Republic of Madagascar: Staff Report for the 2017 Article IV Consultation, First Review Under the Extended Credit Facility Arrangement, and Requests for Waiver of Nonobservance of Performance Criterion, Modification of Performance Criterion, and Augmentation of Access—Debt Sustainability Analysis

2017 Article IV Consultation, First Review Under the Extended Credit Facility Arrangement


2017 Article IV Consultation, First Review Under the Extended Credit Facility Arrangement

article image

Madagascar’s risk of external debt distress is assessed to be ‘moderate,’ in line with the last DSA at the time of the ECF program request in July 2016. Debt dynamics have improved slightly since then, mainly because of more favorable financing assumptions following a successful donor conference in late 2016. The public DSA suggests that the dynamics of Madagascar’s total public and publicly-guaranteed (PPG) debt are sustainable, although weak fiscal revenue generation, possible exchange rate shocks, and contingent liabilities related to state-owned enterprises remain potential sources of vulnerability.


1. This joint DSA has been prepared by IMF and World Bank staff. It is based on the framework for LICs approved by the respective Executive Boards. The framework takes into account indicative thresholds for debt burden indicators determined by the quality of the country’s policies and institutions.2 The assessment comprises a baseline scenario and a set of alternative scenarios.

2. This DSA includes public debt and guarantees of the central government. The DSA does not include the debt of local government or state owned enterprises (other than through direct guarantees provided by the central government). The measure of debt is on a gross rather than net basis. And the residency criterion is used to determine the split between external and domestic debt.

Recent Developments and Current Debt Situation

3. The trend toward an increasing reliance on domestic debt was reversed in 2016 (Figure 1). A smaller amount of loans from development partners during the 2008-13 crisis made the government more dependent on domestic borrowing to finance budget deficits. Domestic debt, including domestic budgetary arrears, grew from 7.6 percent of GDP in 2008 to 12.9 percent in 2015. With the government regaining the confidence of the international donor community, external financing has become more readily available and is reducing the need for domestic borrowing and domestic debt had declined to 11.7 percent of GDP in 2016. Total public debt rose from around $2.7 billion (31 percent of GDP) in 2008 to $3.7 billion (41 percent of GDP) in 2015 and then remained stable at that amount (a decline to 39 percent of GDP) in 2016 (Table 1). These debt levels are substantially less than the pre-HIPC peak of 95 percent of GDP. The debt service to revenue ratio has been trending upward.

Figure 1:
Figure 1:

Debt Level and Service Ratios

Citation: IMF Staff Country Reports 2017, 223; 10.5089/9781484310656.002.A003

Table 1:

Break-down of Total PPG Debt (end-2016)

article image

4. A stronger than expected exchange rate, on average more favorable borrowing terms, and a delayed commercial loan led to a better-than-expected outcome in 2016. A spike in vanilla prices boosted export revenues and the real effective exchange rate appreciated by about 3½ percent in 2016 (January to December).3 The 2016 debt level was further reduced by a change in IDA lending terms that led to the front loading of grant disbursements (part of which will be offset in subsequent years) and a commercial loan (with an AfDB guarantee) that was delayed and is now projected to be disbursed in 2017. The authorities have also largely refrained from borrowing externally on non-concessional terms, which helped support debt sustainability.

5. The majority of external debt is owed to multilateral creditors on highly concessional terms (Table 1). About one-third of total debt is held by domestic creditors mainly in the form of treasury bills and debt to the central bank 4. Domestic arrears have declined over 2016 to about 2½ percent of GDP from around 3½ percent in the previous year. The vast majority of external debt is held by multilateral creditors, in particular the World Bank and African Development Bank.

6. Private external debt is mainly issued by local subsidiaries of multinational companies. According to the authorities, external debt owed by domestically owned companies and households is negligible. However, there are a number of multinational companies (in mining, banking, telecommunications) with wholly-owned local subsidiaries that have accumulated external debt. While the authorities do not have comprehensive data on such obligations, the largest of these debtors is the Nickel/Cobalt mining company Ambatovy with external debt just under $2bn (20 percent of GDP). This obligation caused total external debt to increase from 24 percent of GDP in 2007 to 39 percent in 2010, with a more gradual increase after that. It is projected that this commercial loan will be fully repaid by 2030.

