in terms of dollars after the unification of the exchange rate at the end of 1949 may then be analyzed in terms of this formula. The relevant information is given in Table 5.
Factors Determining Expenditures on Imports of Peru, 1950–541
Based on data from International Monetary Fund, Balance of Payments Yearbook and International Financial Statistics (Washington).
Estimated from the annual average money supply by assuming the income velocity of money for 1954 to be the same as for 1953.
Index of home goods prices deflated by the average certificate rate of the year concerned.
A combination of the unit value of U.S. exports to Latin America (supplied by the U.S. Department of Commerce), the over-all export unit value of the United Kingdom, and the over-all export unit value of continental OEEC countries published in Organization for European Economic Cooperation, Statistical Bulletins, General Statistics (Paris), weighted according to the value of Peru’s imports from these three main sources of supply. The index of import prices published by the Peruvian authorities, which consists of unweighted geometric averages of the domestic wholesale prices of 30 import commodities, when deflated by the relevant certificate rates, would be as follows: 1950, 100; 1951, 116; 1952, 117; 1953, 104; 1954, 101. While this index corresponds closely with the index of import prices adopted for this study up to 1952, the severe drop of 12 per cent from 1951 to 1952 was far in excess of the minor decline shown by the latter. The Peruvian index is believed to be less reliable, since it is unweighted and of rather limited coverage.
Imports c.i.f. as recorded by customs less gold plus unrecorded trade and government imports. From International Financial Statistics, February 1956, p. 230.
According to the Balance of Payments Yearbook, there was, instead of a positive direct investment, a net repatriation of foreign direct investment of $8.2 million for the year 1950. There is reason to believe, however, that some direct investment was hidden in the abnormal increase of import collections for that year and that net direct investment might still be positive though not very large for 1950. See footnote 7 in text.
Factors Determining Expenditures on Imports of Peru, 1950–541
| 1950 | 1951 | 1952 | 1953 | 1954 | ||
|---|---|---|---|---|---|---|
| 1. National income | ||||||
| In millions of soles | 11,984 | 14,506 | 16,238 | 17,205 | 18,6542 | |
| In millions of U.S. dollars (at average certificate rate for year) | 807 | 963 | 1,053 | 1,009 | 963 | |
| 2. Price index of domestic goods valued in U.S. dollars3 | 100 | 103 | 103 | 97 | 94 | |
| 3. Price index of imports valued in U.S. dollars4 | 100 | 117 | 119 | 116 | 114 | |
| 4. Item 3 ÷ item 2 | 100 | 114 | 116 | 120 | 121 | |
| 5. Imports c.i.f. (in millions of U.S. dollars)5 | 178 | 264 | 301 | 300 | 258 | |
| 6. Foreign direct investment and drawings on Export-Import Bank Loans (in millions of U.S. dollars) | _6 | 30 | 57 | 49 | 21 | |
| 7. Estimate of imports not directly financed by foreign investment (in millions of U.S. dollars) | 178 | 234 | 244 | 251 | 237 | |
Based on data from International Monetary Fund, Balance of Payments Yearbook and International Financial Statistics (Washington).
Estimated from the annual average money supply by assuming the income velocity of money for 1954 to be the same as for 1953.
Index of home goods prices deflated by the average certificate rate of the year concerned.
A combination of the unit value of U.S. exports to Latin America (supplied by the U.S. Department of Commerce), the over-all export unit value of the United Kingdom, and the over-all export unit value of continental OEEC countries published in Organization for European Economic Cooperation, Statistical Bulletins, General Statistics (Paris), weighted according to the value of Peru’s imports from these three main sources of supply. The index of import prices published by the Peruvian authorities, which consists of unweighted geometric averages of the domestic wholesale prices of 30 import commodities, when deflated by the relevant certificate rates, would be as follows: 1950, 100; 1951, 116; 1952, 117; 1953, 104; 1954, 101. While this index corresponds closely with the index of import prices adopted for this study up to 1952, the severe drop of 12 per cent from 1951 to 1952 was far in excess of the minor decline shown by the latter. The Peruvian index is believed to be less reliable, since it is unweighted and of rather limited coverage.
Imports c.i.f. as recorded by customs less gold plus unrecorded trade and government imports. From International Financial Statistics, February 1956, p. 230.
According to the Balance of Payments Yearbook, there was, instead of a positive direct investment, a net repatriation of foreign direct investment of $8.2 million for the year 1950. There is reason to believe, however, that some direct investment was hidden in the abnormal increase of import collections for that year and that net direct investment might still be positive though not very large for 1950. See footnote 7 in text.
