Journal Issue
Share
Article

Turkey

Author(s):
International Monetary Fund
Published Date:
May 2005
Share
  • ShareShare
Show Summary Details

I. Introduction

1. The 2004 Article IV consultation takes place at an opportune time. First, the current program is drawing to an end, and the authorities will soon decide whether they wish to request a successor arrangement. Second, later this year the European Union will formally assess Turkey’s compliance with the Copenhagen criteria and its eligibility to begin accession negotiations. Finally, recent tensions in financial markets have brought out a number of difficult short-term policy challenges that the authorities need to address—a reminder that, despite recent progress, Turkey is still highly vulnerable to shifts in sentiment towards emerging markets. The Article IV consultation therefore offers a well-timed opportunity to take stock of what has been achieved under the program, to set out reform priorities ahead (which could provide useful input for any new program), and to gauge Turkey’s challenges in commencing EU accession negotiations.

II. From Economic Crisis to Economic Recovery

A. The Economic Crisis

2. For Turkey, the 1990s was a decade of economic mismanagement. Although economic growth averaged 4 percent, it was extremely volatile. Inflation was variable and high, averaging close to 80 percent and peaking in 1994 at more than 100 percent. High inflation raised nominal interest rates, and with it the public sector borrowing requirement, which increased from 10 percent of GNP in 1991 to more than 20 percent in 1999. High deficits and inflation uncertainty pushed up real interest rates and slowed growth, and doubled the public debt ratio to 60 percent of GNP.

3. The collapse of the 2000 exchange rate based disinflation program magnified these weaknesses. Although the crawling peg had reduced inflation and interest rates significantly, competitiveness deteriorated. With domestic demand booming and fiscal policy failing to tighten, the current account deficit worsened. Vulnerabilities in the financial sector increased, as poorly supervised banks bet on the success of the program by borrowing overnight and in foreign currency, and investing in longer-term TL paper, anticipating further reductions in interest rates. When the exchange rate came under pressure in late 2000 and again in early 2001, interest rates increased, highlighting these weaknesses in the financial system. When the exchange rate was finally allowed to float in February 2001, it lost more than half of its value. This pushed inflation back up to 70 percent by the end of the year, and increased the government’s net debt ratio to more than 90 percent of GNP (75 percent using centered GNP). Parts of the banking system faced huge losses both because of the collapse of the exchange rate, and the high interest rates and borrowing costs that preceded it.

B. The Economic Recovery Strategy

4. With the devaluation of the exchange rate and the sharp rise in interest rates, the outlook for disinflation, debt sustainability, and renewed growth appeared bleak. Although the possibility of debt restructuring loomed, it was quickly ruled out.

5. The reforms launched in May 2001, and extended into the current program in early 2002, were a strong attempt to address Turkey’s underlying vulnerabilities. With debt restructuring ruled out, the first priority was to ensure debt rollover and debt sustainability. Notwithstanding the severe recession, this meant committing to high and sustained primary surpluses. For 2001, the target was 5½ percent of GNP; this was increased to 6½ percent of GNP in 2002 and beyond. The exchange rate was allowed to float, its role as nominal anchor abandoned. Instead, to reduce inflation the central bank was freed from any obligation to finance the deficit, and given operational independence to reach its new primary objective of achieving and maintaining price stability. To do this, the staff advocated the early introduction of formal inflation targeting, though this was perhaps premature given the high inflation rate and fiscal dominance problems. However, given Turkey’s severe balance of payments and government financing problems, these efforts at macroeconomic stabilization would not have had the chance to succeed without substantial outside financial assistance. This was supplied by the Fund.

6. The program also strongly emphasized structural reform, largely to signal a break from the failed economic policies of the past. To reduce a key source of vulnerability and quasi-fiscal activity, the banking system was restructured and recapitalized (first the state banks, then the private banks), at a cost of US$47 billion (more than 30 percent of GNP). A wide range of other structural reforms was also introduced, from privatization to governance reform, in an attempt to change the way the economy functioned and to demonstrate this to markets.

C. Macroeconomic Performance under the Program

7. The program strategy has delivered a dramatic improvement in financial market confidence (Figure 1). Though markets were volatile through much of 2001, the February 2002 augmentation and relaunching of the program helped assuage the financing concerns that had resurfaced after September 11, 2001. Despite interruptions from domestic political tensions (early elections in 2002) and outside shocks (the Iraq War), interest rates have declined substantially and debt maturities have lengthened, helping the government’s debt rollover. The program’s focus on generating large primary surpluses and on reducing inflation have been integral to this improvement. But outside factors—the availability of substantial financial assistance: first from the Fund, later (potentially) from the United States, and the extended period of low world interest rates—have also been important.

Figure 1.Turkey: Financial Indicators, 2001-2004

(in percent, unless otherwise indicated)

Source: Data provided from the Turkish authorities.

8. Recovery has been rapid and powerful (Table 1). After contracting by nearly 10 percent in 2001, GDP grew by close to 8 percent in 2002 and 6 percent in 2003, well above program projections (Figures 2 and 3). Once political stability emerged and the authorities demonstrated their commitment to keep fiscal policy on track, real interest rates declined sharply in 2003, below program projections, setting the stage for a rapid increase in consumer credit. With consumer and investor confidence picking up too, the recovery has become increasingly led by domestic demand. The strength of the exchange rate has fuelled private consumption and consumer goods imports, and this year’s pension and minimum wage increases have added to this. Despite banks’ reluctance to lend to indebted corporates, improved profitability and higher capacity utilization have stimulated private investment.

Table 1.Turkey: Selected Indicators, 2000-05
Projections
200020012002200320042005
(In percent)
Real sector
Real GNP growth rate6.3-9.57.95.95.05.0
GNP deflator50.955.344.422.511.510.1
Nominal GNP growth rate60.440.555.829.717.115.6
WPI (12-month, end-of-period)32.788.630.813.914.28.0
CPI (12-month, end-of period)39.068.529.718.412.08.0
Average nominal treasury bill interest rate38.099.163.544.124.819.0
Average ex-ante real interest rate 1/-9.535.530.328.612.812.1
(In percent of GNP)
Central government budget
Primary balance 2/3/4.64.82.45.05.05.0
Net interest payments 4/15.824.717.516.113.610.4
Overall balance-11.2-20.0-15.1-11.1-8.6-5.3
Consolidated public sector
Primary balance 3/3.05.54.16.36.56.5
Net interest payments 5/21.926.616.116.113.510.3
PSBR (including CBT profits)18.921.112.09.87.03.7
Operational balance-6.9-4.7-4.4-4.8-2.9-1.0
Net debt of public sector57.493.978.870.570.364.7
Net external19.037.732.122.222.720.1
Net domestic39.356.246.748.347.744.6
Net debt of public sector (in percent of centered GNP)652.475.668.765.965.758.9
External sector
Current account balance-4.92.4-0.8-2.8-3.6-3.0
Gross external debt59.079.072.161.853.251.2
Net external debt38.853.854.645.140.739.7
Short-term external debt (by remaining maturity)21.722.717.917.516.216.1
Monetary aggregates
Seignorage 7/1.81.11.01.21.40.5
Nominal growth of M2Y broad money (in percent)40.287.525.413.020.813.5
(In billions of U.S. dollars, unless otherwise indicated)
Privatization proceeds3.32.80.50.33.03.0
Net external financing of central government4.1-2.3-1.4-1.4-0.1-0.3
Amortization6.27.86.88.98.68.6
Gross borrowing10.35.55.37.58.58.3
Of which: Eurobond issues7.52.23.35.35.05.0
GNP201.3144.0182.7238.5
GNP (in quadrillions of Turkish lira)125.6176.5275.0356.7417.6482.8
Sources: Data provided by Turkish authorities; and IMF staff estimates.

Average of monthly nominal interest rate divided by 12-month ahead CPI inflation. With average maturity of newly issued debt less than one year, and with FRNs paying quarterly coupons, this measure overstates the effective real interest rate when inflation is declining.

On a commitment basis, excluding profit transfers from the CBT, interest receipts, and privatization proceeds

For 2003 and 2004, program projections.

Interest payments minus interest receipts plus profit transfers from the central bank.

Interest payments minus interest receipts plus CBT profits before transfers to the government..

Defined as the sum of quarterly GNP in the last two quarters of the year and in the first two quarters of the following year.

Change in reserve money in percent of GNP, where reserve money is defined as currency issued plus reserve requirements.

Sources: Data provided by Turkish authorities; and IMF staff estimates.

Average of monthly nominal interest rate divided by 12-month ahead CPI inflation. With average maturity of newly issued debt less than one year, and with FRNs paying quarterly coupons, this measure overstates the effective real interest rate when inflation is declining.

On a commitment basis, excluding profit transfers from the CBT, interest receipts, and privatization proceeds

For 2003 and 2004, program projections.

Interest payments minus interest receipts plus profit transfers from the central bank.

Interest payments minus interest receipts plus CBT profits before transfers to the government..

Defined as the sum of quarterly GNP in the last two quarters of the year and in the first two quarters of the following year.

Change in reserve money in percent of GNP, where reserve money is defined as currency issued plus reserve requirements.

Figure 2.Turkey: Macroeconomic Indicators, 1999-2004

Source: Fund staff estimates and projections.

Figure 3.Turkey: Output and inflation, 2001-2004

(in percent, unless otherwise indicated)

Source: State Institute of Statistics; and Central Bank of Turkey.

Turkey: Real Interest Rate on Domestic Treasury

Turkish Lira Borrowing, 2002-04 1/

Citation: 2005, 163; 10.5089/9781451838145.002.A001

1/ 12-month ahead CPI used in real interest rate calculations.

9. Inflation too has outperformed expectations. The contribution of primary surpluses towards achieving debt sustainability, the CBT’s adherence to the base money program in 2002 and 2003, and the resulting improvement in financial market confidence and stabilization of the exchange rate, have led to a dramatic decline in inflation to levels not seen since the 1970s. Were it not for the recent exchange rate depreciation and the likely pass-through from higher world oil prices, inflation would have been on track to fall to single digits by end-year, below the CBT’s 12 percent target.

Reserves Indicators

(In US$ billions)

Source: Data provided by the Turkish authorities.

10. The strength of the recovery and the appreciation of the real exchange rate has pushed the current account deficit above the original program path (Table 2). Instead of stabilizing at around 1 percent of GNP, by 2003 the current account deficit had widened to close to 3 percent of GNP. Higher capital inflows—mainly the drawdown of residents’ foreign currency holdings—have more than compensated, allowing the CBT to accumulate reserves well above the projections made in 2002, but financing has been skewed toward short-term portfolio inflows. The CBT has responded to these developments by intervening to build up reserves.

Table 2.Turkey: Balance of Payments, 2001–08 1/(In billions of U.S. dollars)
20012002200320042005200620072008
Current account balance3.4-1.5-6.6-10.1-8.8-7.2-7.7-9.1
Trade balance-4.5-8.3-13.7-19.1-18.8-17.9-18.9-20.8
Exports (f.o.b.)34.439.851.062.968.675.182.490.5
Of which:
Exports (f.o.b.) in trade returns31.335.847.158.664.470.978.186.2
Shuttle trade3.04.14.04.34.34.34.34.3
Imports (f.o.b.)-38.9-48.1-64.7-82.1-87.5-93.0-101.3-111.4
Of which:
Imports (c.i.f.), incl. non-monetary gold-41.4-51.2-68.8-87.2-93.0-98.8-107.6-118.4
Energy imports (c.i.f.)-8.3-9.2-11.4-14.5-14.7-14.5-15.0-15.7
Services and Income (net)4.13.35.16.87.68.18.48.9
Services and Income (credit)18.817.321.325.427.428.930.131.9
Of which:
Tourism receipts8.18.513.216.116.817.618.519.5
Services and Income (debit)-14.7-13.9-16.2-18.6-19.8-20.9-21.7-23.0
Of which:
Interest-7.1-6.4-6.9-7.6-8.4-8.8-8.9-9.2
Private transfers (net)3.63.01.71.91.92.02.12.2
Official transfers (net)0.20.50.30.30.40.60.60.6
Capital account balance-14.61.45.911.59.411.311.411.7
(including errors and omissions)-16.31.310.711.79.411.311.411.7
Direct investment 1/2.80.90.11.61.82.12.53.1
Portfolio investment in securities-4.6-1.21.22.31.61.31.51.7
Public sector (central & local governments & EBFs)-1.90.4-0.70.71.62.42.01.7
Bonds (net)0.11.01.51.22.73.22.62.1
Eurobond drawings2.13.35.35.05.55.75.75.7
Eurobond repayments-2.0-2.3-3.8-3.7-2.8-2.5-3.1-3.6
Loans (net)-2.0-0.7-2.2-0.6-1.0-0.8-0.6-0.4
Loan disbursements1.62.31.02.82.12.22.32.3
Loan repayments-3.6-3.0-3.2-3.3-3.2-3.0-2.9-2.7
Central Bank of Turkey, (Excl. reserve assets, liabilties)0.81.40.60.10.00.00.00.0
Domestic money banks (net)-9.4-1.83.05.01.31.61.51.5
Domestic money banks (FX deposits abroad, -: accumulation)0.90.60.72.1-0.2-0.3-0.3-0.3
Domestic money banks (other, net)-10.3-2.42.32.91.51.91.81.8
Other private sector (net)-2.31.81.71.83.13.83.83.7
Other private sector (medium and long term, net)0.32.71.52.42.32.82.82.9
Other private sector (short term, net)-2.6-0.90.2-0.60.71.11.00.9
Errors and omissions 2/-1.7-0.14.80.20.00.00.00.0
Overall balance-12.9-0.24.11.50.54.23.72.6
Overall financing (NIR change excl. ST liabilities, + denotes decline)12.90.2-4.1-1.5-0.5-4.2-3.7-2.6
Change in net international reserves (+ denotes decline)12.90.2-4.1-1.5-0.5-4.2-3.7-2.6
Change in gross official reserve assets (+ denotes decline)2.7-6.2-4.01.36.37.3-1.9-1.8
Change in reserve liabilities (IMF)10.26.4-0.1-2.8-6.9-11.5-1.8-0.9
Purchases11.312.51.71.80.70.00.00.0
Repurchases-1.1-6.1-1.7-4.7-7.6-11.5-1.8-0.9
Memorandum items:
Trade in goods and services
As percent of GNP
Current account balance, incl. shuttle trade2.4-0.8-2.8-3.6-3.0-2.3-2.3-2.5
Trade account balance, incl. shuttle trade-3.2-4.6-5.8-6.8-6.5-5.8-5.6-5.8
Exports of goods and non-factor services36.131.030.031.132.833.033.233.6
Imports of goods and non-factor services32.230.631.033.134.133.834.134.7
Percent change
Value growth in exports of goods (incl. shuttle trade)11.915.828.123.39.19.49.79.9
Value growth in exports of goods (excl. shuttle trade)12.814.131.624.69.810.110.210.4
Value growth in imports of goods-26.823.734.526.76.66.38.910.0
Volume growth in exports of goods16.611.716.314.810.110.79.99.5
Volume growth in imports of goods-22.921.117.815.47.47.68.99.4
Terms of trade1.70.0-0.8-1.20.50.70.40.3
Reserve and debt indicators
Gross foreign reserves (Central Bank of Turkey)
In billions of U.S. dollars19.828.135.233.927.620.322.224.0
Months of goods & NFS imports4.45.45.24.03.12.12.22.1
External debt (end-of-period)
In billions of U.S. dollars113.8130.9147.3149.6148.3144.0148.9154.4
Percent of GNP79.072.161.853.251.246.344.542.9
Percent of exports of goods & NFS218.8232.5205.5170.9156.3140.2133.9127.6
Net external debt (end-of-period) 3/
In billions of U.S. dollars77.599.1107.5114.3115.1106.2107.1109.7
Percent of GNP53.854.645.140.739.734.132.030.4
Short-term debt (end-of-period)
In billions of U.S. dollars16.416.422.926.229.032.135.238.4
Ratio to end-period foreign reserves82.658.465.177.2104.9158.3158.7160.1
Short-term debt plus MLT repayments
In billions of U.S. dollars32.732.541.845.446.649.453.757.9
Ratio to foreign reserves164.7115.4118.8133.9168.8243.5241.9241.6
Debt service ratio 4/41.337.334.929.826.524.623.823.1
Sources: Data provided by the Turkish authorities; and Fund staff estimates and projections.

Including privatization receipts.

For 2004 includes reported data for January-March.

Nonbank external debt less the NFA of the banking system.

Interest plus medium- and long-term debt repayments as percent of current account receipts (excluding official transfers).

Sources: Data provided by the Turkish authorities; and Fund staff estimates and projections.

Including privatization receipts.

For 2004 includes reported data for January-March.

Nonbank external debt less the NFA of the banking system.

Interest plus medium- and long-term debt repayments as percent of current account receipts (excluding official transfers).

11. Although extreme poverty remains low, income inequality and unemployment are serious problems. In 2001, at the peak of the crisis, less than 2 percent of the population had per capita consumption under US$1 per day, little changed since 1994. However, inequality remains high, mainly reflecting extreme differences between regions, with the Southeast in particular much poorer than the rest of the country. The strong economic recovery has done little to reverse the increase in unemployment during the crisis (Chapter X of the Selected Issues paper). The official unemployment rate has risen from 6½ percent in 2000 to 10½ percent in 2003, and is especially high for educated youth, at almost 30 percent. Although the overall rate seems to be leveling off, this reflects declines in labor force participation—employment has actually fallen. With population growth in the 20-54 age group of around 2 percent per annum, generating a sustained recovery that lowers the unemployment rate will be a major challenge.

