Czech Republic: Selected Issues
Author:
International Monetary Fund
Search for other papers by International Monetary Fund in
Current site
Google Scholar
Close

This Selected Issues paper examines recent developments in inflation and its key determinants in the Czech Republic, with particular focus on the role of wages. A simple analytical framework is presented that relates inflation to wages, import prices, and money, and the interaction of inflation with these variables is then examined empirically in the context of a vector autoregression model. The findings confirm the critical influence of wages, exchange rate changes, and money growth. The paper also analyzes developments in the public finance.

Abstract

This Selected Issues paper examines recent developments in inflation and its key determinants in the Czech Republic, with particular focus on the role of wages. A simple analytical framework is presented that relates inflation to wages, import prices, and money, and the interaction of inflation with these variables is then examined empirically in the context of a vector autoregression model. The findings confirm the critical influence of wages, exchange rate changes, and money growth. The paper also analyzes developments in the public finance.

VI. External Sector Developments85

136. Economic developments since the mid-1990s in the Czech Republic have been characterized by a rapid widening of the current account imbalance through early 1997, which eventually led to mounting uncertainty in the foreign exchange market and a foreign currency crisis in May 1997. Specifically, the current account deficit grew from 2 percent of GDP (US$0.7 billion) in 1994 to 7½ percent of GDP in 1996 (US$4.3 billion), and to 8½ percent of GDP in the first quarter of 1997. This outcome reflects both an increase in domestic investment and a decline in savings, and is partly attributable to huge capital inflows, which peaked at 16 percent of GDP in 1995. These inflows financed fixed investment and fueled increases in wages—and hence consumer spending.

137. Following the currency crisis, the current account deficit narrowed to about 5 percent of GDP in 1997 H2, as the combined effects of a tightening in financial policies dampened domestic demand and import growth, and a sharp recovery in exports—due to improved demand conditions abroad, earlier foreign direct investments coming on stream, and a somewhat depreciated exchange rate—took hold. Notwithstanding the loss of almost US$1½ billion in official reserves in 1996 and US$3 billion in 1997, reserve cover is still prudent at 3½ months of current account payments and about 1⅓ times the level of short-term external debt. Debt service remains relatively low. The ratio of gross external debt to GDP has risen rapidly to reach 40 percent of GDP—the level commonly applied as a prudential threshold. Overall, however, the Czech Republic remains a small net external creditor, when direct investment is excluded, which together with a floating exchange rate and a narrowing external deficit should help reduce its vulnerability to shocks.

A. External Developments Since 1993

138. This section traces the evolution of the current account, with a view to understanding its key determinants and assessing its sustainability, through developments in (a) its main components (exports and imports of goods and nonfactor services and net factor income), (b) its saving-investment counterpart, and (c) the composition of its financing.

Trade

139. The deterioration in the current account has closely mirrored that of the trade balance (Table 19; Figure 31). During 1993-96, the share of exports in GDP declined by almost 3 percentage points, to 38½ percent in 1996, while that of imports increased by more than 6 percentage points to 49 percent of GDP. In 1997, the export and import ratios are estimated to have risen by 4 and 2 percentage points, respectively (the increase reflects in part the valuation effects of the koruna depreciation).

Table 19.

Czech Republic: Balance of Payments, 1993-97 1/

(In percent of GDP)

article image
Sources: Data provided by the Czech National Bank; and staff estimates.

Includes transactions in convertible and nonconvertible currencies, and transactions with Slovakia; based on new customs methodology.

The transfer to Slovak citizens of shares in enterprises privatized in the first wave of voucher privatization and the corresponding offsetting portfolio investment have been netted out.

FIGURE 31
FIGURE 31

CZECH REPUBLIC: CURRENT ACCOUNT AND ITS COMPONENTS, (1993-97)

Billions of U.S. Dollars

Citation: IMF Staff Country Reports 1998, 036; 10.5089/9781451810011.002.A006

Sources: Czech National Bank and Fund Staff estimates.

140. Exports have broadly tracked developments in domestic and foreign demand and in external competitiveness, while some of the volatility in the series may be attributed to special supply factors (Table 20; Figure 32).86

Table 20.

Czech Republic: Trade Volumes and Demand, 1993-97

(In percent change, in real terms)

article image
Source: Czech Statistical Office, WEO, and Fund staff calculations.

Foreign demand of partner countries of the Czech Republic, excluding Slovakia.

Excluding oil.

Growth in the sum of domestic demand plus real exports.

Full-year projection.

Figure 32
Figure 32

CZECH REPUBLIC: MAJOR DETERMINANTS OF EXPORT VOLUMES

Citation: IMF Staff Country Reports 1998, 036; 10.5089/9781451810011.002.A006

Sources: Czech Statistical Office, WEO, and Fund staff estimates.1/ 1997ql through 1997q3.
  • Export volume growth during 1993-95 averaged 11 percent per annum, and was almost double that of foreign demand. This reflected in part the impact of the “cushion” of price competitiveness established by the nominal depreciation of the koruna in 1990-91, which helped exporters to successfully divert goods from traditional (Eastern European, CMEA) to new (Western European, EU) markets (Table 21).87 These factors dominated the negative influence of a rapid acceleration in domestic demand (from about 1 percent in 1993 to 8-9 percent in 1994-95)88 that added to a build-up of pressure on domestic resources.

Table 21.

Czech Republic: Export Market Shares, 1993-96

(Share, in percent of total)

article image
Source: Czech Statistical Office, IMF Direction of Trade Statistics, and Fund staff calculations.
  • Exports stagnated in 1996. This is probably explained by a combination of factors. With domestic demand continuing to rise much faster than real GDP for a third consecutive year, tradables—especially in traditional/basic goods sectors—were probably diverted to the home market to an even greater extent.89 Meanwhile, foreign demand growth—dominated by Germany, which alone accounts for more than one-third of Czech exports—decelerated by 3 percentage points, to 6 percent in 1996.90 The growth of foreign investment demand—a more pertinent variable given the commodity composition of Czech exports—decelerated even faster (Table 22 and Table A38, SM/98/30, Sup. 1, (1/30/98)). It also seems likely that by 1996, external competitiveness had been significantly eroded: the ULC-based REER had appreciated by another 4½ percent in 1996, which probably removed the remaining “cushion” of competitiveness (Figure 34; see Box 2 for an discussion of competitiveness).

  • Export volume growth reached 10.3 percent during Q1-Q3 1997, although the turnaround only began in April 1997. The recovery in Germany and the sharp deceleration in domestic demand in the Czech Republic were important, and other (supply) factors also were at play. However, the exchange rate depreciation was not likely to have played an important role, in part owing to relatively long lags which are normally associated with the export supply response function.91 New foreign direct investments and joint ventures coming on stream further contributed to the “jump” in exports in 1997.92 The revealed comparative advantage of the Czech Republic is based on its low level of local wages, a skilled industrial labor force and its location at the center of Europe.93 Sectors that benefited from foreign investment include road vehicles, electrical machinery, and chemicals (Table A40, SM/98/30, Sup.1).94 In the case of road vehicles, the launch of a new Škoda-Volkswagen sedan model helped enhance volume growth of road vehicles (SITC 78);95 other notable investments coming on stream included investments by Siemens and Matsushita. It should be noted however that the Czech value added of many of these “new” exports is low since they rely more heavily on the use of imported inputs as local firms become more integrated with suppliers abroad.

Table 22.

Czech Republic: Export Composition, 1993-97

article image
Source: Data provided by the Czech Statistical Office and Fund staff estimates.

Based on merchandise exports denominated in koruny.

