THE JAPANESE BALANCE OF PAYMENTS in the postwar period began to show a tendency to fluctuate in 1953, when it moved abruptly into deficit. Between 1952 and 1958, there were two cycles in Japan’s balance of payments. These cycles were examined in considerable detail by Narvekar,1 who brought out, on the basis of annual data, the interrelations between the domestic and international factors that were reflected in these movements in Japan’s external balance. The present paper traces the fluctuations in Japan’s balance of payments since 1959 and examines some of their characteristics. For this purpose, it draws on quarterly data on Japan’s balance of payments now available for the period beginning 1961 and estimated by the authors for 1959 and 1960. However, it does not attempt to relate these fluctuations in detail to developments in domestic or international demand and supply conditions.
Over the period covered by Narvekar, short-term capital movements in Japan on an annual basis were not very significant2 and his studies were therefore based on data relating to Japan’s over-all balance of payments. Since 1959, however, short-term capital movements into and out of Japan have grown enormously, probably because of the widespread relaxation of payments restrictions during 1958 in the leading industrial countries, as well as Japan. As a result, the basic and over-all balances in Japan have tended to differ greatly from one another.3 Any analysis of Japan’s balance of payments for the period since 1959 must take this fact into account. Possibly, short-term capital flows were quite large on a quarterly basis even before 1959 but tended to be offset within the year, thus appearing small when taken on an annual basis. If so, short-term capital flows should be considered in studying fluctuations in Japan’s balance of payments even before 1959. However, quarterly data are not available for the earlier period.
The basic balance of payments comprises all the current transactions, transfers, and long-term capital flows. The results achieved on the basic balance may therefore be said to reflect the underlying strength or weakness of the economy from the viewpoint of its external relations. When the basic balance is in substantial deficit, action is usually taken to eliminate the deficit in order to protect official reserves. However, if an inflow of short-term capital helps to offset that deficit and to produce a smaller over-all deficit (and sometimes even an over-all surplus), there would tend to be less strain on resources available for official financing of the deficit and, consequently, less danger of a disturbance on the exchanges. A large basic surplus may then finance an outflow on short-term capital account instead of increasing net official resources available for external financing. These offsetting movements of the basic balance and short-term capital movements would moderate fluctuations in the over-all balance and thus economize on the use of official resources for financing external imbalances. It is possible, however, that short-term capital movements may accentuate fluctuations in the over-all balance and thus have a destabilizing rather than a stabilizing effect. Therefore, it is important to know whether short-term capital movements have tended to moderate fluctuations in the over-all balance of payments or to accentuate them.
If short-term capital flows play a stabilizing role in the sense just described, the over-all balance could still show surpluses, or relatively small deficits, even when the basic balance is already in relatively large deficit. Again, the over-all balance may show deficits, or relatively small surpluses, even when the basic balance has already moved into relatively large surplus. Depending, therefore, on whether the offsetting movements of short-term capital flows are sufficiently large relative to the basic surplus or deficit, the over-all balance may sometimes have the same sign as the basic balance and sometimes the opposite sign. It follows that even when there is a definite pattern in the movements of the basic balance of payments, this pattern may not be apparent in the movements of the over-all balance. If attention is concentrated on the over-all balance of payments, significant underlying patterns in the country’s external balance may thus be missed. In looking for cyclical and other patterns in the movements of the Japanese balance of payments over the period under review, greater attention will therefore be paid to the data on the basic balance.
Narvekar’s studies of fluctuations in Japan’s balance of payments between 1952 and 1958 were based on annual data, quarterly data not being available. For the purposes of cyclical analysis, such data cannot be considered very satisfactory. For instance, annual data may show a large surplus in a year in which a distinct movement toward a deficit has already begun to manifest itself during the latter part of the year. The reverse may also be true. To obtain a better picture of cyclical patterns and of the turning points in them, as well as to relate these to official policy measures, it is necessary to use quarterly data. The latter data also make possible a more detailed analysis of the relationship between movements in the basic balance, on the one hand, and flows of short-term capital, on the other. Quarterly data for recent years are now available and are used in this study. For continuity with Narvekar’s studies, the present paper briefly presents results obtained from the annual data since 1959 before turning to the more detailed examination made possible by the availability of quarterly data.
Fluctuations Shown by Annual Data
On the basis of annual data (Table 1 and Chart 1), it appears that Japan has passed through two additional cycles after the 1952–55 and 1955–58 cycles covered by Narvekar’s studies. However, the growth of short-term capital flows over this period and particularly the direction of such flows relative to the direction of the fluctuations in Japan’s basic balance of payments have produced a situation in which the cycles sometimes appear to cover different periods if cycles are defined from peak surplus to peak surplus, depending on whether the basic balance or the over-all balance is used as the criterion. Thus, the third cycle began in 1959 on the criterion of basic balance, but in 1960 on the criterion of over-all balance because a substantial inflow of short-term capital offset the fall in the basic surplus in 1960 to maintain the over-all surplus of this year at the high level of the preceding year. The dip in the over-all surplus thus began a year later than the dip in the basic surplus.
|Private Short-Term Capital, Including Net Errors and Omissions|
|1966 3||432||−482||−50|Chart 1.Japan: Balance of Payments, Annual Data, 1953–66
In general, the annual data show clearly that the flows of short-term capital have tended to fluctuate in a direction opposite to that of the fluctuations in the basic balance of payments. Thus, the short-term capital inflow increased substantially as the basic surplus fell sharply from 1959 to 1960 and increased further still as the basic balance went into heavy deficit in 1961. This latter increase in the inflow of short-term capital, however, was not sufficient to prevent the over-all balance from going into a large deficit in that year (1961). In 1962 the basic balance of payments showed a small surplus, which was supplemented by a small inflow of short-term capital to produce a rather larger over-all surplus. In 1963 the basic balance went into deficit again, but the inflow of short-term capital increased, yielding a small over-all surplus. In 1964, however, the basic deficit became so great that the inflow of short-term capital, though large, did not prevent an over-all deficit. The sharp swing in the basic balance in 1965 to a large surplus coincided with a marked outflow of short-term capital, yielding a definite but smaller over-all surplus in 1965. In 1966 a slightly diminished basic surplus but a larger outflow of short-term capital resulted in a small over-all deficit.
