This Background Paper examines sources of net international reserve inflows for Poland. It shows that once estimate of unrecorded border trade is included, both the current and capital accounts are in surplus. The paper illustrates the methods through which estimates of unrecorded trade can be obtained. It offers a complementary view of Poland’s inflation problem from a medium-term perspective. Specifically, the paper shows that relative price adjustment, combined with indexation and inertia, has been part and parcel of Polish inflation.

Abstract

This Background Paper examines sources of net international reserve inflows for Poland. It shows that once estimate of unrecorded border trade is included, both the current and capital accounts are in surplus. The paper illustrates the methods through which estimates of unrecorded trade can be obtained. It offers a complementary view of Poland’s inflation problem from a medium-term perspective. Specifically, the paper shows that relative price adjustment, combined with indexation and inertia, has been part and parcel of Polish inflation.

I. Sources of Net International Reserve Inflows

1. Introduction

Poland has been experiencing significant surpluses in the balance of payments over the past two years. Excluding valuation effects, net international reserves (NIR) increased by US$2.2 billion in 1994 (2.4 percent of GDP), and are estimated rise by US$7.6 billion in 1995 (6.3 percent of GDP). 1/ In designing the appropriate policy response to inflows of such magnitude, it is critical to identify the nature of the underlying inflows. With the recent structural transformation of Poland’s economy and rapid development of the private sector, the statistical apparatus to record external transactions has become insufficient to appropriately classify all transactions. In particular, the sources of inflows recorded as private transfers--in the current account--and as net purchases of foreign exchange in the capital account--are not identified. These inflows gained importance over the past three years, reaching the equivalent of 4.4 percent of GDP in 1994 and an estimated 6.7 percent of GDP in 1995.

The purpose of this chapter is to attempt to identify the nature of the underlying transactions recorded under private transfers and net purchases of foreign exchange. The presentation of balance of payments data in this chapter will be on a cash basis; it also excludes official transfers from the current account, since the bulk of these transfers are related to debt restructuring operations. As a result, the presentation deviates from the standard balance of payments format reported in the staff report and the statistical tables.

2. Overview

Poland’s official current account before transfers and on a cash basis has been in deficit since 1993 (Table I.1). However, private transfers have been significantly positive through 1994, and net purchases and sales of foreign exchange in the form of banknotes--classified in short-term capital transactions in the official statistics--have skyrocketed over the past two years. The inflows underlying these two balance of payments items consist mainly of transactions involving tourism, border trade, workers’ remittances, and other transfers. Data on the geographical distribution of purchases of foreign exchange in the so-called kantor market (formed by small foreign exchange bureaus) are consistent with the hypothesis that tourism and border trade are dominant components of the inflows.

Table I.1.

Poland: Balance of Payments--Modified Presentation, 1991-95

article image
Sources: Data provided by the Polish authorities; and staff estimates.

Actual through September 1995, and projected for the fourth quarter.

A reasonable starting point for estimating the “true” current account is the assumption that during 1994 all inflows from the kantor market and into foreign exchange accounts reflected current account transactions. This assumption is supported by the relatively stable interest differential between the zloty and foreign currency throughout 1994, the absence of any shifts in the timing of payments for trade transactions, and lingering uncertainty on the commercial bank debt restructuring that was only finalized in late October 1994. For other years, developments in border crossings by individuals can be used to gauge the magnitude of the underlying current account transactions.

Taking into account the estimated current account component included in private transfers and net purchases of foreign exchange, Poland’s underlying current account--rather than being in deficit--registered surpluses of 2.9 percent of GDP in 1994 and 2.7 percent of GDP in 1995 (Table I.1). With a small capital outflow in 1994, current account transactions were responsible for the entire buildup in NIR during 1994 (Chart I.1). In the first half of 1995, current and capital account inflows contributed roughly equally to the NIR build-up. In the second half of 1995, the underlying current account--while still in surplus--is expected to contribute significantly less than the capital account. The increased contribution from the capital account stems mainly from an increase in foreign direct investment, the new phenomenon of inward portfolio flows that is estimated to reach the equivalent of 0.9 percent of GDP, and other short-term inflows, part of which remains unexplained.

CHART I.1
CHART I.1

POLAND: ADJUSTED CURRENT AND CAPITAL ACCOUNTS, 1991–95 1/

(In percent GDP)

Citation: IMF Staff Country Reports 1996, 019; 10.5089/9781451831788.002.A001

Sources: Polish authorities and staff estimates.1/ Second half 1995 is a projection.

