Information about Europe Europa
Journal Issue

Bulgaria: Selected Issues

International Monetary Fund
Published Date:
December 2007
  • ShareShare
Information about Europe Europa
Show Summary Details

II. An Assessment of Bulgaria’s External Stability Risks5

Core Questions and Findings

  • Is the present level of Bulgaria’s current account deficit sustainable? This is highly unlikely. Maintaining the current account deficit at about 20 percent of GDP—the projected level for 2007—would eventually drive the country’s external financial liabilities and future external repayment obligations to unsustainable levels. Estimates of Bulgaria’s equilibrium current account deficit as a share of GDP center around 8 percent, with a range of 5-10 percent reflecting different estimation methodologies and whether or not future EU capital grants are taken into account.
  • Is the present level of the real effective exchange rate significantly overvalued? A range of assessment methodologies, including estimates of the equilibrium real exchange rate, indicate no clear evidence of overvaluation. Moreover, export competitiveness, despite some recent weakening, seems to remain robust.
  • Does the large present current account deficit signal near-term external stability risks? Given the assessment of the present exchange rate level as broadly competitive, this is unlikely. Moreover, staff’s analysis points to a one-off private investment boom as the main driver of the large external imbalance. Once this boom tapers off, the current account deficit is projected to approach equilibrium levels under present policy settings.
  • What are the main downside risks to the staff’s baseline scenario over the medium term? Key risks are that present prudent policy settings are not maintained or that investors’ expectations about future risks and returns on Bulgarian investment projects undergo an abrupt change for the worse. Additional risks arise from a more prolonged investment boom or a fast pick-up in consumer goods imports following EU accession.
  • Does the structure of Bulgaria’s external balance sheet signal near-term external stability risks? Only to a moderate degree. With less than a third of external debt at short-term maturities, and these more than fully covered by international reserves, rollover risks remain low. Also, domestic foreign exchange liabilities, although rising, remain relatively modest at some 30 percent of GDP, and banks’ foreign-exchange exposure to the booming real estate sector remains limited.

A. Background

28. Bulgaria’s rapidly widening external imbalance is raising concerns about stability risks. The current account deficit more than tripled during the last three and a half years from 5 percent of GDP in 2003 to a projected 20 percent of GDP in 2007. Although the rising deficit has mostly been financed by foreign direct investment (FDI) inflows, the external debt stock has also surged to over 80 percent of GDP, notwithstanding rapid pay down of external public debt (Figure II.1). The large current account deficit and mounting external liabilities have raised concerns about Bulgaria’s external vulnerabilities, not least in light of the relative price adjustment constraints imposed by the currency board arrangements. Responding to the requirements of the 2007 Surveillance Decision, this chapter evaluates whether Bulgaria’s balance of payments position—as reflected in assessments of the current account balance, the real exchange rate, and the structure of the external balance sheet—are consistent with maintaining external stability (see IMF 2007a).

Figure II.1.Bulgaria: Current Account Deficit, FDI and External Debt, 2003-07

Source: Bulgarian National Bank (BNB).

Data for 2007 reflects rolling 12-month developments until August.

B. An Assessment of the Equilibrium Current Account Balance

29. This section estimates Bulgaria’s equilibrium current account (CA) balance using various approaches. The projected level of Bulgaria’s CA deficit in 2007—some 20 percent of GDP—is much higher than what would seem to be warranted by savings-investment fundamentals or external sustainability considerations. But estimates of Bulgaria’s sustainable current account balance range widely and are subject to uncertainties.

Macroeconomic Balance Approach

30. The macroeconomic balance approach estimates an equilibrium relationship between CA balances and a set of fundamentals that determine a country’s savings and investment positions using panel econometric techniques. The equilibrium CA balance (CA norm) for any individual country is then computed from this relationship as a function of the levels of fundamentals projected to prevail in the medium term. The gap between the estimated CA norm and the underlying CA balance, i.e. the current account balance stripped of temporary factors and adjustment lags, then gives a measure of disequilibrium in the CA. In Bulgaria’s particular situation, staff’s analysis shows that the ongoing FDI-driven investment boom, which has been the main force behind the expanding CA deficit (see section D of this chapter), mostly reflects a temporary phenomenon driven by a one-off reassessment of the country as an investment location (see Chapter I—Bulgaria’s Investment Boom: Drivers and Pay-offs). As such, the relevant underlying CA is deemed to be the medium-term CA balance which is reached with unchanged policies as FDI slows down to more sustainable levels.

31. The macroeconomic balance approach suggests a CA deficit norm of about 5 percent of GDP (Figure II.2). Using a panel of 38 industrial and European emerging market economies for the period 1992-2006, we estimate the CA norm as a function of fiscal balance, demographic variables, FDI, reserves assets, and energy balance. Based on these coefficients, Bulgaria’s CA deficit norm is calculated at little over 5 percent of GDP (Appendix 1). This estimate is similar to the one obtained from the CGER coefficients, which were estimated using a sample of 54 industrial and emerging market economies for the period 1973-2004.6

Figure II.2Bulgaria: Estimated Current Account Norm and the Projected Adjustment Path under the Medium-Term Scenario

Sources: IFS, WDI, BNB, staff estimates.