D. Remaining Vulnerabilities

12. Turkey’s economy is still vulnerable, and at the heart of this problem lies the magnitude of public debt, its short maturity, and currency composition (Table 3). Although the corporate sector has used the recovery to improve its financial health, its high leverage and unfavorable debt structure still leaves it vulnerable (Box 1). Banking sector balance sheets have improved, although much of this has come at the government’s expense. It has socialized many of the private sector’s risks, by taking on higher foreign currency and floating rate debt (Chapter III of the Selected Issues paper). Although the government has gradually lengthened maturities of new debt, average maturity remains low, so that rollover risk and vulnerability to increases in short-term interest rates are still concerns. And despite the shift to floating exchange rates, foreign currency risk is also a problem. Dollarization is extensive—around half the government debt and bank deposits are foreign currency denominated—and short-term external debt by remaining maturity exceeds gross international reserves.

Table 3.Turkey: Indicators of External Vulnerability, 2000–05 1/(In percent, unless otherwise noted)
Projections
200020012002200320042005
CPI inflation (end year)39.068.529.718.412.08.0
Public sector borrowing requirement (percent of GNP)18.921.112.09.87.03.7
Net debt of the public sector (percent of GNP)58.393.978.870.570.364.7
Export volume (percent change)6.616.611.716.314.810.1
Import volume (percent change)29.5-22.921.117.815.47.4
Current account balance, in percent of GNP-4.92.4-0.8-2.8-3.6-3.0
Capital account balance (in billions of US$)6.8-16.31.310.711.79.4
Of which: Foreign direct investment0.12.80.90.11.61.8
Foreign portfolio investment-5.2-4.6-1.21.22.31.6
Gross official reserves, in billions of US$ 2/23.219.828.135.233.927.6
In months of imports of goods and NFS4.04.45.45.24.03.1
In percent of broad money27.426.834.432.630.122.8
Gross total external debt, in billions US$118.7113.8130.9147.3149.6148.3
In percent of GNP59.079.072.161.853.251.2
In percent of exports of goods and NFS225.0218.8232.5205.5170.9156.3
Gross short-term external debt, in billions US$ 3/43.732.732.541.845.446.6
In percent of gross total external debt36.828.724.828.430.431.4
In percent of gross official reserves188.4164.7115.4118.8133.9168.8
Debt service 4/36.941.337.334.929.826.5
REER appreciation (CPI based, period average)11.1-18.29.58.9
REER appreciation (CPI based, end of period) 5/15.5-21.25.013.41.4
Capital adequacy ratio 6/17.315.325.329.6
State banks7.934.050.259.8
SDIF banks..-17.8-7.6-9.6
Private banks18.39.019.622.3
Foreign banks29.441.048.454.9
Nonperforming loans (in percent of total) 6/9.229.317.615.3
Real broad money, percentage change 7/0.811.2-3.3-6.67.45.1
Real credit to the private sector, percentage change 7/24.5-27.5-16.516.96.95.1
Banks’ net foreign asset position, in billions of US$-5.82.43.12.0
Banks’ net open exchange position, in billions of US$ 8/-5.4-0.1-0.4-0.3-0.5
Spread on Turkish dollar Eurobonds (in basis points) 9/800707693309463
Sources: Data provided by the Turkish authorities; and IMF staff estimates and projections.

For 2004-05, program projections.

As of June 18, 2004, reserves stood at US$34.0 billion (at actual exchange rates).

By residual maturity.

Interest plus medium- and long-term debt repayments as percent of current account receipts (excl. off. transfers).

For 2004, as of May 2004.

For end-2001 Pamuk Bank is treated as a private bank, for 2002 as an SDIF bank.

Deflated by the CPI.

For 2004, as of June 18, 2004

For 2004, as of June 21, 2004.

Sources: Data provided by the Turkish authorities; and IMF staff estimates and projections.

For 2004-05, program projections.

As of June 18, 2004, reserves stood at US$34.0 billion (at actual exchange rates).

By residual maturity.

Interest plus medium- and long-term debt repayments as percent of current account receipts (excl. off. transfers).

For 2004, as of May 2004.

For end-2001 Pamuk Bank is treated as a private bank, for 2002 as an SDIF bank.

Deflated by the CPI.

For 2004, as of June 18, 2004

For 2004, as of June 21, 2004.

Maturity of Newly Auctioned Debt, 2001-04 (in months)

Composition of Public Debt, 2003 1/

1/ Consolidated “budget debt stock” (central government).

Box 1.Corporate Sector Vulnerabilities—During and After the Crisis1

Turkish corporations were highly exposed to exchange rate and interest rate risk at the eve of the crisis. Similar to banks, corporations had significantly increased their foreign currency liabilities during 2000. Given that an increasing share of the domestic banks’ lending was going to the government, corporations relied heavily on borrowing from abroad. And a large share of their foreign currency debt, including that owed to domestic banks, had short maturities, even if it was not necessarily trade-related. In addition, corporations had signed (off-balance sheet) forward contracts with domestic banks, to which they were often closely linked, allowing these banks to meet the regulatory limits regarding their open foreign currency positions (banks reported a positive net forward position of US$9 billion by end-2000).

Foreign currency assets and liabilities of companies traded on the ISE
19992000200120022003 (Sept.)
In billions of US$
Assets2.12.33.14.14.4
Liabilities8.110.18.610.99.7
Net position-6.1-7.9-5.5-6.8-5.3
No. of companies217217217214208
Source: CBT, ISE (Istanbul stock exchange)
Source: CBT, ISE (Istanbul stock exchange)
External debt of non-financial private sector
19992000200120022003
In billions of US$
Short-term (original maturity)9.19.77.78.410.4
Medium and long-term22.624.624.327.229.5
Source: CBT, Balance of Payments data (International Investment Position)
Source: CBT, Balance of Payments data (International Investment Position)

The fall of the lira and the surge in interest rates created large financial losses in 2001, which the depreciation-induced export gains could not compensate.2 Corporations’ natural hedge (improved export competitiveness from a weaker exchange rate) showed some effect in 2001. However, the rise in exports was much less than would have been needed to offset the parallel collapse in the value of domestic sales. A lack of financing—as external credit was cut and domestic banks were in distress—is likely to have limited the full exploitation of export opportunities. Since the weak lira also implied a sharp fall in corporations’ domestic input costs, however, corporations managed to maintain positive operating profits during the 2001 crisis. Yet, the massive surge in financing expenses caused by a depreciated lira and rising interest rates turned these operating profits into net losses.

Corporations’ gross sales

(in billions of US$)

Corporations’ operating and net profits

(in billions of US$)

The financial health of corporations has much improved in tandem with the economic recovery. Profits swiftly recovered during 2002, and profitability ratios relative to capital, sales and financial obligations all improved, often showing stronger values than in 1999, before the crisis. Even with annual 2003 results not yet available for the larger sample, the further reduction in interest rates, the regained access to external credit, and the resolution of the banking crisis, should in principle have helped corporations to further strengthen their balance sheets. Yet, some 2003-data from larger companies traded on the ISE also indicates that the appreciation of the lira has lowered the margins of export-oriented companies.

Selected profitability ratios of Turkey’s corporate sector
1999200020012002
Relative to capital
Net Profit / Own Funds7.58.00.08.9
Net Profit / Total Assets 1/2.32.70.03.0
Relative to sales
Operating Profit / Net Sales3.82.53.94.8
Net Profit / Net Sales 1/1.21.00.02.0
Cost of Goods Sold / Net Sales90.593.091.087.5
Interest Expenses / Net Sales5.12.27.04.6
Relative to financial obligations
Profit Before Interest and Tax / Interest Expenses (Interest Coverage Ratio)146.8182.3103.0170.4
Net Profit and Interest Expenses / Interest Expenses123.8146.089.3143.1

Profits in 2001 were negative, but the profitability ratios are reported as zero.

Source: CBT

Profits in 2001 were negative, but the profitability ratios are reported as zero.

Source: CBT

Relatively high leverage ratios and the unfavorable structure of the debt continue to pose risks. Leverage ratios of Turkish corporations remain relatively high on average, leaving them with only a limited equity buffer against balance sheet shocks. Although it is estimated that between 10 to 20 percent of the funds recorded as debt on corporate balance sheet are funds provided by the owners and shareholders of the relevant company, it remains questionable to what extent such loans can provide an equity-like function in times of financial stress. Importantly, the short maturity of the debt financing and its extensive denomination in, or indexation to, foreign currencies implies that corporations still have considerable exposure to interest rate and foreign currency liquidity risk. Given that this composition of corporations’ debt is to a large extent simply the reflection of the banks’ need to create assets that balance their own foreign currency and interest rate exposures, the further strengthening of the domestic financial system will be key to moving such risks off the corporations’ balance sheets. This includes also a deepening of local capital markets and better access to equity financing.

Selected financial ratios of Turkey’s corporate sector
1999200020012002
(in percent)
Liquidity
Current assets / Short-term liabilities (Current Ratio)117.8121.1113.4118.5
Liquid assets (incl. marketable securities)/short term liabilities (Cash Ratio)25.725.623.525.4
Financial position
Total Loans / Total Assets (Leverage Ratio)69.466.571.366.5
Total Loans / Own Funds (Debt-to-Equity Ratio)226.4198.5248.2198.9
Short-Term Liabilities / Total Liabilities49.147.548.544.6
Short-Term Liabilities / Total Loans70.771.468.167.1
Source: CBT
Source: CBT
1 For an overview of Turkey’s sectoral balance sheets see Chapter III of the Selected Issues paper.2 The following analysis is based on annual aggregate corporate balance sheet data of over 8000 Turkish companies provided by the CBT (http://www.tcmb.gov.tr). End-2003 data will only be available later in 2004.

13. Recent financial market turbulence highlights these vulnerabilities. For much of 2003, the exchange rate appreciated and domestic interest rates and spreads on external borrowing declined sharply. But in May this year, the benchmark bond rate rose sharply, Eurobond spreads increased considerably, and the exchange rate depreciated by around 15 percent against the dollar. In part these shifts reflected changing expectations over the future path of US interest rates, and have been shared with other emerging markets. But in Turkey’s case the reaction has been more severe, reflecting the greater vulnerabilities in its economy, and also concerns over the widening current account deficit. Any exchange rate adjustment needed to boost net exports would also tend to exacerbate the debt burden.

III. Short-Term Macroeconomic Policy Challenges

14. The economy is on track to meet or exceed this year’s targets of 5 percent growth and 12 percent inflation, but faces problems of excess demand and a growing current account deficit (Figure 4). In the first few months of the year, growth of consumer credit, retail sales and imports have all been very strong (Figure 5). Indeed, it appears that domestic demand and growth have been running above the program targets, widening the current account deficit. In the staff’s view, current indicators pointed to growth well in excess of the target, the authorities thought any revision of the growth outlook for 2004 should await the first quarter outturn. (The first quarter national accounts released since the mission indicate that year-on-year growth in the first quarter was about 10 percent. An analysis of these data will be presented in the supplement to the staff report).

Figure 4.Turkey: External indicators, 2000-04

(in billions of U.S. dollars; unless otherwise indicated)

Sources: State Institute of Statistics; and Central Bank of Turkey.

Figure 5.Turkey: Recent Credit Developments, 2000-04

Source: Central Bank of Turkey.

15. Though closely monitoring the risks to the current account, the authorities felt that subsequent developments had mitigated these somewhat. Recent real exchange rate depreciation would depress incomes and make imports more expensive. Higher secondary market interest rates had been passed on by banks to their borrowers, and there were signs that the growth in consumer credit was slowing. The government had also cautioned state banks to curtail consumer credit growth, and had reduced tax incentives for new car purchases. And although it had worsened the current account, the oil price increase would, if fully passed on, reduce domestic demand by lowering real incomes.

16. Even so, it was not clear whether these factors would suffice to contain domestic demand and stabilize the current account. While a slowdown was likely, it would take time to gauge whether it would be sufficient. Financial market conditions had already started to improve again, so the effects of the earlier correction might dissipate. The staff therefore urged the authorities to let the automatic fiscal stabilizers work and to save the recent demand-driven over performance of tax revenues, at least until the risks to the current account had clarified. It was also important to accept any exchange rate adjustment, even if this might have the short-run effect of raising prices. Historically, the current account had been a bell-weather of the need for policy correction and needed to be watched closely. The authorities accepted much of this, agreeing to let the stabilizers operate, but wishing to review the situation later in the year to see if some of the budget over performance could be spent on social priorities.

IV. Sustaining and Advancing Recent Gains: Medium-Term Policy Challenges

Discussions took place against a backdrop of consensus on the key medium-term challenge: generating sustained and rapid growth with low inflation. Rapid growth would provide the most promising avenue for achieving lasting remedies to Turkey’s social challenges, especially unemployment. The government had been extremely successful in achieving macroeconomic stabilization, with fiscal consolidation supporting low inflation and growth, and this augured well for the future. The record was more mixed as regards the composition of the adjustment and the pace of structural reforms. While rooted in social concerns, spending initiatives inconsistent with the aims of the program and delays in the implementation of reforms raised questions about the sustainability of fiscal consolidation and growth. Looking ahead, the government needed to build on its record of fiscal consolidation by demonstrating greater ownership of structural reforms. The political conditions for doing this had rarely been better. Unlike the coalition governments of the past, the single party government had the opportunity to introduce fundamental reforms (such as in social security or public expenditure) where short-term (political) costs may be high, but where long-term benefits to the country would be substantial.

A. Reassessing the Medium-Term Macroeconomic Framework

17. Historically, Turkey has gained much from economic liberalization but paid dearly for policy indiscipline. Since 1980, greater openness to trade and financial market liberalization have helped boost economic growth. However, this growth has been extremely uneven, and prone to interruption by economic and political crisis. The problem is that liberalization has gone hand in hand with increased fiscal, monetary and financial indiscipline that has effectively negated its beneficial effects. This has depressed economic growth and made it more volatile.

18. With continued liberalization and a refocusing on microeconomic reforms, Turkey has the potential for sustained rapid growth. Although the current program’s 5 percent medium-term growth path has seemed ambitious, it is not that far above the 4 percent historical average. Indeed, in the medium term Turkey has the potential for even higher growth. If the emergence of greater political stability can be used to contain demands for greater spending and translated instead into less discretionary fiscal policy, and if the recent reduction in inflation can be sustained, growth should pick up considerably (Box 2). The analysis in Chapter I of the Selected Issues paper indicates that growth could then be as much as 2 percentage points above its historical average of 4 percent. This stable environment would also be conducive to TFP-led growth, contrasting with the investment-led growth of the past, which has often run into current account constraints.

19. Despite this potential upside, the baseline scenario assumes a more conservative growth path (Table 4). For the medium term, growth is still assumed to be 5 percent, with inflation staying in single digits. Domestic demand is assumed to slow, lowering the current account deficit to more sustainable levels. Despite the reform program and passage in 2003 of the Foreign Direct Investment Law, FDI has averaged less than US$1 billion (½ percent of GNP) each year, very low compared with new member states and countries acceding to the EU. Without such long-term flows, a current account deficit above 3 ½-4 percent of GNP in Turkey is unlikely to be sustainable, and would likely result in an exchange rate correction (Box 3). However, the staff stressed that achieving the desired current account adjustment and sustained low inflation would depend on tight fiscal policy and wage restraint, and success in containing the growth in domestic demand. If the authorities were unable to implement policies that delivered such a slowdown, a more abrupt exchange rate adjustment might result instead, with adverse implications for inflation and debt.

Table 4.Turkey: Medium-Term Scenario, 2000-09(percentage change, unless otherwise indicated)
2000200120022003200420052006200720082009
Real GNP6.3-9.57.95.95.05.05.05.05.05.0
Real GDP7.4-7.57.95.85.05.05.05.05.05.0
Real domestic demand9.8-18.59.39.36.74.34.35.15.45.4
Consumption6.3-9.12.55.66.43.83.94.64.64.6
Private6.2-9.22.16.67.54.44.35.05.05.0
Public7.1-8.55.4-2.4-3.0-2.00.00.00.00.0
Fixed investment16.9-31.5-1.110.012.28.88.08.09.29.2
Private16.0-34.9-5.320.317.59.08.08.09.59.5
Public19.6-22.08.8-11.5-2.68.08.08.08.08.0
Change in stocks 1/1.1-4.07.13.0-0.3-0.3-0.30.00.00.0
Exports19.27.411.116.012.68.48.98.38.08.0
Imports25.4-24.815.827.114.67.17.58.79.09.0
Contributions to growth7.4-7.57.95.85.05.05.05.05.05.0
Real domestic demand10.3-19.88.88.96.64.34.35.05.35.4
Net exports-3.012.4-0.9-3.1-0.70.70.80.0-0.4-0.4
Statistical discrepancy0.10.00.10.0-1.00.00.00.00.00.0
Savings and investment (percent of GDP)
Domestic saving19.419.320.720.219.420.320.920.921.021.3
Foreign saving 2/4.9-2.40.82.83.63.02.32.32.52.5
Gross investment24.317.021.523.023.023.323.223.223.523.8
Consumer prices
Period average54.954.445.025.311.810.36.44.53.53.0
End-period39.068.529.718.412.08.05.04.03.03.0
Nominal GNP (in TL quadrillion)125.6176.5275.0356.7417.6482.8539.2591.4642.4694.8
Nominal GNP (in US$ billion)201.3144.0182.7238.5
Sources: SIS; SPO; CBT; and IMF staff projections.

Contribution to the growth of GDP.

Balance of payments basis.

Sources: SIS; SPO; CBT; and IMF staff projections.

Contribution to the growth of GDP.

Balance of payments basis.

Box 2.Turkey’s Growth Record and its Determinants, 1960–20001/

Turkey’s relative economic growth performance has improved since the 1980s. Although during the last four decades real per-capita GDP expanded on average only slightly faster than in the rest of the world, this masks pronounced differences in growth performance during the two subperiods 1960–80 and 1981–2000. In the first two decades Turkey trailed the rest of the world but then surged ahead in the last two decades, when growth decelerated around the globe but not in Turkey. During 1981–2000, Turkey posted per-capita growth rates about one percentage point higher than in the rest of the world. And while Turkey did not escape the worldwide productivity slowdown, it was much less drastic than in the rest of the world. Notably, it managed to step up rates of capital accumulation when investment faltered elsewhere.