Figure 33
Figure 33

CZECH REPUBLIC: TRADE BALANCE, TRADE VOLUME, AND TERMS OF TRADE

Citation: IMF Staff Country Reports 1998, 036; 10.5089/9781451810011.002.A006

Sources: Czech Statistical Office, Czech National Bank and Fund staff estimates.
Figure 34
Figure 34

CZECH REPUBLIC: EXCHANGE RATE INDICATORS

Citation: IMF Staff Country Reports 1998, 036; 10.5089/9781451810011.002.A006

Sources: Czech authorities; IMF, International Financial Statistics; and Fund staff estimates.1/ An increase denotes real appreciation.

External Competitiveness of the Czech Industry

During the period 1991-May 1997, when the koruna was pegged to a currency basket,1 external competitiveness and enterprise profitability were gradually eroded as rapid increases in real wages fed through into rising prices and unit labor costs, which coupled with nominal appreciation pressures during 1996 and 1997 Q1, caused the real exchange rate to appreciate further (Figure 34; Table A.44). The koruna was pegged to a basket in 1991, after the nominal effective exchange rate had depreciated by about 80 percent: The initial overshooting helped to establish a “cushion” for competitiveness. During the period of the peg the real effective exchange rate (REER) had steadily appreciated by between 55 percent (based on consumer prices, CPI) and 60 percent (based on unit labor costs, ULC), and the nominal effective exchange rate (NEER) by about 3 percent. During May- September 1997, the NEER depreciated by 3¾ percent, and the ULC-based REER depreciated by about 12 percent, restoring external competitiveness to the level of 1995. It may be noted, however, that such exchange rate based measures may not accurately reflect developments in competitiveness beyond broad trends in light of statistical data deficiencies (see in particular Chapter II on the issue of likely overestimation of productivity growth in industry).

1 From 1991 until the adoption of a managed float on May 27, 1997, the central rate was pegged to a hard currency basket which (from 1993) comprised the Deutsche mark (65 percent) and the U.S. dollar (35 percent). At end-February 1996, the CNB widened the exchange rate band from ±½ percent to ±7½ percent around the central rate.

Anecdotal evidence suggests that stocks were run down in less advanced sectors in 1997. One-off exports, including the sale of decommissioned aircraft, also helped boost export revenues.

141. The evolution of imports has been closely related to developments in total (domestic and export) demand as well as the real effective exchange rate as, until recently, consumers found imports increasingly affordable (Table 20; Figure 35).96

Figure 35
Figure 35

CZECH REPUBLIC: MAJOR DETERMINANTS OF IMPORT VOLUMES

Citation: IMF Staff Country Reports 1998, 036; 10.5089/9781451810011.002.A006

Sources: Czech Statistical Office, WEO, and Fund staff estimates.1/ 1997ql through 1997q3.
  • Import volume growth rose steadily from 13 percent in 1993 to 27 percent in 1995, which was well in excess of growth in domestic or total demand, suggesting that the real currency appreciation made foreign products increasingly attractive and affordable to Czech consumers and investors. Over this period, the composition of imports (in terms of consumer, intermediate and investment goods) changed very little (Table 23).

  • Import volume growth slowed markedly in 1996, to 11 percent, but remained relatively high in relation to domestic and, especially, total demand given stagnant exports. Import volume growth decelerated much less than domestic demand in 1997, apparently reflecting increased demand for imported inputs for exportables.97

Table 23.

Czech Republic: Composition of Imports, 1995-97

article image
Sources: Czech National Bank, and Fund staff calculations.

Services, Factor Income and Transfers

142. Traditionally, the services balance has been in surplus, helping to offset the merchandise trade deficit (Figures 31 and 36). Total services receipts remained fairly stable at around 14 percent of GDP during 1993-97. Travel receipts had shown a steady increase between 1993 and 1996 (rising by about 3 percentage points of GDP, to 7½ percent), as visitor numbers grew and the sector expanded. Tourism fared less well in 1997 as visitor numbers stabilized—partly because of emerging capacity constraints. Services payments showed similar patterns, with total payments broadly constant at about 11 percent of GDP, while travel payments increased by almost 4 percentage points between 1993 and 1996, to 5½ percent of GDP, aided by real wage gains and the koruna appreciation.98 Travel payments declined in 1997 as the economy slowed and uncertainty increased (following the exchange rate crisis), while a series of bankruptcies among travel agents and the floods also depressed foreign travel.

FIGURE 36
FIGURE 36

CZECH REPUBLIC: TRADE IN NONFACTOR SERVICES

(Billions of U.S. Dollars)

Citation: IMF Staff Country Reports 1998, 036; 10.5089/9781451810011.002.A006

Source: Czech National Bank

143. As the Czech Republic has become more reliant on debt financing and joint ventures have come on stream, the factor income balance has grown more negative (Table 24). Almost one-third of the deterioration of the current account in 1996 was attributed to the deterioration in factor income and unrequited transfers.99 Indeed, 1996 was the first year that the Czech Republic incurred a deficit on profits and other remittances, as joint ventures and foreign workers repatriated more profits/income abroad. Interest receipts stabilized, while payments increased as external debt obligations rose (Table 25).100

Table 24.

Czech Republic: Income and Transfers Balances, 1993-97

(In millions of U.S. dollars)

article image
Sources: Data provided by the Czech National Bank, and Fund staff estimates.
Table 25.

Czech Republic: Foreign Assets and Liabilities, 1993-97

article image
Source: Data provided by the Czech National Bank, and Fund staff estimates.

Savings and Investment

144. The deterioration in the current account reflects the relative strength of fixed investment and a concurrent weakening of domestic savings. Between 1993 and 1996, gross domestic investment as a ratio to GDP rose by 5 percentage points to 33 percent, while the gross national savings ratio fell by 4 percentage points to 25½ percent of GDP (Table 26; Figure 37). Preliminary estimates for 1997 indicate decreases in savings of 1½ percentage points, and investment, of 3 percentage points.101

Table 26.

Czech Republic: Savings and Investment Balances, 1993-97

article image
Sources: Data provided by Czech Statistical Office; and Fund staff estimates.

External current account deficit (+). A small discrepancy between the national accounts and the balance of payments occurred in 1995.

Equal to gross domestic investment (excluding statistical discrepancy) minus foreign savings.

Equal to gross national savings minus net factor income and transfers from abroad.

Figure 37
Figure 37

CZECH REPUBLIC: CURRENT ACCOUNT AND ITS MAJOR DETERMINANTS

Citation: IMF Staff Country Reports 1998, 036; 10.5089/9781451810011.002.A006

Sources: Czech Statistical Office, National accounts, and Fund staff estimates.1/ Gross national saving, including statistical discrepency.2/ Through Sept. 1997.
  • The rising investment ratio largely reflected transition-related infrastructural and environmental investment. Fixed investment in manufacturing was somewhat sluggish and its share in total investment fell from 27 percent (1992-94 average) to 21½ percent in 1995. Of the 5 percentage point increase in overall investment to GDP ratio, only 1 percentage point was attributable to government investment (which had been stable at 4.7 percent in every year between 1994 and 1997), and the remainder resulted from the steady increase in private investment. The budget cuts announced in April-May 1997 are expected to lower government investment by 1¼ percentage points to 3¼ percent of GDP, while tight monetary policy accompanied by the slowdown in activity is expected to lower private investment by 1½ percentage points to 27 percent.102

  • Developments in national savings were largely driven by the private sector; the government traditionally ran a conservative fiscal policy, maintaining government saving at 3-3½ percent of GDP during 1994-96. The sharp increases in real wages during 1994-96 gradually eroded enterprise savings and fueled consumer spending on the expectation that these gains would sustained.103 Estimates for 1997 indicate that government savings fell to 1¼ percent of GDP, consistent with the slowdown in the economy, while private savings increased to 23.7 percent of GDP, below the peak of 25.2 percent, reached in 1993 and 1995.