32. While these estimates provide useful benchmarks, caution is warranted in their application. For an economy such as Bulgaria, which is undergoing major structural changes and income catch-up, history may be of limited relevance. In addition, for EU members like Bulgaria, it can be argued that the receipt of annual capital grants in the range of 2-3 percent of GDP allows for a higher current account deficit over and above the estimated norm without posing sustainability concerns. This would imply that the estimated CA norm under the macrobalance approach could be as high as 7-8 percent of GDP.

External Sustainability Approach

33. From an external sustainability perspective, one would need to look at the implications of a rapidly deteriorating net International Investment Position (IIP). Like other new EU member countries, Bulgaria has experienced a sharp deterioration in its net IIP during the last few years as a result of large FDI inflows and borrowing by the private sector (Figure II.3). The projected income and interest payments associated with these inflows will need to be taken into account when deciding on the appropriate level of net IIP.

Figure II.3.Bulgaria: Net International Investment Position (IIP)

Sources: IFS and BNB.

34. The benchmark level of net IIP is an important element in the assessment of the CA norm, but this choice involves a difficult trade-off. While a lower level of net IIP clearly carries less external vulnerabilities and repayment obligations, a higher level may be necessary given a country’s development needs. In staff’s medium-term baseline scenario (Table II.1), despite a soft-landing, the financing needs remain sizable enough to cause the net IIP to deteriorate to -80 percent of GDP by 2012 from -65 percent in 2007. Staff considers this to be a viable medium-term outlook provided current policy prudence continues. While this level of net IIP is higher than the present average for non-industrial countries (-51 percent), the following two factors would justify a higher net negative IIP for Bulgaria: (i) the initial large needs for capital upgrading unique to a transition economy, particularly in the services sector, which has been the main force behind the rapid decline in the net IIP during 1998-2006; and (ii) EU accession, which has favorably reassessed the desirability of Bulgarian assets to foreigners. To stabilize the net IIP at this benchmark of-80 percent of GDP, the CA deficit norm would have to be at 8 percent of GDP or, including EU capital grants, at 10 percent of GDP.

Table II.1:Key Ratios and Assumptions Underlying the Baseline Medium-Term Current Account Projections
Key Ratios (in percent of GDP)20072012Adjustment
Current Account Deficit-20.2-8.311.8
Merchandise Trade balance-26.2-17.29.0
Non-factor Services balance4.48.64.2
Income Balance-0.7-1.7-1.1
Employees compensation4.34.50.2
Interest Payments2.14.01.9
Income payments4.34.0-0.2
Current transfers2.32.0-0.3
Private transfers receipts2.52.0-0.5
EU current transfers receipts1.51.4-0.1
Transfer payments1.71.5-0.3
Capital transfers1.32.00.7
Stock of FDI70.169.1-1.0
Stock of External Debt87.693.35.7
Net IIP-65-80-15
Gross FDI inflows/CAD77.460.7-16.7
Stock of Gross Reserves, billions of Euro11.720.18.4
Key assumptions (2008-12):
Exports volume growth13.2
Imports volume growth10.6
Growth in tourism receipts15.1
Rate of return on FDI equity investment10.4
Rate of reinvestment of FDI-related earnings30.0
Interest rate on FDI-related debt2.2
Implicit interest rate on other debt4.7
Average growth in workers’ remittances6.8
Average growth in employees’ income12.6
Bulgaria’s real GDP growth6.5
Foreign demand for imports growth4.3
Export volume elasticity with respect to income1.4
Export volume elasticity with respect to REER-0.5
Import volume elasticity with respect to income1.4
Import volume elasticity with respect to REER-0.3
Import volume elasticity with respect to FDI inflows0.2
(only applied to investment goods and parts of raw materials imports)
Cumulative real appreciation of leva (GDP deflator based)13.8

35. The above estimates indicate a wide range for Bulgaria’s CA deficit norm centering around 8 percent of GDP. Without adjusting for EU capital grants, the deficit norm ranges between 5-8 percent of GDP. However, taking into account the authorities’ projected receipts of annual EU capital grants of 2 percent of GDP, the deficit norm CA range can be adjusted upward to be between 7-10 percent of GDP.

Medium-Term (underlying) CA Balance

36. While the CA norm varies considerably, it is clear that a substantial reduction in the CA deficit is called for to ensure sustainability. Such a reduction is possible under the currency board regime (Figure II.2) provided (i) the investment and credit boom start to slow down over the medium term; (ii) strong fiscal prudence continues to neutralize revenue windfalls from the domestic absorption boom and pursues public sector wage growth broadly in line with the productivity growth; and (iii) strong buffers against external liabilities in the form of international reserves are maintained. A slowdown in the investment boom in the near future is anticipated given that all major privatizations have been completed, returns on investment have significantly come down in recent years, excess capacity exists in the tourism sector, and pressure on the labor force is visible in certain sectors of the economy.

37. Under this baseline scenario, the required ambitious adjustment in the CA balance is being driven by the trade and services sectors. More specifically, the following are the key drivers of adjustment (Table II.1.): (i) a slowing down in the growth rate of imports reflecting a slowdown in the pace of FDI inflows from their current very high levels (ii) a modest pick-up in exports volume growth benefiting from added refining capacity in the copper sector and a payoff from the large-scale FDI that has taken place during 2002-06 (euro 13.7 billion), almost half of which are in manufacturing and export-enhancing services sectors, and (iii) a strong pick-up in services exports, led by the tourism sector. However, one cannot rule out a more prolonged investment boom, in which case, continued large FDI inflows and credit boom could slow down the envisaged reversal in the CA deficit, saddling the economy with a much larger stock of external liabilities (Box 1).