The results of the above-mentioned aggregate approach are broadly in line with independent information from surveys of border trade and tourism. These surveys--which are only available for 1994 (on an annual basis) and the first quarter of 1995--are not comprehensive, but can explain about three quarters of the inflows in 1994 and in the first quarter of 1995. If one were to take evidence only from these surveys and assume that the residual consists of capital account transactions, most of the build-up in international reserves in 1994 and in the first quarter of 1995 would still have resulted from current account inflows. However, all these estimates ignore the role of leads and lags in trade payments.

During the first half of 1995, the officially recorded trade deficit may have been somewhat understated because of a possible advance in the timing of export receipts in the first half of 1995 in response to the widely anticipated appreciation of the zloty. To the extent that this can be seen as a disguised capital account transaction, it reinforces the conclusion that capital inflows have recently become more important in explaining the build-up in reserves. There was no evidence of delays in import payments.

3. The adjusted current account

Poland’s official balance of payments is recorded on a settlements basis using information from the banking system. Transactions without payments equivalent, except for arrears on debt service, are not recorded. Since the purpose of this chapter is mainly to analyze the sources of the increase in NIR, several adjustments are made to the official presentation in order to derive the underlying current account: (i) only interest paid is included as the bulk of the interest arrears were subsequently rescheduled or canceled; (ii) official transfers are moved to the capital account as they consist almost exclusively of transfers related to debt restructuring; and (iii) private transfers--which are measured as the change in the balance of the foreign exchange accounts of private citizens--and net purchases of foreign exchange in the kantor market were lumped together both because of the fungibility of these transactions and to remove currency substitution from the balance of payments. The key issue is to estimate the current account component included in the sum of these two categories of transactions.

a. Official trade in goods and services

Trade in goods and services is recorded accurately in the official current account on a settlement basis and corresponds with information from partner countries. After registering small surpluses, Poland’s official trade and services account swung into a sizable deficit in 1993, equivalent to 2.7 percent of GDP (upper panel of Chart I.2). It improved to a deficit of 1.6 percent in 1994, and is estimated to reach a deficit of 2 percent of GDP in 1995 reflecting the gradual real appreciation of the zloty and further import liberalization.

CHART I.2
CHART I.2

POLAND: OFFICIAL TRADE, 1991–95

Citation: IMF Staff Country Reports 1996, 019; 10.5089/9781451831788.002.A001

Sources: Polish authorities, Direction of Trade Statistics and staff estimates.

Even though there is a full export surrender requirement, the timing of trade payments may be shifted to take advantage of expected changes in the relative returns between domestic and foreign currencies, imparting a bias on the estimate of the underlying official trade balance. Comparing partner country customs data with Polish settlements data shows that there have been episodes of shifts in the timing of payments (lower panel of Chart I.2), but that in most cases such shifts were short lived. With the exception of 1992 (for exports) and the period around the exchange rate adjustment in August 1993 (mainly for imports), settlements data have generally been below customs data as the latter include trade without payments equivalent or without payment through the Polish banking system. The difference is more pronounced on the import side, owing to foreign direct investment in kind and reprocessing activities. During 1994 no shifts in timing in either exports or imports could be observed, but during the first half of 1995 export receipts seem to exceed what could be expected on the basis of customs data, perhaps in anticipation of the exchange rate appreciation that took place in May 1995. The official trade balance for the first half of 1995, measured on a settlements basis, may have overstated the underlying trade performance by some US$0.6 billion, or 1 percent of GDP at an annual rate.

b. Private transfers and net purchases of foreign exchange

Flows recorded under private transfers and net purchases of foreign exchange are best treated as a single item because of the interaction between the kantor market--the source of net purchases of foreign exchange--and foreign exchange accounts of residents--which are used to estimate the bulk of private transfers. There is no requirement to declare the source of foreign exchange deposited in foreign exchange accounts or sold to the kantors. Like private citizens, kantors may hold foreign exchange accounts. Hence, for example, a sale to kantors of foreign exchange withdrawn from an account could be registered as an outflow in the current account and an inflow in the capital account, and vice versa. As there is considerable turnover in private foreign exchange accounts, such transactions could be widespread (Chart I.3). Moreover, during much of 1995, when expectations of zloty appreciation prevailed, private residents were no longer building up foreign exchange balances but instead preferred to sell the underlying inflows. Considering private transfers and net purchases of foreign exchange together eliminates spurious transactions between residents and facilitates historical comparisons.