External Financing

145. The level and composition of net capital inflows has changed significantly since 1993 (Figure 38). Strong confidence in the value of the currency, together with a relatively large domestic-foreign interest rate differential, attracted increasing (debt and nondebt) capital inflows, amounting to 16 percent of GDP in 1995 alone. The subsequent decline in the attractiveness of the koruna—owing both to the large current account deficit and the increased flexibility of the koruna (following the adoption of wider margins in February 1996 and of a managed float in May 1997)—contributed to a large decline in net capital inflows to 6 percent of GDP in 1996 and even less in 1997. Since 1996, the financing of the current account has necessitated a drawdown in official reserves (Figure 12, in Chapter IV), albeit from high levels.

Figure 38
Figure 38

CZECH REPUBLIC: CAPITAL FLOWS (1993-97)

(Billions of U.S. Dollars)

Citation: IMF Staff Country Reports 1998, 036; 10.5089/9781451810011.002.A006

Sources: Czech National Bank and Fund Staff estimates.
  • Net nondebt capital inflows have averaged 3-3½ percent of GDP per year since 1993, barring the peak of 6¾ percent in 1995 (Table 27). The greater part of these inflows is accounted for by foreign direct investment, which rose from 1½-2 percent of GDP in 1993-94 to 5 percent in 1995—when a large stake in SPT Telecom was sold to foreign investors,104 before being halved to 2½ percent in 1996. FDI is estimated to have fallen further in 1997, to about 2 percent. Portfolio equity inflows oscillated from year to year, also reflecting progress in privatization and, more recently, declining foreign investor confidence. Equity inflows reached about 1¾ percent of GDP in 1993 and 1995, and about 1 percent in 1994, 1996, and 1997.

Table 27.

Czech Republic: Capital Account, 1993-97

article image
Source: Czech National Bank.
  • While the exchange rate was fixed, and the interest differential with abroad was rather high, borrowers had an incentive to raise funds from abroad (Figure 19, in Chapter IV). Inflows of medium- and long-term (MLT) capital also peaked in 1995, rising from around 2½ percent of GDP in 1993-94 to 6¾ percent.105 A leveling in net MLT inflows occurred in 1996, although this represented a fall of 1¼ percentage points of GDP. Data on the stock of external debt indicate that the increase in borrowing came from commercial banks and corporations, while the government was reducing its external debt. In order to better gauge the nature of the borrower and the maturity of the loans, illustrative data on loan commitments from abroad confirm that in 1995-96 the banks were the largest borrowers from abroad (most often at maturities of 1-5 years), followed by the utilities/local authorities/private finance companies (Table 28).106 In the new exchange rate environment post-May 1997, and with interest rate spreads increasing across emerging markets, MLT borrowing was more than halved from 1996 levels—to 2¼ percent of GDP (1997 Q1-Q3)—notwithstanding continued activity in the offshore market (see Box 3).

  • Since 1996 there has been a net outflow of short-term capital, prompted by policy measures to discourage short-term capital inflows and the widening of the exchange rate band in February 1996. The short-term net capital outflow in 1996 amounted to 1½ percent of GDP (approximately US$1 billion). The outflow continued at a slightly higher level in 1997 (2 percent of GDP) following turbulence in the foreign exchange market.107

  • Capital inflows were insufficient to cover the current account deficit in 1996 and Q1-Q3 1997—in part because the CNB chose not to intervene in the foreign exchange market to prevent an appreciation of the koruna through early 1997—and official reserves were drawn down by 1½ percentage points of GDP in each period. By end-September 1997 reserve cover had fallen to about 5 months of merchandize imports (when official reserves were close to US$11 billion), from a peak of 6.7 months as end-1995 (when official reserves almost US$14 billion). By end-December 1997, preliminary official reserves figures indicated a further decline to US$9¾ billion, or 4½ months of estimated imports.

Table 28.

Czech Republic: Loan Commitments from Abroad

(In millions of U.S. dollars)

article image
Source: Capital Data, Czech National Bank.

Includes local authorities’ public corporations, utilities, and private finance companies.

As reported in the balance of payments.

Eurokoruna Bonds

A growing offshore market for Eurokoruna bonds has facilitated bank and enterprise borrowing, as issuers on-lent koruna receipts in the Czech Republic directly or London swap houses placed koruna receipts in Czech bonds earning a higher rate of interest. From the inception of the Eurokoruna bond market in September 1995 to end-1997, about US$4 billion was raised by issuers in the offshore market. The largest volume of issues was recorded in Q1 1997, and reached US$1.7 billion.1 Issuers have primarily been German commercial banks, which have on-lent the koruny through subsidiaries in the Czech Republic (allowing Czech entities to refinance their external liabilities in domestic currency), and supranational and sovereign issuers, which have used their strong credit rating to swap koruna receipts into other currencies at lower cost. London swap houses, which have entered as the counter- parties to such transactions have, reportedly, often placed the koruna receipts in Czech bonds in order to hedge their obligations. Both types of transactions should facilitate borrowing from abroad, and although they may not be fully reflected in the balance of payments, they do impact on the spot exchange rate.

1 By comparison, at end-September 1997 the total issue to date of Euroforints and Eurozlotys had reached US$363 million and US$614 million, respectively.

146. The Czech Republic’s positive net foreign asset position has gradually eroded, declining from US$8 billion at end-1993 (23 percent of GDP) to US$4.4 billion at end-September 1997 (8½ percent of GDP). Meanwhile, the stock of debt (in convertible currencies) rose from 24.7 percent of GDP (US$8½ billion) to 40.3 percent of GDP (US$21 billion) between end-1993 and end-September 1997; the maturity structure remained broadly unchanged, with the share of short-term debt remaining at around 28 percent.

B. Vulnerability of the External Position

147. The events leading up to the May 1997 currency crisis reveal the extent to which the Czech Republic’s external position had become vulnerable to a change in investor confidence, as the market judged that the prevailing exchange rate was not likely to be sustainable, in the light of deteriorating fundamentals (a large current account deficit combined with weak output and export performance) and perceived political constraints to a further tightening of policies. Confidence in the exchange rate was shaken when the Czech National Bank came under pressure to loosen monetary policy as activity slowed, while the government’s first package of fiscal and structural measures (in mid-April) was considered to be inadequate by the markets—to the extent that a widening fiscal deficit was not consistent with a narrowing in the current account deficit and the preservation of the exchange rate peg. Contagion from Southeast Asia helped trigger the speculative attack: Investors perceived that there were certain parallels with Thailand, where the crisis was preceded by a period of political uncertainty and differences between the government and the central bank over the conduct of policies were aired in public. However, in contrast with the experiences in Mexico in 1994-95 and in Southeast Asia in 1997, the crisis in the Czech Republic did not appear to have any elements of a banking or debt crisis: There was no evidence of diminished public confidence in the banking system such as deposit withdrawals (other than a limited shift from koruna to foreign-currency-denominated deposits), and the Czech Republic’s investment grade sovereign credit rating was confirmed soon after the crisis by the CNB’s ability to borrow from abroad on favorable terms.

148. Indicators of external vulnerability are in some respects better for the Czech Republic than they were for Mexico and countries in Southeast Asia before they experienced serious financial difficulties, and broadly comparable to those for neighboring transition economies (Table 29).108 The current account deficit of the Czech Republic was still unsustainably large in 1997, at 6-6½ percent of GDP, though it is targeted to decline over the medium term. According to a simple rule of thumb, the maximum sustainable current account deficit of the Czech Republic is about 2½—3 percent of GDP (on the assumption that the ratio of external debt to GDP does not exceed 40 percent, and that annual nominal dollar GDP rises by 6-8 percent).109 However, the maximum could be somewhat higher to the extent that the Czech Republic can rely on nondebt capital inflows (likely to amount to about 1½—2 percent of GDP over the medium term). The level of investment in the Czech Republic is high relative to comparator transition economies (Hungary and Poland) as well as Mexico in 1993. But as the recent Southeast Asian experience shows, where investment rates exceeding 40 percent are common, this is not necessarily a reliable indicator of sustainability. The distinction between public and private savings postulated by the “Lawson doctrine” (which considers private sector dissaving to be “benign” since disequilibria, in an efficient market, will be corrected through changes in interest rates) proved to be of limited predictive value, given that the private sector was primarily responsible for the observed dissaving in the Czech Republic—as it had been in Mexico.