Thus, it can be said that short-term capital flows helped to moderate the fluctuations in Japan’s over-all balance of payments, compared with the fluctuations in its basic balance, and thus played an anticyclical role in Japan’s balance of payments during 1959–66.
Fluctuations Shown by Quarterly Data
We may now turn to an examination of the quarterly data, which are available only from the beginning of 1961 for all items in the balance of payments. To obtain a complete series of quarterly data beginning in 1959, the authors have estimated the basic balance and short-term capital flows on a quarterly basis in 1959 and 1960 (see Appendix I, p. 427). Also, the published series of quarterly data for the private short-term capital flows and the over-all balance from the first quarter of 1961 through the second quarter of 1963 have been adjusted. Certain short-term capital flows from U.S. banks to Japanese banks, as well as their repayment from 1961 to 1963, are recorded in the statistics as changes in official liabilities because it was the Bank of Japan which arranged the credits with the U.S. banks. Japanese banks used these funds, together with a large proportion of other private short-term capital flows, to finance imports from the United States. In our opinion, therefore, these funds are more satisfactorily treated as private short-term capital flows. The published data were adjusted accordingly (see Appendix II, p. 430). The partly estimated, partly adjusted quarterly data obtained in this way are reproduced in Table 2 and presented in Chart 2.
|Trade Balance||Other Goods and Services and Private Transfers||Current Account (Cols. 1+2)||Private||Official||Total (Cols. 4+5)||Basic Balance (Cols. 3+6)||Private Short-term Capital, Including Net Errors and Omissions2||Over-All Balance (Cols. 7+8)2||Changes in Official Reserves 3|
|IV||770||−272 6||498||−280||−39||−319||179||−152||27||30|Chart 2.Japan: Balance of Payments, Quarterly Data, 1959–66
For the purposes of this analysis on the basis of quarterly data, a cycle was defined as a fluctuation in the balance of payments covering a period of more than four quarters, measured from the first quarter following a peak surplus to the next peak surplus. In determining the peak surplus in the balance of payments, attention was concentrated on the basic balance and the term “peak” was interpreted somewhat liberally.
On this basis, the third cycle in Japan’s balance of payments extended from the first quarter of 1961 through the fourth quarter of 1962. This cycle began in the first quarter of 1961 with an abrupt swing in the basic balance to a deficit, which worsened in the second and third quarters. Some improvement in the fourth quarter was followed by a new adverse movement in the first quarter of 1962. Thereafter, the basic balance improved steadily in each quarter, producing a large surplus in the third quarter and a peak surplus in the fourth quarter.
The fourth cycle in Japan’s balance of payments extended from the first quarter of 1963 through the fourth quarter of 1964. This cycle, like the previous one, began with a sharp adverse movement in the first quarter. Although the deficit was substantially reduced in the second quarter, somewhat larger deficits appeared once again in the third and fourth quarters of 1963. In the first quarter of 1964 a sharp adverse movement created the heaviest quarterly deficit recorded at any time during this period. Thereafter, however, as in 1962, the situation improved in each quarter, until a large surplus was produced in the third quarter and a peak surplus was reached in the fourth quarter.
These two cycles, each covering eight quarters, thus extended over the period from the first quarter of 1961 through the fourth quarter of 1964. This leaves the two years 1959–60, before the beginning of the third cycle described above, and the two years 1965–66, after the end of the fourth cycle. During these four years there were also fluctuations which are not treated in this study as cycles because each of them does not cover more than four quarters. They do, however, exhibit a very definite pattern. In each of these years, there was a sharp adverse movement in the basic balance in the first quarter of the year compared with the preceding quarter.4 This adverse movement resulted in a relatively small surplus in the first quarter of 1959 and a deficit in the first quarters of 1960, 1965, and 1966. Thereafter, in all four years, the data indicate steady improvement until the third quarter; in 1959 and 1960, further improvement occurred in the fourth quarter, and in 1965 and 1966 the surplus became somewhat smaller. Since the fluctuations during these years cover a period of only four quarters each, it is probable that these fluctuations are primarily seasonal. Inspection of the data for these four years also makes it evident that these seasonal fluctuations are superimposed in all these years on a situation of fundamental surplus in the basic balance of payments. Apparently the cyclical forces that operated on Japan’s balance of payments from time to time were more or less dormant during these years.
Broadly the same seasonal pattern may be traced in the second half of each of the two cycles covered here (i.e., in 1962 and in 1964), though in an exaggerated or enlarged form. The term cycle, as it has been used here to describe the fluctuations of 1961–62 and 1963–64, covers any fluctuation which is spread over more than four quarters and is produced by the combined impact of seasonal and cyclical forces. In view of the great importance of seasonal factors in the Japanese balance of payments, an attempt was made, in spite of the limitations involved, to eliminate seasonal variations in the basic balance data directly.5 The results obtained indicate the presence of stable seasonality at the 1 per cent level. The seasonally adjusted data are presented with the original data in Chart 3. As the chart shows, the seasonally adjusted data confirm the existence of two cycles covering the 1961—62 and 1963–64 periods. During 1965 and 1966, the seasonally adjusted data show irregular movements. Trend-cycle data for 1965 and the first two quarters of 1966, obtained by eliminating both the irregular and the seasonal movements, show an almost constant surplus. This result supports the earlier conclusion that fluctuations during these two years were primarily seasonal. During 1959 and 1960, the seasonally adjusted data indicate a tendency for surpluses to fall from the beginning of 1959. This tendency, however, may be due to weaknesses in the estimated quarterly data for this period. Therefore, there does not appear to be any need to modify the conclusions drawn earlier about the existence of a primarily seasonal fluctuation over this period also.
Chart 3.Japan: Balance of Payments, Actual and Seasonally Adjusted Quarterly Data, 1959–66
Admittedly the detailed quarterly movements shown over the two cycles by the seasonally adjusted data do not agree exactly with the description of the cycle given earlier on the basis of actual data. However, the main conclusions about the periods covered by the two cycles remain unaffected, particularly if the uncertainties associated with the seasonal adjustment process are kept in mind. An examination of Chart 3 suggests that in years of primarily seasonal movements, the seasonally adjusted basic balance data may give some advance hint of the beginning of a new cycle. To read more than this into the seasonally adjusted data would not seem to be justified.