Factors Determining Expenditures on Imports of Peru, 1950–541
| 1950 | 1951 | 1952 | 1953 | 1954 | ||
|---|---|---|---|---|---|---|
| 1. National income | ||||||
| In millions of soles | 11,984 | 14,506 | 16,238 | 17,205 | 18,6542 | |
| In millions of U.S. dollars (at average certificate rate for year) | 807 | 963 | 1,053 | 1,009 | 963 | |
| 2. Price index of domestic goods valued in U.S. dollars3 | 100 | 103 | 103 | 97 | 94 | |
| 3. Price index of imports valued in U.S. dollars4 | 100 | 117 | 119 | 116 | 114 | |
| 4. Item 3 ÷ item 2 | 100 | 114 | 116 | 120 | 121 | |
| 5. Imports c.i.f. (in millions of U.S. dollars)5 | 178 | 264 | 301 | 300 | 258 | |
| 6. Foreign direct investment and drawings on Export-Import Bank Loans (in millions of U.S. dollars) | _6 | 30 | 57 | 49 | 21 | |
| 7. Estimate of imports not directly financed by foreign investment (in millions of U.S. dollars) | 178 | 234 | 244 | 251 | 237 | |
Based on data from International Monetary Fund, Balance of Payments Yearbook and International Financial Statistics (Washington).
Estimated from the annual average money supply by assuming the income velocity of money for 1954 to be the same as for 1953.
Index of home goods prices deflated by the average certificate rate of the year concerned.
A combination of the unit value of U.S. exports to Latin America (supplied by the U.S. Department of Commerce), the over-all export unit value of the United Kingdom, and the over-all export unit value of continental OEEC countries published in Organization for European Economic Cooperation, Statistical Bulletins, General Statistics (Paris), weighted according to the value of Peru’s imports from these three main sources of supply. The index of import prices published by the Peruvian authorities, which consists of unweighted geometric averages of the domestic wholesale prices of 30 import commodities, when deflated by the relevant certificate rates, would be as follows: 1950, 100; 1951, 116; 1952, 117; 1953, 104; 1954, 101. While this index corresponds closely with the index of import prices adopted for this study up to 1952, the severe drop of 12 per cent from 1951 to 1952 was far in excess of the minor decline shown by the latter. The Peruvian index is believed to be less reliable, since it is unweighted and of rather limited coverage.
Imports c.i.f. as recorded by customs less gold plus unrecorded trade and government imports. From International Financial Statistics, February 1956, p. 230.
According to the Balance of Payments Yearbook, there was, instead of a positive direct investment, a net repatriation of foreign direct investment of $8.2 million for the year 1950. There is reason to believe, however, that some direct investment was hidden in the abnormal increase of import collections for that year and that net direct investment might still be positive though not very large for 1950. See footnote 7 in text.
In each year except 1950, there was a considerable inflow of direct investment from the United States to U.S. firms operating in Peru, supplemented by loans granted to these firms by the Export-Import Bank of Washington. The inflow of direct investment took the form primarily of imports of investment goods and materials by the investing firms. Since such imports are not associated with national demand and expenditure in the same way as are other imports, they are best regarded as “exogenous,” i.e., independent of Peru’s normal domestic demand for imports. A part of the foreign direct investment must have been used to finance the local currency expenditures of the firms’ capital formation, but there is no way to separate this part of direct investment from the major part used directly to finance imports of investment goods. In the absence of precise information, it is assumed here that the entire amount of direct investment was used directly to finance imports of investment goods or other goods used in connection with the investment projects, and the entire amount of direct investment and specific loans for development investment is deducted from total imports to obtain the estimates of those imports which may be regarded as generated by the general internal demand for imports.
As shown by Chart 3, the broad movements in the dollar value of imports, adjusted in this way, can be explained to a large extent by changes in the national income in dollars. For 1952, however, imports appear to be far below what would be expected from the level of national income in dollars alone. This deviation may be associated with the fact that inventories of imported goods accumulated in 1951, at a time when fears of war-time scarcity were active, were liquidated in 1952, when these fears had been replaced by the expectation of a decline in world market prices. Actually, the prices of Peru’s imports in terms of dollars did turn downward after 1952. The fact that the demand for imports in 1951 and 1952 taken together should have diverged from the “normal” relationship may have something to do with the relative level of domestic and import prices, but this cannot be established in view of the doubtful character of the available price data.
National Income and Imports of Peru, Annually, 1950–54
(In millions of U.S. dollars)
Citation: IMF Staff Papers 1957, 001; 10.5089/9781451960211.024.A003
1 Estimates of imports not directly financed by foreign investment.2 At average certificate rate of the year.At any rate, the enormous increase in imports (excluding those directly financed by foreign direct investment) in 1951 and 1952, compared with 1950, must be attributed chiefly to the internal expansion of income. Broadly speaking, it was the balance between the increase in export proceeds, resulting from the favorable world demand for Peru’s exports, and the increase in demand for imports, owing to internal expansion, that maintained the remarkable exchange rate stability during the first three years of the experiment with fluctuating rates.13
Depreciation of the Sol in 1953
The real test of the workability of Peru’s flexible rate system under adverse trade conditions came in 1953, when the prices of its chief exports suffered severe setbacks on the world market, while the momentum of internal inflation still continued.