Table 29.

Indicators of External Vulnerability

article image
Source: Fund staff estimates, Recent Economic Developments, Bank for International Settlements.

Central Government Debt.

Official data.

BIS data.

CPI based.

149. The crises in Mexico and Southeast Asia have highlighted the importance of the level and composition of external liabilities in assessing the sustainability of current account deficits and the country’s ability to withstand shocks. Unlike these economies however, the Czech Republic has relied on a broad base of both debt and non-debt financing, and the share of short-term debt in total debt is moderate. The stock of gross external debt reached the widely accepted prudential limit of 40 percent of GDP by 1997 Q3, but was still relatively low when measured against exports of goods (at around 100 percent—where the commonly applied threshold is 200-220 percent) and exports of goods and services (at 73 percent) (Table 30, and Table A41, SM/98/30, Sup.1).110 Short-term debt has also been kept to within manageable bounds, and was equivalent to about three fourths of official foreign reserves and 40 percent of banking system foreign reserves at end-1997. From a comparative perspective, debt ratios in the Czech Republic are somewhat below Mexico and Thailand (particularly for short-term debt). The Czech Republic is also relatively well placed when compared with larger transition economies—since the country avoided foreign debt during the 1980s. This has helped the debt service position: Total debt service (in convertible currencies) remained at 15-16 percent of exports in 1996-97, without any evident bunching of maturities (Table A42, SM/98/30, Sup.1). Debt service was a little above that in Thailand but considerably below Mexico’s.111

Table 30.

Czech Republic: External Debt Indicators, 1993-97

article image
Sources: Czech National Bank; and staff estimates, and reports on Recent Economic Developments for Thailand and Mexico.

150. The integration of the Czech Republic with international financial markets makes it more sensitive to changes in market sentiment. The Czech Republic moved quickly to integrate with international financial markets in 1995, when it acceded to the IMF Article VIII requirements for current account convertibility and the OECD codes on capital liberalization. By April 1997, global turnover in the Czech koruna was more than ten times greater than in other Central European transition economies, and not much less than the Thai baht, despite the much larger size of the Thai economy (Table 31). This reflects active onshore and offshore trading in koruna securities.

Table 31.

Foreign Exchange Turnover in Emerging Market Currencies, 1995-97

(In billions of U.S. dollars)

article image
Source: BIS (1997).

Estimates as reported by the respective central banks, net of double-counting unless otherwise specified, for a period as near as possible to April. For Thailand, 1995 second half and 1996 annual averages. For Argentina, annual average.

Citibank estimates, net of double-counting.

On a gross basis.

The Central Bank Survey of Foreign Exchange and Derivatives Market Activity 1995 reports a grand total (including South Africa) of US$1,136.9 billion.

C. Conclusion

151. The current account has progressed through three phases since 1993: from a period of surplus/sustainable external deficits during 1993-95; to a period when the deficit was unsustainable, which resulted in a currency crisis in April-May 1997; followed by a gradual return toward sustainability, as policies have been tightened and the exchange rate has been allowed to depreciate. The deficit became unsustainable once it had become apparent that the prevailing policy configuration had left exports increasingly uncompetitive, not least because sizable real wage increases had raised unit labor costs while fueling demand for imports. After the stock adjustment in foreign investment in 1995, investors’ demand for koruna-denominated assets waned. A rapid downturn in investor confidence took place as the economic fundamentals weakened, and questions arose about the authorities’ resolve to pursue prudent fiscal policies and deliver on promised structural reforms. It was precisely at this time that the international environment toughened markedly, in light of developments in Southeast Asia. As a result, the authorities were obliged to float the exchange rate in May 1997. The tide appears to have turned for the current account deficit, following the tightening of fiscal policies in April-May and renewed efforts by the authorities to further contain the fiscal deficit, improve wage discipline in the public sector, and accelerate capital market reform. But as long as the deficit is large, and capital flows to emerging economies are subject to fads and fashions, the economy will remain exposed and fragile.

References

  • Begg, David (1997), “Pegging out: lessons from the Czech exchange rate crisis”, mimeo, Birkbeck College, London, December.

  • Frankel, J.A., and A. Rose (1996), “Currency Crashes in Emerging Markets: An Empirical Treatment,Journal of International Economics, Vol. 41, Nos. ¾, November 1996.

    • Search Google Scholar
    • Export Citation
  • Goldstein, Morris, 1996, “Presumptive Indicators/Early Warning Signals of Vulnerability to Financial Crises in Emerging Market Economies.” mimeo (Washington: Institute for International Economics).

    • Search Google Scholar
    • Export Citation
  • Halpern, L. and C. Wyplosz (1996), “Equilibrium Exchange Rates in Transition Economies”, IMF Working Paper WP/96/125 (Washington: International Monetary Fund, November 1996).

    • Search Google Scholar
    • Export Citation
  • Jack, W. (1996), “East-West Exports During Transition—The Relative Performance of Central and East European Countries in Accessing Western European Markets,” Czech Republic Recent Economic Developments (Washington: International Monetary Fund, November 1996).

    • Search Google Scholar
    • Export Citation
  • Kaminsky, G., S. Lizondo and C. Reinhart (1997), “Leading Indicators of Currency Crises”, IMF Working Paper WP/97/79 (Washington: International Monetary Fund, July 1997).

    • Search Google Scholar
    • Export Citation
  • Krajnyák Kornélia, and J. Zettelmeyer (1997), “Competitiveness in Transition Economies: What Scope for Real Appreciation?IMF Working Paper WP/97/149 (Washington: International Monetary Fund, November 1997).

    • Search Google Scholar
    • Export Citation
  • Milesi-Ferretti, Gian-Maria, and Assaf Razin, 1996, “Current Account Sustainability.” Princeton Essays in International Finance, #61, October.

    • Search Google Scholar
    • Export Citation
  • Roubini, N., and P. Wachtel (1997), “Current Account Sustainability in Transition Economies”, NYU Stern School of Business, mimeo, June.

    • Search Google Scholar
    • Export Citation
  • Williamson, John, 1996, “Comment [on Giovannini],” in Private Capital Flows to Emerging Markets After the Mexican Crisis,” ed. by G.A. Calvo, M. Goldstein and E. Hochreiter (Washington: Institute for International Economics).

    • Search Google Scholar
    • Export Citation

ANNEX I: Analytical Framework for the Study of Inflation

1. Consider the following simple model of a small open economy with two sectors—traded and nontraded goods—and money market clearing, where nominal wages are assumed to be exogenous and the exchange rate is assumed to be floating:1

π = α 1 π NT + α 2 π T ; π T = e + π * , ( 1 )
y d NT = β 1 [ π NT π T ] + β 2 y ; 0 < β 2 < 1 , ( 2 )
y s NT = δ 1 [ w π NT ] + δ 2 [ π NT π ] , ( 3 )
y s NT = y d NT , ( 4 )
m + v = π + y ; v = μπ e , ( 5 )
y s T = λ 1 [ w π T ] λ 2 [ π NT π T ] , ( 6 )

and

y = y s NT + y s T , ( 7 )

where π denotes domestic inflation, πe expected inflation, π* foreign inflation, e the rate of depreciation, w the rate of nominal wage growth, m the rate of nominal money growth, V the change in velocity, ydNT and ysNT the change in demand and supply of nontraded goods, ysT the change in the supply of traded goods, and y the change in total income, π, πNT, πT, ydNT, ydNT, ysT, ysT, and are endogenous, while w, πe, and m are exogenous.

2. Equation (1) determines inflation as a weighted average of nontraded and traded goods inflation, equations (2)-(4) describe the nontraded goods market, equation (5) is the money market equilibrium, and equations (6)-(7) determine total income.

