## _wp1312

## Source details

**Canonical URL:** [_wp1312](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1312.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1312.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1312.pdf.json)

---

### Role and urgency of bond market development in sub-Saharan Africa
- African Development Bank intends to launch a new bond program for infrastructure to raise up to US$40 billion for investments in projects such as ports and airports, highlighting the growing role for bond markets in financing development in sub-Saharan Africa.
- Bond markets in these countries are at a nascent stage of development and there is a strong need to promote their development.
- Key motivations for promoting bond markets:
  - Reliance on external grants and concessional loans: western donors face substantial fiscal challenges and donor flows to sub-Saharan Africa may be scaled back significantly.
  - Financial stability: deeper bond markets helped many Asian economies since the 1997 Asian financial crisis and aided their responses to the recent global financial crisis by providing funding sources for fiscal stimulus packages.
  - Intermediation of savings: although Africa “needs money,” Africa is a net capital exporter to the rest of the world (IMF, 2012); this is mainly because of a lack of effective intermediate channels to absorb capital.
  - Financial structure improvement: the African financial sector is dominated by banks; non-banking sector and bond markets (public and private) are in their infancy. Bond markets are complementary to banks—banks excel at short-term working capital while bond markets have comparative advantages in financing government deficits, infrastructure investment, and providing longer-term capital to companies.
  - Monetary policy implementation: deeper bond markets would provide central banks with a wider, more effective range of instruments beyond short-term government debt for sterilization and monetary policy operations.

### Empirical focus, data, and methodology
- Objective: Investigate empirically the determinants of local currency bond markets in sub-Saharan Africa, with emphasis on ability to overcome “original sin.”
- Data and sample:
  - Local currency government securities market capitalization for 36 countries, over the years 1980–2010.
  - Newly developed corporate bond market capitalization database covering 24 countries active over 1980–2010.
- Methodology:
  - Econometric approach draws on Eichengreen and Luengnaruemitchai (2004), Claessens, Klingebiel, and Schmukler (2007), Adelegan and Radzewicz-Bak (2009).
  - Uses generalized method of moments estimation (system GMM) to address possible endogeneity among variables relevant to bond market development.
- Dependent variables:
  - Gdebt: government domestically issued and marketable securities (percent of GDP).
  - Cdebt: corporate bonds outstanding (percent of GDP).
- Endogenous controls in GMM: fiscal balance (fiscal), interest rate variability (intvol), bank interest rate spread (spread), and the dependent variables.
- Independent variables include: econsize, tradeopen, credit, gdpcap, spread, intvol, xrvol, capopen (Chinn-Ito Index), fiscal, laworder, corruption, invprofile, bureaucracy, comprisk, legalorigin (English dummy), and area.
- Data sources: IMF WEO and IFS; World Bank ADI; ICRG for institutional variables; multiple primary national sources for corporate bond data.

### Stylized facts and market evolution (government and corporate)
- Aggregate capitalization metrics (2010):
  - Government securities market capitalization as a percent of GDP: 14.8 percent in 2010 for sub-Saharan Africa.
  - Corporate bonds capitalization as a percent of GDP: 1.8 percent in 2010 for these countries.
  - Share of government securities in total local currency bond market capitalization: 89.2 percent in 2010.
  - Share of corporate bonds in total bonds: 10.8 percent in 2010.
- Temporal trends:
  - Government securities market capitalization: fell from 18.7 percent of GDP in 2006 to 14.1 percent in 2009; rose to 14.8 percent in 2010 amid widening fiscal deficits.
  - Corporate bond market capitalization: grew as a share of GDP from 1 percent in 2006 to 1.8 percent in 2010.
  - Corporate bond market capitalization climbed from approximately zero in 1989 to over 1.3 percent of GDP in 2010 (aggregate, Figure 4 statement).
  - Share of corporate bonds in total bonds increased from 5.1 percent in 2006 to 10.8 percent in 2010.
- South Africa as outlier:
  - South Africa’s corporate bond market capitalization declined from around 25 percent of GDP in 1989 to less than 15 percent by 2000, then recovered to almost 20 percent of GDP by 2010.
  - In 2010 South Africa’s government securities market capitalization was 31.2 percent of GDP and corporate 20.0 percent of GDP.
- Cross-regional heterogeneity:
  - International sovereign bonds issued by sub-Saharan countries were estimated at US$5 billion (Bloomberg International) compared to local currency government securities markets of US$135 billion (2010 estimate).
  - Comparative corporate bond capitalization examples: Canada 26.5 percent of GDP; United States 98.6 percent of GDP.
  - Middle-income countries have the most developed corporate bond market capitalization among subgroups, followed by WAEMU countries.
- Data coverage and limitations:
  - Analysis restricted to 36 countries between 1980–2010 due to missing, insufficient, inconsistent, or unreliable data for some countries.
  - Corporate debt database covers 24 countries that had active corporate bond markets over the period 1980 to 2010.

### Key empirical findings — Government securities markets (GMM and FE evidence)
- Consistent negative correlates with government securities market development:
  - Spread (bank interest rate spread): negative and significant.
  - Fiscal (three-year moving average of overall fiscal balance): negative and significant.
  - Exchange rate variability (xrvol): negative and significant in many specifications.
  - Trade openness and capital account openness: negative and significant in several specifications.
  - Area (land area): negative and significant in GMM and some fixed effects specifications.
- Positive correlates:
  - Legalorigin (English origin): positive and significant in many specifications.
  - Comprisk (composite lower risk / better institutions) and Laworder: positive and significant.
  - Interest rate volatility (intvol): positive in some GMM specifications.
- Notable contrasts:
  - GMM B-specification finds positive effect of econsize; some FE estimates show econsize negative.
  - GDP per capita (gdpcap) becomes negative and significant in GMM after controlling for risk in some specifications.
- Diagnostics:
  - Hansen and Arellano-Bond AR tests typically do not reject instrument validity and estimation consistency in reported specifications.