The index of export prices in U.S. dollars dropped by 20 per cent from 1952 to 1953 and by 33 per cent from the boom year 1951 to 1953. On the other hand, the demand for imports, in terms of dollar value, as a function of the exchange rate shifted upward from 1952 to 1953. As may be seen from Tables 2 and 4, national income in soles rose by 7 per cent, and the price of home goods by 4 per cent, from 1952 to 1953,14 whereas the price, in dollars, of Peru’s imports fell by 3 per cent.
In 1953, therefore, the certificate rate was under pressure from both the demand and the supply side. At first, the monetary authorities showed some reluctance to allow too rapid a rise of the certificate price; this is evidenced by the Central Reserve Bank’s net sales of some $12.5 million (exclusive of $10.0 million received under a loan from the International Petroleum Company) during the first half of 1953. The certificate rate was thus prevented from rising rapidly in the first half of the year. From the end of 1952 to the end of June 1953, the certificate rate rose from 15.60 soles per dollar to only 16.13 soles.
During the latter half of 1953, however, the authorities seem to have given up major attempts to intervene and to have resigned themselves to letting the rate find its own level. The net sales of foreign exchange by the Central Bank to the private sector during the second half of the year amounted to no more than a fraction of a million dollars, although government requirements ($4.3 million) continued to be met out of the Central Bank’s exchange holdings. In these circumstances, the certificate rate rose rapidly from 16.13 soles at the end of June 1953 to 19.89 soles at the end of December 1953 and 21.63 soles at the end of January 1954.
This significant experience of sharp exchange depreciation should be examined in relation to three important questions: whether speculation in the foreign exchange market tended to be destabilizing; whether the exchange rate depreciation necessary to maintain balance of payments equilibrium in the short run was much in excess of the change that was necessary in the long run; and whether the fluctuation in the exchange rate tended to intensify the instability of internal prices.
Such evidence as is available in regard to the first question indicates that, on the whole, speculative capital movements during 1953 were on balance not destabilizing. This was not because the exclusion of unauthorized capital transactions from the certificate market and the limitation of the validity of the certificates to 15 days or less had rendered the certificate rate immune from the destabilizing influence of speculation; for, as already pointed out above, the draft market, where capital movements are permitted freely, and the certificate market are in fact linked through the possibility of arbitrage operations by foreign investors. Consequently, speculation on the draft market would certainly have repercussions on the certificate market. Moreover, speculation in the form of postponing exports and speeding up imports cannot be prevented and certainly can exert considerable pressure on the certificate rate. Finally, an anticipation of impending increases in the certificate rate would make importers less willing to import on credit that would add to their future commitments in foreign exchange and hence would add pressure to the spot certificate market. Nevertheless, in spite of all these possibilities, no evidence has been found of large-scale speculative operations in the direction of destabilization.
From the balance of payments statement in Table 1, it may be seen that private short-term capital exports by Peruvian residents, in the form of net increases in foreign assets held by Peruvian residents through Peruvian banks (item 12) and in dollar balances of Peruvian residents in the United States (item 13), decreased from $14.1 million in 1952 to $5.3 million in 1953, and then recovered to $11.1 million in 1954, despite the fact that in 1954 there was some appreciation and finally a complete stabilization of the exchange rates.
Moreover, the outstanding amount of import credits received from foreign suppliers (item 14 of Table 1) continued to increase in 1953, although the rate of increase was smaller than in 1952 ($3.3 million, compared with $7.9 million). The fact that import credits continued to increase indicated that importers as a whole had not attempted to reduce their future foreign exchange commitments because of the expectation of further depreciation of the sol. On balance, there appears to have been a much smaller net outflow of private short-term capital in 1953, the critical year of exchange depreciation, than in 1952, a year of stability, and in 1954, a year of appreciation and stabilization.15 Thus there was no evidence of a speculative capital flight generated by the anticipation of, or the actual continuous depreciation of, the sol during 1953.
Nor is there sufficient evidence to justify the belief that imports in 1953 were artificially inflated by speculative anticipation of further depreciation. After making a deduction for imports arising out of foreign direct investment and developmental loans to Peru, the value of imports attributable to domestic demand is estimated to have increased from $244 million in 1952 to $251 million in 1953 (see Table 5).