Growth Accounting, 1961-2000(Annual percent changes)
Per-capita GDP 1/TFPPhysical Capital 2/Human Capital 2/
1961-2000
Turkey2.330.921.770.40
Middle East2.751.271.710.41
Uppermiddle income countries2.450.971.760.40
All countries1.980.731.560.34
1961-1980
Turkey2.321.101.690.37
Middle East3.711.942.090.39
Uppermiddle income countries3.071.412.010.43
All countries2.621.071.920.33
1981-2000
Turkey2.350.731.860.43
Middle East1.790.611.340.43
Uppermiddle income countries1.830.531.510.36
All countries1.330.371.200.34
Note: The table covers 73 countries in our sample for which Bosworth & Collins (2003) provide physical and human capital stock data. GDP data are from the Penn World Tables (6.1), TFP is calculated as the residual, as described in the text.
Note: The table covers 73 countries in our sample for which Bosworth & Collins (2003) provide physical and human capital stock data. GDP data are from the Penn World Tables (6.1), TFP is calculated as the residual, as described in the text.

Turkey’s changing growth record reflects the influence of conflicting forces. Econometric analysis of a cross-country dataset covering 92 economies helps identify the factors chiefly responsible for the difference in growth experience in 1981–2000 and 1960–80. The main determinants turn out to be (i) trade openness, (ii) fiscal volatility, (iii) inflation, and (iv) financial development. This specification is robust to the inclusion of other potential explanatory variables, such as external volatility, government size, or measures of political constraints.

Increased trade openness was the main engine of the post-1980 growth performance. Between the two subperiods trade as a share of GDP rose from 13 to 38 percent—a much larger change than elsewhere. According to the regression results, this added some 2.3 percentage points to growth during 1981–2000, mainly through stimulating investment. Indeed, had trade and financial liberalization been the only policy changes, Turkey would have posted growth rates similar to those in east-Asia.

Growing fiscal volatility and inflation, however, acted increasingly as a drag on growth. After 1980 fiscal policy became more discretionary, as measured by unexpected, cyclically-adjusted deviations from the past fiscal policy stance. Even though this development was not associated with larger government, it still knocked 0.5 percentage points off Turkey’s annual growth rate, mainly by discouraging investment. Rising inflation reduced growth by another 0.7 points, mainly by lowering productivity.

Turkey’s recent advances in combating inflation augur well for the growth outlook especially if fiscal discretion can also be scaled back. With inflation rates firmly heading toward single digits one key ingredient for better future growth is already falling into place. While there is some evidence that fiscal discretion might be positively correlated with deregulation, and indeed trade liberalization, a strengthening of institutions, more reliance on rules, and adherence to pre-set targets appear to offer the best way forward.

1/ This box summarizes the main findings of Chapter I of the Selected Issues Paper.

Cumulative FDI Per Capita, 1996-03

(in U.S. dollars)

Source: IMF, World Economic Outlook

B. Fiscal Policy: Safeguarding Debt Sustainability

Fiscal adjustment to date

20. There was strong agreement that the strategy of targeting a high primary surplus had been critical in reducing debt and stabilizing the economy. This had helped improve policy credibility, supported the CBT’s disinflation efforts, and delivered lower interest rates. By alleviating worries over debt sustainability, fiscal discipline had also helped re-establish consumer and investor confidence.

Turkey: Summary of Primary Surplus, 1999-2003(In percent of GNP)
Current SBA
20022003Changes
199920002001Prog.Act.Prog.Act.2001-031999-2003
Consolidated public sector primary surplus-1.43.05.56.54.16.56.20.77.6
Central government1.44.64.85.42.45.15.10.33.7
Rest of public sector-2.7-1.60.81.11.71.41.10.33.8
Social Security0.00.0-0.10.00.00.00.00.10.0
Unemployment insurance fund0.30.60.30.40.30.3-0.30.0
Extra-budgetary funds-0.5-0.20.1-0.2-0.1-0.20.10.00.6
SEEs-2.1-1.50.11.11.11.20.70.62.8
Local Governments-0.2-0.20.10.00.10.00.0-0.10.2
Revolving funds0.00.10.10.00.10.00.10.00.1
Memorandum item:
Social security transfers3.52.62.93.04.13.34.51.61.0
Source: Fund staff estimates.
Source: Fund staff estimates.

Box 3:Competitiveness and the Current Account Balance

Turkish exports have performed well since the 2000 crisis, underpinned by a more competitive exchange rate, market diversification, productivity gains, and a reduction in real wages. However, the exchange rate appreciated as the economy recovered, imports accelerated and the current account deficit widened from a position of surplus to a projected deficit of nearly 4 percent GNP in 2004. This has ignited concerns that Turkey’s recovery is generating excessive external financing needs. In particular, the focus is on short-term debt financing and portfolio flows that may be curtailed by tighter international liquidity conditions, as well as longer-term considerations of debt sustainability.

Turkey: export market share, 1997-2003

(1997-100)

Turkey: Share of International Tourist Arrivals

(percent)

Short-term external financing has increased significantly, though not to levels seen before earlier crises. Short-term external debt rose 40 percent in 2003, though remains some 20 percent below its peak of $28 billion in 2000. While this undoubtedly poses a higher risk of a credit reversal and currency mismatch, the stock of trade credit is below 20 percent of annual import value, compared to more than 40 percent before the 1994 and 2000 crises. Similarly, banks’ foreign exchange credits are well below historical peaks and, more importantly, their on balance sheet open foreign currency positions have been maintained in the range of zero to US$2 billion and within prudential limits. Nonetheless, short-term external debt remains high by international standards in relation to gross official reserves, underlining the importance of continued reserve accumulation in coming years.

Over the longer term, the path of the current account deficit will need to be consistent with a level that will stabilize a prudent level of external debt to output. Starting from a debt stock of 60 percent of GNP, and 5 percent GNP growth, Turkey’s debt ratio will continue to fall gradually as long as the current account deficit stays below 4 percent of GNP. But to stabilize at more prudent debt/GNP levels of 40 percent or lower, the current account will need to fall below 3 percent of GNP. An increase in non-debt creating inflows would allow higher deficits. Rapid productivity growth consistent with EU convergence will also help lower the ratio of debt/GNP through the channel of real exchange rate appreciation.

What exchange rate adjustment would be needed to move Turkey to the long-run debt stabilizing current account deficit? Using estimates of trade responsiveness to exchange rate changes, we can calculate the “equilibrium” exchange rate that will deliver a debt stabilizing current account deficit when the Turkish economy and trade partners operate at full output potential. While this assessment is sensitive to the target current account balance, estimates of potential output gaps and assumed trade price and output elasticity’s it provides some indication of exchange rate divergence relative to the desired long-run equilibrium. Staff estimates show that the equilibrium real exchange rate has appreciated significantly since 2000, indicating that the Turkish economy has become more competitive, i.e. any given current account deficit can be sustained at a more appreciated real exchange rate than before. In addition, market forces have appreciated the actual exchange rate some 10 percent above the long run equilibrium rate. This divergence can be explained by factors such as a positive market assessment of future productivity gains associated with EU accession negotiations, higher real interest rates and the risk preferences of investors.

Turkey: Actual and Equilibrium Real Effective Exchange Rate, 2000-2004

21. While fiscal consolidation had been impressive, the quality of adjustment raised concerns. Staff analysis presented to the authorities showed that the fiscal effort since 2001 had been high and more enduring than previous episodes of adjustment in Turkey (Chapters VI and VII of the Selected Issues paper). However, the quality of the fiscal adjustment under the program had been mixed. Adjustment at the state enterprise level had been supported by increasing prices, lowering costs, and explicit targets for reducing employment. However, current spending at the central government level had risen, financed by tax increases and cuts in capital expenditure. This raised questions about the sustainability of the fiscal adjustment, especially as the international experience suggested that adjustments based on lasting reductions in expenditure were more likely to be sustainable (Chapter VII of the Selected Issues paper).

22. Short-term policy choices had at times also aggravated the quality of the adjustment. Unplanned spending initiatives outside the budget cycle, such as the pensions and minimum wage increases earlier this year, had undermined the budget process and required compensatory or corrective measures. And while corrective steps helped put the primary surplus targets back on track, they had often come at the expense of the quality of the fiscal adjustment. Investment, for example, had borne the brunt of budget cuts. Frequent ad hoc policy changes (proposals for selective tax reductions, incentives and tax amnesties) had also undermined the credibility of the budget process.

23. The authorities thought that the achievement in 2003 of the highest primary surplus on record had substantially enhanced policy credibility. They stressed that the markets no longer doubted the government’s determination to deliver on fiscal consolidation. Strong fiscal policy had also gained increasing acceptability in the population at large as it was credited with helping to restore growth and reduce inflation. However, the authorities emphasized that sustaining the fiscal adjustment necessitated paying sufficient attention to social concerns. The increases in pensions and the minimum wage had been necessary to restore them to their pre-crisis levels and maintain social consensus for reform. However, they agreed that such decisions should and would be considered and acted upon within the budgetary framework.

24. There has also been progress on structural fiscal reforms. A major improvement in fiscal management had been the approval after much delay of the Public Financial Management and Control Law which established a clear delineation of responsibilities in the preparation and formulation of the budget, introduced a medium-term budget framework, and established new systems for control and auditing of government operations. Tax administration reform had recently gained momentum; this should help with efficiency of collection and begin to tackle the unregistered economy. Indirect taxes had also been reformed: simplifying and reducing the number of taxes, and consolidating excises and VAT. However, the reform of direct taxes had fallen short of the original objectives.

Structural challenges to sustained fiscal consolidation in the medium term

25. Looking ahead, the authorities saw an overhaul of the taxation system to reduce distortions and fight tax evasion as a top priority. Although Turkey’s tax rates did not stand out as too high compared with the EU and the OECD, they were very high compared to other emerging markets. The personal income tax system was too complicated, with multiple rates and deductions. High social security taxes imposed a large cost on employment. The headline corporate income tax rate in Turkey was high, particularly compared to EU accession countries, and yet the effective yield was much lower because of generous investment tax credits and exemptions. The authorities were also concerned that VAT avoidance was high. Staff estimates indicated that the VAT yield was not out of line with international experience but collections accounted for about 70 percent of the calculated potential VAT.

Cross Country Tax Rate Comparison
VATCITPITSocial Security 1/
Turkey18, 8&130 2/20-45 3/38.5
OECD (average standard rates)17.8
OECD Europe19.9
OECD Pacific9.4
Poland22, 7&31919-4037.8-45.5
Hungary25&121820-4057.2
Bulgaria2023.518-2954.7
Romania192518-4045-62
Greece18,4&835
Croatia2220
Czech Republic22&531
Korea102710-409
Thailand7305-378.2-10
Chile19175-4010--12
Source: OECD, FAD databases and TA reports; VAT rates as of 2001; Turkey’s are as of 2004.

Including, employers, employees and unemployment insurance contribution.

Excluding the 10 percent temporary surcharge.

PIT on wage earners is 15-40 percent.

Source: OECD, FAD databases and TA reports; VAT rates as of 2001; Turkey’s are as of 2004.

Including, employers, employees and unemployment insurance contribution.

Excluding the 10 percent temporary surcharge.

PIT on wage earners is 15-40 percent.

26. However, the authorities acknowledged that any tax reform needed to take account of an expected structural decline in revenues. As interest rates come down, the yield of intermediation taxes and the withholding tax on interest income (more than 1 percent of GNP) will fall. Indeed, because of their distortionary nature, the government has rightly declared its intention to phase them out. Also dividends, turnover taxes, and profits from SEEs, which currently contribute some 1 percent of GNP to the primary surplus, will diminish as these firms are privatized. Of course, in the long run this should be at least in part offset by higher private sector tax receipts, but in the interim it would lower revenues. Finally, the government’s plans for greater fiscal decentralization posed a significant challenge to maintaining fiscal discipline. If not handled carefully, this could seriously worsen public finances in the medium term.

27. Public expenditure reform is also needed. Non-discretionary spending accounts for some three-quarters of central government budget spending, with social security transfers and wages making up two-thirds of primary spending. Despite an attempt at reform in 1999, central government social security transfers had continued to grow and were now close to 5 percent of GNP, in part because the principles of the 1999 reform, such as no ad hoc increases in benefits, had not been observed. Arguably, the program could have focused on social security reform earlier and more forcefully, but the authorities saw this as a sensitive area of reform which needed strong political will and consensus, which was only now being formed. The authorities also agreed that wage outlays were high by international comparison; however, they thought this could only be addressed gradually. In contrast, they thought that the real decline in public investment over the last three years needed to be reversed in the medium term.

Turkey: Indicators of Government Employment and Wages
General Government Employment (in % of population)Average Central Government Wage to Per Capita GDP 1/Ratio of Public to Private Sector WagesGeneral Govt. Wages and Salaries (in % of GDP)1/Central Govt. Wages and Salaries (in % of revenues)
Turkey3.22.41.010.232.3
EU10.5….….5.413.3
Central and Eastern Europe1.30.75.114.4
Latin America & Caribbean2.50.9
OECD 2/….1.60.910.1….
Low-income countries2.35.722.6
Middle-income countries4.36.022.1
High-income countries10.55.915.6
Source: FAD guidance note on civil service reform; most data are for the second half of the 1990s, whereas in Turkey it refers to 2003.

In the case of Turkey the ratio is to GNP

Includes 15 OECD and 12 EU countries.

Source: FAD guidance note on civil service reform; most data are for the second half of the 1990s, whereas in Turkey it refers to 2003.

In the case of Turkey the ratio is to GNP

Includes 15 OECD and 12 EU countries.

Avenues to sustained debt reduction

28. There was agreement on the need to elaborate a medium-term debt reduction strategy and an agenda of reforms intended to make consolidation sustainable. Structural reforms were needed to improve the underlying expenditure composition of the budget, improve tax collection, bring the unregistered economy within the tax base and eventually reduce the tax burden. These were essential to ensure that the adjustment of the last few years could be maintained in the medium-term. A medium-term debt reduction objective and strategy was needed to guide fiscal policy.

29. On the structural front, a large agenda needed to be tackled. Health and pension reform were clearly urgently needed given the widening social security deficit. There was also agreement on the importance of pressing ahead with tax administration reform to improve compliance and the efficiency of tax collection, with staff urging civil service reforms as an avenue to lower the tax burden. To reduce tax distortions, and pre-empt pressures to create new ones, the staff and the authorities agreed that work should begin on a major reform of tax policy, but careful attention had to be paid to maintaining the revenue base of the budget (technical assistance in this area has already been initiated). There was agreement that this agenda needed to be prioritized. The staff saw social security and tax administration reforms as priorities. Expenditure reform also needed to begin in the context of the 2005 budget, and the legal framework for decentralization carefully developed at an early stage. While broadly agreeing with this prioritization, the authorities also thought that tax policy reform needed to be tackled upfront but perhaps in a staged manner.

30. With public indebtedness at the core of Turkey’s vulnerability, debt reduction was seen as the necessary focal point of medium term fiscal strategy, but it needed to be articulated with care. Given the sensitivity of debt and fiscal variables to financial market conditions, legislative adoption of conventional rules such as a balanced budget rule, a borrowing rule, or a debt rule did not seem feasible at this stage. Instead, to guide fiscal policy over the next few years and ensure debt sustainability, the staff and the authorities saw merit in adopting a multi-year debt reduction plan to build government consensus, manage expectations, and guide policy. One option discussed was targeting a specific reduction in the debt (say 12–15 percentage points in the debt to GNP ratio over a four-year period) with an explicit commitment to adjust policies every year to achieve the targeted debt reduction. With a view to strengthening growth and limiting vulnerabilities to financing shocks, the medium-term aim should be to lower gross debt substantially, for example, to the average level of the recent EU accession countries (40 percent of GNP). If debt diverged from the target path, the primary surplus would be adjusted to achieve the targeted reduction in debt.

Gross Public Debt, 2003

(in percent of GDP)

Source: IMF, World Economic Outlook.

31. Staff noted that, for an unchanged pace of debt reduction, there did not seem to be room within such a framework for easing the primary surplus target. Recent market tensions and uncertainties over the current account deficit suggested that real interest rates and the exchange rate are subject to a great deal of uncertainty. Therefore, any debt reduction strategy, the staff thought, needed to build in substantial safety margins on the one instrument that the government does control: the primary surplus. Indeed, recent hints by some government members that the 6½ percent of GNP primary surplus target may be relaxed next year had already troubled the markets. A lower primary surplus should only be considered once greater credibility of fiscal policy has been established and real interest rates have been lowered on a sustained basis.

32. While not ruling out the 6½ percent target beyond 2004, the authorities thought there may be a trade-off between the quality and the size of fiscal adjustment. To meet the primary surplus target, government investment had declined in real terms and needed to be restored to ensure adequate infrastructure spending, particularly in health and education. Undertaking a major tax reform and ensuring continued decline in financial intermediation taxes would also be difficult, unless some decline in revenues was accommodated. After several years of fiscal consolidation there was also strong pressure from spending ministries for higher resources in priority areas.

33. The staff argued that a looser fiscal stance would slow debt reduction and could have a severe cost in the form of higher real interest rates. Indeed, surprising the markets by maintaining a stronger than expected fiscal stance was likely to be the most effective way of lowering real interest rates. Provided that the authorities were prepared to initiate expenditure policy reforms (FAD has provided TA to facilitate this), the staff presented a three year fiscal adjustment scenario that indicated it was possible to cope with the expected revenue decline, reverse the real fall in investment, and make room for a major tax policy reform within the constraints of the existing primary surplus targets (Table 5).