3. Substituting (1)-(3) and (5) in (4) yields the following semi-reduced form for inflation in terms of notably the rate of depreciation, money, and wage growth:

A B e + A B π * + C B m + C B μπ e + D B w , ( 8 )

where

A = -α1β1 - α21 + δ12) < 0,

B = δ21-1) - α1β2 - β1 - δ1 < 0,

C = -α1β2 < 0,

D = -α1δ1 < 0, and

dπ/de > 0; dπ/dm > 0; dπ/dw > 0; dπ/dπe > 0.

4. Thus, as one would expect in line with the discussion above, inflation depends positively on the rates of depreciation, money growth, and wage growth as well as inflation expectations.

ANNEX II: Data Properties and VEC Estimation in Inflation Study

1. The efficient estimation of the vector autoregression (VAR) model requires that the time series be stationary. Table 1 below reports the outcome of Augmented Dickey Fuller and Phillips-Perron tests of the null hypothesis that variables have a unit root (are non-stationary) and shows that—with the exception of core inflation—this cannot be rejected for the variables in levels (annual growth rates). However, first differences appear to be stationary. In principle therefore, one needs to estimate the model in differences rather than levels.

Table 1.

Czech Republic: Test for Stationarity of Variables

article image

Augmented Dickey-Fuller tests.

Phillips-Perron tests.

Note: Two lags of the variables were included in the ADF test. (c) and (t) denote the inclusion of an intercept and a trend, respectively. *, **, and *** denote significance at the 10, 5, and 1 percent levels, respectively.

2. It is also necessary to test for any long-run relationship between the variables to ensure that the estimated model makes use of all available information. To the extent that such a relationship exists, the VAR needs to be supplemented with an error correction term that captures the dynamic response to deviations from the long-run equilibrium growth path; otherwise, a VAR in first differences is adequate. The most powerful procedure for testing such a long-run or cointegrating relationship is the Johansen test, which examines the existence of cointegration within the VAR framework and simultaneously determines the number of cointegrating vectors. It is assumed that the series have means and linear trends, but that the cointegrating equations have only intercepts. The test is relatively sensitive to the number of lags included in the VAR model. Estimations of an unrestricted VAR suggest including as many lags as possible (four in this case, given the short data period). The results of the Johansen test are shown in Table 2 below and generally indicate the presence of at least one cointegrating vector.1 Cointegration also exists between core inflation and subsets of these variables, albeit less robustly so (results not reported).

Table 2.

Czech Republic: Test for Co-integration of Variables

article image
Note: * and ** denote rejection of hypothesis at the 5 and 1 percent significance levels, respectively.

3. On this basis, a restricted VAR or vector error correction (VEC) model of the following form is estimated:

Δy t = A + Σ i = 1 K B i Δy t i + γ ( α + βy t 1 ) + t ( 1 )

where yt is a vector of variables [πt, mt, wt, and et] with πt, denoting core inflation, mt, growth in broad money, Wt, growth in industrial unit labor costs, and et, changes in import prices; A and B are vectors (matrices) of coefficients; and α+βyt-1 denotes the deviation from long-run equilibrium in the previous period (the error correction term). The results are as follows (with t-statistics reported in parenthesis):

π = 0.02 + 0.47 ( 20.8 ) w + 0.32 ( 13.5 ) m + 0.38 e ( 9.2 ) ( Long run cointegrating relationship ) ( 2 )
Δπ = 0.20 ( 1.8 ) + 0.55 ECT t 1 ( 3.1 ) ( Short run dynamic relationship ) ( 3 ) 0.12 Δ ( 0.39 ) π t 1 0.23 Δ t 2 ( 1.5 ) + 0.09 Δ ( 0.7 ) π t 3 0.21 Δ ( 1.9 ) π t 4 + 0.25 Δ ( 1.6 ) w t 1 0.55 Δ ( 3.7 ) w t 2 0.89 Δ ( 4.6 ) w t 3 0.01 Δ ( 0.1 ) w t 4 + 0.65 Δ ( 5.0 ) m t 1 + 0.21 ( 1.2 ) Δm t 2 + 0.34 Δ ( 3.1 ) m t 3 0.08 ( 0.8 ) Δm t 4 + 0.30 ( 3.5 ) Δe t 1 + 0.17 Δ ( 2.1 ) e t 2 + 0.07 Δ ( 0.7 ) e t 3 + 0.05 Δ ( 0.7 ) e t 4 R squared = 0.95

4. As can be seen from the estimated VEC model, the unit labor cost variable is the most important factor in explaining the long-run behavior of inflation, followed by import prices and money in that order. However, the interpretation of such a long-run relationship between inflation, wage growth, money growth, and changes in import prices (the rate of depreciation) is not clear.2 It does suggest, nevertheless, that in the long run the key macroeconomic nominal variables move in tandem (e.g., if wages, money, and import prices are growing by 10 percent, inflation would be around 12 percent). While the short-run dynamic (error correction) model has surprisingly good explanatory power and indicates that inflation responds positively to money and import price shocks in previous periods, the positive coefficient to the error correction term is odd and would seem to suggest that the model was dynamically unstable. Also, the negative coefficients to wage changes in previous periods are difficult to interpret.

5. The impulse response functions traced in Figure 1 point to broadly the same qualitative conclusions as those derived from the unrestricted model in levels discussed in the main text. Inflation responds positively to wage, money, and import price growth shocks, including an immediate and strong reaction to changes in the rate of depreciation that peaks after 5–7 months. Further, wage growth reacts quickly and strongly to changes in the rate of inflation, while faster wage growth in turn has a strong and persistent impact on money growth. Money growth responds negatively to higher inflation consistent with the findings earlier of nonaccommodation. As discussed in the main text, however, all these results need to be interpreted with great caution.

Figure 1
Figure 1

CZECH REPUBLIC: IMPULSE RESPONSES IN VEC MODEL

Citation: IMF Staff Country Reports 1998, 036; 10.5089/9781451810011.002.A006

Table 3.

Czech Republic: Pairwise Granger Causality Tests

article image
Note: Eight lags were included in the tests.

ANNEX III: Testing the Sustainability of the Exchange Rate Band

1. Interest rates, the forward exchange rate, and the direction of capital flows reveal information about market confidence on the exchange rate band. In a setup of capital mobility, a credible band implies that domestic and foreign interest rates may not differ by more than the bandwidth: a larger difference leaves room for arbitrage profits, which in turn build up pressure for changing either the central parity or the bandwidth.1 If, however, domestic and foreign rates differ by less than the bandwidth, the test is inconclusive. Similarly, a credible band implies that the forward exchange rate should be constrained within the fluctuation limits; if not, there are arbitrage profits to be made, which would put pressure on the exchange rate. This annex considers evidence from interest rate differentials and the forward exchange rate.

Evidence from Interest Rate Differentials

2. The annualized koruna yield from investing in a foreign asset for m months is given by Rt=(1+it*)(St+m/St)12/m1 where it* stands for the annual return of the foreign asset and St for the spot exchange rate of the koruna per unit of foreign currency. If SU and SL denote, respectively, the upper and lower limits of the fluctuation band, then a credible exchange rate band implies that the spot exchange rate m months ahead would be limited within SU≥St+m≥SL and the koruna yield of the foreign asset would be bounded by:

( 1 + i t * ) ( S U / S t ) 12 / m 1 R t ( 1 + i t * ) ( S L / S t ) 12 / m 1 ( 1 )

The interval implied by equation (1) is wider for shorter maturities as the effect of any given exchange rate change per unit of time increases; this tends to make the test inconclusive for shorter maturities.