To judge from the relationship between national income, import expenditures, and the exchange rate shown in Chart 3, this increase of $7 million (about 2 per cent) in imports represented merely a return to the normal relationship between income and imports after the strong downward deviation of 1952. The drawing upon inventories of import goods in the preceding year must have come to an end when the prices of imports (in dollars) had fallen and no further declines were in prospect. Thus, the increase in adjusted import expenditures in dollars, in spite of a decline of national income in dollars, can be accounted for without reference to speculative anticipation of further depreciation. It might be thought that, in the absence of speculation, the import rates should have declined on account of the effect of the depreciation (approximately 10 per cent on the average between 1952 and 1953) upon the relative prices of imports. This effect, however, was probably insignificant, since the depreciation was offset to a very large extent by a rise in home prices (4 per cent in wholesale prices and 8 per cent in retail) and a decline in dollar prices of imports (3 per cent) (see Tables 2 and 5).
The absence of large-scale destabilizing speculation on the exchange market might be attributed in part to the restraint on credit expansion applied by the Peruvian authorities soon after the exchange market showed signs of weakness. On April 20, 1953, the Government imposed supplementary reserve requirements of 50 per cent on new demand deposits and 25 per cent on new time deposits of all commercial banks. This measure almost completely arrested the rapid expansion of bank credit that had been taking place in the first four months of that year. Subsequently, on November 16, 1953, the basic reserve requirement for commercial banks on demand deposits as of April 20, 1953 was raised from 20 per cent to 22 per cent, and that on time deposits was increased from 10 per cent to 11 per cent. This had the effect of mopping up, to a large extent, excess reserves in commercial banks. As a result, although the domestic loans and investments of commercial banks increased by 427 million soles during 1953 as a whole, only 64 million soles of the increase occurred after April of that year. The market rate of interest as represented by the government bond yield rose from a monthly average of 7.50 per cent in December 1952 to 8.16 per cent in April, 8.82 per cent in December 1953, and 9.02 per cent in January and February 1954.16 The extent to which the restraints on credit expansion and the consequent rises in interest rates affected the cost calculations of speculators in foreign exchange and hence the international flow of short-term capital is hard to assess. But the timely restraints on credit must certainly have tended to strengthen confidence in the sol, and hence to encourage speculation in a stabilizing direction.
The second question is more difficult to answer conclusively on the basis of the available statistics, although, a priori, the short period change in the exchange rate would clearly have to be greater than the ultimate change required. In any case, this question cannot be answered until the stabilization of the rate in 1954 is studied. For the time being, it may be noted that the rise in the certificate dollar rate brought about by free market forces during the second half of 1953 was smaller than the rough estimate of the increase in the average cost of earning dollar exchange through exports. According to Table 4, the decline in export prices in foreign markets combined with the increase in domestic cost would, on the assumption of no improvement in productivity, probably have raised the average cost of producing a dollar’s worth of exports by some 37 per cent from 1952 to 1953. The certificate rate at the end of 1953, on the other hand, was only 28 per cent higher than at the end of 1952.17 No doubt, productivity in export industries must have increased somewhat from 1952 to 1953 with the completion of some of the investment projects in mining and agriculture. Even when due allowance is made for the increase in productivity, however, the certificate rate does not seem, up to the end of 1953 at least, to have been pushed up much in excess of the cost of earning dollars through producing for export.
An examination of the third question must also be deferred until the stabilization of 1954 is studied.
Stabilization in 1954
The certificate rate continued to rise rapidly during January 1954, and by the end of the month it was 9 per cent higher than at the end of 1953. This sharp rise coincided with an even more abrupt seasonal decline in export proceeds, which dropped by 19 per cent, from US$19.2 million in December 1953 to US$15.6 million in January 1954. In the face of such an abrupt drop of exchange proceeds, the 9 per cent rise (or even a much greater rise) of the certificate rate could easily have been engendered merely by the reluctance of private speculators to supplement out of their own holdings the temporary deficiency of supply in the exchange market, as the elasticity of demand for imports in the very short run is likely to be quite small.
The authorities appear to have done little, if anything, to tide over the temporary shortage of foreign exchange in January. In February, however, the President of Peru announced a stabilization program which included inter alia (1) the reduction of government expenditures to the level of current receipts, particularly by slowing down public works, (2) the checking of bank credit expansion, and (3) more active regulatory intervention by the Central Bank in the exchange market, using, if necessary, a newly acquired $30 million stabilization credit, for the purpose of keeping the rate stable in the short run while continuing to allow it to be determined by long-run market forces. This credit was available in terms of stand-by arrangements made with the International Monetary Fund ($12.5 million), the U.S. Treasury ($12.5 million), and a private New York bank ($5 million). The President’s announcement brought about an almost immediate improvement in the certificate market. The certificate rate declined from 21.63 soles per dollar at the end of January to 18.73 soles at the end of February. The rate then fluctuated between the levels reached during November and December 1953, until October 1954 when things had so improved that the Central Bank had to intervene by buying surplus supplies of dollars in order to stabilize the rate at 19.00 soles for the rest of the year.