Payments Fluctuations and Monetary Policy
Throughout the period under review, the Japanese authorities tended to use monetary policy in an extremely flexible manner in order to influence economic activity. Alternate tightening and relaxing of monetary policy could not but be intimately related to fluctuations in the Japanese balance of payments examined above. Changes in Japanese monetary policy are outlined below against the background of the balance of payments fluctuations in order to throw some light on this relationship. For this purpose, the eight years 1959–66 are divided into four equal periods of eight quarters each:
Two seasonal years: First quarter 1959-fourth quarter 1960
Cyclical fluctuation: First quarter 1961-fourth quarter 1962
Cyclical fluctuation: First quarter 1963-fourth quarter 1964
Two seasonal years: First quarter 1965-fourth quarter 1966
The two seasonal years (first quarter 1959–fourth quarter 1960)
It has been shown already on the basis of estimated quarterly data that the basic balance during each period of four quarters during these two years showed purely seasonal movements. In each of these periods, there was a steady improvement from a small surplus (1958) or a deficit (1959) in the first quarter to a relatively large surplus in the fourth quarter. Monetary policy developments during this period fell into three distinct subperiods, the first of which began before the period under consideration.
Monetary ease (June 1958–August 1959)
As the economy returned to a more normal situation after the balance of payments difficulties of 1957, the Bank of Japan embarked on a policy of monetary ease by reducing its official discount rate by 0.73 per cent in June 1958 from 8.395 per cent per annum. This move was followed by a further reduction of 0.365 per cent in September 1958. The rates for sale (to local banks) of bills held by the Bank of Japan were also reduced on each occasion. As the economy, in the view of the authorities,6 continued to enjoy a fair degree of prosperity without inflation at the beginning of 1959, the Bank reduced its discount rate by a further 0.365 per cent in February 1959, bringing it down to 6.935 per cent per annum.
Monetary restraint (September 1959–July 1960)
When earlier relaxations of monetary policy began to produce an overexpansion in the economy, the Bank of Japan introduced a degree of monetary restraint in September 1959 by implementing for the first time the Reserve Deposit Requirement System and raising its selling rates for bills by 0.365 per cent. These measures were followed in December 1959 by an increase in official interest rates by 0.365 per cent to 7.300 per cent per annum. This was intended as a precautionary measure but was maintained during the first half of 1960 in view of the deficits on basic balance which appeared during this period. These deficits were not, however, very large and were in fact purely seasonal.
Monetary ease (beginning August 1960)
As a favorable balance was achieved in July 1960, the Bank of Japan lowered its discount rates in the latter part of August by 0.365 per cent, thus restoring the rates that had prevailed before the upward revision of December 1959.
Cyclical fluctuation (first quarter 1961–fourth quarter 1962)
The cyclical fluctuation under review began with the development of a large deficit in the basic balance during the first quarter of 1961; this deficit persisted into the second and third quarters and declined only slightly in the fourth quarter. It increased again in the first quarter of 1962, but thereafter there was a steady recovery until a large surplus resulted in the fourth quarter. Monetary developments during this cyclical fluctuation fell again into three subperiods.
Monetary ease (August 1960–June 1961)
It has been mentioned above that the Bank of Japan began this sub-period of monetary ease by lowering its discount rates by 0.365 per cent in August 1960. In January 1961 the Bank cut its discount rates by another 0.365 per cent to 6.570 per cent and followed this step by reducing the maximum limits of interest rates on deposits with financial institutions in April 1961. The rapid growth of the economy following these measures produced, during the earlier part of 1961, the sharp adverse movement in the basic balance already described as the beginning of the cyclical fluctuation of 1961–62.
Monetary restraint (July 1961–September 1962)
Noting these developments in Japan’s external payments situation, the Bank of Japan reversed its monetary policy. Discount rates were raised twice, in July and September, by 0.365 per cent to a level of 7.300 per cent; the penal rates system for bank borrowings from the Bank of Japan was tightened in September; and the reserve requirements were raised in October 1961. These measures had a considerable impact on business activity, but only after a lag. The basic balance began to improve in the second quarter of 1962. By the third quarter of 1962, the basic balance showed a substantial surplus, which improved further during the fourth quarter. Monetary restraint during this period can thus be related to the recovery phase of this cycle.
Monetary ease (beginning October 1962)
In response to the improvement in the basic balance which had begun already in the second quarter of 1962, the Bank of Japan returned to a policy of monetary ease by reducing official discount rates by 0.365 per cent in the latter part of October 1962. In November 1962 the Bank sharply reduced the reserve requirements and also abolished the penal interest rates system, replacing it with a new system of credit ceilings. Further, in the latter part of November 1962, the Bank made another cut of 0.365 per cent in official interest rates, bringing the rate down again to 6.570 per cent.
Cyclical fluctuations (first quarter 1963–fourth quarter 1964)
The fourth cycle in Japan’s balance of payments began in the first quarter of 1963 with the appearance of a large deficit in the basic balance. Although there was some improvement in this balance in the second quarter, the situation worsened in the third quarter again, and the fourth quarter also showed a deficit, contrasting sharply with the large surplus which seasonal factors tend to produce in that quarter. By the first quarter of 1964, the deficit in the basic balance had become the largest during 1959–66; thereafter improvement was steady, large surpluses being earned in the third and fourth quarters of 1964. Monetary policy developments during this period again fell into three subperiods.
Monetary ease (October 1962–autumn 1963)
As already noted, the Japanese authorities, finding an improvement in the external balance from the second quarter of 1962, took a number of measures to relax monetary policy in October and November 1962. The combined effect of these measures on the state of monetary ease was undoubtedly quite substantial. However, the Bank further reduced its official discount rates twice, in March and April 1963, by 0.365 per cent on each occasion from the already low levels reached in November 1962, thus establishing a rate of 5.840 per cent. In March 1963 the penal addition to the interest rate applicable to Bank of Japan loans exceeding the limit for loans at the official rate but within the credit ceiling (the so-called primary penal rate) was reduced from 1.095 per cent to 0.365 per cent. A few months later, in June 1963, the discount rate applicable to sale or purchase of short-term securities was lowered from 6.023 per cent to 5.658 per cent per annum. Then, in July 1963, the primary penal rate was abolished altogether. It is not surprising that all these measures of monetary relaxation between October 1962 and July 1963 were associated with an adverse reaction on the country’s basic balance of payments. In fact, a new cycle in the basic balance began from the first quarter of 1963.