38. There are substantial downside risks to this baseline scenario arising from the export side as well. If exports continue to rely heavily on sectors that are highly import dependent, it will hinder the needed fast turnaround in the CA deficit. The share of exports in labor- and resource-intensive manufacturing sectors remain high in Bulgaria at 81 percent compared with the average for other new EU members at 56 percent. One of the reasons behind this is that only a modest part of FDI, 10 percent of the total stock, has gone into non-resource intensive manufacturing sub-sectors. Investors seem to view cheap energy prices (particularly, electricity) as a key attraction for locating in Bulgaria while the shortage of workers with mid-level skills hinders investment in higher value-added sectors. The low share of FDI in manufacturing in general and even lower share in higher value-added sectors somewhat undermine the economy’s ability to bring in a quick turn-around in the CA deficit.

39. Additional downside risks come from a possible pick-up in consumption-driven imports in the aftermath of EU accession and changes in sentiment among investors. Imports growth in the first eight months of 2007 show an increase in consumer goods of 33 percent (y-o-y), compared with a growth rate of 22 percent experienced during the same period in 2004-06. Other risks to the current account recovery arise from a possible loss of investor confidence in the economy’s growth potential that could trigger a faster repatriation of profits, or from a more prolonged turbulence in the international financial market that could increase investors’ risk aversion and tighten credit conditions making debt servicing more expensive.

Box 1.A More Prolonged Investment Boom—Implications for the CA Adjustment

Assuming FDI inflows of euro 4 billion a year experienced during 2005-06 continue until 2011, the adjustment in the CA balance would take place at a much slower pace. It would take about 10 years for the overall deficit to reach within the sustainable range while net IIP deteriorate to -110 percent of GDP.

The basic assumptions are similar to those in the baseline scenario, (i) the current policy mix continues; (ii) real exchange rate appreciates in line with productivity growth of the economy as wage growth remains moderate; (iii) imports growth is mostly driven by raw materials and investment goods; and (iv) exports growth remain robust reflecting payoff from foreign investment and continued structural reforms. FDI inflows are expected to taper off relatively quickly after 2011 with the stock as a share of GDP peaking at 82 percent in 2011, then declining to 60 percent by 2018. As a bottom line, reserves accumulation is expected to ensure a coverage of 4 months of imports.

Box Table 1:Key assumptions underlying the scenario with a more prolonged investment boom
Key assumptions2008-2018
Volume of Exports Growth11.3
Volume of Imports Growth10.3
Cumulative real appreciation23.4
Real GDP growth5.8
GDP deflator growth3.9

Box Figure 1:Bulgaria, CAD and net IIP with a more prolonged investment boom

C. An Assessment of the Real Effective Exchange Rate

40. A large CA deficit is much less sustainable if accompanied by real appreciation that reflects a misalignment. As such, the 2007 surveillance decision also calls for an assessment of the level of the real exchange rate in addition to that of CA developments. This assessment is particularly relevant for Bulgaria given the fixed exchange rate regime with no change in the rate since July 1997.

Real Exchange Rate Developments

41. Based on a number of deflators, Bulgaria’s real effective exchange rate (REER) appreciation seems to be in line with that experienced by other new members of the EU (Figure II.4). Between 2000Q1-2007Q3, Bulgaria’s CPI-based REER appreciated by 37 percent compared with 27 percent experienced by other new members (excluding Bulgaria). However, the unit labor cost-based REERs have appreciated by much less at 5 percent (economy-wide) and 18 percent (manufacturing), which compare favorably against the average for EU new members at 32 percent and 29 percent, respectively.

Figure II.4.Bulgaria: Real Effective Exchange Rates, 1990=100

Source: Eurostat

42. Purchasing-power-parity and dollar wage-cost comparisons in industry across countries yield no strong evidence of overvaluation. Controlling for per-capita income, Bulgaria’s exchange rate path during 2004-06 as captured by the PPP-based equilibrium exchange rate measure suggests that the leva is still somewhat undervalued. Similarly, using dollar wages as a proxy for real exchange rate also shows no signs of misalignment (Figure II.5).

Figure II.5Bulgaria: PPP-based Equilibrium Real Exchange Rates

Sources: IFS; and Fund staff estimates.

Equilibrium Exchange Rate Approach

43. Given that the PPP-based equilibrium assessment have well-known drawbacks, we also estimate Bulgaria’s equilibrium real effective exchange rate (EREER). Earlier estimates of Bulgaria’s EREER show minor, if any, misalignment (See Égert 2005, Chovanov and Sorsa 2004). We estimate Bulgaria’s EREER for the period 1996-2006 using a model developed in Alberola and others (1999) and used in Burgess and others (2004) for the Baltics. This model estimates EREER as a function of relative productivity and the net foreign asset (NFA) position capturing internal and external equilibrium, respectively. Internal equilibrium implies clearance in the non-tradable goods market. Assuming that the Balassa-Samuelson hypothesis holds, if Bulgaria’s relative productivity in the tradable sector is higher compared to trading partners’ relative productivity in the tradable sector, one would expect the EREER to appreciate over time. External equilibrium is reflected in the clearance of the tradable goods market where any CA deficit is financed by a sustainable level of capital inflows characterized by the achievement of a desired NFA position. If the stock of NFA drops below this desired level, one would expect the EREER to depreciate ensuring an improvement in the trade balance to compensate for the lower foreign income.