The immediate strengthening of the certificate rate at the end of February was no doubt due mainly to the confidence inspired by the stabilization program and the $30 million stabilization credit and the general belief that the Government would not hesitate to use the credit to force down the rate if it was considered too high. Confidence encouraged speculators to supply the temporary excess demand for foreign exchange without asking for too great a risk premium; the sol was thus strengthened at once, although the seasonal dip in export proceeds was not over until April.
This psychological stimulus to confidence, however, cannot be regarded as the cause of the lasting improvement in the balance of payments and of the certificate rate in the year 1954 as a whole. The major component in the improvement in Peru’s over-all balance of payments in 1954 was the trade balance, which changed from a deficit (f.o.b.) of $25 million in 1953 to a surplus of $37 million in 1954 (Table 1). The balance of private short-term capital movements, which were most susceptible to speculative and psychological influences, actually deteriorated considerably during 1954 compared with 1953. As already pointed out, the amount of private short-term capital outflow in 1954 (in the form of increases in foreign assets held through Peruvian banks and in dollar balances of Peruvian residents in the United States), at $11.1 million, was more than double the outflow of $5.3 million in 1953.
Thus the eventual stabilization of the sol in 1954 must have been due to improvements in the basic conditions of imports and exports. The internal stabilization policy, however, could hardly make any appreciable contribution to the improvement of the trade balance and to the stability of the sol until late in the second half of the year. Although it was announced in the stabilization program that the Government would attempt to balance the budget and check the expansion of bank credit, neither of these aims was achieved in 1954. The net increases in aggregate bank credit to the private sector and to the development banks were larger both in the first half and in the second half of 1954 (S/. 368 million and S/. 100 million, respectively) than in the corresponding periods of 1953 (S/. 355 million and S/. 83 million, respectively).18
Moreover, although the total deficit of the government budget was smaller in 1954 than in 1953 (S/. 148.8 million, against S/. 294.6 million), the part of the deficit that was financed in an inflationary way was actually larger. Of the 1953 deficit, 159.6 million soles was financed by a $10 million loan from the International Petroleum Company, which enabled the Central Reserve Bank to sell that much more foreign exchange to the public and to absorb the corresponding amount of soles into the Bank from the private sector. Another 121 million soles was financed by the issue of Road Bonds and Public Work Bonds, which, to the extent they were taken up by commercial banks,19 are not as inflationary as the treasury paper discounted with the Central Bank with which the rest of the deficit was financed. In contrast, practically all of the government deficit in 1954, i.e., 131.5 million soles out of 148.8 million, was financed by treasury paper discounted with the Central Bank and other banks. Thus during 1954 as a whole, the money supply expanded further by 221 million soles, or by more than 6 per cent. It was not until August 1954 that the money supply finally leveled off.
In spite of the continued expansion of bank credit and the money supply, there was a remarkable improvement in the trade balance. This was the combined result of a decrease in imports (f.o.b.) of $34.6 million20 and an increase in exports (f.o.b.) of $27.3 million. A great part of the decrease in imports, however, was a direct consequence of the decline in direct investment by foreign companies and in the utilization of specific foreign loans for development ($21.3 million in 1954 against $48.8 million in 1953). On the assumption, as above, that the whole amount of direct investment and specific development loans was used to finance imports of investment goods and materials used in connection with the investment projects concerned, the decrease between 1953 and 1954 in imports not directly financed by foreign investments may be put at somewhere around $14 million c.i.f. (Table 5). Detailed statistics of recorded trade show that from 1953 to 1954 imports of capital goods declined the most, viz., by $28 million, or 30 per cent, which was very close to the $27.5 million decrease in direct investment and the utilization of specific loans.21 Imports of industrial materials and fuels declined by $3 million, or 3.8 per cent, while imports of consumer goods declined by $12 million, or 14 per cent.22 It would not be far wrong, therefore, to estimate at somewhere around $14 million the decrease in 1954 in imports not directly induced by the decrease in foreign investment.
A re-examination of Table 5 and Chart 3 indicates that this internally induced decrease in imports in 1954, which took place in spite of an increase of 221 million soles in the money supply, is to be attributed chiefly to the depreciation of the sol in the latter months of 1953. Since the depreciation took place chiefly in those months, the average certificate rate for 1954 (expressed in soles per dollar) was 13 per cent higher than in 1953, even though in 1954 there was some strengthening of the sol.