Monetary restraint (autumn 1963–November 1964)
The adverse developments in the balance of payments during the first three quarters of 1963 led the Bank of Japan in autumn to adopt a less easy monetary policy. It did this first by making adjustments in its buying and selling operations in securities and by lowering its credit ceilings. This move was followed in December 1963 by a sharp increase in reserve requirements, which had already been extended to cover mutual banks and credit institutions in April 1963. During the first quarter of 1964, the Bank of Japan carried out operations in securities with the object of tightening the money market and took measures to restrict lending by commercial banks. In March 1964, official discount rates were raised by 0.730 per cent to 6.570 per cent. All these measures produced a considerable tightening of the monetary situation, and the basic balance improved steadily from the second quarter of 1964, with large surpluses being earned in the third and fourth quarters. Monetary restraint during this period can thus be related to the recovery phase of this cycle just as monetary restraint from July 1961 to September 1962 can be related to the recovery phase of the previous cycle.
Monetary ease (beginning December 1964)
The Bank of Japan introduced its relaxation of monetary policy by halving the required ratios for nontime deposits for all classes of banks.
Seasonal years (first quarter 1965–fourth quarter 1966)
During 1965 the basic balance showed a small deficit in the first quarter, a small surplus in the second quarter, and large surpluses in the third and fourth quarters. During 1966, this pattern was repeated over the four quarters of the year.
In 1965 the Bank of Japan continued the policy of monetary ease which it began by reducing required reserve ratios in December 1964. First, it reduced its official discount rates in January 1965 by 0.365 per cent. As the fundamentally surplus tone of the external balance strengthened, the Bank reduced its official discount rates again by 0.365 per cent in early April 1965 and by another 0.365 per cent in late June. This brought the rate down to 5.475 per cent per annum. Then, it discontinued its measures to check increases in commercial bank lendings and in mid-July 1965 further reduced the required reserve ratios. Monetary conditions eased significantly after the tight-money measures were relaxed; nevertheless, the economy continued to be sluggish, and business conditions failed to improve during the year. The external balance thus showed the usual seasonal pattern, with a stronger surplus tone than normal. In 1966, also, this tone continued. The authorities then used fiscal measures to stimulate the economy, and there was some evidence toward the end of the year that the economy had begun responding to these measures, with possible implications for the balance of payments during 1967.
Summary of monetary policy, 1959–66
This review of monetary policy changes indicates an intimate relationship between changes in Japanese monetary policy and fluctuations in Japan’s basic balance of payments during 1959–66, as is vividly brought out by Chart 4. In this chart, showing fluctuations in the basic balance and in short-term capital flows, periods of monetary ease have been shaded while periods of monetary restraint have been left unshaded. Although this chart cannot provide a full indication of the degree and detailed timing of monetary tightening and relaxation, it does show that alternating periods of monetary ease and monetary restraint have generally been accompanied, after a lag of varying duration, by alternating periods of deficit and surplus in Japan’s basic balance of payments.
Chart 4.Japan: Balance of Payments and Monetary Policy, Quarterly Data, 1959–66
Note: Shaded areas represent periods of monetary ease.
Evidently, both the eight-quarter cycles during this period followed the relaxation of monetary policy at the end of the year preceding the cycle. In both cycles, monetary policy was tightened during the third quarter of the cycle but did not succeed in facilitating an improvement in the basic balance until the sixth quarter. In both cycles, also, as the basic balance continued to improve in the seventh quarter, monetary policy was relaxed again. Depending on the strength of the underlying demand and supply conditions and on the extent and timing of relaxation, the next year marked either the beginning of another cycle or the generation of a purely seasonal fluctuation.
During the eight-quarter period 1959–60, when fluctuations in the basic balance were primarily seasonal, monetary policy passed through an intermediate period of monetary restraint extending from September 1959 to July 1960. In this respect, this period was similar to the two eight-quarter cycles of 1961–62 and 1963–64. However, during the eight-quarter period of 1965–66, in which the fluctuations were also mainly seasonal, the entire period was characterized by monetary ease. The explanation for this difference between conditions during 1959–60 and 1965–66 falls outside the purview of this study.
Role of Short-Term Capital Flows
Attention was drawn earlier to the remarkable inverse relationship between the basic balance, on the one hand, and short-term capital flows on the other, in the annual data from 1959 to 1966. This inverse relationship is also evident in the quarterly data.7 During 1959 and 1960 the inverse relationship existed in terms of the direction of change from quarter to quarter, but not necessarily in terms of sign. Since 1961, however, the basic balance and short-term capital flows, with minor exceptions, have been inverse in both sign and direction (see Chart 2).
Short-term capital flows can thus be said to have performed, particularly since 1961, a compensatory role on a quarterly basis. The compensation has not been complete, and the over-all balance has continued to fluctuate in broad cyclical and seasonal patterns covering the same periods as those covered by the basic balance. However, the detailed patterns within each of these periods have been less regular for the over-all balance of payments than for the basic balance.
The high inverse correlation between the basic balance of payments and short-term capital movements on both an annual and a quarterly basis suggests fairly close links between the elements in the basic balance and those in short-term capital flows. The most obvious is a possible link between trade and trade finance. Japanese banks generally finance Japan’s exports, while foreign banks generally finance its imports. In these circumstances, trade balances and net usance finance received should show an inverse relationship. Although this relationship is in fact evident, it is not as close as that between the basic balance and short-term capital flows, probably because some part of the trade financing does not take the form of usance finance.8 Direct interbank borrowing for the purpose of financing trade probably plays an important role. Suppliers’ credits, whether provided by Japanese institutions to finance exports or by institutions abroad to finance imports into Japan, must also have played some part in strengthening the close connection between trade balances and short-term capital flows.