Table 5.Turkey: Public Sector Primary Balances, 2000-07
20002001200220032004200520062007
Est.Est.Est.Prog.Projections with Reforms
(In trillions of TL)
Public Sector3,7609,74111,15622,42027,26631,44234,53337,588
Central government 1/5,8318,4206,64417,95420,98524,27426,64829,497
Total revenue30,34644,76663,36990,445104,425116,306126,227136,738
Tax revenue26,51439,76859,63484,33597,927114,201124,320135,203
Direct taxes10,84916,08020,07727,80029,92033,47334,78237,257
Indirect taxes15,66523,68739,55756,53568,00780,72889,53897,946
Nontax revenue 1/5,4637,9169,40214,44616,62813,80315,31116,375
Tax rebates-1,632-2,918-5,666-8,335-10,130-11,697-13,403-14,841
Non-interest expenditure24,51536,34656,72672,49283,44092,03299,579107,240
Personnel9,98215,20423,16030,20035,10138,30040,41742,432
Other current1,1931,4312,5102,5483,4325,2795,5195,968
Defense and security2,3273,5914,4855,6685,8956,4916,8287,132
Transfers8,76312,21819,28926,91132,11332,82436,20839,617
Social security institutions3,3205,9109,94616,22819,46819,87421,56723,108
State enterprises and banks 2/1,2801,7772,2451,9711,5001,5001,5001,500
Agricultural subsidies3591,0331,8682,8053,2153,8014,8135,822
Other transfers3,8043,4995,2305,9077,9307,6498,3289,187
Investment2,2513,9027,2827,1656,8999,13810,60712,092
Rest of the public sector-2,0711,3214,5134,4666,2817,1687,8858,090
EBFs-225149-24939049000
Unemployment insurance fund3341,0989621,2281,7832,0102,2072,389
Local governments-28494338-1530000
SEEs-1,9201493,1392,4874,0994,6975,2175,241
Social insurance institutions24-170-85530000
Revolving funds 3/9995407461350461461461
(In percent of GNP)
Public Sector3.05.54.16.36.56.56.56.5
Central government 1/4.64.82.45.05.05.05.05.1
Total revenue24.225.423.225.325.024.223.623.5
Tax revenue21.122.521.823.623.523.723.323.2
Direct taxes8.69.17.37.87.27.06.56.4
Indirect taxes12.513.414.515.816.316.816.816.8
Nontax revenue 1/4.44.53.44.04.02.92.92.8
Tax rebates1.31.72.12.32.42.42.52.5
Non-interest expenditure19.520.620.720.320.019.118.618.4
Personnel7.98.68.58.58.48.07.67.3
Other current2.82.82.62.32.22.42.32.2
Transfers 2/7.06.97.17.57.76.86.86.8
Investment1.82.22.72.01.71.92.02.1
Rest of the public sector-1.60.71.71.31.51.51.51.4
EBFs-0.20.1-0.10.10.00.00.00.0
Unemployment insurance fund0.30.60.40.30.40.40.40.4
Local governments-0.20.10.10.00.00.00.00.0
SEEs-1.50.11.10.71.01.01.00.9
Social insurance institutions0.0-0.10.00.00.00.00.00.0
Revolving funds 3/0.10.10.10.10.10.10.10.1
Memorandum item:
Social Security deficit (percent of GNP)2.63.33.6 #4.54.74.14.04.0
Source: Turkish authorities; and staff estimates.Note: from end-2003 the figures include special revenues and expenditures. From 2004 the authorities have moved to the GFS 2001 classification.

Excluding privatization proceeds, transfers from the CBT, and interest receipts.

Excluding recapitalization of state banks; including net lending to the private sector.

Added to the public sector balance for 2002. Not included in the 2001 primary surplus calculation.

Source: Turkish authorities; and staff estimates.Note: from end-2003 the figures include special revenues and expenditures. From 2004 the authorities have moved to the GFS 2001 classification.

Excluding privatization proceeds, transfers from the CBT, and interest receipts.

Excluding recapitalization of state banks; including net lending to the private sector.

Added to the public sector balance for 2002. Not included in the 2001 primary surplus calculation.

C. Monetary Policy—Strengthening the Operating Framework

Sustaining the reduction in inflation

34. Reducing inflation has been central to the economic program. This was one of the key objectives of the original 1999 program. Lower inflation has become a gauge of the success of monetary policy and also the fiscal and structural reforms that support it. After the abandonment of the crawling peg, under the revised program the nominal anchor for reducing inflation has been base money. In both 2002 and 2003 the CBT met every one of its base money targets (Tables 6 and 7). As inflation has fallen and debt rollover concerns lessened, real interest rates have fallen dramatically. Disinflation has been associated with recovery rather than recession, and has been a major boost to the CBT’s credibility. Success in disinflation has also reduced exchange rate pass-through, and increased the significance of forward-looking rather than backward-looking inflation in price-setting behavior (Chapter II of the Selected Issues paper).

Table 6.Turkey: Monetary Aggregates, 2000-05(In quadrillions of Turkish lira)
2004
2000200120022003Mar.Jun.Sep.Dec.2005
Projected
Broad money (M2Y)56.8106.6133.7151.0153.8165.8174.5182.4207.1
Lira broad money (M2)31.947.261.982.792.697.1102.6106.9121.1
Foreign exchange deposits 1/24.959.371.868.361.368.771.875.586.0
Repos6.02.82.83.13.13.13.23.43.9
Broad liquidity62.8109.4136.4154.1156.9168.9177.7185.9211.0
Base money5.87.810.414.916.918.820.920.923.2
Net foreign assets 1/2.8-1.9-6.3-4.0-4.6-1.30.2-2.6-4.2
(in billions of U.S. dollars)4.1-1.3-3.9-2.9-3.5-0.90.1-1.6-2.4
Net domestic assets54.1108.5140.0155.0158.4167.1174.2185.0211.3
Net claims on government31.689.7122.8138.9141.4149.5154.8163.3184.2
Claims on business sector 2/31.738.742.058.059.961.564.367.677.0
Turkish lira claims22.323.124.539.742.844.546.649.055.7
Foreign exchange claims (est.) 1/9.415.617.518.417.117.017.818.721.3
Other items (net)-9.1-20.0-24.8-41.9-42.9-43.9-44.9-45.9-49.9
Memorandum items:(annual percent change)
Broad money (M2Y)40.287.525.413.016.327.826.720.813.5
Lira broad money (M2)42.548.031.033.745.143.236.929.313.3
Foreign exchange deposits 1/37.3137.921.0-4.9-10.611.014.510.613.8
Claims on business sector 2/73.022.38.438.427.329.926.016.513.8
(billions of U.S. dollars)
Broad money (M2Y)84.674.081.8108.2117.3108.1110.9112.6121.1
Lira broad money (M2)47.532.837.959.370.663.365.266.070.8
Foreign exchange deposits37.141.243.948.946.744.845.646.650.3
Net claims on government47.062.375.199.5107.897.598.4100.8107.7
Credit to the private sector47.126.925.741.645.740.140.941.845.0
(in percent share)
Base money/GNP 3/4.64.43.84.24.64.95.35.04.8
Broad money (M2Y)/GNP 3/45.360.448.642.342.043.544.043.742.9
Lira broad money (M2)/GNP 3/25.426.822.523.225.325.425.925.625.1
Private credit/GNP25.221.915.316.316.416.116.216.215.9
Foreign currency deposits/M2Y43.955.753.745.239.841.441.241.441.5
Money multiplier
Broad money (M2Y)9.813.712.810.29.18.88.48.78.9
Lira broad money (M2)5.56.15.95.65.55.24.95.15.2
Sources: Central Bank of Turkey and Fund staff projections.

Monetary authorities and deposit money banks; evaluated at current exchange rates.

Includes credit to local governments and state economic enterprises.

Evaluated as percent of nominal GNP over previous four quarters.

Sources: Central Bank of Turkey and Fund staff projections.

Monetary authorities and deposit money banks; evaluated at current exchange rates.

Includes credit to local governments and state economic enterprises.

Evaluated as percent of nominal GNP over previous four quarters.

Table 7.Turkey: Central Bank Balance Sheet, 2000-04 1/
20002001200220032004
Dec ActualDec ActualDec ActualDec ActualMar ActualApr ActualAug ProjectedDec Projected
1. Central Bank Balance Sheet(in quadrillions of Turkish lira) 1/
Net foreign assets3.3-12.73.98.88.09.212.810.7
Gross foreign assets15.628.637.742.540.541.242.538.4
Gross foreign liabilities12.341.233.833.632.532.029.727.7
International reserve liabilities3.120.711.59.88.38.05.83.5
Other reserve liabilities 2/4.910.113.214.514.914.914.914.9
Banks’ FX deposits with CBT4.310.49.19.39.39.09.09.3
Net domestic assets2.520.66.56.09.08.57.710.2
Base money5.87.810.414.916.917.820.520.9
Currency issued3.85.37.610.711.613.014.814.7
Banks’ lira deposits at the CBT2.02.52.84.25.34.85.76.1
(in billions of U.S. dollars)
CBT gross international reserves23.219.826.229.528.128.629.626.7
at current cross rates:19.828.135.2
CBT gross international liabilities18.328.623.523.422.622.220.719.2
CBT net foreign assets4.9-8.82.76.15.56.48.97.4
plus CBT forward position0.00.00.00.00.00.00.00.0
plus other reserve liabilities7.37.19.210.110.410.410.410.4
minus Dresdner one year deposits0.60.71.42.12.22.22.22.2
minus defence fund0.40.40.40.40.40.40.40.4
CBT net international reserves11.1-2.910.013.613.314.216.715.2
Treasury net international reserves 3/-1.3-14.7-14.2-14.1-14.6-15.2-15.8
Net international reserves (Treasury plus CBT)-4.2-4.6-0.5-0.8-0.41.5-0.5
(in quadrillions of Turkish Lira, program exchange rate)
Net foreign assets (Treasury)-1.8-21.1-20.4-20.3-21.0-21.8-22.7
Net foreign assets (Treasury plus CBT)-14.5-17.2-11.6-12.4-11.7-9.0-12.0
Net domestic assets (Treasury) 4/1.821.120.420.321.021.822.7
Net domestic assets (Treasury plus CBT)22.427.626.429.329.529.532.9
Base money (Treasury plus CBT)7.910.414.916.917.820.520.9
Exchange rate (TL per US dollar, in millions)0.671.441.631.401.311.44
Sources: Central Bank of Turkey; Fund staff projections. Although program targets for base money and NDA are five day averages, all observations in this table are end of period.

Except for 2000, all foreign currency aggregates are valued at end-December 2001 exchange rates (program exchange rates).

Mainly Dresdner deposit liabilities.

Equals borrowing from IMF plus short-term foreign currency denominated liabilities.

Since the Treasury cannot create base money, equals negative of Treasury net foreign assets.

Sources: Central Bank of Turkey; Fund staff projections. Although program targets for base money and NDA are five day averages, all observations in this table are end of period.

Except for 2000, all foreign currency aggregates are valued at end-December 2001 exchange rates (program exchange rates).

Mainly Dresdner deposit liabilities.

Equals borrowing from IMF plus short-term foreign currency denominated liabilities.

Since the Treasury cannot create base money, equals negative of Treasury net foreign assets.

Inflation Credibility Gap 1/

(In percent)

1/ Expected annual inflation minus CBT’s announced inflation target.

35. Despite this success, monetary policy faces a number of challenges, both short-and medium-run. In the short run, recent exchange rate depreciation is likely to add to inflation. So far, the pass through into prices from this adjustment has been modest and, with seasonally adjusted inflation having fallen to single digits over the last six months, the 12 percent inflation target should still be met. Higher oil prices could also threaten the target, though cuts in excises have so far limited the impact on domestic prices. However, the staff warned against using administered prices to meet the inflation target, especially given Turkey’s fiscal constraints and the need for relative prices to change as part of the adjustment mechanism. Instead, some one-off impact on the price level might need to be accepted, but any second-round effects onto higher wages and prices would need to be resisted.

36. The main challenge for the medium term is to put in place a more transparent monetary framework. With monetary targets starting to be missed, the program’s formal nominal anchor—base money—is showing signs of strain. The overshoots are in part benign, as they mirror the program’s success in reducing inflation and restoring confidence in the currency. But more recently they also seem to reflect rapid growth in credit and domestic demand. The CBT agreed that base money targeting was becoming problematic, but argued that this was of little consequence since in practice it had adopted an implicit inflation targeting framework instead. The CBT nonetheless still considered the adoption of formal inflation targeting as inadvisable because of continuing concerns over fiscal dominance, weaknesses in the financial system, uncertainties over the transmission mechanism, upcoming changes in the CPI basket, and uncertainties associated with EU accession.

37. The staff suggested that a strengthening of the monetary framework would help guard against the risks that concerned the CBT. For example, a strong monetary framework was needed regardless of EU accession process, even more so if this stalled. Fiscal dominance was a concern, but the track record of primary surpluses and debt reduction had arguably made this less so. The CBT’s credibility would only be boosted by being tested, not by waiting for the perfect moment to launch inflation targeting. The staff therefore encouraged the CBT to adopt more of the features of its ultimate objective—formal inflation targeting—through steps to enhance the transparency of monetary policy. These could include a greater role for the monetary policy council, and more predictable timing of interest rate decisions.

38. Foreign exchange intervention policy could be strengthened, with the goal being the steady accumulation of reserves. Frequent changes to the foreign exchange purchase auctions had confused the markets and undermined the CBT’s intervention framework. With reserve accumulation essential both to reduce vulnerability and to prepare for external debt repayments, the staff argued for a steadier and less reactive approach. However, the staff conceded that the occasional discretionary foreign exchange purchases of the last year had been extremely helpful in building reserves. The CBT explained that it had carefully gauged market developments prior to any change in its foreign exchange purchase auctions; however, market movements had recently became too volatile. They agreed, however, with the need for a more steady intervention approach in the future.

D. Foundations of an Effective Banking Sector

The strategy to deal with a severe banking crisis: how has it fared?

39. The key objective of banking reforms has been to address the long-standing weaknesses that were brought into the open by the financial crisis of 2000–01 (Table 8). Years of poor governance and excessive risk taking by banks, abuse by their owners, a weak regulatory and supervisory framework, lack of market discipline, and extreme macroeconomic volatility had left the banking system vulnerable. The financial crisis of late 2000 and early 2001, and the economic slowdown that followed, magnified these weaknesses. Banking sector profits and capital adequacy ratios fell sharply even before the full effect of credit losses was recognized. State banks, which had financed large “duty losses” (associated with politically directed lending) through short-term borrowing, experienced massive losses when short-term rates soared. Many private banks also had unviable balance sheets and a negative net worth.

Table 8.Turkey: Banking System—Selected Indicators, 1999-2004 1/(in trillions of Turkish lira)
1999200020012002200320032004
Dec.Dec.Dec. 1/Dec.Sep.Dec.Mar.
Banking System
Total assets79,763117,649179,675212,681225,097249,693255,974
Cash and claims on CBT5,0976,23512,55813,87213,86614,96214,941
Claims on other banks7,76613,59919,87115,40111,07615,14112,159
Securities portfolio22,04031,25170,02686,10596,129106,844113,159
Loans, net22,60135,78941,05852,93258,76467,21072,407
Other assets22,25830,77536,16244,37145,26145,53643,308
Total liabilities79,763117,649179,675212,680225,097249,693255,974
Deposits48,27268,143110,298137,973141,872155,312158,756
Borrowing from banks10,07015,99623,79821,96722,32925,91825,508
Repos7,64512,84310,7766,1617,28411,24110,698
Other liabilities9,53912,20720,52721,35121,07721,68322,134
Shareholders’ equity (incl. profits)4,2348,46114,27625,22832,53435,53938,878
Memorandum items:
Capital adequacy ratio (%)17.315.325.331.430.932.1
NPLs (%) total loans9.79.229.317.613.811.510.2
Provisions (%) NPLs61.959.847.164.277.588.589.8
Net profit (loss) after tax-305-888-9,9102,3364,4105,678921
ROA (%)-0.4-0.8-5.51.12.02.30.4
ROE (%)-7.2-10.5-69.49.313.616.02.4
Share in assets (%)100100100100100100100
Share in deposits and repos (%)100100100100100100100
Private Banks
Total assets42,16556,17997,930119,471125,925142,270142,621
Cash and claims on CBT2,7073,3628,4349,3569,1409,8689,488
Claims on other banks5,3869,76910,4947,6235,2556,3695,420
Securities portfolio14,33513,49127,14639,81944,47151,48550,174
Loans, net12,44519,58726,50635,75240,72946,40249,759
Other assets7,2929,97125,35026,92126,33128,14627,781
Total liabilities42,16556,17997,930119,471125,925142,270142,621
Deposits23,16030,82767,22380,62979,75888,18085,518
Borrowing from banks6,50810,04515,58513,70316,08018,15818,141
Repos4,6543,9181,8034,0744,3968,1037,870
Other liabilities3,0363,5035,7795,8716,9906,8727,902
Shareholders’ equity (incl. profits)4,8067,8867,54015,19418,70120,95823,191
Memorandum items:
Capital adequacy ratio (%)18.39.019.6423.523.524.7
NPLs (%) total loans3.53.527.68.97.76.56.2
Provisions (%) NPLs62.263.031.053.067.580.081.1
Net profit (loss) after tax1,6181,276-7,3832,4102,0242,917772
ROA (%)3.82.3-7.52.01.62.10.5
ROE (%)33.716.2-97.915.910.813.93.3
Share in assets (%)52.947.854.556.255.957.055.7
Share in deposits and repos (%)49.742.957.058.856.457.855.1
State Banks
Total assets27,10440,65557,58367,83176,14083,13489,335
Cash and claims on CBT1,9992,5883,5444,0004,2834,5894,912
Claims on other banks1,0881,6395,0963,9962,8095,3653,894
Securities portfolio4,6718,13932,75639,24544,30247,71655,104
Loans, net6,52110,0259,1778,80410,15812,20214,325
Other assets12,82518,2657,01111,78614,58913,26311,099
Total liabilities27,10440,65557,58367,83176,14083,13489,335
Deposits19,20427,60637,25848,48955,11559,86266,098
Borrowing from banks6301,3392,3812,2301,4722,3382,103
Repos1,9224,9493,8441,0229351,018897
Other liabilities4,3175,6599,7079,3439,63310,3429,825
Shareholders’ equity (incl. profits)1,0311,1014,3936,7478,9869,57410,411
Memorandum items:
Capital adequacy ratio (%)7.934.050.262.056.356.2
NPLs (%) total loans9.111.137.337.431.926.223.5
Provisions (%) NPLs35.130.36374869897.8
Net profit (loss) after tax284-177-6811,0561,4711,790182
ROA (%)1.0-0.4-1.21.61.92.20.2
ROE (%)27.6-16.1-15.515.716.418.71.7
Share in assets (%)34.034.632.031.933.833.334.9
Share in deposits and repos (%)37.840.233.934.437.636.639.5
SDIF Banks
Total assets5,48012,91211,0359,3106,9197,0756,825
Cash and claims on CBT2481034562435235
Claims on other banks211535874619307456545
Securities portfolio2,5118,5728,4514,6554,9094,9644,942
Loans, net1,0522,5336021,8891,042910655
Other assets1,4581,1691,0642,085618693647
Total liabilities5,48012,91211,0359,3106,9197,0756,825
Deposits5,3638,8273,5665,7704,1904,1333,994
Borrowing from banks2638192,0201,274479837837
Repos1,4263,9935,0231,0241,8532,0251,896
Other liabilities9681,3298142,3381,164927995
Shareholders’ equity (incl. profits)-2,540-2,056-388-1,096-767-847-897
Memorandum items:
Capital adequacy ratio (%)-17.8-7.6-16.6-21.6-25.6
NPLs (%) total loans61.941.467.369.458.153.828.6
Provisions (%) NPLs75.380.389.160.570.875.467.9
Net profit (loss) after tax-2547-2314-2,344-1,6773212726
ROA (%)-46.5-17.9-21.2-18.04.63.80.1
ROE (%)-153.0-41.8-32.1-0.7
Share in assets (%)6.911.06.14.43.12.82.7
Share in deposits and repos (%)12.115.87.14.74.13.73.5
Foreign and Investment Banks
Total assets5,0147,90313,12616,06816,11217,21317,194
Cash and claims on CBT143182535454400454506
Claims on other banks1,0821,6573,4083,1642,7062,9512,300
Securities portfolio5231,0491,6732,3862,4482,6802,939
Loans, net2,5833,6454,7736,4876,8357,6957,668
Other assets6831,3702,7383,5773,7233,4343,781
Total liabilities5,0147,90313,12616,06816,11217,21317,194
Deposits5468822,2523,0862,8103,1373,145
Borrowing from banks2,6703,7923,8124,7614,2984,5854,427
Repos-357-17105401009536
Other liabilities1,2181,7164,2273,7983,2903,5423,412
Shareholders’ equity (incl. profits)9371,5292,7304,3835,6145,8546,173
Memorandum items:
Provisions (%) NPLs31.351.881.269.384.885.586.0
Net profit (loss) after tax340328498548594698-39.2
ROA (%)6.84.13.83.43.74.1-0.2
ROE (%)36.321.418.212.510.611.9-0.6
Share in assets (%)6.36.77.37.67.26.96.7
Share in deposits and repos (%)0.31.11.92.22.01.91.9
Sources: Data provided by Turkish authorities; and Fund staff estimates