3. If the domestic interest rate exceeds, say, the upper limit, (1+it*)(SU/St)12/m1, arbitrage profits can be made by borrowing abroad and lending the proceeds of the loan in the domestic market; the pressure would show up in the accumulation of official foreign exchange reserves (i.e., interventions to prevent the currency from appreciating). This was precisely the situation in the Czech Republic until the beginning of 1996 when interest rates of maturity longer than two months exceeded the bounds implied by equation (1). Indeed, the widening of the fluctuation band was intended to discourage nonresident investors from investing in koruna denominated assets, thereby reducing the need for sterilization operations. In the subsequent period until late-April 1997, all interest rates with a maturity of up to one year were within the implied bounds (Figure 1). The test is not sharp enough to detect the degree of market confidence in the exchange rate band during this period.

Figure 1
Figure 1

CZECH REPUBLIC: EXCHANGE RATE, INTEREST RATE, AND OFFICIAL RESERVES

Citation: IMF Staff Country Reports 1998, 036; 10.5089/9781451810011.002.A006

Sources: Czech National Bank and Fund staff calculations.

4. A domestic interest rate above the upper limit, which is unaccompanied by capital inflows, suggests expectations of depreciation. This was the case from late-April 1997 until the May 1997 attack on the koruna when the 12-month interbank interest rate moved outside the implied bounds: the higher interest rate differential represented an exchange rate risk premium that rendered arbitrage unprofitable. Although concerns about the sustainability of the band were voiced publicly already in March 1997, the test suggests that the overall market sentiment started to move against the koruna about a month later. Interest rates with a maturity of less than a year did not move outside the implied bands. Besides indicating the low power of the test, this suggests also that the market did not consider the depreciation imminent.

5. Two aspects need close scrutiny in conducting the test: the degree of capital mobility and the risk comparability of the financial assets involved. In the first case, the differential would incorporate the cost of the impediments to capital mobility. In the second case it might reflect differences in the default risk of the underlying assets. In such circumstances, the test should be performed after correcting interest rates for these two factors.

Evidence from the Forward Exchange Rate

6. The forward discount on the koruna did not reflect the increasing market concern about the sustainability of the exchange rate band. Except for periods of turbulence, the koruna has traded at a discount in the forward market,2 which has broadly reflected the differential between domestic and foreign interest rates (Figure 2). This suggests not only the dominance of arbitrage in the determination of the forward exchange rate, but also the practice of currency traders to set forward rates on the basis of interest rate differentials and to hedge their positions with offsetting lending and borrowing operations. Deviations from covered interest parity (CIP) have been almost ½ of 1 percentage point prior to the May 1997 depreciation but they have doubled since then, probably owing to higher uncertainty. The deviations from CIP widened substantially immediately after the May 1997 depreciation of the koruna, as well as during the turbulence in foreign exchange markets in November-December 1997. The higher deviations reflect the relatively larger role of speculation during these periods as well as the shortage of funds for arbitrage (owing to the tightening of monetary policy), the effect of higher uncertainty,3 data imperfections,4 and the fact that during the May 1997 episode several derivative instruments temporarily ceased to be quoted.

Figure 2.
Figure 2.

CZECH REPUBLIC: FORWARD EXCHANGE RATE

Citation: IMF Staff Country Reports 1998, 036; 10.5089/9781451810011.002.A006

Sources: CNB and staff calculations.

7. These observations are supported by an error correction (EC) regression of the discount on the three-month forward exchange rate (FR) on the differential between the domestic and foreign interest rate (Rdif), all in annualized terms. To account for differences in market attitude toward the koruna, uncertainty, and the like, the model was estimated separately for the pre- and post-depreciation periods. In the pre-depreciation period, the cointegrating equation:

FR = 0.19 ( 0.55 ) + 1.098 ( 0.087 ) Rdif

had a coefficient for the interest rate differential insignificantly different from unity. In the post-depreciation period, however, which is characterized by higher uncertainty and recurrent speculative attacks, the cointegrating relation was:

FR = 0.24 ( 0.43 ) + 1.113 ( 0.035 ) Rdif

The (statistically significant) greater than unity coefficient of the interest rate differential suggests a higher role of speculation whereas the higher constant term reflects the higher uncertainty.

8. The forward rate systematically underpredicted the depreciation of the koruna in May 1997. Despite deteriorating fundamentals, continued weakening of the koruna, and increasing market concern about the sustainability of the exchange rate band, the discount on the koruna remained fairly constant until about one week before the May 1997 attack (this probably reflects a negative correlation between expected depreciation and exchange rate uncertainty). Moreover, the forward rate of the koruna against the basket5 remained within the fluctuation band up until the eve of the depreciation, implying that the market was caught by surprise by the abandonment of the fluctuation band, which is the expected behavior in an efficient market (otherwise, there would have been unexploited profitable arbitrage opportunities).

ANNEX IV: Events Related to Exchange Rate Developments in 1997

February 11. The koruna reaches a high of 5.5 above its central parity, following a wave of koruna-denominated Eurobond issues, prompting the CNB to “talk down” the koruna.

March 12. The Czech Statistical Office (CSO) releases industrial output data showing economy slowing. Prime Minister Klaus openly criticizes monetary policy stance as tight.

March 25. Prime Minister Klaus rules out devaluation or administrative measures to address widening trade deficit, and again criticizes monetary policy stance as tight.

April 11. The CNB announces a reduction of the reserve requirement ratio from 11.5 percent to 9.5 percent, effective May 8.

April 14. The exchange rate of the koruna reaches a ten-month low against the currency basket, trading close to parity, in advance of the expected announcement of a policy package.

April 16. Coalition announces a wide-ranging package of stabilization measures and structural reforms, including budget cuts of 1¾ percent of GDP. Prime Minister Klaus warns that the economy is slowing and urges a loosening of monetary policy.

April 28. The CNB denies rumors of imminent cut in interest rates.

May 2. The CSO revises its 1997 GDP growth projection to 2.9-3.5 percent from 4-5 percent.

May 12. The CSO releases 1997 Q1 industrial output data, confirming the slowdown of activity. Rumors of ministerial resignations.

May 14. The Thai baht comes under strong pressure. The Financial Times publish a negative country survey for the Czech Republic. Opinion poll shows Civil Democrats’ (ODS) popularity at all-time low.

May 15. CNB intervenes in the foreign exchange market for the first time in over one year, as koruna weakens to 5 percent below its central parity.

May 16. The CNB raises the repo rate by 50 basis points to 12.9 percent; and, in light of continued pressure on the koruna during the day and need for intervention, it also raises the Lombard rate from 14 percent to 50 percent per annum. Notwithstanding continued interventions (US$ 1¼ billion over the two days) the koruna fails to recover to its pre-crisis level.

May 19. The CNB sets a maximum rate of 45 percent for its repos (to withdraw liquidity). The overnight repo rate in CNB auctions rises to 30 percent by mid-day and reaches 45 percent later in the day. Interbank rates double. The koruna strengthens about 2.8 percent below parity despite the absence of foreign exchange sales by the CNB.

May 20. The CNB limits access to the Lombard window to an unspecified portion of a bank’s capital; in practice, this discriminated against branches of foreign banks, which have no capital of their own and are reportedly more active in taking positions against the koruna. As a result, overnight interbank rates rises to almost 100 percent, while three-month interbank interest rates (bid-offer) rise from 16-27 percent to 20-35 percent. Two major state-controlled banks doubled their prime lending rates to 25 percent.

May 21. Coalition discussions on Cabinet reshuffle continue. The CNB injects some liquidity with reverse repos at an average interest rate of 106 percent; it also continues to withdraw liquidity at an average interest rate of 57 percent. The koruna strengthens temporarily but the market remains very thin. The overnight interest rate remains close to 100 percent, although rates of 500 percent are also reported.

May 22. The koruna comes under renewed pressure, depreciating to 6.2 percent below parity early in the morning. In response, the CNB intervenes; closes the Lombard facility; raises the one-week repo rate to 75 percent; and instructs major domestic banks to limit access by nonresidents to the Czech money market. As a result the koruna recovers to 3-4 percent below parity and the overnight interest rate reaches 200 percent. However, public confidence in the koruna weakens and residents start changing korunas into foreign currency. Polish zloty weakens. The National Bank of Slovakia informs larger banks they are not required to quote crown rates to nonresidents.