CHART I.3
CHART I.3

POLAND: TURNOVER IN PRIVATE FOREIGN EXCHANGE ACCOUNTS, 1991–95 1/

Citation: IMF Staff Country Reports 1996, 019; 10.5089/9781451831788.002.A001

Source: Polish authorities.1/ In millions of US dollars.

The sum of the these two categories of transactions moved from a small deficit in 1991 to surpluses of US$1.4 billion in 1992, US$2.6 billion in 1993, and US$4.2 billion in 1994 (equivalent to 4.5 percent of GDP). In 1995, total net inflows are expected to reach US$7.3 billion, equivalent to 6 percent of GDP.

c. Estimating the underlying current account component

In the absence of a declaration requirement on the source of the foreign exchange, it is difficult to determine directly what share of these transactions should be classified in the current or capital account. With all of these inflows exclusively in the form of banknotes and with kantors prohibited to purchase foreign exchange from the banking system or the interbank market through mid-May 1995, it would seem implausible and unwieldy to conduct capital account transactions on a significant scale through this market. However, it is possible that not all of the proceeds from border trade and tourism were immediately repatriated or deposited in the banking system.

The geographical distribution of foreign exchange purchases shows that a large portion of purchases of foreign exchange stems from border areas and other regions attracting tourism (Table I.2). The Warsaw region--which has always registered the largest share of the purchases of net foreign exchange--appears to stand out as an exception. However, it should be borne in mind that Warsaw is the largest business and financial center, harbors a large outdoor market attracting shoppers by bus from Belarus, and is a prime location for small-scale industries producing mainly for border trade.

Table I.2.

Poland: Purchases of Foreign Exchange, 1994-95: Geographical Composition

(In millions of U.S. dollars)

article image
Source: National Bank of Poland.

Less than 4 percent of total, precise data not available.

Looking at macroeconomic conditions, it can be argued that prior to 1994, general uncertainty as regards the value of the zloty and the stability of the banking system may have discouraged the repatriation and conversion into domestic currency by residents of proceeds from border trade, worker remittances, tourism, and other private transfers. During 1994, however, the relative rate of return between zloty and foreign currency denominated assets was fairly stable, while at the same time general economic conditions improved and uncertainty was reduced (but not quite eliminated). Under these circumstances, it is plausible that inflows into foreign exchange accounts and through the kantor market during 1994 consisted of current account transactions only. Since the beginning of 1995, however, expectations of an appreciation of the zloty likely induced some capital inflows, additional repatriation, or conversion of foreign exchange held in cash through these channels.

For years other than 1994 some independent indicator needs to be used to estimate the underlying current account inflows. Since inflows into foreign exchange accounts and through the kantor market are likely to be linked to border trade and tourism, as indicated by independent surveys (see below), developments in the number of border crossings could be used as such an indicator (Chart I.4). Since 1991, the balance of border crossings started to show a surplus of visits by foreigners, estimated to reach 50 million person-crossings in 1995.

CHART I.4
CHART I.4

POLAND: BALANCE OF BORDER CROSSINGS, 1990–95 1/

(Millions of person-crossings)

Citation: IMF Staff Country Reports 1996, 019; 10.5089/9781451831788.002.A001

Sources: Polish authorities and staff estimates.1/ Number of visitors from abroad minus number of Polish residents traveling abroad.

This methodology leads to estimates of implied current inflows that increase gradually from US$1.6 billion in 1991 (equivalent to 2 percent of GDP) to US$5.7 billion in 1995 (equivalent to 4.8 percent of GDP). Adding these inflows to the official current account (before transfers) shows that the underlying current account has been in surplus throughout 1991-95. Owing to a deterioration in the official current account, the adjusted surplus fell to less than 1 percent of GDP in 1993 but rebounded to nearly 3 percent of GDP in 1994 and is estimated to fall slightly to 2.7 percent of GDP in 1995. Residually these estimates imply that there were considerable unrecorded capital outflows in 1991-1992. This may have taken the form of accumulation of assets abroad or the build-up of foreign exchange in cash by Polish residents.