Data for December 2001 onward reflect the results of the audits conducted during the first half of 2002.

Sources: Data provided by Turkish authorities; and Fund staff estimates

Data for December 2001 onward reflect the results of the audits conducted during the first half of 2002.

40. To address these weaknesses, a comprehensive reform program was introduced. The staff reviewed with the authorities the reform strategy that essentially had two key components:

  • Enhancing the operating environment. To contain the immediate crisis and protect against panic driven deposit withdrawals, the authorities introduced a blanket guarantee of all bank depositors and creditors (blanket guarantee has been removed and replaced by a limited deposit guarantee scheme as of July 5, 2004). The independent Banking Regulation and Supervision Agency (BRSA), established too late to prevent the crisis, would bring the supervisory and regulatory framework closer in line with international standards. Prior to this, banking supervision had been fragmented between the Treasury and the Central Bank, and dominated by a semi-autonomous group of on-site bank examiners.

  • Strengthening the balance sheet of banks. State banks were recapitalized at a cost of close to US$30 billion, and a far reaching operational restructuring program was initiated. Their “duty losses” and overnight borrowing positions were eliminated; new management was installed; the housing bank Emlak was shut; 800 branches of Ziraat and Halk were closed and some 30,000 employees laid off; and a time table was developed for privatizing the state banks. Further, the core private banking system was strengthened: unviable banks were intervened; banks’ foreign exchange risk was reduced by the CBT’s sales of foreign exchange and by swapping Turkish lira for foreign currency denominated government paper; and the BRSA’s three stage audits forced banks to disclose NPLs and for their capital needs to be assessed uniformly. Recapitalization took place mainly by injecting private funds. Finally, to facilitate corporate debt restructuring, the bankruptcy law was overhauled and a voluntary framework for debt restructuring—the Istanbul Approach—was introduced.

Banking system profitability(in percent)
199819992000200120022003
Return on assets (ROA)1.9-0.4-0.8-5.51.12.3
Private banks2.83.82.3-7.52.02.1
State banks 1/0.51.0-0.4-1.21.62.2
SDIF banks-19.0-46.5-17.9-21.2-18.03.8
Return on equity (ROE)23.1-7.2-10.5-69.49.316.0
Private banks27.433.716.2-97.915.913.9
State banks 1/14.327.6-16.1-15.515.718.7
SDIF banks-153.0-32.1

State banks include Ziraat, Halk and Vakif

State banks include Ziraat, Halk and Vakif

41. The current and previous authorities, and the banking industry acknowledged that the above strategy was successful in substantially strengthening the system. Banking practices and the regulatory environment had gone through a sea change over the last few years. The banking system was properly capitalized, bank profitability had been restored and NPLs were falling. There had also been a consolidation in the banking system with the number of banks reduced from 80 to 49. Despite intervention in some 22 private banks, other private banks had increased their market share such that the role of state and SDIF banks in the banking system had gradually declined since 2000.

Turkey: Number of Banks

Turkey: Banks’ Market Share

Source: Data from the Turkish authorities.

42. However, there was also broad agreement that while the initial pace of reform was dramatic, sustaining it and preserving its achievements had been challenging. Although the regulatory framework had improved markedly, program conditionality had failed to recognize sufficiently early on deficiencies in the Banking Act and weaknesses in the judicial system. Former bank owners had been able to challenge BRSA decisions, long after their banks have been intervened and resolved, even in cases where the banks appear to have been clearly insolvent. Staff and board members of the SDIF and BRSA had insufficient protection against law suits for actions performed while carrying out their duty. As a result, supervisors were unwilling or reluctant to make decisions, and asset disposition had been slow (since staff are potentially liable if assets are sold at too low a price). Resolution of bad assets, had been very disappointing and slow and, the staff thought, not sufficiently transparent. Failure to split the Boards of the BRSA and SDIF upfront, and lack of experienced staff with banking experience were also seen by the banking community as contributing to the above problems.

43. Many in the banking industry also pointed to a number of remaining future challenges. Financial intermediation in Turkey was very low compared to other countries. This was in part because of distortionary intermediation taxes, which the government was committed to reducing but only gradually given its fiscal impact. Macroeconomic volatility and Turkey’s high debt also crowded out private lending in favor of government bonds. State bank assets consisted almost entirely of government securities, while the private banks also invested heavily in government paper. It was widely believed that Turkey was still over- banked and that further consolidation was inevitable, indeed desirable (see Chapter VIII of the Selected Issues paper). Increased foreign bank participation, which had already risen in the last three years, might facilitate this process. Further restructuring and eventual privatization of the state banks was another focus of discussions, with many arguing that the initial momentum of reform had been lost.

Low Financial Intermediation by Turkish Banks

(in percent)

Number of Branches per Bank

Tackling medium-term challenges

44. The authorities agreed that the unfinished agenda of financial sector reform needed to be tackled forcefully over the next few years. Although the 2000-01 financial crisis now seemed to have been successfully resolved, this came at the cost of a huge increase in the government debt, which had put a severe constraint on the conduct of economic policy and continues to raise concerns regarding debt sustainability. It thus needed to move more forcefully with a medium-term banking reform agenda. The staff and the authorities identified four key areas where further reform or an acceleration of reform was needed: state bank restructuring and privatization, the supervisory regime, asset resolution, and reducing financial disintermediation.

45. The mission stressed the importance of an ambitious but realistic strategy for moving the state banks to the private sector. The staff indicated that the excess capitalization of state banks, together with their subsidized funding base and their lack of a clear profit motive distorted the competitive environment with private banks. The authorities explained that while the state banks have been restructured and were now profitable, a realistic assessment of whether and how they can be privatized is only now being made in the context of the World Bank’s PFPSAL III loan and with the assistance of international consulting firms. They remained committed to privatization of Halk and thought that its integration with Pamuk, with its private sector culture, should facilitate this process. However, they thought that Ziraat’s privatization was a more long-term objective given its size, its reliance on government securities, and the social role it played by having branches in rural areas. The staff emphasized that an arms length relationship between these banks and the government was the best way of ensuring their effective restructuring in the long run.

46. While the supervisory environment had improved, the staff noted many shortcomings that needed to be addressed in the coming years. The ambiguous relationship between the BRSA and the Sworn Bank auditors (which have the sole right to conduct on-site supervision) was also a problem, preventing off-site supervisors or outside technical expertise being brought in to examine banks. This artificial split in responsibilities was one of the factors that in the staff’s view had contributed to the failure to detect the Imar bank scandal. The existing split in the supervision of banks (BRSA) and nonbank financial institutions (Treasury) remained a source of vulnerability, especially given the extensive interlinkages between corporates and financial intermediaries in Turkey. Supervision also needed to better enforce existing prudential requirements to ensure that banks have the appropriate risk management systems and capital cushions to withstand recent sharp movements in interest rates and exchange rates.

47. The authorities were confident that many of these supervisory shortcomings would be addressed in the new banking law currently under preparation (Chapter IX of the Selected Issues). They also saw the new law as an opportunity to clarify the responsibilities of the BRSA and the SDIF and provide adequate legal protection for their staffs in order to accelerate asset recovery. In this regard, the staff also noted that setting realistic price expectations and explaining these to the public, and transparent negotiations with former bank owners, were equally important. The authorities also noted that judging by recent court decisions, new legislation by itself might not be sufficient to ensure a predictable environment for banking. The new Banking Act will try to address some of the legal shortcomings that had led to recent court decisions against the BRSA but, ultimately, more far-reaching judicial reform may also be needed. However, this was an extremely sensitive and divisive issue that might best be undertaken with reforms needed to facilitate EU accession.

48. There was broad agreement that distortions in the taxation of the financial system needed to be reduced further. Intermediation taxes fundamentally undermined banking sector profitability, and encouraged banks to concentrate assets in government paper instead of extending loans. The taxes may also have reinforced dollarization and encouraged transactions to move off-shore where they can also escape regulation. The authorities emphasized that although the cost of these distortions was now more clearly recognized, fiscal constraints slowed their removal; nonetheless they remained committed to a phased reduction. There was also agreement that money and foreign exchange markets needed to be developed further in Turkey to facilitate improved risk management by banks. These markets were extremely thin in Turkey, with the CBT the primary counterpart to most transactions.

E. Sustaining Economic Growth by Strengthening the Investment Climate

49. The authorities have made some progress in improving Turkey’s investment climate. Achieving macroeconomic and political stability has been foremost among these. Together with structural reforms, the creation of independent regulatory institutions and increased transparency had also helped. In addition, the 2003 Foreign Direct Investment Law established the principle of equal treatment for foreign and domestic investors, allowed foreign investors to purchase real estate, and streamlined investment procedures. Other actions include: a new government procurement law which limits supplies of Turkish origin to a 15 percent price preference, governance reform (for both civil servants and SEEs), and the inaugural Investment Advisory Council (IAC) meeting in March 2004.

Factors Determining Investment in Turkey 1/(Percentage weight assigned to general categories)
CategoryWeight (in %)
Political environment25
Macroeconomic environment19
Labor17
Taxes/incentives17
Infrastructure12
Energy8
Research and development2
Total100

YASED (foreign investor association) survey results.

YASED (foreign investor association) survey results.

Turkey’s Place in International Competitiveness Rankings
RankingPercentile
UNCTAD FDI potential86 (out of 140)61
Hertiage Foundation Index of Economic Freedom106 (out of 152)70
PRS International Country Risk Guide101 (out of 140)72
World Economic Forum Global Competitiveness Report65 (out of 102)64
IMD World Competitiveness Yearbook55 (out of 60)92
Transparency International Corruption Perceptions Index77 (out of 133)58
Turkey (median percentile ranking)67
Bulgaria49
Czech Republic35
Hungary30
Poland40
Romania67
Slovakia40

50. Despite this, further reforms were needed to improve the business environment. Foreign direct investment is low, consistent with Turkey’s poor ranking in several widely- used measures of international competitiveness. To address these concerns, the Investor Advisory Council advocated several reforms to improve the investment climate, including accelerated privatization, a clearer judicial framework, less red tape, more efficient customs, and bringing standards and regulations in line with the EU. The staff noted that a key test will be to ensure these recommendations are prioritized and acted upon. It would also be important to stick to rule-based decision making and to deal with these defects directly, rather than trying to compensate for them through distortionary and costly tax incentives.

51. Progress in privatization has been disappointing. From the outset, the program has stressed the importance of privatization as a test of the authorities’ commitment to changing the way of doing business in Turkey. The focus of the program has been on bringing state enterprises to the point of sale (by improving management and developing privatization plans, often in conjunction with the World Bank) and on agreeing on realistic overall targets for privatization receipts, rather than requiring specific enterprises to be sold by a specific date. While the coalition government arguably failed this test, the new government’s greater enthusiasm for privatization has not shown much success. All of the large-scale privatizations of the last year—PETKIM (sale to the Uzan group fell through), TEKEL tobacco unit (price rejected as too low), and TÜPRAŞ (overturned by the local Administrative Court)—have failed.

52. Greater emphasis needs to be placed on judicial reform. In particular, problems have arisen with the scope of judicial review of actions of the BRSA and SDIF concerning the banking sector and, more broadly, in the area of bankruptcy and commercial proceedings. On several occasions, administrative courts have overturned BRSA decisions with respect to insolvent banks relying on broad judicial review powers. These court actions make it difficult for BRSA to fulfill its supervisory mandate. The lack of judicial predictability is now extended to the government’s ability to sell assets, bringing its privatization program into question. While the authorities have taken steps to address some of these problems in the recently adopted Bankruptcy Law and in the new draft Law on Credit Institutions, the authorities recognize the need for more fundamental judicial reform to support the functioning of Turkey’s market economy, and they are of the view that such reform would best be undertaken as part of EU accession.

53. Turkey will need to continue with tariff liberalization, particularly in the area of agriculture where barriers remain high. Key challenges will be to resist pressures to raise tariffs closer to their binding limits, but also to avoid overuse of trade remedies such as anti-dumping, where the number of cases has risen rapidly (Box 4). Addressing these challenges will help maintain trade flows and improve the investment climate. The scheduled removal of quotas on clothing and textiles imports in 2005 (in Turkey as well as the U.S. and EU) will represent a significant liberalization of Turkey’s import regime, and will be a challenge for low value added clothing exports to the EU and U.S. markets (Box 5).

54. The authorities are also taking steps to improve the quality and transparency of their operations. A new fiscal ROSC is likely to take place in 2005, once the Public Financial Management and Control Law is in effect. The authorities are also likely to request an FSAP in 2005, to take stock of the banking reforms made under the current program. Procedures against anti-money laundering had been reviewed twice (1994 and 1998) by FATF without any criticism and a third review is expected in 2006. A new law is expected to be passed before the end of the year to bring procedures in full compliance with new FATF regulations introduced last year. Macroeconomic statistics have some shortcomings, notably as regards coverage of the public sector, but are usually published in a timely fashion. Turkey subscribes to the Fund’s Special Data Dissemination Standard (SDDS).

Box 4.Trade Policy—Recent Developments

Turkey’s trade policies are framed by commitments under World Trade Organization agreements and, since joining the EU customs union in 1996, the acquis communitaire. This box highlights selected trade policy issues discussed with the Turkish authorities.1

Tariffs have remained stable since 1998. The tariff regime is generally liberal, with EU customs union and free trade arrangements with ECSC and EFTA covering about half of imports excluding basic agricultural products. However, agricultural goods, especially processed food, are highly protected owing to low agricultural productivity while tariffs on industrial products remain low. Average tariff protection increased slightly in 2004 reversing decreases in previous years. This likely reflected increased competitive pressures from a more appreciated exchange rate. Increases were focused on some agricultural products. The number of tariff spikes (lines with tariffs greater than 15 percent) continues to decrease. EU enlargement in May 2004 had no discernible on tariff preferences, as Turkey had FTAs with 8 of the 10 new members (all except Cyprus and Malta). The authorities have no plans for major changes of tariffs (or NTBs) in 2005.

Structure of MFN Tariffs in Turkey(percent)
199820032004
Simple average tariff rate12.411.812.6
Industrial products tariff rate5.34.44.2
Agricultural products tariff rate53.155.455.6
Tariff peaks (>15 percent, share of all tariff lines)16.715.514.7
Standard deviation of applied rates23.425.827.2

The use of anti-dumping measures is on the rise, with 51 measures in force in mid-2004, one third against China. Turkey is ranked seventh in the WTO in terms of anti-dumping cases initiated in 2003. The authorities explained that these measures represent cases of material injury to domestic industry, and goods affected include yarns and fibers and manufactures (lighters, tires and tubes, etc.). The expanded use of anti-dumping may also be indicative of competitive pressures from exchange rate appreciation since 2002. Safeguard legislation was amended in May 2004 to bring it fully in line with the WTO agreement on safeguards. Turkey also faces 19 measures against exports, mostly for iron and steel products. These measures were later removed.