May 23. The koruna trades at about 4 percent below parity but with interventions after the fixing it strengthens to about one percent below parity. The CNB continues to withdraw liquidity at almost 75 percent with one- and two-week repos.

May 26. Market relatively calm due to a holiday in the U.S. and in currency markets in many London. The koruna opens at 1 percent below parity. Trade data showing an improvement relative to previous months fail to impress the market. Subsequently the koruna weakens to 3 percent below parity but, following CNB interventions, it strengthens again to 2 percent below parity. The interest rates in CNB’s overnight and one week repos remain unchanged at almost 75 percent. A major state-controlled commercial bank reduces its prime rate by 4 percentage points to 21 percent. At 7:30 pm, the CNB Governor (in a joint press conference with the Prime Minister) announces that as of Tuesday (i) the ±7½ target band would be replaced by managed floating and (ii) the koruna would be stabilized vis-à-vis the DM but without officially binding limits. Later, the CNB made public an indicative range of CZK 17-19.5 per DM (which is equivalent to a range -11 to +1 from the central parity). The discount rate is raised by 2.5 percentage points to 13 percent; other CNB rates are kept unchanged. The Governor asks for additional macroeconomic (fiscal) tightening; the Prime Minister stresses the need to avoid wage indexation.

May 27. The koruna opens at 8 percent below parity, rapidly depreciates to almost 12 percent below parity (prompting CNB interventions and the rise of the discount rate from 10.5 percent to 13 percent), but subsequently strengthens to 10.7 percent below parity. The market remains extremely thin and margins large; interbank interest rates increase by 3 percentage points. Slovak crown heavily sold, reaches bottom of its ±7 percent band.

May 28. The koruna trades at about 11 percent below parity in a very thin market. There are no reports of intervention; the CNB supplies CZK 3 billion of liquidity at an average interest rate of 165 percent. Three month interbank interest rates decline by 1 to 4 percentage points. In the evening, the three parties in then ruling coalition agreed on a policy package to revive the economy: Budget cuts and freeze of public sector wages to balance the 1997 budget and attain a slight surplus in 1998; restrictions on imports financed by the budget; negotiations with the unions to stem wage growth; and additional measures (including higher tariffs) would be considered in the second half of the year if the external situation does not improve. After modest intervention succeeds in stabilizing the zloty, Polish Finance Minister declares zloty safe from speculative attack. Slovak crown strengthens after NBS persistent fixing above market levels.

June 2. CNB lowers repo rate from 75 percent to 45 percent.

June 7. The CNB announces that it had asked four banks (Chase Investment Bank Ltd, Commerzbank AG, JP Morgan Securities Ltd, and SBC Warburg) to arrange a US$1.5 billion one-year credit facility as a standby to strengthen its reserves.

June 9: Austerity measures agreed by the ruling coalition parties.

June 10. Coalition wins Parliamentary no confidence vote, by a single vote.

June 11. Taking advantage of the improved political situation after the confidence vote, the CNB reduces the maximum rate on its one-week repos from 39 percent to 31 percent and on two-week repos from 31 percent to 29 percent. Market participants expect the koruna to trade at 18.5–19 against the DM in the immediate future. Komercni reduces its prime lending rate by 4.4 percentage points (from 21.7 percent to 17.3 percent).

June 12. Access to the Lombard facility is reopened.

June 13. The maximum rate on one-week CNB repos is reduced from 31 percent to 29 percent.

June 17. Access by nonresidents to the Czech money market restored.

June 18. Government adopts the package of additional stabilization measures announced immediately after the floating of the koruna.

June 19. CNB reduces the repo rate from 29 percent to 25 percent

June 20. CNB signs one-year US$2 billion revolving syndicated credit at 10 basis points over LIBOR; the cheapest credit granted to a sovereign transition borrower to date.

June 27. CNB lowers its Lombard rate from 50 percent to 23 percent.

July 2. The Thai authorities float the baht. The Philippine peso comes under pressure.

July 14. The Polish zloty depreciates by 5 percent in volatile trading.

July 21. The koruna temporarily weakens to 11½ percent below its former parity on fears that monetary policy will be loosened in the wake of extensive flooding.

July 24. Currency crisis in Southeast Asia; the Thai baht plummets by 25 percent.

October 23. The Czech Foreign Minister resigns; doubts about the viability of the coalition government. The CNB abolishes the limits of the short koruna position of banks and limits on the short position of banks vis-à-vis nonresidents.

October 23-28. Hong Kong stock market declines by 25 percent.

November 29-30. PM Klaus resigns. The koruna weakens to almost CZK 20 per DM 1.

December 17. CNB Governor Tošovský is appointed Prime Minister with the mandate to lead the country to early elections in 1998; the koruna strengthens.

ANNEX V: Proximate Determinants of the Intermediation Spread

1. The intermediation spread, the difference between the average lending and average deposit interest rates, depends on market efficiency (competition depresses operating costs and profitability), credit/default risk, and the extent of regulation (high and nonremunerated required reserves raise the cost of intermediation). The spread is also a key determinant of bank profitability in environments such as that of the Czech Republic, where lending and deposit-taking dominate banking activity, although its importance has been waning as banks rely increasingly on foreign borrowing and deposit substitutes (bonds, certificates of deposit) for their funding, engage in nonlending activities (e.g., leasing) and raise significant revenue from fee-services.

2. The balance sheet and income statement of banks provide a useful framework for analyzing the determinants of the spread. A stripped-down version of the balance sheet:

L + ρ . D = D + A , ( 1 )

has on the asset side domestic currency lending to nonbanks, L, and required reserves, ρ.D, (where ρ stands for the required reserves ratio); and, on the liability side, deposit liabilities to nonbanks, D, and net other liabilities, A (including foreign currency lending, borrowing from the interbank market, and own funds). The corresponding income statement is:1

K = R L L R D R A A + I C Q , ( 2 )

where K stands for before-tax profits and Rj denotes the interest rate on item j. The first two terms in equation (2) capture net interest income, while the third, - RA A, gives the net cost of net other liabilities. The remaining terms measure, respectively, noninterest revenue, I, operating costs, C, and the buildup of provisions for bad loans, Q. Scaling down equation (2) by lending to nonbanks, L, yields:

k = R L - R D d - R A a + i - c - q

where the lower case letters denote ratios of the corresponding item in equation (2) to bank lending. Separating out the intermediation spread (RL - R D) and the cost of required reserves evaluated at the average deposit interest rate, -RL d ρ, and collecting terms yields:

k = ( R L R D ) R D d ρ [ R A R D ] a + i c q . ( 3 )

3. The first term on the RHS of equation (3) corresponds to the net interest income when lending is financed entirely by deposits. The second term captures the cost to banks from the requirement to keep nonremunerated reserves with the central bank. The third term captures the differential effect on profitability from funding part of lending from sources other than deposits (bank bonds, foreign borrowing, interbank borrowing, etc.); the overall impact of this depends on the extent to which (i) the cost of alternative funds, RA, exceeds deposit interest rates, RD, and (ii) banks fund their lending operations from alternative sources.2 The last two terms capture the contribution of noninterest income (a relatively high value indicates resilience of banks’ revenue to interest rate variability) and provisioning, respectively.

4. The terms on the RHS of equation (3) are not independent. When, say, interbank interest rates increase (reflected in a higher RA), banks that borrow from the interbank market are forced to raise their spread so as to mitigate the negative impact on their profitability. Although aggregate profits may increase, the impact on bank profitability will not be uniform: net lenders in the interbank market (mainly Ceska Sporitelna) would benefit, but net borrowers would experience a weakening in profitability.