Another explanation for the close connection between the basic balance and short-term capital flows in Japan may be found in the Japanese banks’ practice of using foreign funds to meet their needs for yen finance for domestic purposes.9 Borrowing and repayment of these funds from time to time have thus been influenced by the exigencies of local money market conditions, but these conditions, in turn, have tended to reflect either variations in the external balance or more generally variations in underlying economic conditions responsible for these variations in external balances. Some parts of these capital movements that are not directly connected with trade financing have consisted of funds borrowed from or repaid to the Euro-dollar market, which usually are somewhat more volatile, responding more sensitively even than other flows to interest differentials and to the strength or weakness of the Japanese balance of payments. In Japan, however, variations in interest differentials have been less important in this context for two reasons: interest rates have generally been much higher in Japan than elsewhere, and variations in the differentials have been relatively small.
The flow of Euro-dollar funds to Japan may have been restrained to some extent by unwillingness among international bankers to allow a large proportion of their funds to be held in Japan rather than by unwillingness among Japanese banks either to absorb larger amounts or to pay sufficiently high interest rates. It is possible also that the strength or weakness of the Japanese balance of payments had some effect on the volume of Euro-dollar funds available to Japanese banks.10 “Guidance” by the Japanese authorities, however, has probably had a more important influence on such flows. In any case, Euro-dollar funds form only a small proportion of short-term funds flowing into Japan. Therefore, even if any direct relationship between Japanese surpluses or deficits and the flow of Euro-dollar funds exists, it was probably swamped by the inverse relationship which other short-term flows, much larger in amount, had with the Japanese basic balance.
In this context, it is evident that the Japanese authorities have been keenly aware of the importance of short-term capital flows.11 They have used various measures, including “guidance” provided to Japanese banks, to influence such flows in an effort to see that net flows of short-term capital are in the right direction and of the right magnitude. The nature of these measures can be illustrated by reference to the steps taken during the 1961–62 and 1963–64 cycles. In 1961, inflows of short-term capital were substantial, resulting partly from the measures of exchange liberalization adopted during earlier years and partly from Japan’s large trade deficits which underlay its basic balance deficits during this period. By the latter part of the year, the Japanese authorities began adopting measures designed to reduce the inflow while rectifying the basic imbalance. Thus, in September 1961, the Bank of Japan put into operation a System of Loans Against Foreign Exchange Assets and fixed the interest rate for this purpose at the extremely low level of 2.555 per cent per annum. In June 1962 the Bank put into operation a Foreign Exchange Reserve Requirement System, under which foreign exchange banks were required to hold a prescribed proportion of foreign exchange assets against their foreign exchange liabilities. At the same time, it made additional types of foreign exchange assets eligible for Bank of Japan loans. All these measures tended to reduce the net inflow of funds by making it both necessary and profitable to hold more foreign assets.
As the basic balance went into heavy deficit once again at the beginning of 1963, the Japanese authorities encouraged short-term capital inflows once again. They did this at first mainly by giving necessary guidance to authorized foreign exchange banks.12 Later, in July 1963, the Bank of Japan increased its interest rates on loans against foreign exchange assets to offset the effects of higher interest rates abroad. By mid-1964 the basic balance had improved substantially. It therefore became necessary to reduce inflows of short-term capital and encourage outflows instead. In July 1964 the Bank of Japan laid down “general guidelines for the maximum amounts of short-term foreign capital to be held by each of Japan’s twelve class ‘A’ authorized foreign exchange banks.”13 In August, it followed this step by raising the reserve ratio under the Foreign Exchange Reserve Requirement System. These measures generated an outflow of short-term capital. The outstanding example of the extent to which the authorities were willing to go to ensure the appropriate amount and direction of short-term capital flows is provided by the arrangements which the Bank of Japan made in 1961 and 1962 to obtain funds from certain U.S. banks for use by Japanese banks. As pointed out earlier, the flow of such funds is reported in the statistics as changes in official liabilities, but should correctly be treated as private short-term capital flows.
As shown above, the special measures which the Japanese authorities took during this period to influence short-term capital flows have generally reinforced the inverse relationship between these flows and the basic balance. The same thing cannot be said, however, about the broad monetary policies adopted by the authorities to deal with the fluctuations in the basic balance. The influence of these policies has sometimes operated in the opposite direction, though without much apparent effect on the inverse relationship between basic balances and short-term capital flows (see Chart 4). Thus, policies of monetary relaxation should encourage outflows of short-term capital. Yet, periods of monetary ease not only have covered quarters of basic surpluses and short-term capital outflows, but have extended over quarters of basic deficits. Despite the influence of monetary policies, there was a compensating inflow during these quarters. Similarly, policies of monetary tightness should tend to encourage inflows of short-term capital. Nevertheless, periods of monetary tightness not only have covered quarters of basic deficits and short-term capital inflows, but have extended over quarters of basic surpluses. Again, there was a compensating outflow during these quarters. These circumstances indicate that the financial patterns in Japan, combined with the special measures adopted by the authorities and the “guidance” provided by them to affect short-term capital flows, exerted more influence on these flows than general monetary policy did.
Payments Fluctuations During 1953–59
The examination above of the fluctuations in Japan’s balance of payments on a quarterly basis during 1959–66 clearly reveals certain broad patterns in these fluctuations. It indicates that in certain years (1959, 1960, 1965, and 1966) these fluctuations consisted of seasonal variations superimposed on basically surplus situations. On the other hand, in 1961–62 and 1963–64, the fluctuations took on characteristics which may be described as cyclical. Cyclical fluctuations generally extended over eight quarters, from the first quarter of one year through the fourth quarter of the second. During each cycle, large basic deficits occurred in the first five quarters, and improvements began with a reduction in the deficit during the sixth quarter. The last two quarters of each of these eight-quarter periods then showed basic surpluses.