Anti-dumping measures in force

The authorities have made considerable efforts to promote regional trade focused on buyers missions and trade fairs. With the passage of UN Security Council resolution 1483 in May 2004, trade sanctions against Iraq were lifted. Trade has expanded quickly, reflecting close trade linkages before sanctions were imposed and exports are expected to reach US$1.7 billion in 2004. With the border with Armenia closed trade flows are redirected through neighboring third countries.

Increasing Regional Trade

(Share of total trade with 12 neighboring countries, percent)
1 For an in-depth description of trade-related legislation see 1999 2000 2001 2002 2003 Trade Policy Review Turkey, 2003, World Trade Organization, WT/TPR/S/125.

Box 5.Textiles and Clothing—Implications of Quota Removal

The elimination of remaining U.S. and EU textiles and clothing quotas in January 2005 envisaged under the Agreement on Textiles and Clothing poses a significant challenge to Turkish exporters. 2003 Turkish textiles and clothing exports to EU and U.S. markets total US$11.4 billion (4.7 percent GNP) in 2003. Significant price or volume reductions would have macroeconomic effects as well as dis-proportionately affecting employment owing to the labor- intensive nature of clothing production. In the U.S. market, Turkey will gain on product lines where its exports are presently quota constrained. But in the EU, Turkey already benefits from quota protection and will inevitably face price declines and volume reductions in mass market segments.

Textiles and Clothing Exports, 2003

(US$ billions)

Lessons from the Phase III quota elimination in 2002 in the United States, which eliminated 15 percent of remaining quotas, suggest that gains (from elimination of constraining quotas) almost compensated losses (from unit price declines) for Turkey. U.S. imports from Turkey on products where quotas were removed fell 1 percent in value in 2002 against a total import value change of 2 percent, and rose 4 percent in volume. With 43 percent of 2003 U.S. imports from Turkey quota constrained (80 percent+ quota fulfillment), Turkey may well again hold its ground against low cost producers.

Change in U.S. Import Value 2002, for Clothing and Textiles Categories Where Quotas Eliminated

(in percent)

The challenge will be greater in the EU where Turkish exports are protected by quotas. Exporters will face price declines where they compete with hitherto quota constrained competitors, e.g. China, India, and Pakistan. Just over half Turkish exports of textiles and clothing are in quota constrained categories. Average price declines may amount to 5 percent, as experienced in the US in 2002. While price declines are likely to be much higher in the basic mass market segment, the effect is moderated by higher value lines less affected by quota removal (some 30 percent of Turkish clothing exports are in mass market segments). The extent of volume reductions will depend in large part on the cushion existing profit margins provide. As margins have been squeezed with TL appreciation, volume declines could amount to between 10 and 20 percent, or most of the mass market sector. Assuming textiles and clothing exports are similarly affected, the potential Turkish export loss with average prices down 5 percent and 10 percent volume declines amounts to US$810 million, and with a 20 percent volume decline some US$1,340 million. The higher end of these estimates corresponds to estimates of export losses of US$1,316 million calculated by the American Textiles Manufacturers Association. While this will be a significant setback in 2005 and may cut GNP growth by 0.3- 0.4 percent, the Turkish authorities believe that quota removal will accelerate the shift of production towards higher quality fashion items, higher margin own brand items, and that Turkey will retain the geographical advantage necessary for fast moving clothing items.

Turkey will also remove its own quotas on textile and clothing imports, which will lower prices in the domestic market. Quota constrained imports apply to 19 countries and amount to 13 percent of total textile and clothing imports. The authorities believe that quota removal will not have a destructive impact on the domestic market, while producer losses will likely be offset by consumer gains from lower prices.

V. Financing Outlook: External and Government Debt

55. The authorities were examining the financing outlook in an environment that was becoming more challenging. They were well aware that markets were increasingly focusing on the financing outlook beyond the expiration of the program. Turkey’s medium-term external financing requirements were challenging because of the high current account deficit and increasing external debt payments (Tables 9 and 10). A coherent medium-term financing and macroeconomic framework was desirable not only to respond to this outlook but also to prepare for uncertainties ahead. Rising world interest rates and the possible reduction of flows to emerging markets made the financing picture uncertain, especially given Turkey’s significant reliance on short-term external financing. While the authorities were hopeful of a positive EU accession decision, and in fact the markets had already priced in a favorable outcome, this could not be taken for granted and added to financing uncertainties.

Table 9.Turkey: External Financing Requirements and Sources, 2000 -08(In billions of U.S. dollars)
200020012002200320042005200620072008
Gross financing requirements39.939.037.445.949.949.849.352.856.9
Current account deficit (excluding official transfers)10.0-3.22.06.910.59.37.88.39.7
Amortization on debt securities (bonds)1.72.12.73.93.73.22.63.23.7
Of which:
Public sector1.42.02.33.83.72.82.53.13.6
Medium and long-term debt amortization (loans)13.814.313.415.015.514.514.715.315.8
Of which:
Public sector 1/3.63.63.03.23.33.23.02.92.7
Short-term debt amortization14.325.819.320.020.222.924.226.027.7
Public sector (net)1/-1.01.00.00.00.00.00.00.00.0
Trade credits 2/20.217.917.922.323.524.225.527.329.0
Banks and other private (net)-4.96.91.5-2.3-3.2-1.3-1.3-1.3-1.3
Available financing39.939.037.445.949.949.849.352.856.9
Foreign direct investment (net)0.12.80.90.11.61.82.12.53.1
Debt securities (bonds)3.4-1.74.27.97.68.18.38.38.3
Public sector7.52.13.35.35.05.55.75.75.7
Medium and long-term debt (loans)18.113.215.714.116.114.815.416.116.9
Of which:
Public sector 1/3.43.22.90.72.72.12.12.22.3
Short-term trade credits, currency and deposits17.913.216.022.925.725.227.028.930.7
Official transfers0.20.20.50.30.30.40.60.60.6
Other 3/-2.8-1.7-0.14.80.20.00.00.00.0
Net reserves (+/- = decrease/increase)3.012.90.2-4.1-1.5-0.5-4.2-3.7-2.6
Accumulation of gross reserves-0.42.7-6.2-4.01.36.37.3-1.9-1.8
IMF (net)3.410.26.4-0.1-2.8-6.9-11.5-1.8-0.9
Purchases3.411.312.51.71.80.70.00.00.0
Repurchases-0.1-1.1-6.1-1.7-4.7-7.6-11.5-1.8-0.9
Memorandum item:
Net public sector financing (incl. IMF, excl. reserves9.410.27.8-0.7-1.9-4.9-8.50.81.4

General government and Central Bank of Turkey.

Series reflects gross flows of short term trade credits, and stocks of credits to the banking sector.

Errors and omissions.

General government and Central Bank of Turkey.

Series reflects gross flows of short term trade credits, and stocks of credits to the banking sector.

Errors and omissions.

Table 10.Turkey: Indicators of Fund Credit, 2000–08 1/
200020012002200320042005200620072008
Outstanding Fund credit (end of period)
In billions of SDRs3.211.216.216.214.39.61.90.70.1
In percent of quota3331,1651,6851,6821,4831,000195687
In percent of exports of G&NFS82737322415310
In percent of GNP210121075100
In percent of public sector external debt72024242316410
In percent of overall external debt41316151410210
In percent of end-period foreign reserves1872756462521440
Repurchases of Fund Credit
In billions of SDRs0.10.94.91.23.25.17.81.20.6
In percent of quota79051012732853180612662
In percent of exports of G&NFS02112581121
In percent of GNP014123410
In percent of public sector external debt service193611273951147
In percent of overall MLT external debt service0522615233273
In percent of start period foreign reserves0532613224294
In percent gross public sector ext. financing 2/02221719394374
Net Fund Resource Flows 3/
In billions of SDRs2.57.74.4-0.7-2.5-5.4-8.1-1.3-0.6
In percent of quota259798455-73-258-556-844-133-64
In percent of exports of G&NFS61910-1-4-8-12-2-1
In percent of GNP2730-1-3-4-10
In percent of public sector external debt service338033-6-21-41-53-14-7
In percent of overall MLT external debt service164220-4-12-24-33-7-3
In percent start period foreign reserves144229-3-10-23-44-9-4
In percent gross public sector ext. financing 2/1819319-4-15-41-45-7-4
(In percent of total)
Public debt to preferred creditors/public debt13273231
Collateralized & securitized public debt/public debt353028293136444749
Debt service: public debt to preferred creditors/public debt8184421
Debt service: collateralized & securitized public debt/public debt283024403829255057

Projected on an expectations basis, except repurchase expectations on purchases made between February 2002 and July 2003 falling due in 2004-05 which are projected on an obligations basis.

Consolidated govt. and CBT. Includes reserve accumulation before repurchases.

Net purchases less repurchases and charges.

Projected on an expectations basis, except repurchase expectations on purchases made between February 2002 and July 2003 falling due in 2004-05 which are projected on an obligations basis.

Consolidated govt. and CBT. Includes reserve accumulation before repurchases.

Net purchases less repurchases and charges.

Table 11.Turkey: Social and Demographic Indicators, 1998-2002
Unit of Measurement199820002002
Natural resources
AreaThousand sq. Km769.63
Agricultural landPercent of land area31.4
Access to improved water sourcesPercent of population82.0
Human resources
Total populationIn millions65.1667.4269.63
Urban populationPercent of population64.965.866.6
Population growthPercent per annum1.771.691.59
Life expectancy at birthYears69.669.9
Infant mortality ratePer thousand live births38.035.0
Labor force (ages 15-64)Millions30.732.333.7
Health and education
PhysiciansPer 100,000 population120130130
Child malnutritionPercent of population under 5 years88
Immunization ratePercent of population under 12 months81.085.078.0
Primary school enrollmentPercent of school age group91.9
Secondary school enrollmentPercent of school age group73.3
Illiteracy rate (age 15 and above)Percent of population16.213.514.0
Income and poverty
GDP per capita (PPP basis)US dollars annually6,0006,2506,390
Share of top 20 percentPercent of income46.7
Share of bottom 20 percentPercent of income6.1
Population below poverty linePercent of population2.0
Sources: State Planning Organization; and World Bank, World Development Indicators.
Sources: State Planning Organization; and World Bank, World Development Indicators.

Gross Public External Debt Disbursements

(US$ billions)

56. The authorities financing scenarios pointed to a tight outlook. These scenarios were of course highly sensitive to underlying assumptions such as private market access, FDI, privatization receipts, real interest rates, the primary surplus target, and external reserves build up.

Gross Public External Debt Amortization

(US$ billions)

57. The authorities noted the role of increased net public sector external repayments—principally to the Fund—as a cause of the high rollover rates in 2005–06. These public sector debt repayments required both higher domestic rollover rates and a drawdown of official foreign reserves from present levels. The authorities argued that relying on external market borrowing to address these financing constraints would be difficult given Turkey’s sub-investment grade rating and tightening liquidity in global financial markets. They also noted the possibility of requesting an extension of payments falling due to the Fund on an expectations basis in 2006; this would postpone almost US$4 billion beyond the peak repayment period.

Net External Debt Disbursements

(US$ billions)

58. While the staff acknowledged that the financing outlook was challenging, it also noted that the Fund’s exposure to Turkey was extremely high (accounting for some one- quarter of total outstanding credit). There was a need, therefore, to plan on significant net repayments to the Fund in the coming years now that the financial crisis was behind us.

59. Against this background, the staff and authorities explored different avenues that could strengthen the financing situation. The staff emphasized that the amount of market financing that Turkey could expect would depend on the perceived quality of policies. Strong policies could act as a catalyst for potential capital inflows and also enable Turkey to tap international capital markets more aggressively than currently projected. Correspondingly, a higher primary surplus would reduce the public sector borrowing requirement; and a more aggressive privatization program might help alleviate financing constraints, though progress here remained slow. If market conditions improved, the authorities could also over borrow domestically, reinstate their pre-announced foreign exchange purchase auctions, and move aggressively to extend maturities. A positive EU decision on accession negotiations could also boost financing prospects, although this could not be counted on.

60. Given these complexities, it was agreed that discussions on relations with the Fund after the current program expires would need to continue. The authorities, however, stressed that no decision had yet been taken on whether to approach the Fund for a successor program, but expected a decision to be taken by September, in tandem with work on the 2005 budget.

61. While Turkey’s debt sustainability outlook has improved in recent years, risks remain (Appendix II—Debt Sustainability).

  • The standard debt sustainability analysis continues to show that sustainability is reached in the baseline, and is robust to individual shocks, but that it could be jeopardized by a combination of large shocks. However, external financing needs are rising on account of the larger current account deficits and increasing external public repayments.

  • Staff also presented nonstandard debt sustainability analysis to the authorities. Under a “sudden stop” scenario—which involves both the external and domestic sides—external capital flows decline, and the exchange rate depreciates to bring about the required adjustment in domestic savings. Real growth declines sharply, automatic stabilizers are allowed to operate, while real interest rates are driven higher as Treasury reliance on the domestic market increases. If the float is allowed to operate freely, and strict macroeconomic policies are preserved, the reserve position is safeguarded and external debt and reserve indicators improve modestly. The main impact would be expected to take place on the domestic financing side. Here, the debt burden increases, and would require strong policies to ensure that the situation remains manageable.

62. Against this background, the authorities emphasized that the Treasury continues to improve the structure of its debt. It has used favorable market conditions to lower the share of riskier foreign exchange debt from almost two-thirds at end-2003 to less than half, and has also lowered the proportion of floating rate instruments in its debt (see Chapter V of the Selected Issues paper). Consistent with its internal risk analysis, Treasury intends to further reduce its reliance on domestic foreign exchange borrowing and floating rate notes.

63. Nevertheless, Turkey’s debt situation poses risks. For a start, Turkey’s debt burden remains high, in a situation where there are concerns that emerging markets may have over- borrowed given their vulnerabilities (see Chapter IV of the Selected Issues paper). The debt structure also remains risky: average maturities are low, and the share of foreign exchange and floating rate notes remains too high.

VI. Staff Appraisal

64. Economic performance since the last Article IV consultation has been impressive. Growth has been sustained and rapid, while inflation has been lowered dramatically, reaching single digits. Large primary surpluses have reduced government debt ratios and improved market confidence. The result has been a significant reduction in real interest rates and an end to decades of continuous exchange rate depreciation. The economy has also become much more robust to shocks, with the Iraq conflict and the recent market turbulence having had only a limited adverse impact on economic performance.

65. Despite these considerable achievements, Turkey’s economy is still very vulnerable. At the heart of this vulnerability lies the high public debt. Its short maturity and large foreign currency component make Turkey vulnerable to interest rate and exchange rate shocks. Recent market turbulence and concerns regarding a widening current account deficit are a reminder of these vulnerabilities. Related to this, the quality of the fiscal adjustment and the pace of structural reforms raises questions about the sustainability of the fiscal consolidation and growth prospects.

66. The most pressing concern for the short run is that domestic demand may now be growing too rapidly. Although this year’s targets of 5 percent growth and 12 percent inflation should be met or exceeded, the authorities need to ensure that the rising current account deficit does not jeopardize macroeconomic stability. While these will help, it is not clear that the recent exchange rate and interest rate adjustment, the slowdown in consumer lending, and the reduction in tax incentives for new car purchases will slow domestic demand sufficiently.

67. The authorities need to be ready to take additional policy measures to contain domestic demand. With the inflation target on track and with fiscal dominance constraints, tighter monetary policy is unlikely to be appropriate. Instead, the authorities need to rely on fiscal and exchange rate policy. At a minimum, the authorities need to let the automatic fiscal stabilizers work and to over perform on the primary surplus targets, at least until the risks to the current account are clarified. Further tightening of the fiscal stance should not be ruled out if domestic demand continues to be strong. Incomes policy can also help, and while the authorities’ efforts to restrain government sector wages have been helpful here, repeated increases in the minimum wage have not. Finally, the authorities need to continue to allow the exchange rate to adjust to contain the current account imbalance.

68. To sustain growth over the medium-term, the authorities need to maintain their record of fiscal consolidation. Fiscal consolidation—the move towards sustained high primary surpluses and greater prospects of ensuring debt sustainability—has been key to the improved economic performance of recent years, and will also be key to its continuation. The reduction in real interest rates that resulted has boosted growth. And the lessening in concerns over fiscal dominance has given the Central Bank greater freedom to follow a monetary policy that supports disinflation. The authorities should resist pressures to relax the primary surplus targets as any benefits are likely to be far outweighed by higher real interest rates, which will slow growth. Indeed, with considerable public sector debt repayments coming due in the next few years, continued high primary surpluses are needed to help forestall potential financing problems.

69. Improving the quality and composition of the adjustment will be critical to sustaining this fiscal consolidation. Though the authorities deserve credit for their fiscal consolidation, too often they have had to introduce temporary revenue measures and cuts in investment spending to pay for large increases in current spending—pensions and the minimum wage. This has exacerbated short-run demand pressures and constrained the budget’s flexibility. Looking ahead, the authorities face an even steeper hill in sustaining this consolidation. As interest rates fall, revenues from intermediation taxes are set to decline; revenues from SEEs will fall as these firms are privatized; and plans for decentralization pose a risk to fiscal discipline. Although contribution rates are high, the social security system is in large deficit, and again faces the need for reform, while reform of tax administration is also needed to combat widespread tax evasion. To make room for increased capital and social spending in the coming years, the authorities will face the challenge of having to reduce current non-discretionary spending and to reform the tax system.

70. A less discretionary environment for fiscal policy would also be helpful. Too much attention has been given to ad hoc initiatives such as VAT cuts for selective sectors, tax holidays for preferred investments, amnesties for tax and social security arrears, adjusting excises to insulate the economy from changes in world oil prices, or increasing social spending beyond what had been agreed. This tendency towards discretion and short-run remedies rewards those that lobby for tax relief or directed spending, and undermines the long-term effectiveness and credibility of fiscal policy. Instead, the authorities need to consider a more fundamental reform of the tax system that reduces tax rates and simplifies the system for those that do comply.