5. The contribution of provisioning in equation (3) may differ significantly from the credit risk premium, an important element in setting lending interest rates. It overstates the premium in the period before provisions reach their steady-state level and understates it subsequently.

85

Prepared by David Bendor.

86

A priori, one would expect exports to be positively related to foreign demand, and negatively related to domestic demand (which acts to divert exportables to the home market when it increases).

87

The unit labor cost (ULC) based real effective exchange rate (REER) appreciated rapidly between 1990 and 1993, as output fell without an equivalent labor shedding. When output recovered and industry continued to shed labor, the rate of REER appreciation slowed down. This notwithstanding, exporters were generally able to increase their prices (partially through product and quality enhancements); meanwhile import prices were kept low (relative to domestic prices) such that the overall terms of trade steadily improved (Figure 33).

88

The growth of domestic demand in 1993 is based on the national accounts series at constant 1984 prices, which is currently under revision.

89

The shares of iron and steel and electrical machinery exports declined in 1996, likely as a result of buoyant fixed investment; competition from lower cost Eastern European producers may also account for the weakening performance of iron and steel.

90

Market re-orientation, seen in 1993-95, was less evident however, as the Czech share of world trade also fell. See also Jack (1996) for a historical perspective on market share developments.

91

Basic materials and manufacturing sectors, including textiles and metals, may have benefited from the depreciated exchange rate.

92

While manufacturing has typically accounted for less than half total FDI, there is a lagged relationship between investment in machinery and equipment, and chemicals, and these exports.

93

A recent survey reported by the Czech investment agency (Czechlnvest) found that there are over 600 joint-ventures in the manufacturing sector, which account for 40-50 percent of all Czech exports. Of these, 40 percent are working three daily shifts (presumably close to full capacity) and a similar percentage are taking on new employees and are re-investing profits.

94

The extent to which supply constraints have negatively impacted on exports, especially in 1996, is unclear. The CSO measure of capacity utilization indicates that total industrial capacity utilization remained unchanged at 83 percent in 1996, but key export sectors (including chemicals and motor vehicles) increased utilization rates by up to 5 percentage points, in the period January 1996—July 1997.

95

Škoda-Volkswagen unit exports to Germany rose by 39 percent in 1997, in line with the total unit increase.

96

The Czech National bank’s classification of each category is based on 3-digit Standard International Trade Classification categories and needs to be interpreted with care, the investment share is too high. Such disaggregated sectoral data is unavailable prior to 1995, when the new methodology trade data was introduced.

97

Government concerns about the growing trade imbalance led to the imposition of a 6-month noninterest-bearing deposit equal to 20 percent of import value on consumer and agricultural goods between April and August 1997, which resulted in some temporary compression of these items. The scheme was to have been in place for one year, but was abolished sooner—once signs that the trade balance was improving were confirmed. (The weighted average import tariff in 1997 was about 5 percent, ranging from 4 percent on primary goods to 10 percent on finished goods; tariffs were generally applied uniformly.)

98

The travel item includes tourism and “shuttle” or cross-border trade with neighboring countries. The latter was first estimated in the balance of payments in 1994: the Czech Republic has more informal imports (e.g., from Poland) than exports. This level change distorts the series for the real growth of imports of services in 1995.

99

In 1995, net interest payments were US$-134 million while profits were US$28 million; in 1996 these were US$-475 million and US$-248 million respectively.

100

Interest payments in convertible currencies, on medium- and long-term debt, rose across a wide range of instruments and borrowers—notably commercial bank credit and bonds, as well as corporations’ trade credits (Table A42, SM/98/30, Sup. 1). Interest payments may also have risen as Czech entities turned to koruna financing (see Box 3); CNB statistics have included koruna denominated foreign assets and liabilities from September 1996.

101

The national accounts suffer from weaknesses in coverage and reporting, and need to be treated with caution. Savings and investment, in relation to GDP, are unusually high by international standards. The Czech Statistical Office practice is to include the statistical discrepancy item with stocks, which likely overstates non-government savings. Hence gross domestic investment here is taken to be fixed investment, while the statistical discrepancy (which is assumed to comprise measurement errors in consumption) is included in gross national savings.

102

Industry accounts for about 30 percent of GDP in the Czech Republic. By comparison, industry accounts for 25 percent of investment (27 percent of GDP) in Poland, and 27 percent of investment (31½ percent of GDP) in Hungary.

103

Developments in total employment, average nominal wages and nominal GDP would suggest that the labor share in output has risen at the expense of capital.

104

The Telecom sale raised US$1½ billion (3 percent of GDP) providing a one time boost to foreign direct investment.

105

The Czech Republic accepted the IMF’s Article VIII and OECD conditions for the liberalization of current payments and capital flows in 1995.

106

There are discrepancies between the commitments data and the data on medium- and long-term capital inflows from the balance of payments (reported as a memo item), since the balance of payments flows include other forms of debt financing (e.g., suppliers’ credits) and refer to actual disbursements.

107

The balance of payments reports short-term capital on a net basis. Banking system returns indicate that the commercial banks’ short-term net foreign asset position has improved markedly in the period following the currency crisis.

108

Recent cross-country studies include Milesi-Ferretti and Razin (1997); Goldstein (1996); Frankel and Rose (1996) and Kaminsky et al (1997).

109

Williamson (1996) provides a theoretical underpinning for this.

110

The shares of enterprise and commercial bank debt in total debt increased by 5 percentage points during 1995-97, to 35 percent and 25 percent respectively; while that of the government halved to around 6 percent.

111

Official figures showing the currency composition of external debt do not exist. It is believed, however, that the currency composition broadly mirrors export receivables, which should not pose additional servicing risks. Data for the Czech Republic in the World Bank’s Global Development Finance report (1997) generally confirm this.

1

This model follows closely that of Coorey et al. (1996) but abstracts from relative price changes.

1

The inclusion of four lags points to the presence of four cointegrating vectors, but in the analysis below we assume that the rank is one.

2

Coefficients are not elasticities as the model could not be estimated in logs due to the presence of negative observations (changes in import prices).

1

L. Svensson, “The simplest test of target zone credibility”, Staff Papers, IMF, Vol. 38, (September 1991), pp. 655-65.

2

The market for foreign exchange forwards and swaps developed rapidly soon after the widening of the fluctuation band to ±7½ percent in February 1996; the large issues of eurokoruna bonds in late 1996/early 1997 have also provided an impetus. By April 1997 the volume of forwards and swaps had risen to ⅓ of the spot foreign exchange market. The rates in the domestic and offshore markets have been close, owing to the freedom of capital mobility and the internationalization of Czech financial markets.

3

During periods of volatility in foreign exchange markets traders raise the premium above the interest rate differential to cover the risk of having to hedge their position at interest rates substantially different than the ones prevailing at the time of negotiating the forward contract.

4

The major problem is that interest and exchange rates are sampled at different times of the day; moreover, they may be quotations rather than actual transaction rates.

5

The forward rate of the basket is calculated as the koruna cost of purchasing forward the amount of DM and U.S. dollars in the currency basket.

1

The presentation is based on the current accounting practice of treating as regular income interest accruals on loss-loans given by B.RL where B stands for outstanding loss loans. If loan loss provisions fall short of the corresponding interest accruals, this practice can lead to the decapitalization of banks (via the distribution of paper profits) and to the overestimation of banks’ capital adequacy. The payment of profit tax on accrued interest puts additional pressure on bank profitability.

2

A negative value of A in equation (1) means that deposits exceed lending and, therefore, banks use deposits to finance nonlending activities, while -[RA - R D]a captures the extra revenue (in excess of the deposit interest rate) that banks generate from their nonlending activities. The term a=1-d(1-ρ) measures the extent to which deposits are used for purposes other than lending or for meeting the minimum reserve requirement.

  • Collapse
  • Expand
Czech Republic: Selected Issues
Author:
International Monetary Fund