If these patterns prevailed also during the period 1953–59, for which no quarterly data are available, it is possible to speculate that the two cycles which occurred before 1959 were also eight quarters in length each. These two cycles thus probably covered the period from the first quarter of 1953 to the fourth quarter of 1954 and again from the first quarter of 1957 to the fourth quarter of 1958. If this is correct, there should have been purely seasonal fluctuations during each of the two years 1955 and 1956. The annual data for these two years do show large surpluses, which are consistent with the idea of a purely seasonal movement—from deficit to surplus on basic balances—superimposed on a fundamentally surplus situation. Without detailed quarterly data, it is of course difficult to be sure that these two years exhibited the kind of purely seasonal pattern which has been traced during 1959–60 and 1965–66. It is even more difficult to be sure that the two cyclical fluctuations of the earlier period in fact followed the eight-quarter pattern. What evidence is available of the pattern within the relevant years is clouded by the fact that it is based on over-all balance data rather than data on basic balance.
With regard to the first cycle in Japan’s balance of payments, Narvekar has pointed out that the crisis was at its height by mid-1953 and that the recovery from that crisis began in the second quarter of 1954.14 It is true that the basic balance for the whole year of 1954 shows only a very small surplus equivalent to US$22 million, but the small size of the surplus may well have been due to a fairly large deficit in the earlier part of the year, particularly in the first quarter. In this context, it is of interest that the Japanese authorities inaugurated a tight-money policy in October 1953, and made it more stringent in January 1954 and again in March 1954. These dates correspond closely, if a cycle operating over the eight quarters of 1953–54 is assumed, to the phasing of monetary policy in the two cycles of 1961–62 and 1963–64.
The second cycle in Japan’s external balance seems to have begun with a sharp adverse movement in the early part of 1957. Recovery apparently set in, on the basis of over-all payments data, near the end of 1957.15 In any case, full recovery had taken place by the end of 1958. During this cycle, monetary policy was first tightened as early as March 1957, but only slightly and without much effect. Further tightening was resorted to in May 1957. As this measure succeeded in improving the balance of payments, monetary policy was relaxed again from June 1958. Again, considered against the background of the phasing of monetary policy in the 1961–62 and 1963–64 cycles, this pattern of monetary policy is broadly consistent with a cycle of eight quarters covering the years 1957–58.
The study of Japanese balance of payments during 1959–66 has shown that the years 1959–60 and 1965–66 exhibited what may be described as primarily seasonal movements in the basic balance superimposed on fundamentally surplus situations. This seasonal pattern was one in which there was a sharp adverse movement during the first quarter of each year and a steady improvement during the subsequent three quarters. The study has shown also that in both of the cycles 1961–62 and 1963–64, the second half of the cycle (i.e., the four quarters of each of the two years 1962 and 1964) exhibited exactly the same seasonal pattern, though in an exaggerated or enlarged form. Further, it has shown that the first half of each cycle (i.e., the four quarters of each of the two years 1961 and 1963) represented in a sense the failure of the usual seasonal pattern to make itself felt. Thus, in the 1961–62 cycle, the second and third quarters of 1961 failed to show an improvement over the results produced by the sharp adverse movement during the first quarter, while the improvement in the fourth quarter succeeded only in producing a reduced deficit rather than a surplus. In the 1963–64 cycle, the second quarter of 1963 did show an improvement over the first quarter, but there was a worsening in the third and fourth quarters. Seasonally adjusted data broadly confirm the existence of these cyclical and seasonal fluctuations over this period.
In both the cycles described above, the Japanese authorities adopted restrictive monetary policies around the third quarter of the cycle and tightened them further during the fourth and fifth quarters. Recovery from the adverse position of the basic balance began during the sixth quarter of the cycle and was completed by the eighth quarter. In both cycles also, the authorities relaxed monetary policy during the seventh quarter of the cycle, inducing either purely seasonal movements in the next year or the beginning of another cycle, depending on the extent of the relaxation and the strength of the underlying forces of demand and supply.
It is known that there were two cycles in Japan’s balance of payments during the period from 1953 to 1959. On the assumption that the patterns that have operated in the balance of payments since 1959 also prevailed before 1959, and on the basis of annual data and other available evidence, it was found that these two cycles probably occurred over the eight-quarter periods of 1953–54 and 1957–58. Primarily seasonal movements probably occurred during the four quarters of each of the years 1955 and 1956, which intervened between these two cycles. Table 3 lists the various cycles and purely seasonal fluctuations in Japan’s balance of payments that apparently occurred over the entire period from 1953 to 1966.
|From First Quarter of||Through Fourth Quarter of||Number of Quarters||Nature of Fluctuations||Evidence|
|1953||1954||8||First cycle||Annual data and speculation|
|1959||1959||4||Seasonal||Estimated quarterly data|
|1961||1962||8||Third cycle||Quarterly data|
The study of Japanese balance of payments fluctuations has also focused attention on the existence of a remarkable inverse relationship between the basic balance and short-term capital flows. Fluctuations in Japan’s basic balance are accompanied by offsetting fluctuations in short-term capital flows into and out of Japan. This inverse relationship is found to operate from quarter to quarter even on the basis of unadjusted data, though the relationship is improved when adjustment is made in the data for the period from the first quarter of 1961 through the second quarter of 1963 for credits arranged by the Bank of Japan with private U.S. banks for the benefit of private Japanese banks financing external trade transactions. The inverse relationship is explained partly by the fact that the flows of short-term capital depend on the Japanese patterns of financing trade. However, the Japanese authorities have used various policy instruments to influence short-term capital flows with the object of reinforcing the impact of the financing patterns. The combined influence of Japanese patterns of trade financing and the special policy measures used to influence short-term capital flows has been much greater than that of general monetary policy, which has sometimes operated in the opposite direction.
As a result of the inverse relationship between the basic balance and short-term capital flows, fluctuations in the Japanese over-all balance of payments since 1959 have been much smaller than fluctuations in the basic balance. Short-term capital movements have therefore helped to moderate the fluctuations which need official financing. They have moderated rather than accentuated fluctuations in Japan’s balance of payments.
As quarterly data on Japan’s balance of payments for 1959 and 1960 are not available in published form, they have been estimated for this study. Balance of payments data for the first half of 1959 are available in the International Monetary Fund’s Balance of Payments Yearbook, Volume 12, and for the first half of 1960 in Volume 13. The figures for the second half of each of the years 1959 and 1960 were derived by deducting the figures for the first half of each year from the annual figures of each year. The method adopted for dividing the figures of the first half into those for the first and second quarters differs from that adopted for dividing the second-half figures into figures for the third and the fourth quarters. This difference is due to differences in the availability of data for estimation.