71. To guide fiscal policy over the next few years and ensure debt sustainability, the authorities need to adopt a multi-year debt reduction plan. With financing conditions for emerging market countries such as Turkey susceptible to shocks, the medium-term aim should be to lower gross debt to pre-crisis levels, say, to the 40 percent average for EU accession countries. If debt dynamics diverge from the target, the primary surplus path would be adjusted to achieve the targeted reduction in debt. However, to maintain the pace of debt reduction under such a framework, there does not seem to be room to ease the 6½ percent primary surplus target. This should only happen once greater credibility of fiscal policy has been established and real interest rates have fallen on a sustained basis.

72. Both the CBT and the government deserve credit for reducing inflation to single digits. The CBT’s skilful use of monetary policy has been central to this success, but so has fiscal discipline and public sector wage restraint. To continue the disinflation process, commitment to medium-term debt reduction through sustained primary surpluses and wage moderation remain critical, including for minimum wages. The CBT’s independence also needs to be preserved and its monetary framework enhanced, by adopting more of the features of its ultimate objective—formal inflation targeting. As first steps, this could include giving the monetary policy council more of a role in advising the Governor on interest rate decisions, and timing these decisions to coincide with monetary policy council meetings.

73. Foreign exchange intervention policy could also be improved. Given the need to reduce vulnerability and to prepare for external debt repayments, the Treasury should have made greater use of last year’s more favorable market conditions to build up reserves. On the CBT’s side, a steadier and less reactive approach to the foreign exchange purchase auctions would have been less confusing to markets, and might also have limited the risks of a sharp exchange rate appreciation. Clearer mechanisms for sharing the costs of sterilizing foreign exchange purchases, including by the Treasury building up deposits at the CBT, could make intervention policy more effective.

74. Confidence in the banking system is largely restored but formidable challenges remain. Creation of an independent Banking Regulation and Supervision Agency, tightening of supervisory practices, introduction of the blanket guarantee, recapitalization and restructuring of the state banks, and the triple audit of the private banks have all strengthened the banking sector. However, weaknesses in the Banking Act and fundamental deficiencies in the court system have allowed former owners of insolvent banks to contest the loss of their banks and undermine the effectiveness of the supervisory regime.

75. Financial sector reform needs to be tackled forcefully. The first priority should be to monitor removal of the blanket guarantee, in particular developments in individual banks, and to have contingency plans ready in case weaknesses emerge. A realistic strategy for privatizing the state banks needs to be developed and, in the interim, distortions to competition with the private sector reduced. For the medium-term, the reform agenda needs to improve the environment for private banking. To improve banking supervision, the Sworn Bank auditors’ exclusive right to conduct on-site supervision should be removed, leaving the BRSA free to send off-site supervisors or outside technical experts to examine banks. To boost profitability and to encourage diversification of assets, a clear timetable for removing tax distortions in the financial system needs to be established.

76. Asset recovery and legal reforms also need to be pursued. The SDIF should announce a clear timetable for future loan auctions, otherwise bidders will not be interested, and more realistic asset valuations prepared and presented to the public. The recent takeover of assets owned by former bank owners will prove difficult to resolve and the SDIF will need outside assistance if it is not to be overwhelmed. Consideration also needs to be given to legal reforms. This should include passage of a new Banks Act, more closely in line with EU standards, with appropriate “fit and proper” criteria for bank owners, legal protection of BRSA and SDIF staffs and clear delineation of responsibilities between BRSA and SDIF. But more far-reaching judicial reform that ensures that judges who consider financial sector cases have sufficient expertise in the area and are aware of the systemic impact of the decisions they make, may also be needed.

77. To allow the private sector—foreign and domestic—to spearhead economic growth, Turkey needs to further improve the investment environment. While maintaining stable macroeconomic and financial conditions is critical here, other impediments also need to be tackled. Implementing rules fairly, strengthening the independence of regulatory institutions, and reforming the judicial system are all key to improving the business climate. The government should move ahead expeditiously to put in place the steps recommended by the inaugural Investment Advisory Council meeting in mid-March 2004. These reforms should also facilitate privatization of state assets, where the government’s commitment is encouraging, but which has been undermined by the legal system. To turn this commitment into reality, the government will need to persevere in its efforts, ensure impeccable preparation to avoid legal challenges, and be ready to accept realistic market prices.

78. Turkey faces an historic opportunity. Macroeconomic conditions are the best they have been in decades. The government needs to show that, unlike in the past, this period of stability and hope is not followed by complacency and reversal. Instead, it needs to seize this unique opportunity to reduce vulnerabilities; press ahead with fiscal, financial, and structural reforms; and thereby put Turkey firmly on an irreversible path towards convergence with European economies.

79. Finally, although the Fund-supported program has contained the key elements of success, there are also lessons for future program design. Adoption of the floating exchange rate regime, comprehensive banking system restructuring, and maintenance of a high primary surplus through relentless, and often difficult, fiscal adjustment have all been essential for success. However, with the benefit of hindsight several aspects of the program might have been better formulated or sequenced. Inflation targeting may have been promoted too early, given concerns over fiscal dominance. Arguably, social security reform could have been included in the program at the outset; even if it was thought at the time that the 1999 reform would be sufficient to bring the deficit under control. Deficiencies in the judicial process and the Banking Act were also not recognized early enough. Many of these issues will need to be addressed as the authorities move ahead with economic reform.

80. It is proposed that the next Article IV consultation be held on the standard 12- month cycle. However, this could be switched to a 24-month cycle if there is a successor arrangement.

APPENDIX I Turkey: Fund Relations

(As of April 30, 2004)

I. Membership Status: Turkey became a member of the Fund on March 11, 1947. It has accepted the obligations of Article VIII, Sections 2, 3, and 4 as of March 22, 1990.

II. General Resources Account:

Millions of SDRsPercent of Quota
Quota964.00100.00
Fund holdings of currency16,436.151,704.99
Reserve position in Fund112.7811.70

III. SDR Department:

Millions of SDRsPercent of Allocation
Net cumulative allocation112.31100.00
Holdings148.03131.81

IV. Outstanding Purchases and Loans:

Millions of SDRsPercent of Quota
Stand-by Arrangements15,494.551,607.32
First credit tranche90.389.38

V. Latest Financial Arrangements:

TypeApproval

Date
Expiration

Date
Amount

Approved
Amount

Drawn
In millions of SDRs
Stand-By02/04/0202/03/0512,821.2011,460.40
Stand-By12/22/9902/04/0215,038.4011,738.96
Of which: SRF12/21/0012/20/015,784.005,784.00
Stand-By07/08/9403/07/96610.50460.50

VI. Projected Payments to Fund (Expectations Basis)1

(In millions of SDRs; based on existing use of resources and present holdings of SDRs)

Forthcoming
20042005200620072008
Principal2,190.175,158.077,556.20595.4285.05
Charges/Interest432.87452.17164.309.830.78
Total2,623.045,610.247,720.50605.2685.83

Projected Payments to Fund (Obligations Basis)2

(In millions of SDRs; based on existing use of resources and present holdings of SDRs)

Forthcoming
20042005200620072008
Principal284.385,225.576,930.552,463.95595.42
Charges/Interest449.34547.50267.2244.469.53
Total733.725,773.077,197.772,508.41604.95

VII. Safeguard Assessments:

Under the Fund’s safeguards assessment policy, the Central Bank of the Republic of Turkey (CBT) is subject to a full safeguards assessment with respect to the SBA arrangement, which was approved on February 04, 2002 and is scheduled to expire on December 31, 2004. A safeguards assessment was completed on April 4, 2002 and staff’s findings and recommendations are reported in Section IV of IMF Country Report No. 02/137. The CBT authorities have implemented all of the measures recommended by staff.

VIII. Exchange Rate Arrangement:

For the period January 1, 2000–June 30, 2001, the lira was to have depreciated against a basket comprising US$1 and € 0.77 along a daily path pre-announced by the central bank. The preannouncement was for the 12-month period, and was updated quarterly. There would not be an exchange rate band around the pre-announced path during the first 18 months of the program. Thereafter—that is, from July 1, 2001—a symmetrical intervention band was to have been introduced around the central parity rate, with the total width of the band increasing gradually at a rate of 15 percentage points per year. This exchange rate arrangement was in place until February 22, 2001, when the government decided to float the currency.

Turkey has taken measures to implement United Nations sanctions that affect international payments and transfers and can give rise to exchange restrictions subject to Fund jurisdiction. In particular, Turkey has implemented financial sanctions against persons and entities associated with the Al-Qaida network, Usama bin Laden, and the Taliban. A list of the persons affected is published as Council of Ministers Decrees by the Ministry of Foreign Affairs. Turkey will notify these measures shortly in accordance with Decision 144.

IX. Article IV Consultations:

The 2002 Article IV staff report (IMF Country Report No. 02/137) was issued on April 4, 2002, and the accompanying Statistical Appendix (IMF Country Report No. 02/138) was issued on April 8, 2002. Board discussion took place on April 15, 2002.

X. ROSCs

Standard or Code AssessedDate of IssuanceDocument Number
Fiscal TransparencyJune 26, 2000SM/00/139
Corporate Governance prepared by the World Bank)December 11, 2000
Data ROSCMarch 14, 2002Country Report No. 02/55

XI. Technical Assistance: (1993–present)

DepartmentTimingPurpose
MAEJuly 1994Banking sector reform
MAEJuly 1995Inflation accounting
FADSeptember 1995Taxation of petroleum products
FADOctober 1995Assistance to IBRD Public Financial Managing Project; 8 FAD missions since 1994, assignment of 5 resident experts, mainly focused on customs modernization
STAFebruary 1997Balance of payments compilation
PDR/EU1/MAEDecember 1998Short-term debt monitoring
MAEJune 1999Basel Core Principles
MAEAugust 1999Debt management policies
MAEOctober 1999Banking sector reform
MAEMarch 2000Banking sector reform
FADApril 2000Fiscal transparency
FADApril 2000Tax policy
MAEApril 2000Banking sector reform
MAEMay 2000Banking sector reform
MAEJuly 2000Inflation targeting
STASeptember 2000Balance of payments statistics
MAESept. 2000–April 2001Banking sector reform
MAEApril 2001Debt management
FAD/STAMay 2001Fiscal accounting and reporting
MAE/RESSeptember 2001Inflation targeting
STAOctober 2001Data ROSC
STAApril 2002National accounts statistics
MAE/RESApril/May 2002Inflation targeting
STAJuly 2002Public finance statistics
FADJuly 2003Social security
FAD/MFDSeptember 2003Taxation of Financial Intermediation Direct Tax Reform
MFDDecember 2003Banking legislation
FADDecember 2003Informal sector and tax administration reform
MFDMarch 2004Currency reform
STAApril 2004Consumer and wholesale price indices
STAMay 2004National account statistics
FADMay 2004Public expenditure analysis
APPENDIX II Turkey: Assessing Debt Sustainability

This Appendix analyzes public sector debt and external sustainability and considers the impact of several shocks proposed as part of the debt sustainability framework. Although domestic and external debt sustainability can be achieved even under a range of plausible shocks, a combination of large shocks could jeopardize this. A nonstandard test involving a sudden stop of external capital flows would likely be manageable, as long as good policies remain in place. Any prolonged deviation from sound policies could, however, undermine Turkey’s public debt sustainability.

Public sector debt sustainability

1. Under the program’s baseline projections, Turkey’s public sector debt ratio declines substantially over the medium term (Table 12). Turkey’s public debt burden has fallen sharply in 2002-03 from its end-2001 peak of 94 percent of GNP. Helped by strong fiscal performance, declining real interest rates, the recovery of economic growth, and above all, by the appreciation of the real exchange rate, Turkey’s net public sector debt stock is estimated to have fallen to about 71 percent of GNP at end-2003. Although still high, public debt is set to decline even further under baseline projections—to 58 percent of GNP by 2009 (modestly higher than Seventh Review projections given the recent real exchange rate weakening).

Table 12.Turkey: Public Sector Sustainability Framework, 1999-2008I. Baseline Medium-term Projections
Projections
t-5t-4t-3t-2t-1tt+1t+2t+3t+4
1999200020012002200320042005200620072008
2Public debt/GNP (staff baseline)61.057.591.078.770.570.366.563.060.959.5
3Change in public debt/GNP17.3-3.636.5-15.1-8.3-0.2-3.8-3.5-2.1-1.4
4Net debt-creating flows/GNP (lines 5+8+11)-1.61.0-11.0-9.0-2.8-2.8-3.4-3.1-2.7
5Primary deficit/GNP2.0-2.4-5.6-4.1-6.0-6.5-6.5-6.5-6.5-6.5
8Minus net non-debt creating inflows/GNP-0.8-3.4-1.8-1.1-1.0-1.1-1.0-1.0-0.9
9Change in reserve money/GNP0.81.41.51.10.70.60.50.50.4
10Privatization receipts/GNP0.01.90.30.10.40.50.50.50.5
11(r-g-(pi+gpi))/(1+g+pi+gpi))debt/GNP (lines 13/12)1.610.0-5.1-1.84.74.84.14.44.8
12Adjustment factor: 1+g+pi+gpi1.61.41.61.31.21.21.11.11.1
13(r-g-(pi+gpi))debt/GNP (lines 14+15+16)2.614.0-8.0-2.35.55.64.74.95.2
14r (interest rate) times debt/GNP39.637.444.521.117.616.613.612.011.5
15minus g (real GNP growth rate) times debt/GNP-3.85.4-7.5-4.6-3.5-3.5-3.3-3.2-3.0
16minus (pi+gpi) (pi = GNP deflator, growth rate) times debt/GNP-33.2-28.8-45.0-18.8-8.5-7.5-5.6-4.0-3.2
17Residual, incl. change in assets/GNP (lines 3-4)-2.035.5-4.20.72.6-1.0-0.11.01.2
Key Macroeconomic and Fiscal Assumptions
Real GNP growth-6.16.3-9.47.95.95.05.05.05.05.0
Nominal GNP deflator (in local currency, change in percent per year)55.851.255.344.422.511.510.18.06.05.0
Effective nominal interest rate on government debt64.965.429.324.225.621.519.418.819.0
Benchmark bond rate106.238.099.163.544.129.027.724.622.922.9
memo: Ex-ante real rate (using GNP deflator) 1/36.4-11.237.933.529.217.218.217.617.017.0
Nominal TL rate on dollar debt112.132.77.36.221.413.912.412.412.4
memo: Nominal dollar rate on dollar debt8.08.08.07.07.07.07.07.07.0
Share of FX-linked debt (average)36.350.760.952.544.944.942.739.037.0
Effective real interest rate on government debt (using GNP deflator)6.114.55.511.414.012.512.613.113.3
A. Alternative Scenarios
A1. Key variables are at their historical averages in 2004-08 2/76.480.585.593.4103.7
A2. Primary balance under no policy change in 2004-0870.867.564.563.062.2
A3. Country-specific shock in 2004, with reduction in GDP growth (relative to baseline) of one standard deviation84.281.378.777.877.6
A4. Selected variables are consistent with market forecast in 2004-09 3/71.368.666.265.465.3
B. Bound Tests
B1. Real interest rate is at historical average plus two standard deviations in 2004 and 200576.987.893.793.894.9
B2. Real GDP growth is at historical average minus two standard deviations in 2004 and 200584.297.996.396.697.9
B3. Primary balance is at zero in 2004 and 200576.880.077.376.276.0
B4. Combination of B1, B2, B3 using one standard deviation shocks88.5112.3117.6119.4122.5
B5. One time 30 percent real depreciation in 200480.675.375.073.573.0
B6. 10 percent of GDP increase in other debt-creating flows in 200480.377.274.473.172.6
C. Tailored Tests
C1. Sudden stop: real depreciation, higher real interest rates, lower growth and smaller primary surplus. 4/70.379.575.876.979.4
Sources: Data provided by the Turkish authorities; Fund staff estimates; Consensus Economics Inc.

Real interest rate assumptions used for 2005 -2008 include safety margin.

The key variables include real GDP growth; real interest rate; and primary balance in percent of GDP.

Assumes growth of 5.0 % and primary balances of 5.5% of GNP over medium term. Real interest rates are assumed to be those in the program as consensus forecasts of interest rates are now close to those of program.

Assumes zero growth in 2005, 3% in 2006 and 7% in 2005; a primary surplus of 4.5% of GNP in 2005 and 5.5% of GNP in 2006, returning to 6.5% of GNP in 2007; a 15 percentage point real interest rate increase in 2005 and a 10 percentage point increase in 2006; and a 15% real depreciation in 2005 followed by a recovery in 2006-07.

Sources: Data provided by the Turkish authorities; Fund staff estimates; Consensus Economics Inc.

Real interest rate assumptions used for 2005 -2008 include safety margin.

The key variables include real GDP growth; real interest rate; and primary balance in percent of GDP.

Assumes growth of 5.0 % and primary balances of 5.5% of GNP over medium term. Real interest rates are assumed to be those in the program as consensus forecasts of interest rates are now close to those of program.

Assumes zero growth in 2005, 3% in 2006 and 7% in 2005; a primary surplus of 4.5% of GNP in 2005 and 5.5% of GNP in 2006, returning to 6.5% of GNP in 2007; a 15 percentage point real interest rate increase in 2005 and a 10 percentage point increase in 2006; and a 15% real depreciation in 2005 followed by a recovery in 2006-07.

2. For the purposes of the debt sustainability exercise, assumptions remain broadly unchanged from recent reviews. Specifically, it is assumed that annual economic growth will be 5 percent over the medium term, the primary surplus will reach 6.5 percent of GNP each year, and average real interest rates on domestic debt will remain at above 17 percent. (Real interest rates are currently about 15 percent.) The real exchange rate path is, however, slightly weaker, reflecting recent currency movements.