7. The government may face some contingent liabilities with respect to state-owned enterprises including the nonbank financial sector, while the banking sector is less likely to generate direct fiscal costs. While the recapitalization of Air Madagascar is part of the baseline assumptions and thus reflected in projected debt dynamics, contingent liabilities from other state-owned enterprises are not included.5 The electricity utility, JIRAMA, had long-term debt corresponding to ½ percent of GDP and short-term debt (suppliers’ credits, overdrafts etc.) corresponding to 5 percent of GDP at end-2014.6 The postal savings scheme and possibly the Madagascar Savings Fund (Caisse d’Epargne de Madagascar, CEM) may need future recapitalization (probably less than 1 percent of GDP combined). While the government is a minority shareholder in several commercial banks, most banks have financially solid foreign majority shareholders and bank liabilities are mainly composed of deposits that exceed loans. Dollarization of deposits and credits is not pronounced and banks generally maintain foreign assets that are larger than their foreign liabilities.

Underlying Assumptions

8. Besides the increasing current account deficit, most key variables driving debt dynamics are forecast to improve over the coming years (Box 1 and Table 2). The DSA projections are consistent with the authorities’ plan to scale-up much needed infrastructure investment and social spending. A big part of this investment will be financed through concessional external borrowing and grants.

Table 2:

Madagascar; Baseline Macroeconomic Assumptions

article image
Source: World Bank and IMF staff projections.

Baseline Macroeconomic Assumptions

Real GDP growth. Growth is projected to peak temporarily just below 6 percent in 2019. Compared to the 2016 DSA, the acceleration is more dynamic in the short-run, based on higher foreign financed investment. Medium-term growth remains roughly unchanged at 5 percent, driven by improved confidence, further reengagement of development partners, and increased mining exports.

Current account. While the 2016 current account was substantially stronger than expected, it was mostly due to temporary factors. Consistent with faster investment-led growth and additional reconstruction-related efforts, imports have been revised up, which leads to larger current account deficits in the short- and medium term.

Fiscal variables: More financing, including to finance the reconstruction efforts and the restructuring of Air Madagascar in 2017, will allow the government to run slightly higher primary deficits in the short run. While revenue projections have remained unchanged, donor grant support has been revised up substantially, particularly in the short-run.

9. Following a successful donor-conference in late 2016, expected financing conditions have improved. At the time of the donor conference in Paris in December 2016, the authorities received pledges of $6.4 billion (over 60 percent of 2016 GDP). While not all of this amount is available for investment in the near-term, concrete pledges for project financing have nevertheless exceeded assumptions in previous frameworks in 2017-2020.

10. Semi-concessional and very limited non-concessional borrowing is envisaged throughout the forecast horizon. Consistent with the ceiling in the program, non-concessional borrowing (with a negative average grant element of 12 percent7) is foreseen at $100 million in 2017, of which $55 million will be disbursed before end-June 2017 under a long-delayed loan with an AfDB guarantee and an additional $45m is expected under similar conditions to finance the restructuring of Air Madagascar. Over the medium term, the importance of semi-concessional borrowing is expected to increase, reducing the average grant element of new borrowing from above 45 percent in 2019 to roughly 37 percent in 2037.

11. The main risks to these assumptions relate to political instability, revenue generation, the exchange rate, and the persistence of donor grant support. Political instability could weaken economic confidence, with negative implications for key macroeconomic variables, such as growth, the exchange rate and donor support. Continued weak revenue performance would accelerate debt accumulation and a faster-than-expected depreciation of the Ariary would increase the real value of the existing debt stock. Additionally, while the outlook on donor grant support is positive, lack of reform progress going forward could undo these gains. There are reasons to expect that the risks from natural disasters not already incorporated in the baseline to be rather limited (Box 2). Risks are assessed to be symmetric since the exchange rate may continue to surprise on the upside and revenue generation has a significant upside potential given the low base. This would boost the ability to service higher debt levels while structural fiscal reforms could stimulate higher-than-expected donor support also in the medium-to longer-term.

External DSA

Baseline scenario

12. The level of PPG external debt was roughly $2.5 billion in 2016 and is projected to grow gradually throughout the forecast period. PPG external debt is forecast to increase from 27 percent of GDP in 2016 to a peak of 36 percent of GDP in 2022 (Table 3). A temporarily higher trade deficit and outflows from the mining sector8 are balanced by increasing transfer inflows and a moderate increase in net FDI inflows9, consistent with the authorities’ National Development Plan. As domestic debt markets deepen (see below), PPG external debt is projected to decline to 25 percent of GDP by 2037.

Table 3.