In response to growing Inflows in items for which there is no declaration requirement, the Polish authorities have started conducting surveys of border trade and border tourism. On the basis of these surveys--which are available on an annual basis for 1994 and quarterly for the first quarter of 1995 only--and using some additional assumptions on spending by Polish travelers abroad and business expenses in cash by foreigners visiting Poland, it is possible to derive an alternative estimate of the underlying current account inflows. Plausible assumptions can attribute up to US$3 billion out of US$4.2 billion to current account transactions for 1994 (compared to the aggregate assumption that all of the US$4.2 billion consists of current inflows). Similarly, for 1995, these transactions could amount to US$5.2 billion out of an estimated US$7.1 billion (compared to the US$5.7 billion as derived with the methodology based on the aggregate assumption; Table I.3). These estimates should be treated with caution as they are sensitive to some key assumptions and the surveys are not comprehensive. In addition, there is no plausible interpretation for the unexplained residual, especially in 1994.

Table I.3.

Poland: Survey-Based Estimation of Underlying Current Account Inflow

article image
Sources: Data provided by the Polish authorities; and staff estimates.

4. Capital account

Developments in the capital account during 1991-94 were marked by external debt restructuring and general political and macroeconomic uncertainties. Medium- and long-term inflows, foreign direct investment, and inward portfolio inflows were all relatively insignificant. Officially recorded short-term capital, including errors and omissions but excluding purchases of foreign exchange from the kantor market, registered significant deficits during 1991-93 and a small surplus in 1994. Taking into account the underlying current account inflows as estimated in the previous sectors, implied capital outflows for 1991-92 were larger than the officially recorded flows, while in 1993 they were roughly similar. In sum, the adjusted capital account is estimated to have made a strong negative contribution to the build-up in net international reserves in 1991 (equivalent to 3.9 percent of GDP), while its contribution was negative by 1 percent of GDP in 1992 and by about 0.5 percent of GDP per year in 1993-94 (Table I.1). With strengthening foreign direct investment, large portfolio inflows, and sharply increasing short-term inflows (part of which residually estimated from the derivation of the underlying current account), the contribution of the capital account has been increasing significantly throughout 1995. Under the aggregate approach used above, the capital account is estimated to contribute more than half of the build-up in NIR in 1995, registering a surplus of about 3.5 percent of GDP. With the alternative estimate based on surveys and taking into account possible effects of the shift of timing of exports, this contribution could be as high as 4.9 percent of GDP out of the total buildup of 6.3 percent of GDP.

The medium- and long-term capital account, including all operations related to debt restructuring, was roughly in balance during 1991-93 but registered a large deficit in 1994, taking into account the cost of the London club debt operation. Since the beginning of 1995, medium- and long-term capital inflows have strengthened as Poland regained access to international capital markets with a June bond issue of US$250 million. As a result, medium- and long-term inflows are estimated to contribute to the build up in NIR by US$0.5 billion during 1995.

Foreign direct investment in cash increased gradually from a mere US$0.1 billion in 1991 to over US$0.5 billion in 1993-94 and is expected to increase to about US$0.9 billion in 1995, following renewed efforts at privatization involving foreign investors. More recently, foreign investors have become interested in the equity and treasury bill markets as the downward risk on the zloty virtually disappeared since the beginning of 1995 and the yields on treasury bills looked attractive. After nine months, cumulative net portfolio inflows amounted to US$0.9 billion, and by year-end these flows are expected to total US$1.1 billion.

The remaining contribution from the capital account stems from short-term inflows that are residually estimated on the basis of the methodology used to estimate the underlying current account inflows. These estimates indicate relatively large implied outflows for 1991-92 (US$2.5 billion) and sizable unexplained inflows for 1995 (almost US$2 billion), while for the period in between officially recorded and implied inflows were similar. It is difficult to explain the source of the implied capital inflows in 1995, but there is some anecdotal evidence that savings in foreign exchange in the form of cash held by Polish residents were being converted into zloty to take advantage of the expected zloty appreciation. These savings could have been accumulated during 1991-93. While this is a hypothetical explanation, it is consistent with the observed increase in local currency in circulation, although not to the full extent of the implied inflows.

1/

The underlying increase in NIR in 1994 was even larger, as commercial bank debt restructuring imposed an exceptional net cost of about US$1.4 billion.

Republic of Poland: Background Paper
Author: International Monetary Fund