First half 1959 and first half 1960
Quarterly figures for exports and imports were first calculated from monthly trade returns available for 1959 and 1960 in the Quarterly Bulletin of Financial Statistics, issued by the Ministry of Finance of Japan. The balance of payments figures for exports and imports during the first half of 1959 and the first half of 1960 were divided into figures for the first and second quarters in proportion to the quarterly values of exports and imports according to trade returns. This procedure assumes that the adjustments to trade returns data for freight, insurance, and other items are distributed between the first and second quarters in the same proportion as the trade returns data. Trade balance figures for each quarter of the first half of 1959 and the first half of 1960 were then calculated in the usual way.
No reliable basis for distributing the semiannual figures for services and transfers, private long-term capital, and official long-term capital was available. The semiannual figures for these groups were therefore divided equally among the two quarters in each half. The basic balance for each quarter was then derived by adding transactions for net services and transfers, private long-term capital, and official long-term capital to the trade balance. The over-all balance for each half was divided among quarters in proportion to changes in official reserves, which are available on a quarterly basis in International Financial Statistics. The private short-term capital flows, including net errors and omissions, were then derived as residuals by deducting the basic balance from the over-all balance of each quarter.
Second half 1959 and second half 1960
Balance of payments data for January–September 1959 and for January-September 1960 are available in the Quarterly Bulletin of Financial Statistics, March 1960 and 1961. These figures of balance of payments, however, are not strictly comparable with the Balance of Payments Yearbook data as regards transactions on capital account. The figures for trade balance, net services and transfers, and private and official long-term capital for the first half of each of the years 1959 and 1960 were deducted from the January-September data available for 1959 and for 1960 to derive the figures for the third quarter for each of the years 1959 and 1960. The fourth-quarter figures were derived by deducting the third-quarter figures from the second-half figures available for 1959 and for 1960. By adding the figures thus derived, the basic balances were calculated for the third and fourth quarters of each of the years 1959 and 1960. It is not possible to derive comparable figures for private short-term capital, including net errors and omissions, from the January-September balance of payments data available in the Quarterly Bulletin of Financial Statistics because changes in the short-term capital of the commercial banks are combined in the Bulletin with those of official institutions.
During the third quarter of 1959, the official reserves showed an increase equivalent to US$167 million. During this period, gold worth US$70 million, whose ownership had been in dispute, was transferred to reserves. The figure for the change in official reserves in this quarter was adjusted by this amount. The figure for the over-all balance of payments for the second half of 1959 was then distributed among the third and fourth quarters in proportion to these adjusted figures of official reserves. The over-all balance figures thus derived were adjusted slightly on the basis of additional information available in International Financial Statistics on changes in payments agreements balances. The over-all balance figure for the second half of 1960 was divided between the third and the fourth quarters in the same proportion as quarterly changes in the published official reserves without further adjustment. The private short-term capital, including net errors and omissions, for the third and the fourth quarters of both 1959 and 1960 were then derived as residuals by deducting the estimated basic balance figures from the estimated over-all balance figures for each of these quarters.
The results of these calculations are presented in Table 4.
|Net services and private transfers||14||14||15||24||−23||−23||−4||3|
|Current account balance||1||59||139||233||−90||—||109||205|
|B.||Private long-term capital||14||14||—||−48||−15||−15||−13||−40|
|C.||Official long-term capital||11||11||−14||−11||4||4||−29||−26|
|D.||Basic balance (A + B + C)||26||84||125||174||−101||−11||67||139|
|E.||Private short-term capital and net errors and omissions||85||44||3||−39||146||102||134||28|
|F.||Over-all balance (D + E)||111||128||128||135||45||91||201||167|
|G.||Increases in official reserves||(113)||(131)||(167)||(112)||(39)||(90)||(207)||(166)|
During 1961 and 1962 the Bank of Japan arranged for short-term loans from certain U.S. banks to finance the import of commodities from the United States. These loans were made available to the Japanese commercial banks involved in this activity and were repaid by them during 1962 and the first half of 1963. Because these amounts were formally borrowed by the Bank of Japan, they are treated in the published statistics as changes in official liabilities. However, since they clearly represent flows of private short-term capital from U.S. banks to Japanese commercial banks via the intermediary of the Bank of Japan, they are treated in this study as part of private short-term capital flows.
No data are available on the exact amounts outstanding under these loans at different times. For this study, it was felt that a reasonably close approximation to the correct position would be obtained if the published data on short-term capital movements were adjusted by the whole of the changes in official liabilities over the period 1961 to mid-1963 (Table 5).
|Private Short-Term Capital|
|Adjusted Private Short-Term Capital|
|Adjusted Over-All Balance|
|Changes in Official Reserves|
Fluctuations de la balance des paiements du Japon et rôle des flux de capitaux à court terme, 1959–1966
Les données trimestrielles relatives à la balance des paiements de base du Japon pour la période 1959–1966 font apparaître deux cycles s’étendant sur les périodes de huit trimestres 1961–62 et 1963–64. Pendant chacune des quatre autres années, des mouvements de nature principalement saisonnière se sont superposés à une situation fondamentalement excédentaire. Ces variations saisonnières ont consisté en une évolution défavorable prononcé du solde de base au cours du premier trimestre, suivi d’une amélioration constante durant les trois derniers trimestres de l’année. Les courants saisonniers normaux ne se sont pas manifestés au cours de la première moitié de chaque cycle, et les déficits ont persisté tout au long des années 1962 et 1964. Dans la deuxième moitié de chaque cycle, le mouvement saisonnier a pris une ampleur exagérée, à mesure que le solde de base s’améliorait de trimestre en trimestre. Les pouvoirs publics ont adopté des mesures monétaires restrictives vers le troisième trimestre de chaque cycle et les ont encore renforcées au cours des quatrième et cinquième trimestres. Le mouvement de reprise a commencé au cours du sixième trimestre de chaque cycle et a pris fin au huitième trimestre. A partir du septième trimestre de chaque cycle, les mesures monétaires restrictives ont été assouplies.