One notable fact about the 1953-54 depreciation, however, is that the effect of the 13 per cent depreciation upon the relative prices of imported goods was almost completely obliterated by a 2 per cent decline in the dollar prices of imports in supplying countries and by a 10 per cent rise in the prices of home goods in the domestic market (see Tables 2 and 5). The influence of the depreciation seems to have operated chiefly through the relationship between income and import-expenditure. That is to say, insofar as the expenditure on imports is a function of national income, a depreciation of the exchange rate in excess of the percentage increase in national income would tend to reduce expenditure on imports in terms of foreign exchange, even if the depreciation should bring no significant change in relative prices of imports and home products. Thus, as may be observed from Chart 3, the estimated $14 million decline in imports (excluding those directly financed by capital imports) appears to have been accounted for largely by the depreciation in excess of the percentage increase in national income (i.e., by the decrease in national income in terms of dollars at the current depreciated rate of exchange).
It is not easy to gauge the precise relative importance of various possible causes of the increase of $27.3 million (or 12 per cent) in Peru’s export proceeds in 1954. There was hardly any change in the average dollar prices of Peru’s exports in 1954, compared with 1953.23 With more or less unchanged export prices in world markets, the 13 per cent depreciation of the sol between 1953 and 1954 would certainly increase the profit incentive for exports, for the rises in domestic prices and costs were insufficient to offset such a depreciation. The cost of living increased by 5 per cent, and the wholesale prices of domestic goods by 10 per cent.
It is impossible, however, to separate the effect of the increased current profit incentive from that of the large capital expansion arising out of the Korean boom, which must to a large extent have reached the stage of fructification by 1954. As may be seen in Tables 1 and 5, foreign direct investment and the utilization of specific loans for capital expansion totaled $135.8 million over the three years 1951-53. Such heavy investment outlays would no doubt have considerably increased the production and refining capacity of Peru’s mining industry, which accounted for one third of total exports in 1954.
One extreme example is the newly developed iron mines, which in 1953 produced and exported only 553,000 tons of ore, exported 1,169,000 tons in 1954, and toward the end of that year reached a monthly rate of production of 200,000 tons. Similarly, the more than 7 per cent increase in the volume of petroleum exports in 1954 was probably due mainly to the intensive drilling and prospecting undertaken in the preceding years by a number of foreign and Peruvian companies in the coastal and eastern parts of the country. In agriculture, investment in irrigation and increases in cultivated area under the government development program had also brought about increases in such export crops as cotton, sugar, and rice. Thus, the increase in the dollar value of export proceeds in 1954 must be regarded as the combined result of the increase in productivity brought about by capital expansion in the preceding years and the additional incentive provided by the depreciation.
The quick recovery of the sol, accompanied by the remarkable movement in the trade balance during 1954 without any apparent improvement in Peru’s export markets, seems to indicate that up to January 1954 depreciation had gone further than was required for long-run trade equilibrium, after the depreciation had had time to work out its full effects on exports and imports and the increased productive capacity of export industries that was to be expected from past capital expansion had actually taken place.
This, however, does not mean that the depreciation in 1953 was intensified by exchange speculation of a destabilizing type. On the contrary, as pointed out above, the sharp decrease of 62 per cent in the outflow of private short-term capital between 1952, a year of exchange stability, and 1953, the year of depreciation,24 was probably a result of the normal outflow of short-term capital being partially offset by speculation of a stabilizing sort, i.e., by speculative sales of foreign exchange by those who believed that the depreciation might have proceeded too far. The fact that in 1954, the year of recovery and stabilization of the sol, the outflow of private short-term capital recovered its former strength and rose to more than double that of the preceding year seems to indicate the cessation of speculative sales of foreign exchange and possibly some repurchases by speculators when the expected recovery of the sol was realized. Such behavior of private short-term capital does not conform to the widespread belief, derived from interwar experience, that private short-term capital movements tend to be dangerously destabilizing under any flexible exchange system.
Effect of Fluctuations in the Exchange Rate on the Domestic Income and Price Level
Finally, the effects of fluctuations in the exchange rate of the sol upon domestic income and prices in Peru should be examined. What would the outcome have been had the certificate rate in 1953 been pegged at the level prevailing at the beginning of that year, in the face of severe declines in export prices abroad and continued inflationary tendencies at home? One thing that is certain is that Peru’s exports in 1953 would have been smaller both in quantity and in value. If the same degree of import freedom had been permitted under the pegged rate as under the flexible certificate system, there would certainly have been a far larger deficit in the trade balance. As long as Peru was willing to let the foreign exchange reserve be depleted rapidly, this trade deficit would no doubt have been a major deflationary force and a stabilizing influence on domestic prices. This, however, could not have continued long; for with a foreign exchange reserve25 of only about one sixth of the annual value of imports at the end of 1952, the exchange authorities would certainly not have ventured to use their reserve without limit to peg the exchange rate. Most probably, they would not have drawn upon the foreign exchange reserve to a significantly greater extent than they actually did under the flexible rate system, but would have restricted imports by quota to the available export proceeds and net capital inflow. If this was the most likely course of action under a pegged rate of exchange, then as far as the effect upon aggregate effective demand and money supply is concerned, it would have made little difference whether the exchange rate was pegged or left flexible. The higher income of the export industries resulting from automatic depreciation under the flexible rate, compared with the fixed rate alternative, would be offset exactly by the smaller expenditure of the public on import-competing goods—made possible by the economy’s greater importing power—plus the absence of the abnormal profits which importers would have made under an artificially pegged rate of exchange.