44. Our estimation shows that the actual real exchange rate has appreciated largely in line with the estimated EREER. A vector error correction estimate produces the following long-run relationship between the variables:

EREER = 1.01* Relative Productivity - 0.06* NFA

As expected, the response of the equilibrium exchange rate to relative productivity is positive. A one percent increase in the relative productivity increases EREER by 1.01 percent. The NFA, however, enters the relationship with a negative sign contrary to the prediction of the model. This could be because of the relatively short time span of the data. Typically, studies that find a positive relationship between the NFA and real exchange rate include a much longer time series than the 11 years included in our estimation. However, this could also indicate that in countries like Bulgaria, where there has been an increased confidence in the growth potential in light of the EU accession as well as an upward revision in foreigner’s desired holdings of Bulgarian assets, the decline in the NFA position may actually reflect an equilibrium movement allowing for the coexistence of real appreciation and a declining NFA. The estimation shows that actual REER has mostly been in the neighborhood of its equilibrium value with a tendency of the deviation between REER and EREER to decline in recent years (Figure II.6).

Figure II.6.Bulgaria: Actual and Estimated Equilibrium Real Exchange Rates

Sources: IFS, BNB, staff estimates.

45. Based on the above analysis, there seems to be no strong indication that the real exchange rate is overvalued (Figure II.6). Actual real appreciation has remained contained in the range of 1-5 percent per year since 2000. Estimated equilibrium real exchange rate and the PPP-based comparisons indicate the exchange rate to be fairly valued.

D. Explaining the Disconnect Between CA Balance and Real Exchange Rate

Current account developments in recent years seem to be largely independent of real exchange rate developments, contrary to the assumptions made under the macrobalance approach. Assuming an elasticity of -0.39 for the CA balance with respect to the real exchange rate, the implied CPI-based real appreciation associated with a widening of 10 percent in the CA deficit (26 percent) would have been much higher than what we have actually observed during 2004-06 (9 percent).7 What explains this disconnect between the CA balance and real exchange rate?

Recent large CA Deficits—A Result of Capital Account Driven Private Investment boom

46. The widening of the CA deficit during 2004-06 has mostly been driven by increased imports, which in turn have surged on the back of rising investment goods, fuel and raw materials (Table II.2). In comparison, consumer goods imports, although increasing, played a modest role. The sustained rise in imports has coincided with massive levels of FDI.

Table II.2:Disaggregation of the Current Account Deficit, 2003-06
(As a share of GDP)2003200420052006Difference 2003-06
Current Account Balance-5.5-6.6-12.0-15.7-10.2
Trade balance-13.7-15.1-20.2-22.2-8.5
Exports (fob)37.840.844.147.79.9
Imports (fob)51.555.964.470.018.5
Imports (cif)54.458.567.073.619.2
Investment goods14.015.818.819.35.3
Raw materials21.823.524.626.95.1
Consumer goods8.49.710.510.92.5
Fuel and products9.69.913.816.16.5
Services balance3.
Income balance1.
Memorandum itemsAverage, 2004-06
Exports growth, volume11.512.310.511.310.2
Exports growth, value1019.718.626.618.7
Imports growth, volume18.313.815.31514.7
Imports growth, value14.520.326.925.224.1
Greenfield FDI inflows/GDP8.
Credit growth to the private sector48.348.632.424.635.3
Annual REER appreciation (in percent)6.91.9-0.262.6
Sources: BNB and IMF INS.
Sources: BNB and IMF INS.

Inflows from greenfield investment, which are often considered to have a higher impact on imports relative to privatization inflows, jumped to USD 3.7 billion per year during 2004-06 from an annual average of 0.9 billion during the previous five years. Continued high credit growth also contributed. To the extent the deficit is being driven by imports of investment goods and raw materials, one would not necessarily observe a large contribution from the real exchange rate appreciation to the widening of the CA deficit.

Despite Some Weakening, Exports Grew Strongly and Remain Competitive

47. Exports growth has remained solid albeit with a slight weakening vis a vis the earlier period and greater dependence on import-intensive sectors. Exports volume growth remained strong at 11 percent during 2004-06, above or at par with exports growth experienced by various emerging market country groups (Figure II.7). However, this strong overall growth masks two underlying trends. First, during 2004-06, exports growth depended much more on sectors that have high import content (iron, copper and petroleum products) compared to during 2001-03 when labor-intensive goods played a stronger role (Table II.3). This provides an additional explanation for the surge in imports besides the impetus provided by the FDI. In fact, metal ores and fuel imports contributed, on average, to 35 percent of total imports increase during 2004-06. Second, while this level of export growth is strong, it does show a slight slowdown relative to the growth experienced during 2001-03 (Figure II.7). What is also curious is that exports growth strengthened during the latter period in all emerging market country groups. This relative slowdown, with respect to both the earlier period and other emerging market countries, poses some concerns for competitiveness despite overall increase in market shares and low wages in manufacturing (Figure II.8).

Figure II.7.Bulgaria: Volume of Exports Growth, 200-06

Sources: WEO and BNB.