De 1959 à 1966, une relation inverse accentuée s’est maintenue entre le solde de base et les flux de capitaux à court terme, en raison de l’influence combinée de la structure du financement commercial du Japon et des diverses mesures prises pour influencer les flux de capitaux à court terme. En raison de cette relation inverse de caractère remarquable, les fluctuations du solde global ont été beaucoup moins accentuées que celles du solde de base. Les flux de capitaux à court terme ont ainsi joué un rôle stabilisateur dans la balance des paiements du Japon.
Sur la base des données partielles et des autres éléments d’information disponibles, il apparaît dans ce document que les deux cycles qui se sont produits pendant la période 1953–59 se sont probablement étendus sur les périodes de huit trimestres 1953–54 et 1957–59, tandis que des mouvements de nature principalement saisonnière se sont probablement produits au cours des années 1955 et 1956.
Las fluctuaciones de la balanza de pagos del Japón y el cometido de los movimientos de capital a corto plazo, 1959–66
Los datos trimestrales sobre la balanza de pagos básica del Japón en el período 1959 a 1966 muestran dos ciclos que se han producido en los períodos de ocho trimestres 1961–62 y 1963–64. Durante cada uno de los cuatro años restantes ocurrieron movimientos que fueron primordialmente de índole estacional superpuestos a una situación fundamentalmente superavitaria. Esos movimientos estacionales consistieron en una aguda fase adversa de la balanza básica en el primer trimestre, seguida por una continua mejoría durante los tres trimestres subsiguientes de cada año. El patrón normal estacional no surgió durante la primera mitad de cada ciclo y los déficit persistieron a lo largo de los años 1962 y 1964. En la segunda mitad de cada ciclo, el patrón estacional surgió en forma exagerada a medida que la balanza básica mejoraba de trimestre en trimestre. Las autoridades implantaron políticas monetarias restrictivas hacia el tercer trimestre de cada ciclo y las acentuaron más en el cuarto y en el quinto trimestre. La recuperación se inició durante el sexto trimestre de cada ciclo y quedó completada para el octavo trimestre. En cada período las políticas monetarias fueron atenuadas a partir del séptimo trimestre del ciclo.
Durante 1959–66 existió una fuerte correlación inversa entre la balanza básica y las corrientes de capital a corto plazo, debido a la acción combinada de la configuración del financiamiento del comercio japonés y los diversos instrumentos de política empleados para influir sobre las corrientes de capital a corto plazo. Como resultado de esa notable correlación inversa las fluctuaciones han sido mucho menores en la balanza global que en la balanza básica. Por consiguiente, las corrientes de capital a corto plazo han desempeñado un cometido estabilizador en la balanza de pagos del Japón.
Sobre la base de los datos parciales y otros antecedentes disponibles, el trabajo sugiere que los dos ciclos que ocurrieron durante 1953–59 probablemente comprenden los ocho trimestres de los períodos 1953–54 y 1957–59, en tanto que los movimientos primordialmente estacionales es probable que ocurrieran durante los años 1955 y 1956.
Mr. Ezekiel, Chief of the Financial Studies Division of the Research and Statistics Department, is a graduate of the University of Bombay. He taught economics at St. Xavier’s College, Bombay, and at the University of Bombay, and was Financial Editor of The Economic Times. He has contributed a number of articles to economic journals and is the author of The Pattern of Investment and Economic Development (Bombay, 1967).
Mr. Patel, economist in the Balance of Payments Division, is a graduate of the University of Baroda and of the University of Delhi, and did further work at the Institute of Social Studies, The Hague. Before coming to the Fund, he was Acting Government Statistician with the Government of Tanzania.
P.R. Narvekar, ‘The Improvement in Japan’s Balance of Payments, 1954–55,” Staff Papers, Vol. VI (1957–58), pp. 143–69, and “The Cycle in Japan’s Balance of Payments, 1955–58,” Staff Papers, Vol. VIII (1960–61), pp. 380–411.
The small inflow of such capital during 1956 and 1957 was not sufficient to produce any significant difference between the over-all and basic balances of payments on an annual basis (see fn. 3).
The difference between the over-all and basic balance of payments consists of recorded short-term capital movements and “Errors and omissions,” most of which probably represents unrecorded short-term capital flows. For the purposes of this paper, the “Errors and omissions” item is combined with the “Short-term capital” item and treated as private short-term capital flows. Differences between the concepts of over-all and basic balances of payments are explained in detail by Poul Høst-Madsen, in Balance of Payments—Its Meaning and Uses, International Monetary Fund, Pamphlet Series, No. 9 (Washington, 1967).
It is assumed that a large basic surplus occurred in the fourth quarter of 1958. The evidence points generally to full recovery by that time from the heavy deficits recorded during 1957.
The series was processed by the X-11 variant of the Seasonal Adjustment Program of the U.S. Bureau of the Census. It was assumed that the series could be decomposed into additive components.
Bank of Japan, Annual Report, 1959, pp. 2 and 4.
While this relationship is improved by the adjustment made in the data for short-term capital flows during 1961–63, it is strongly evident even in the unadjusted data.
Some trade may be financed by other means than short-term facilities.
“A substantial part, perhaps as much as one half, of the foreign currency deposits is swapped into yen and used for domestic financing” (Oscar L. Altman, “Recent Developments in Foreign Markets for Dollars and Other Currencies,” Staff Papers, Vol. X (1963), p. 71).
Altman (loc. cit.) quotes a comment of the Nederlandsche Handel-Maatschappji [Netherlands Trading Society] in its Annual Report, 1961 (p. 10) on their investment in Japan, as follows, “At a later date we reduced the latter investments in view of a deterioration in the Japanese balance of payments.”
References to the importance of short-term capital flows appear in many of the annual reports of the Bank of Japan. Illustrative of these is the statement in its Annual Report, 1961 (p. 71): “In connection with the balance of payments and domestic tight money conditions, the Bank paid every attention to the inflow and outflow of short-term foreign capital. …”
Bank of Japan, Annual Report, 1963, p. 4.
Ibid., 1964, p. 48.
Narvekar (1957–58), op. cit., pp. 143 and 159.
Narvekar (1960–61), op. cit., pp. 380 and 401.