Moreover, since the quantity of imports under a controlled fixed rate would probably have been smaller than under the flexible rate, and since it is unlikely that effective price control could have been enforced through rationing imports to final consumers, the prices of imports in the domestic market would have been higher than under the flexible rate system.
While, if the exchange rate had been kept stable in 1953, the prices of export goods in Peru would have been somewhat lower than they actually were, there is no reason to believe that the cost of living as a whole would have been any lower than it actually was. Therefore, the argument that a depreciation of the exchange rate necessarily means a rise in the cost of living and that this, in turn, will give rise to a demand for wage increases, setting off a vicious cycle of increases in wages and prices, does not apply to Peru in 1953.
These conclusions appear to be at least partially borne out by the statistics shown in Chart 4, from which it may be observed that the indices of wholesale prices of all goods, wholesale prices of home goods, and the cost of living show little if any correlation with the certificate rate. Even the sharp depreciation of 1953 and the abrupt improvement in the early months of 1954 had hardly any tangible direct effects upon these indices.
Domestic Price Indices, Money Supply, and Certificate Rate of Peru, Monthly, 1950–54
(Logarithmic vertical scale)
Citation: IMF Staff Papers 1957, 001; 10.5089/9781451960211.024.A003
Mr. Tsiang, economist in the Special Studies Division, is a graduate of the London School of Economics, and was formerly Professor of Economics in the National Peking University and the National Taiwan University. He is the author of The Variations of Real Wages and Profit Margins in Relation to Trade Cycles and several articles in economic journals.
Since this mission was presided over by Mr. Julius Klein, it is commonly referred to as the Klein Mission.
With a pre-reform certificate rate of, say, 17.93 soles (the monthly average certificate rate for October 1949), nonmineral exporters, who had to surrender 45 per cent of their export proceeds at the official rate of 6.485 soles per dollar, received an average of only 12.78 soles per dollar’s worth of their exports. They were thus in effect obliged to pay a discriminatory export tax of 29 per cent ad valorem.
The surrender requirements varied from time to time, however. For instance, on March 21, 1951 the percentage of export proceeds in dollars and French francs that had to be converted into certificates was reduced from 100 per cent to 75 per cent, and on April 23 of the same year it was further reduced to 50 per cent. But it was again raised to 75 per cent on May 10 and 100 per cent on May 16, 1951. The percentage of export proceeds in sterling required to be converted into certificates was reduced to 10 per cent for the period from April 13, 1951 to April 7, 1952.
At first, the period of validity of all certificates was established at 60 days on the recommendation of the Klein Mission. In May 1951, the period of validity for certificates in dollars and French francs was reduced to 15 days. In January 1954, when the sol was under great pressure, the validity period for all certificates except those in Argentine pesos, a perennial surplus currency, was further reduced to only 5 days. Toward the end of the year, when Peru’s balance of payments position had greatly improved and the sol had been stabilized, the validity period of all certificates was established at 10 days, except those in Argentine pesos, which continued to be valid for 60 days.
The types of payment eligible for, and usually permitted in, the certificate market included freight and transit expenses, interest payments and dividends on foreign capital, royalties, agents’ commissions, repayment of commercial debts, payments for insurance related to merchandise trade, remuneration of foreign technicians, etc.
The volume of Peru’s trade is small. The chief exports are cotton, sugar, metals, and petroleum, the markets for which are highly competitive. The prices of these exports in world markets are, therefore, beyond the control of Peru. Similarly, the foreign prices of Peru’s imports are independent of Peru’s actions. Consequently, the argument for a tariff or for discrimination in order to improve the terms of trade is unimportant for Peru.
The negative figure for 1950 may be the result of statistical inaccuracies rather than an actual net withdrawal of foreign investment for that year. In that year there was an unusually large net increase in import collections: $16.3 million, against $6.5 million for 1951, $7.9 million for 1952, $3.3 million for 1953, and $4.2 million for 1954. A considerable proportion of the inflow of direct investment, which consisted largely of imports of capital goods by the investors themselves, had probably been registered as credit for imports and hence inflated the latter to an unusual proportion. From a comparison of this unusually large entry with those of later years, it may be inferred that there may have been a positive though small net inflow of direct investment even in 1950.