Madagascar: External Debt Sustainability Framework, Baseline Scenario, 2014-371

(In percent of GDP; unless otherwise indicated)

article image
Sources: Country authorities; and staff estimates and projections.

Includes both public and private sector external debt.

Derived as [r - g - ρ(1 + g)]/(1+g +ρ+gρ) times previous period debt ratio, with r = nominal interest rate; g = real GDP growth rate, and ρ = growth rate of GDP deflator in U.S. dollar terms.

Includes exceptional financing (i.e., changes in arrears and debt relief); changes in gross foreign assets; and valuation adjustments. For projections also includes contribution from price and exchange rate changes.

Assumes that PV of private sector debt is equivalent to its face value.

Current-year interest payments divided by previous period debt stock.

Historical averages and standard deviations are generally derived over the past 10 years, subject to data availability.

Defined as grants, concessional loans, and debt relief.

Grant-equivalent financing includes grants provided directly to the government and through new borrowing (difference between the face value and the PV of new debt).

Debt Dynamics and Natural Disasters

Madagascar’s history of natural disasters suggests that the effects of these events that are not already incorporated in the baseline scenario would be manageable. The potential effect of natural disasters on debt dynamics are investigated using an event study of the tropical cyclones in 2004 as an example; it was the most destructive year since 1980 (the greater damage by cyclones in the seventies reflected a different economic and social structure). Among standard macro-variables, the effects were most pronounced for inflation and the exchange rate, but modest overall.

Natural disasters are a recurring phenomenon in Madagascar, so their occurrence is incorporated into baseline assumptions. Since 2000, Madagascar has been affected by 30 natural disasters, including droughts, storms, floods, epidemics and insect infestation. Since they are recurring, natural disasters are included into the medium-term macroeconomic forecasts. The starting point for these forecasts are historical averages, which already incorporate the medium-term impact of these recurrent disasters and are then adjusted for the impact of planned policy measures. Nevertheless, the possible occurrence of a more severe event raises the question of its macro-economic implications and its risk to debt dynamics.

The most violent storms in recent history suggest that the macro-economic effects of natural disasters tend to be surprisingly contained. In 2004, Madagascar was hit by two tropical cyclones that caused 400 deaths, nearly 1,000 injured and estimated economic damage equivalent to 5.7 percent of GDP. Macroeconomic variables were less affected than usual volatility, including from political crises. Primary expenditure by the government increased from 17 percent of GDP in the previous year to 22 percent in 2004. However, general government revenues also rose; with 3 of the 5 percentage points of GDP increase coming from additional grant support. Both expenditure and revenues declined again the following year. The primary fiscal deficit thus increased by only 0.3 percent of GDP, from 1.7 to 2 percent of GDP in 2004. In 2005, the primary fiscal accounts were roughly in balance. Despite the disaster, GDP growth was above 5 percent (which was a slowdown from the exceptional 10 percent in the previous year but close to the long-run average). (In early 2017, the country was hit by the biggest tropical cyclone for 13 years, which is also projected to have a very small net impact on growth in that year.) Of greater importance, the cyclone may have contributed to above average inflation reported at 14 percent in 2004 and 18 percent in the following year, compared to an average 8 percent over the previous 5 years. Related to the higher inflation, the currency depreciated in nominal terms by 34 percent in 2004.

A customized debt scenario confirms the robustness of the fiscal debt dynamics. We thus include a 35 percent depreciation of the exchange rate and 10 percent faster inflation in 2018. Consistent with the experience in 2004, the fiscal variables and GDP growth rates for the year of the shock are left unchanged. The effects of the depreciation and the higher inflation roughly offset one another in the case of the PV of Debt-to-GDP ratio. For the other measures in the fiscal scenario, indicators actually improve.

13. Under the baseline projection, all PPG external debt indicators remain below the policy-dependent debt burden thresholds (Figure 2). The present value (PV) of external debt is projected to increase from 15½ percent of GDP in 2016 to 21 percent of GDP by 2022 and then to decline to 16 percent by 2037. This projection is somewhat more favorable than the medium-term DSA forecast at the time of the ECF request.

Figure 2.
Figure 2.