3. The standard alternative scenario tests continue to show that Turkey’s public debt burden is manageable under a variety of scenarios, but setting key parameters to their historical averages gives rise to an unsustainable debt burden over the medium term (Table 12, panel A).

  • Turkey’s public debt ratios continue to remain manageable under the standardized fiscal and growth shocks.

  • Using macroeconomic assumptions from consensus economic forecasts results in a decline in debt ratios. As a rough metric against which to test the staff’s projections, the Eastern Europe Consensus Forecasts and selected market analysis was used to simulate Turkey’s public debt path over the medium term. Although, when the conservative DSA interest rate assumptions are imposed, these are somewhat more pessimistic than staff’s baseline projections—with assumed primary balances of 5½ percent over the medium term—our simulated consensus projections still show Turkey’s debt burden declining over the medium term.

  • A repeat of Turkey’s historical economic performance would, however, leave the debt path unsustainable. With real GNP growth set at the average for the past ten years (2½ percent), a primary surplus of 3 ½ percent of GNP (about half of current targets), and the real benchmark bond rate set at 24 percent (about 10 percentage points higher than current levels), Turkey’s debt ratio increases to more than 100 percent in 2009. This scenario continues to show that a combination of poor fiscal performance, low growth and high interest rates would produce an unsustainable path for public debt.

4. The standardized bound tests show that, while Turkey would be in a position to cope with individual shocks, a combination of large shocks would undermine sustainability (Table 12, panel B). Individual shocks relating to fiscal balances, depreciation and contingent liabilities show manageable debt ratios over the medium term. However, the debt burden increases to more than 95 percent under both the interest rate and growth shock scenarios. As might be expected, the combined shock shows a clearly unsustainable debt position, with the debt burden rising to more than 125 percent of GNP over the medium term.

External debt sustainability3

5. The broad picture of external debt sustainability has slightly weakened since the Seventh review under the program principally reflecting a reversal of confidence (Figure 6). A 10 percent real depreciation relative to the program path during April-May 2004 as well as a modest increase of the projected current account deficit path has pushed up the debt-to-GNP ratio by some 5 percentage points of GNP to 53 percent, and modestly raised gross financing requirements.

Figure 6.Turkey: External Debt Ratio and Gross External Financing Need, 2003-08

Source: Fund staff estimates and projections.

6. External debt remains vulnerable to extreme standardized shocks. The historic volatility of macroeconomic indicators mean that the “standardized” bound debt sustainability tests are particularly demanding in comparison to most countries. Notwithstanding high historical vulnerability, the debt ratio is fairly robust over the medium term to two-year two standard deviation shocks to the current account balance, interest rates, and GNP growth, and to a once-off permanent shock of 30 percent nominal depreciation. More extreme combined shocks and two standard deviation shocks to the exchange rate would significantly test debt sustainability.

Tailored test for both external and domestic debt sustainability

7. A “sudden stop” tailored test illustrates the effect that temporary closure of international capital markets to Turkey might have on both external and domestic debt (Tables 12 and 13, lower panels, see footnotes for further details)4. In this scenario it is assumed that adverse global liquidity developments result in the temporary closure of public international bond borrowing in 2005 and a net outflow of short-term bank and private sector financing. On the assumption that much of the high frequency trade financing would be unimpaired, this would reduce total external financing sources by about a quarter.

Table 13.Turkey: External Debt Sustainability Framework, 2001-09(In percent of GNP, unless otherwise indicated)
ActualProjections
200120022003200420052006200720082009
I. Baseline ProjectionsDebt-stabilizing non-interest current account 7/
External debt79.072.161.853.251.146.244.342.741.4-0.8
Change in external debt20.1-7.0-10.3-8.5-2.1-5.0-1.8-1.6-1.3
Identified external debt-creating flows (4+8+9)19.2-15.8-14.80.3-0.8-1.4-1.3-1.1-1.2
Current account deficit, excluding interest payments-7.3-2.7-0.11.1-0.2-0.8-0.7-0.4-0.4
Deficit in balance of goods and services0.32.73.64.63.62.92.82.92.9
Exports36.931.430.331.433.133.433.634.034.3
Imports37.234.233.936.036.736.336.437.037.2
Net non-debt creating capital inflows (negative)-1.9-0.2-0.4-0.9-0.9-1.0-1.1-1.2-1.3
Automatic debt dynamics 1/28.4-12.9-14.30.10.30.40.50.50.5
Contribution from nominal interest rate5.03.52.92.72.92.82.62.62.5
Contribution from real GNP growth7.9-4.9-3.2-2.6-2.6-2.4-2.1-2.1-2.0
Contribution from price and exchange rate changes 2/15.6-11.5-13.9
Residual, incl. change in gross foreign assets (2-3) 3/0.88.84.5-8.8-1.3-3.6-0.5-0.5-0.2
External debt-to-exports ratio (in percent)214.1229.3203.7169.3154.3138.1131.9125.5120.6
Gross external financing need (in billions of US dollars) 4/40.337.439.853.756.261.555.759.762.3
in percent of GNP28.020.616.710-Year Historical Average10-Year Standard Deviation19.119.419.716.616.616.1Projected Average
Key Macroeconomic Assumptions
Real GNP growth (in percent)-9.57.85.93.36.75.05.05.05.05.05.05.0
GNP deflator in US dollars (change in percent)-20.917.024.02.317.512.3-1.92.32.42.52.53.4
Nominal external interest rate (in percent)6.05.65.36.00.45.25.65.96.26.26.35.9
Growth of exports (US dollar terms, in percent)-1.47.426.611.211.822.38.78.38.28.98.510.8
Growth of imports (US dollar terms, in percent)-22.315.930.412.622.725.15.16.37.99.28.310.3
Current account balance, excluding interest payments7.32.70.12.02.8-1.10.20.80.70.40.40.2
Net non-debt creating capital inflows1.90.20.40.70.50.90.91.01.11.21.31.1
A. Alternative ScenariosII. Stress Tests for External Debt RatioDebt-stabilizing non-interest current account 6/
A1. Key variables are at their historical averages in 2004-08 5/53.248.443.541.539.537.8-1.2
A2. Country-specific shock in 2005, one year “sudden stop” scenario53.255.347.043.341.941.0-1.9
A3. Selected variables are consistent with market forecast in 2004-0853.251.447.045.744.543.3-1.8
B. Bound Tests
B1. Nominal interest rate is at historical average plus two standard deviations in 2004 and 200553.251.847.345.443.842.4-1.8
B2. Real GNP growth is at historical average minus two standard deviations in 2004 and 200553.259.863.361.159.057.3-2.5
B3. Change in US dollar GNP deflator is at historical average minus two standard deviations in 2004 and 200553.274.6103.3100.196.994.4-4.1
B4. Non-interest current account is at historical average minus two standard deviations in 2004 and 200553.255.054.452.550.749.3-1.9
B5. Combination of 2-5 using one standard deviation shocks53.265.980.777.674.972.6-3.1
B6. One time 30 percent nominal depreciation in 200453.265.259.056.954.953.3-2.3

Derived as [r - g - ρ(1+g) + εα(1-r)]/(1+g+ρ+gρ) times previous period debt stock, with r = nominal effective interest rate on external debt; ρ = change in domestic GDP deflator in US dollar terms, g = real GDP growth rate, e = nominal appreciation (increase in dollar value of domestic currency), and a = share of domestic-currency denominated debt in total external debt.

The contribution from price and exchange rate changes is defined as [-ρ(1+g) + εα(1+r)]/(1+g+ρ+gρ) times previous period debt stock. ρ increases with an appreciating domestic currency (ε > 0) and rising inflation (based on GDP deflator).

For projection, line includes price and exchange rate changes

Defined as current account deficit, plus amortization on medium- and long-term debt, plus short-term debt at end of previous period.

The key variables include real GDP growth; nominal interest rate; dollar deflator growth; and both non-interest current account and non-debt inflows in percent of GDP.

The implied change in other key variables under this scenario is discussed in the text.

Long-run, constant balance that stabilizes debt ratio assuming that key variables (real GDP growth, nominal interest rate, and both non-interest current account and non-debt inflows in percent of GDP) remain at their levels of the last projection year. Dollar deflator is assumed to be zero.

Derived as [r - g - ρ(1+g) + εα(1-r)]/(1+g+ρ+gρ) times previous period debt stock, with r = nominal effective interest rate on external debt; ρ = change in domestic GDP deflator in US dollar terms, g = real GDP growth rate, e = nominal appreciation (increase in dollar value of domestic currency), and a = share of domestic-currency denominated debt in total external debt.

The contribution from price and exchange rate changes is defined as [-ρ(1+g) + εα(1+r)]/(1+g+ρ+gρ) times previous period debt stock. ρ increases with an appreciating domestic currency (ε > 0) and rising inflation (based on GDP deflator).

For projection, line includes price and exchange rate changes

Defined as current account deficit, plus amortization on medium- and long-term debt, plus short-term debt at end of previous period.

The key variables include real GDP growth; nominal interest rate; dollar deflator growth; and both non-interest current account and non-debt inflows in percent of GDP.

The implied change in other key variables under this scenario is discussed in the text.

Long-run, constant balance that stabilizes debt ratio assuming that key variables (real GDP growth, nominal interest rate, and both non-interest current account and non-debt inflows in percent of GDP) remain at their levels of the last projection year. Dollar deflator is assumed to be zero.

8. Such a shock would likely affect the economy through multiple channels: the exchange rate would weaken to strengthen domestic saving and pass-through effects would raise domestic inflation. Domestic interest rates (both nominal and real) would increase and growth would slow significantly. On the assumption that growth falls to zero, a 15 percent real depreciation (and higher nominal depreciation to offset higher inflation) would be required to bring about the required current account adjustment to offset the reduction in external financing sources. In subsequent years (2006 and 2007) growth recovers quickly, narrowing the output gap, financing flows return to the baseline scenario and the exchange rate depreciation is reversed.

9. The impact of the sudden stop on external debt dynamics is modest. This reflects both the forced reduction of external financing through the sudden stop itself, an assumed domestic adjustment and a quick reversal of the initial real exchange rate depreciation. Such an outcome depends crucially on both the parameters of the scenario and the authorities’ policy reaction. Notably, if short-term financing were to weaken even further, in a re-run of the 2001 crisis when net bank financing was cut by some US$10 billion, the required exchange rate correction would be even larger and the impact on debt sustainability correspondingly greater. However, providing the exchange rate adjusts promptly (in contrast to the earlier crisis) and does not overshoot, it seems reasonable to assume a much weaker reaction from private creditors, as the external financing situation would be relatively quickly be brought into balance.

10. The impact on domestic debt dynamics would be stronger, but likely remain manageable. With lower external public financing increasing Treasury’s reliance on shallow domestic financial markets, real interest rates would be expected to come under pressure, especially in the first year of the shock. Assuming automatic stabilizers are allowed to work, lower growth and the resulting temporarily lower primary surpluses also add to the domestic debt burden. The end result is that while public debt jumps to almost 80 percent of GNP by end-2006, it stabilizes at this level. Again, the impact of the sudden stop under this scenario depends critically on the policy response, especially on the maintenance of fiscal discipline.

APPENDIX III Statistical Information

1. Macroeconomic statistics have some important shortcomings, but are usually published in a timely fashion. Turkey is a subscriber to the Fund’s Special Data Dissemination Standard (SDDS).

Real sector data

2. Data on wholesale and consumer prices are published monthly, with a very short lag. Monthly data on various production indices are published with lags of some five to six weeks, while quarterly national accounts are published with a three-month lag (although the second and third quarters are more timely—a 70-day lag)). The State Institute for Statistics (SIS) publishes national accounts from the income side. Following technical assistance, an improved consumer price index and a new producer price index—replacing the old wholesale price index—are expected to be published in January 2005. Ongoing assistance is also helping to improve the quality of the national accounts.

3. There is a wide range of data on labor market developments, with the biannual Household Labor Force Survey (HLFS) replaced with a monthly survey at the beginning of 2000. These new data are published quarterly with a three month lag. Coverage of wage developments in the private sector has improved significantly through the use of quarterly surveys of the manufacturing sector.

Fiscal data

4. Budgetary data are published monthly, with a lag of some two to three weeks. However, the coverage of the budget is incomplete, with sizable fiscal operations conducted through extra budgetary funds, for which data are available only with long lags. Fiscal analysis is further complicated by the omission of certain transactions from the fiscal accounts; failure to account for sizable quasi-fiscal operations carried out by state banks and SEEs; and technical problems associated with consolidating the cash-based accounts of governmental entities with the accrual-based accounting of SEEs. The failure to assign a single agency responsible for fiscal statistics also creates reporting weaknesses. It is difficult to reconcile fiscal data with monetary and BOP data, especially in the accounting of external debt flows and central government deposits. Under the IBRD-financed Public Financial Management Project (PFMP), the authorities are to adopt an improved budget coding system, a chart of accounts, and a new debt management database.

5. Turkey reports fiscal data for publication in Government Finance Statistics Yearbook. The latest data available are for 2001 and cover the central government budgetary sector (including annex budget units). Data are not provided for extrabudgetary and social security units. No monthly and quarterly data have been reported for publication in International Financial Statistics.

Monetary data

6. Data on the central bank balance sheet, and provisional data on the main monetary aggregates and total domestic credit, are published weekly, with a one- and two-week lag, respectively. Data on the monetary survey and deposit interest rates are published monthly, with about a two-to-three-month lag. The CBT does not expect to meet the SDDS timeliness requirement for the analytical accounts of the banking sector in the short term due to delays in the preparation of year-end bank balance sheets and ongoing restructuring in the banking system. The CBT reports monthly data to STA with about a three-to-four-month lag.

7. STA and EUR use different measures of the monetary authorities’ net foreign assets, reflecting in large part EUR’s treatment in the program of central bank foreign currency- denominated liabilities to resident banks as foreign liabilities of the central bank. Differences also stem from the use of program exchange rates in EUR’s presentation, while market exchange rates are used in IFS.

External sector

8. In line with SDDS prescriptions Turkey disseminates

  • monthly BOP statistics with two to three months lag;

  • weekly international reserves with a one-week lag;

  • monthly data on the template on international reserves and foreign currency liquidity (reserve template) within one month after the reference period;

  • monthly merchandise trade data with one to two months lag;

  • quarterly external debt with one quarter lag; and,

  • annual international investment position data with a six months lag.

9. The central bank reports quarterly BOP data to STA with about four months lag.

10. The BOP is compiled in broad conformity with the conceptual framework of the BPM5. The CBT periodically reviews the bank’s foreign exchange records to redress, to the extent possible, problems of coverage and misclassification. However, the banks’ foreign exchange records need to be supplemented with estimation techniques and other sources.

Turkey: Core Statistical Indicators(June 30, 2004)
Exchange Rates 1/International ReservesReserve/Base MoneyCentral Bank Balance SheetBroad MoneyInterest Rates 2/Consumer Price IndexExports/ImportsCurrent Account BalanceOverall Government Balance 3/GDP/GNP
Date of latest observationJune 30June 30June 30June 30June 1June 30May 2004Mar. 2004Mar. 2004Apr. 2004Q1 2004
Date receivedJune 30June 30June 30June 30June 15June 30June 3May 31May 31June 15June 30
Frequency of dataDailyDailyDailyDailyWeeklyDailyMonthlyMonthlyMonthlyMonthlyQuarterly
Frequency of ReportingDailyDailyDailyDailyWeeklyDailyMonthlyMonthlyMonthlyMonthlyQuarterly
Source of dataCBT ReutersCBTCBTCBTCBTCBT ReutersSIS Press releaseSISCBTTRE Press releaseSIS Press release
Mode of reportingEmailEmailEmailEmailEmailEmailWebsiteWebsiteEmailFaxWebsite
Confidentiality4/4/
Frequency of publicationDailyWeeklyWeeklyWeeklyWeeklyDailyMonthlyMonthlyMonthlyMonthlyQuarterly

TRE = Undersecretariat of the Treasury; SIS=State Institute for Statistics; CBT = Central Bank of Turkey.

Interest rates for overnight interbank transactions and treasury bill auctions are available on a daily basis; data on deposit interest rates are provided on an end-week basis, with the most recent data available being for November 26, 1999.

Consolidated budget only; data on general government position is available only on an annual basis.

Daily data on international reserves and reserve money are provided to the Fund on a confidential basis.

TRE = Undersecretariat of the Treasury; SIS=State Institute for Statistics; CBT = Central Bank of Turkey.

Interest rates for overnight interbank transactions and treasury bill auctions are available on a daily basis; data on deposit interest rates are provided on an end-week basis, with the most recent data available being for November 26, 1999.

Consolidated budget only; data on general government position is available only on an annual basis.

Daily data on international reserves and reserve money are provided to the Fund on a confidential basis.

This schedule presents all currently scheduled payments to the IMF, including repayment expectations and repayment obligations. The IMF Executive Board can extend repayment expectations (within predetermined limits) upon request by the debtor country if its external payments position is not strong enough to meet the expectations without undue hardship or risk (see repayment schedules and IMF lending for details).

This schedule is not the currently applicable schedule of payments to the IMF. Rather, the schedule presents all payments to the IMF under the illustrative assumption that repayment expectations—except for SRF repayment expectations—would be extended to their respective obligation dates by the IMF Executive Board upon request of the debtor country (see repayment schedules and IMF lending for details). SRF repayment expectations are shown on their current expectation dates, unless already converted to an obligation date by the IMF Executive Board.

External debt in general refers to current, noncontingent claims by non-residents on residents usually in the form of loans, bonds, leases etc. For the purpose of this analysis it is assumed that: (i) securities issued abroad, e.g., Eurobonds, are held by non-residents; (ii) domestically issued securities denominated in foreign currencies are held by residents.

Previous tailored tests illustrate that Turkey’s public debt burden leaves it vulnerable to shocks, especially if policies are allowed to slip (see IMF Country Report No. 04/227, Appendix II, April 2, 2004).

Other Resources Citing This Publication