The theory is sometimes advanced that the disposal of foreign credits in prepayment for exports may constitute another link between the certificate and draft markets. Such prepayments can be either surrendered to the Central Bank for certificates in advance of exports or sold on the draft market. However, an exporter who sells the prepayment for his future shipment of exports on the draft market is obliged to repurchase the same amount of foreign exchange on the draft market when he actually makes the shipment, in order to fulfill the obligation of surrendering the current export proceeds to the Central Bank in exchange for certificates. Therefore, the usual practice is for an exporter who sells spot on the draft market his export prepayments to cover, at the same time, his future obligation with a purchase of futures of the same amount on the draft market. Whether this operation will represent a net gain to the exporters receiving prepayments for future shipments depends upon whether the spot draft rate minus the futures draft rate plus the futures certificate rate is greater than the spot certificate rate (all exchange rates being expressed in soles per unit of foreign currency). The excess of the spot draft rate over the spot certificate rate by itself is neither a necessary nor a sufficient condition for a profitable arbitrage. By the same token, such arbitrage operations would not necessarily tend to equalize the spot draft rate and the spot certificate rate, except when the futures draft rate is expected to be equal to the futures certificate rate.
A simple correlation between the monthly value of imports in soles and the money supply at the end of the preceding month has been attempted for the period January 1949–December 1950. The correlation coefficient is found to be 0.747.
Weighted arithmetic average of the unit values of 7 items that constitute 90 per cent of Peru’s total exports, computed from trade statistics; the weights equal the 1953 relative value of trade. Another series of export prices from a Peruvian source shows the increases in dollar prices of Peru’s exports as follows: 1950,100; 1951,140; 1952,114; 1953,105; 1954,113. This series, however, is less reliable since it is computed as the unweighted geometric average of the wholesale prices of 15 export commodities in the domestic market, and is not directly derived from trade statistics.
The cost of living index is used here because no wage index is available. The cost of living index, however, tends to overestimate the rise in the domestic cost of production for three reasons. First, the index refers only to the cost of living in the capital city of Lima, where prices presumably tend to rise faster than in outlying areas. Second, the wages of miners and agricultural laborers, who are rather unorganized in Peru, tend to lag behind the rise in their cost of living. Third, the index does not allow for the increase in productivity that might be expected from the substantial investments in export industries after the outbreak of the Korean conflict, particularly during the years 1950-53. When these investments came to fruition, the cost of living index would certainly tend to overestimate the cost of production in export industries. Therefore, the estimated indices of the cost of production per dollar’s worth of exports are probably more reliable for the last two years (1953 and 1954) shown in Table 4 than for the preceding years.
There was some direct intervention on the part of the Central Reserve Bank in 1951, during which year it added $9.5 million to its reserves when the prices of Peru’s exports were particularly favorable.
The cost of living rose by 8 per cent from 1952 to 1953.
This statement is true even when allowance is made for the errors and omissions in the balance of payments statement by assuming that the net errors and omissions for these three years occurred entirely in the totals of short-term capital transactions and by adding the former to the latter. See Table 1.
Afterward, it gradually declined to 7.5 per cent toward the end of 1954. The bond yield, however, represents only the long-term rate. How far the short-term rate in the free money market changed over the period concerned cannot be ascertained statistically. The discount rate of the Central Reserve Bank for commercial paper, however, has remained unchanged at 6 per cent since November 1947.
The annual average certificate rate in 1953 was only 11 per cent higher than that in 1952. However, since in 1953 the certificate rate was permitted to rise freely only in the second half of the year, a comparison of the annual average rate in 1953 with that in the preceding year would grossly understate the forces making for the rise of the exchange rate during that year.
The decrease in bank credit to the private sector during the second half of 1954 was more than offset by the increase in credit to the development banks. This shift in the composition of bank credit was due mainly to the transfer of the financing of the current rice crop from the commercial banks to the semiofficial agricultural bank, which was one item in the stabilization program.
None of these bonds seem to have been bought by the public.
Imports c.i.f. decreased by $42 million from 1953 to 1954.
This category of imports, however, included the subgroup, “Vehicles and Transport Equipment,” which in turn included automobiles, the import of which was prohibited from November 1953 to August 1954. It is not known how much of the decrease in this category of imports was due to the prohibition of imports of automobiles in the first seven months of 1954.
Boletín del Banco Central de Reserva del Peru (Lima), March 1954 and March 1955.
See Table 3. According to the Peruvian export price index, however, average export prices increased by more than 7 per cent between 1953 and 1954. This discrepancy is explained in footnote 10.
The net increase in foreign assets held by Peruvian residents through Peruvian banks plus the net increase in dollar balances of Peruvian residents in the United States was $14.1 million in 1952, $5.3 million in 1953, and $11.4 million in 1954.
Including the so-called “untouchable gold reserve.” The foreign exchange and gold reserve of the Central Reserve Bank at the end of 1952 amounted to US$55.7 million, of which US$18.4 million in gold was considered “untouchable.”