Figure II.8.Bulgaria: Export Competitiveness

Sources: Direction of Trade Statistics and ILO.

Table II.3:Sectoral Contribution to Exports Growth
(In percent of total increase in exports)2000200120022003200420052006
Consumer goods217662511479
Clothing and footwear166134337-13
Raw materials and fuels772831746780
Iron and copper29-18-235361035
Petroleum Products21-18-435182424
Investment goods2213018122611
Machinery and vehicles21215103165
Memorandum items:
Share of iron, copper and fuel in42-10-3712273741
total imports
Total exports growth1519461349605131714812546
(in millions of Euro)
Sources: BNB and staff calculation
Sources: BNB and staff calculation

48. To look into export competitiveness more thoroughly, we use Constant Market Share Analysis (CMSA), which, despite notable weaknesses, allows us to gain some useful insights into the anatomy of a country’s exports growth. Specifically, it allows us to analyze if Bulgaria’s exports growth was due to gains in competitiveness, or driven by increased demand in particular export markets or for particular commodities (see Appendix 2 for methodology). We compare Bulgaria’s export performance during 2001-03 and 2004-05 with that of all low- and middle-income countries (LMC) based on product disaggregation at SITC 1-digit level and market disaggregation into four broad groups: EU-15, non-EU high-income countries, low- and middle income countries in Europe and all other low- and middle income countries.

49. The CMSA shows that Bulgaria has experienced some decline in competitiveness in recent years albeit coming on the back of large gains during 2001-03. Bulgaria’s exports grew considerably more rapidly than those of other LMC during 2001 -03, and based on CMSA, this higher export growth was mostly due to gains in competitiveness (Table II.4). As shown by the market growth effect, if Bulgaria’s exports were to grow at the overall growth rate of other LMC, it would have only grown by 52 percent of actual exports growth experienced during 2001-03. This “overperformance” relative to other LMC during 2001-03 is mostly explained by gains in competitiveness, and partly also by higher orientation toward faster-growing markets. For example, 27 percent of Bulgaria’s exports in 2000 went to LMC of Europe, the market group in our analysis absorbing imports at the fastest rate, compared to 7 percent of LMC’s exports going to this market. Bulgaria’s export “overperformance” relative to LMC sharply declines during 2004-05 from 48 percent during 2001 -03 to only 8 percent with a corresponding decline in the contribution of competitiveness to overall export growth. However, overall Bulgaria remains positively competitive compared to other low- and middle-income countries.

Table II.4:Constant Market Share Analysis of Bulgaria’s Export Growth
(In percent of total increase in exports)2001-20052001-20032004-2005
Market Growth Effect675292
(implied “overperformance”)(33)(48)(8)
Commodity Composition Effect0-1-2
Market Distribution Effect12134
Competitiveness Effect21356
Source: UN Comtrade Database and staff calculation.
Source: UN Comtrade Database and staff calculation.

E. An Assessment of the External Balance Sheet

50. Despite a large external debt stock, the composition of the debt and strong buffers would suggest modest near-term stability risks. More specifically, the following factors would mitigate near-term risks that could arise from the structure of a country’s external balance sheet: (i) FDI constitutes more than half of the stock of total foreign liabilities; (ii) 70 percent of the external debt stock are in long-term liabilities, half of which are owed to mother companies in the form of inter-company debt; (iii) the prudent macroeconomic policies pursued since the 1996/97 crisis have resulted in large stock of foreign exchange reserves which more than fully cover all short-term liabilities (remaining maturity) as well as more than half of M3; and (iv) domestic foreign-exchange denominated liabilities remain relatively low at 30 percent of GDP, with limited exposure to the housing sector.

51. However, there are reasons to worry. An external debt stock that is above 80 percent of GDP would be considered risky and high by all conventional benchmarks. For example, recent estimates show that, the conditional probability of a debt crisis rises from 2-5 percent to 15-20 percent when debt reaches above 40 percent of GDP. 8 Although this probability is somewhat dampened by the high share of exports in Bulgarian economy, large debt continues to pose risks to the outlook. Similarly, while the risks of a sudden stop remains low in light of Bulgaria’s relatively low level of foreign-exchange denominated domestic liabilities, should such a stop occur, the growth impact would be significant, a reduction of 8½ percentage points of GDP in the first year after the shock (See Sorsa and others, 2007).

52. Developments in recent years have also increased potential currency and roll-over risks. The share of domestic lending in foreign currency has increased rapidly for both corporate and household sectors, which are at 66 percent and 20 percent, respectively as a share of outstanding loans in these sectors. The exposure is high for the corporate sector where only a quarter of the foreign-currency lending has been channeled into the tradable sector creating concerns for potential currency risks. The roll-over risks have also increased in recent years as the stock of short-term debt (remaining maturity) has increased from less than 9 percent of GDP in 2003 to 24 percent with the composition less in favor of trade credits whose share in the short-term debt has declined from 43 percent to 23 percent.