Madagascar: Indicators of Public and Publicly Guaranteed External Debt Under Alternatives Scenarios, 2017-2037 1/

Citation: IMF Staff Country Reports 2017, 223; 10.5089/9781484310656.002.A003

Sources: Country authorities; and staff estimates and projections.1/ The most extreme stress test is the test that yields the highest ratio on or before 2027. In figure b. it corresponds to a Natural disaster shock; in c. to a Exports shock; in d. to a Combination shock; in e. to a Exports shock and in figure f. to a One-time depreciation shock

14. Private external debt is projected to decline slowly, as the loans related to a major mining project are repaid. Given the exceptional nature of this project, the DSA does not forecast substantial new external borrowing from the private sector. Furthermore, this debt is not assessed to pose a significant threat to external sustainability, as the ultimate liability for these loans is held by the multinational shareholders, rather than resident entities (such as domestic banks or the government).

Alternative scenarios

15. The two standard DSA alternative scenarios are applied to the baseline external PPG debt projections. First, the standard bound tests apply pre-defined shocks to the key macroeconomic variables that drive external debt (summarized in Footnote 1 of Figure 2). Second, a historical scenario where macroeconomic variables are set equal to their average over 2007-16 is imposed on the baseline projection. These shocks are detailed in Table 4.

Table 4.

Madagascar: Sensitivity Analysis for Key Indicators of Public and Publicly Guaranteed External Debt, 2017-37

(In percent)

article image
article image
Sources: Country authorities; and staff estimates and projections.

Variables include real GDP growth, growth of GDP deflator (in U.S. dollar terms), non-interest current account in percent of GDP, and non-debt creating flows.

Assumes that the interest rate on new borrowing is by 2 percentage points higher than in the baseline., while grace and maturity periods are the same as in the baseline.

Exports values are assumed to remain permanently at the lower level, but the current account as a share of GDP is assumed to return to its baseline level after the shock (implicitly assuming an offsetting adjustment in import levels).

Includes official and private transfers and FDI.

Depreciation is defined as percentage decline in dollar/local currency rate, such that it never exceeds 100 percent.

Applies to all stress scenarios except for A2 (less favorable financing) in which the terms on all new financing are as specified in footnote 2.

16. For the standard bound tests, at least one scenario causes a significant breach of the threshold for one or more indicators. A simultaneous slowdown in GDP, decline in exports, a depreciation, and reduction of non-debt creating flows (all a one-half standard deviation; see Table 4) would cause the PV of debt-to-revenue ratio to peak at 243 percent in 2017, above the threshold implied by Madagascar’s CPIA rating. The same shock would also just barely exceed the threshold for the PV of debt-to-GDP ratio. Additionally, a one standard deviation decline in exports would result in the PV of debt-to-exports to increase to 108 percent, above its threshold of 100. The first stress test in particular is sufficient to classify Madagascar as standing at moderate risk of external debt distress.

17. The historical scenario10 projects a rapid increase in all debt metrics and causes a breach for four of the five external debt thresholds. These scenarios cause a substantial breach in three thresholds, especially for the PV of debt-to-GDP and the PV of debt-to-revenue ratios. However, there are good reasons to place less weight on the historical scenario. The large current account deficits in 2008 and 2009 (over 20 percent of GDP) were driven by imports associated with large mining projects, which were principally financed through non-debt creating FDI flows. Thus, these deficits did not lead to a build-up of PPG external debt and this period is not representative of the normal economic environment in Madagascar.

Public DSA

Baseline scenario

18. Domestic PPG debt is projected to decline as a proportion of GDP over the next decade, with authorities replacing domestic financing by concessional external borrowing, given the projected continued improvement in donor relations. The importance of domestic PPG debt is then expected to grow again over the long term, as domestic markets deepen and savings become more abundant. The PV of total PPG debt is projected to remain less than 30 percent of GDP throughout the forecast horizon—well-below the threshold of 38 percent (Figure 3 and Table 5).

Figure 3.
Figure 3.

Madagascar: Indicators of Public Debt Under Alternative Scenarios, 2017-2037

Citation: IMF Staff Country Reports 2017, 223; 10.5089/9781484310656.002.A003

Sources: Country authorities; and staff estimates and projections.1/ The most extreme stress test is the test that yields the highest ratio on or before 2027.2/ Revenues are defined inclusive of grants.
Table 5.

Madagascar: Public Sector Debt Sustainability Framework, Baseline Scenario, 2017-37

(In percent of GDP; unless otherwise indicated)

article image
Sources: Country authorities; and staff estimates and projections.

General government gross debt

Gross financing need is defined as the primary deficit plus debt service plus the stock of short-term debt at the end of the last period.

Revenues excluding grants.

Debt service is defined as the sum of interest and amortization of medium and long-term debt.

Historical averages and standard deviations are generally derived over the past 10 years, subject to data availability.