F. Structural Reforms and External Stability Risks

53. In the absence of any monetary policy and likely political economy constraints for further fiscal tightening, progress on the structural front will be crucial. This would be required not only to increase the economy’s ability to absorb shocks but also to maintain investor confidence, and ensure a productivity-driven growth. Despite low wages compared to other new member states, and a low labor force participation rate, it is disappointing that only a modest share of FDI has gone into manufacturing. Similarly, the structure of exports remains concentrated in relatively low-valued added products undermining the economy’s ability to shake off the imports drag. A simple correlation between share of manufacturing FDI and various structural reforms across a group of new members of the EU shows strong relationships with labor productivity, domestic market size, the ability to provide quality infrastructure and the ease of employing workers (Table II.5). While one must interpret such correlations with high skepticism, given Bulgaria’s small domestic market, it will be important to identify key reforms that are likely to attract FDI into higher value-added manufacturing sectors.

Table II.5:Manufacturing FDI, Possible Determinants
Manufacturing FDI/TotalGDP, in billion USDLabor productivity (PPP GDP per person employed relative to EU-25)Employing workers, Doing Business Indicators 2006 1/Enterprise reform, EBRD Transition Index 2/Infrastructure, EBRD transition country Index 2/
Czech Republic40.1142.50.6850.981.03

Shows ranking among Europe and Central Asian countries, higher score indicates less reform.

Scores as a ratio of EU-10 average, higher score indicates more reform.

Data sources: The Vienna Institute for International Economic Studies FDI database, WEO, Eurostat, EBRD, World Bank.

Shows ranking among Europe and Central Asian countries, higher score indicates less reform.

Scores as a ratio of EU-10 average, higher score indicates more reform.

Data sources: The Vienna Institute for International Economic Studies FDI database, WEO, Eurostat, EBRD, World Bank.
Appendix I: Estimation of the Current Account (CA) Norm Using Macroeconomic Balance Approach

The following explanatory variables were used to estimate the CA norm.

Fiscal balance. A higher government budget balance, in the absence of full Ricardian equivalence, would raise national savings and increase CA balance. This variable is measured by general government balance as a share of GDP. Data source: WEO

Demographics. A higher share of the economically inactive or dependent population reduces national saving and decreases the CA balance. This variable was proxied by two measures: old age dependency ratio, and population growth. Data source: WEO and ILO EPAP database.

Foreign direct investment. Higher FDI inflows should allow a lower CA balance, by directly boosting imports and enhancing future export capacity. This variable is measured as total FDI inflows as a share of GDP. Data source: WEO.

Relative income levels. Economies with lower per capita income would have higher investment needs that should be reflected in a lower CA balance. Data source: WEO.

Energy balance. Sustained increases in energy price represents an exogenous terms of trade shock that would lower the medium-term CA balance for a net energy importer. Data source: WEO.

Foreign-exchange reserve coverage. Higher reserves, by providing a cushion against possible shocks, should allow a lower CA balance. However, higher reserves could also imply larger interest receipts, pointing to a higher CA balance. This variable is measured by stock of foreign reserves in terms of months of imports. Data source: WEO.

The estimation included yearly data for the following 38 countries for 1992-2006: Austria, Australia, Bulgaria, Belgium, Canada, Croatia, Czech Republic, Denmark, Estonia, Finland, France, Greece, Hungary, Iceland, Ireland, Israel, Italy, Japan, Korea, Latvia, Lithuania, Netherlands, New Zealand, Poland, Portugal, Romania, Russia, Singapore, Slovakia, Spain, Slovenia, Turkey, United Kingdom, United States, Norway, Sweden, Germany, and Switzerland.

Table A1.1. Current Account Norm: Estimated coefficients
VariableFixed effects estimation 1/
Fiscal balance0.26 **
FDI-0.07 **
Relative GDP-0.11
Dependency0.34 **
Pop. Growth-0.01 **
Energy Balance0.28 *
Reserve Coverage0.002 **
Adjusted R-squared0.73

*and **indicate significance at 1 and 5 levels.

*and **indicate significance at 1 and 5 levels.

Appendix II. Constant Market Share Analysis

According to the Constant Market Share Analysis (CMSA), the expansion of a country’s exports derived from the overall growth of its competitors is considered the market growth effect. If a country’s exports grow faster than those of its competitors, the CMSA decomposes this “over performance” into three effects:

  • (i) growth due to exports being concentrated in faster growing products, called the commodity composition effect,
  • (ii) growth due to exports being concentrated in faster growing markets, called the market distribution effect, and
  • (iii) growth due to other factors, collectively labeled as competitiveness gains.9

Using this CMSA decomposition, the actual increase in Bulgarian exports between 2001 and 2005, Δ, can be decomposed as:


ΔX=Σ growth effect
+ΣriXiiΣrXii.........(i)commodity composition effect
+ΣiΣjrijXijΣiriXi.........(ii)market distribution effect
+ΔXΣiΣjrijXij.........(iii)“competitiveness effect”
  • r = percent change in the overall exports of competitor countries,
  • ri = percent change in competitors’ exports of SITC product i,
  • rij = percent change in competitors’ exports of SITC product i to market j,
  • xi = Bulgarian exports of product i at the beginning of the period, and
  • xij = Bulgarian exports of product i to market j at the beginning of the period.
  • j1-4 = EU-15, non-EU high income countries, European LMC, all other LMC
Table A2.1.Constant Market Share Analysis with Respect to Other Low- and Middle-Income Countries of the World (LMC)
∑ r Xi∑ ri xi∑∑ rij xijΔx∑ r Xi∑ ri xi∑∑ rij xijΔx∑ r Xi∑ ri xi∑∑ rij xijΔx
SITC 11476799164813913769220SITC 14565236525519793500SITC 180534455906736765720
SITC 2271238242270229086464476SITC 2838476103060220163143SITC 2227823234932224876301334
SITC 3538792462506678937654868SITC 31665568181243500-125328SITC 3223405258487316543780196
SITC 410927130501929924087SITC 433788327106832238SITC 469683268615621848
SITC 5464598579856660705403607SITC 514362017449321352482787SITC 5289959349070367362320820
SITC 63133330413620196SITC 6364414394191419151611336SITC 6940335106445010572871246790
SITC 74420415236358452565SITC 7136648180071315422521295SITC 7501525526068616069686071
SITC 89872137851799054417SITC 83051753493334655670SITC 81100739680931633513453836
SITC 9292888423517356785132777SITC 9905404037767200119906SITC 921732247103348675412871
Market Growth Effect=
Market Growth Effects = [l]/[4]67[l]/[4]52Market Growth Effect= [l]/[4]92
Commodity Composition Effect = ([2]-Commodity Composition Effect = ([2]-Commodity Composition Effect = ([2]-
Market Distribution Effect = ([3] -Market Distribution Effect = ([3] -
Market Distribution Effect = ([3] - [2])/[4]12[2])/[4]13[2])/[4]4
Competitiveness Effect = ([4]-[3])/[4]21Competitiveness Effect = ([4]-[3])/[4]35Competitiveness Effect = ([4]-[3])/[4]6
Appendix III: Estimation of the Equilibrium Real Effective Exchange Rate

Based on Alberola and others (1999), Bulgaria’s EREER (qt) is determined by the relative sectoral productivity between Bulgaria and its trading partners (pnt_tt) and the stock of Bulgaria’s net foreign assets (nfat):


  • qt is the log of the CPI-based multilateral real effective exchange rate calculated by the IMF INS system,
  • pnt_tt is the log of Bulgaria’s ratio of CPI non-tradables and CPI tradables relative to the ratio of CPI and PPI of all trading partners, and
  • nfat are total foreign assets minus total foreign liabilities in billions of US dollar.10

Trend and permanent components in the variables. While the REER trends upward since the trough in 1997, the trend components in the explanatory variables do not necessarily imply an appreciating EREER for the entire period based on the expected relationship (Appendix 3 Figure 1). The relative productivity variable does show an increasing trend during 1996-2006, however with a considerable slowing down after 2000. While there was a cumulative 55 percent increase in the value of this variable between 1996 and 2000, it only increased by 14 percent since then. The NFA variable shows an improvement until end-2003 and a rapid deterioration since then (Appendix 3 Figure 1). Large capital inflows into Bulgaria in recent years have gone hand in hand with a widening CAD and increased indebtedness by the private sector, reflecting a decline in the NFA position. Taken these developments in the fundamental variables, one could expect a sustained appreciation in the equilibrium rate in the first half and after that a more stable rate or some decline assuming the variables demonstrate expected sign of relationship with the REER.

Figure A3.1:Reer and Fundamentals, trend and cyclical components

Data properties and estimation results. All three data series show non-stationarity. We find existence of unit root at the 5 percent critical level using the Augmented Dicky Fuller tests (Table 1). However, Johansen cointegration tests show the existence of one cointegrating relationship between the three variables at the 5 percent critical value indicating a long-run time-varying equilibrium relationship.

Table A3.1.Results of Stationarity Tests and Johansen Co-integration Test 1/
Augmented Dicky-Fuller Tests
Cointegration equation normalized for qt
pnt_tt1.017.67 **
nfat-0.06-2.80 **
Error Correction
qt-0.07-1.71 *
pnt_tt0.174 44 **
nfat-0.49-1.69 *

* and ** denote significance at 10 percent and 5 percent levels, respectively.

* and ** denote significance at 10 percent and 5 percent levels, respectively.

The long-run coefficients are presented in Table 1. Estimation results show that a 1 percent increase in relative productivity causes a 1 percent appreciation in the EREER. The relationship to NFA, contrary to the prediction of macrobalance models, is negative. In summary, the REER has been below the EREER until 2001, after which it remains mostly above the EREER although not dispersing far from its equilibrium value and more recently being below its equilibrium value again. The speed of adjustment by the real exchange rate variable as given by the coefficient of the cointegration equation in the error correction model is negative (-0.07) and significant showing a relatively fast pace of adjustment with half of the movement away from the equilibrium corrected within the first eight months.

Data sources.

  • qt: IMF effective exchange rate facility.
  • pnt_tt: For Bulgaria, prices of tradable goods are proxied by the average of food and non-food prices with the following respective weights: 65 percent and 35 percent, since the PPI series is available only since 2000. Prices of non-tradable goods are proxied by the services’ sector prices excluding administered prices. Data source is Bulgarian National Bank. For trading partners, tradable sector prices are proxied by either monthly PPI or WPI depending on the data availability. Prices of nontradable goods are proxied by monthly CPI series. For EU-25 and Romania, the data source is Eurostat. For others, the IMF IFS database. The following countries were included as partners: EU-25, Romania, Turkey, Macedonia, Albania, Russia, the US, Canada and Brazil. These countries amount to about 75 percent of Bulgaria’s exports and imports.
  • nfat:Bulgarian National Bank.

    Alberola, Enrique; Susana G.Cervero; HumbertoLopez; and AngelUbide (1999) “Global Equilibrium Exchange Rates: Euro, Dollar, “Ins,” “Outs,” and Other Major Currencies in a Panel Cointegration Framework”, IMF Working Paper, WP/99/175 (December).

    • Search Google Scholar
    • Export Citation

    Burgess, Robert; StefaniaFabrizio; and YuanXiao (2004) “The Baltics: Competitiveness on the Eve of EU Accession”, International Monetary Fund (July).

    • Search Google Scholar
    • Export Citation

    Chovanov, Dimitar and PirittaSorsa (2004) “Competitiveness in Bulgaria: An Assessment of the Real Effective Exchange Rate”, IMF Working Paper, WP/04/37 (March).

    • Search Google Scholar
    • Export Citation

    DeBroeck, Mark and TorstenSlok (2001) “Interpreting Real Exchange Rate Movements in Transition Countries”, IMF Working Paper, WP/01/56 (May).

    • Search Google Scholar
    • Export Citation

    Égert, Balázs (2005) “Equilibrium Exchange Rates in South Eastern Europe, Russia, Ukraine and Turkey: Healthy or (Dutch) Diseased?” Economic Systems, No. 29, pp. 205–41.

    • Search Google Scholar
    • Export Citation

    Feldman, RobertA. (1994) “Measures of External Competitiveness for Germany”, IMF Working Paper, WP/94/113.

    International Monetary Fund (2002) “Assessing Sustainability”, (May).

    International Monetary Fund (2006a) “Romania: Selected Issues and Statistical Appendix”, Washington, DC: IMF Country Report No. 06/132 (April).

    • Search Google Scholar
    • Export Citation

    International Monetary Fund (2006b) “Methodology for CGER Exchange Rate Assessments”, (August).

    International Monetary Fund (2006c) “Bulgaria: Selected Issues and Statistical Appendix”, Washington, DC: IMF Country Report No. 06/299 (August).

    • Search Google Scholar
    • Export Citation

    International Monetary Fund (2007a) “Review of the 1977 Decision—Proposal for a New Decision—Companion Paper”, (May).

    • Search Google Scholar
    • Export Citation

    International Monetary Fund (2007b) “Ukraine: Selected Issues and Statistical Appendix”, Washington, DC: IMF Country Report No. 06/299 (August).

    • Search Google Scholar
    • Export Citation

    Kaminski, Bartlomiej (2006) “Bulgaria’s Institutions and Policies: Integrating into Pan-European Markets”, World Bank Policy Research Working Paper, No, 3864 (March).

    • Search Google Scholar
    • Export Citation

    Ko, CarolineD.M. (2006) “What Do the Sources and Uses of Funds Tell Us About Credit Growth in Central and Eastern Europe?”, European Commission, Directorate-General for Economic and Financial Affairs, Occasional Paper No. 26 (November).

    • Search Google Scholar
    • Export Citation

    Landesmann, MichaelA. (2003) “Structural Features of Economic Integration in an Enlarged Europe: Patterns of Catching-Up and Industrial Specialization”, European Commission, Economic Papers No. 181 (January).

    • Search Google Scholar
    • Export Citation

    Learner, EdwardE. and Robert M.Stern (1973) “Constant-Market-Share Analysis of Export Growth” in Quantitative International Economics, Boston: Allyn and Bacon, Inc., pp. 171–l83.

    • Search Google Scholar
    • Export Citation

    SorsaPiritta; BasBakker; ChristophDuenwald; AndreaMaechler; and AndrewTiffin (2007) “Vulnerabilities in Emerging Southeastern Europe—How Much Cause for Concern?” IMF Working Paper, WP/07/236.

    • Search Google Scholar
    • Export Citation

    World Bank (2007) “Accelerating Bulgaria’s Convergence: The Challenge in Raising Productivity”, Washington, DC: World Bank.


Prepared by Jesmin Rahman.


The latter also produces a CA norm of -5 percent. The estimation, however, does not include Bulgaria in the regression, but includes the following countries from central and eastern Europe: Croatia, Czech Republic, Hungary, Poland, Slovak Republic and Slovenia. The methodology is described in IMF (2006b).


The elasticity is calculated based on an import elasticity of 0.92, export elasticity of -0.71, and imports and exports of goods and non-factor services at 83 and 64 percent of GDP, respectively.


Debt crisis is defined as a sharp correction of the debt-to-GDP ratio beyond some de minimis threshold. Over the period 1979-2001, 53 such episodes have been identified. Crisis debt threshold increases to 65% of GDP with exports to GDP ratio greater than 40%. In this case, the conditional probability of a debt crisis is also significantly lower at around 6% although that of a debt correction is high at 20%. See IMF (2002).


The competitiveness measure may reflect one or more of the following price and non-price factors: (i) differential rate of increase in export prices of the same product; (ii) changes in exchange rates which are passed through and not absorbed in the markup; (iii) changes in export taxation or subsidization; (iv) differential rate of productivity growth in different countries; (v) differential rates of quality improvement; (vi) differential rate of improvement in the efficiency with which exports are marketed; and (vii) differential changes in the extent to which export orders can be met promptly.


See Data Appendix for the construction of and data sources for these variables.

Other Resources Citing This Publication