### Key empirical findings — Corporate bond markets (GMM and RE evidence)
- Robust positive correlates:
  - Econsize (GDP, PPP) and Gdpcap (GDP per capita): positive and significant across specifications.
  - Credit (domestic credit to private sector): positive and strongly significant in GMM estimates, suggesting complementarity between credit market development and corporate bond market development.
  - Area (land area) often positive and significant in some specifications.
- Mixed or context-dependent effects:
  - Tradeopen: negative in several GMM specifications.
  - Capopen (Chinn-Ito Index): positive in regressions that control for risk, negative otherwise.
  - Spread: negative in regressions that control for risk (GMM), mixed elsewhere.
  - Intvol (interest rate volatility): mixed effects across specifications.
  - Fiscal: improving fiscal balance tends to be associated with lagging corporate bond market development in some GMM specifications (negative, not always significant).
- Institutional variables:
  - Comprisk tends to be positively related to corporate bond development in GMM (lower risk improves development).
  - Laworder and lower corruption are positively related in GMM results.
- Sample considerations:
  - Corporate bond sample smaller (24 active countries) versus 36 for government securities, affecting precision and significance patterns.

### Sensitivity, robustness, and diagnostics
- Sensitivity analysis uses GMM specification B2 with four endogenous variables and composite risk across sample variants:
  - Variants: baseline, 1993–2010, excluding South Africa, excluding fragile states, excluding WAEMU, excluding CEMAC.
  - Results show a fairly high degree of similarity across different estimations, bolstering robustness of main findings.
- Hausman tests:
  - Favor fixed effects for government securities in many cases and random effects for corporate bonds in several specifications.
- GMM helps address endogeneity concerns present in prior literature.

### Policy implications and recommendations
- Foster economic growth and structural expansion:
  - Growth and larger economic size (econsize, gdpcap) promote corporate bond market development.
- Strengthen macroeconomic policy and institutions:
  - Improve fiscal positions and reduce exchange rate volatility to support government securities markets.
  - Enhance law and order, reduce corruption, and improve institutional quality (comprisk, laworder, bureaucracy) to encourage both government and corporate bond markets.
- Develop credit markets:
  - Deepening domestic credit (credit) complements corporate bond market development—support policies that strengthen banking intermediation while encouraging market-based financing.
- Consider capital account and trade policies carefully:
  - Trade openness and capital account openness exhibit mixed effects; policy design should weigh external access benefits against incentives for domestic market development.
- Support market infrastructure and data collection:
  - Continued efforts to compile reliable corporate bond market data (from stock exchanges, securities commissions, central banks) are necessary to monitor and facilitate market growth.

### Representative exact numeric highlights and descriptive statistics
- Corporate bond capitalization: 1.3 percent of GDP in 2010 (aggregate, Figure 4 statement).
- Share of corporate bonds in total bonds: 5.1 percent in 2006 and 10.8 percent in 2010.
- South Africa corporate bond capitalization: almost 20 percent of GDP as of 2010.
- Sub-Saharan Africa aggregates (2010, Table 4): Government 14.8 percent of GDP; Corporate 1.8 percent of GDP; Contribution of Corporate Bonds to Total Domestic Debt 9.2 percent.
- Descriptive means (1980–2010, Table 3): Gdebt mean 17.36 (percent of GDP); Cdebt mean 2.31 (percent of GDP); Econsize mean 22.24 (US$ billions); Tradeopen mean 31.88 (percent of GDP); Credit mean 18.73 (percent of GDP); Gdpcap mean 2.21 (US$ thousands); Spread mean 12.83.
- Gross official developmental assistance to sub-Saharan Africa amounted to US$49 billion in 2010, accounting for 32 percent of total government consumption expenditure, with 83 percent grants and 17 percent concessional loans (World Bank and OECD data).
- International sovereign bonds issued by sub-Saharan countries estimated at US$5 billion (Bloomberg International) versus local currency government securities markets of US$135 billion (2010 estimate).

*Source: _wp1312*

### 1. Countries in the Sample, Definition of Variables and Sources ......................................25

### 1. Countries in the Sample, Definition of Variables and Sources

### Role and urgency of bond market development in sub-Saharan Africa
- The African Development Bank plans to launch a new bond program for infrastructure to raise up to US$40 billion for investments in projects such as ports and airports, highlighting the growing role for bond markets in financing development in sub-Saharan Africa.
- Bond markets in these countries are at a nascent stage of development and there is a strong need to promote their development.
- Key motivations for promoting bond markets:
  - Reliance on external grants and concessional loans: western donors face substantial fiscal challenges and donor flows to sub-Saharan Africa may be scaled back significantly.
  - Financial stability: deeper bond markets helped many Asian economies since the 1997 Asian financial crisis and aided their responses to the recent global financial crisis by providing funding sources for fiscal stimulus packages.
  - Intermediation of savings: although Africa “needs money,” Africa is a net capital exporter to the rest of the world (IMF, 2012); this is mainly because of a lack of effective intermediate channels to absorb capital.
  - Financial structure improvement: the African financial sector is dominated by banks; non-banking sector and bond markets (public and private) are in their infancy. Bond markets are complementary to banks—banks excel at short-term working capital while bond markets have comparative advantages in financing government deficits, infrastructure investment, and providing longer-term capital to companies.
  - Monetary policy implementation: deeper bond markets would provide central banks with a wider, more effective range of instruments beyond short-term government debt for sterilization and monetary policy operations.

### Empirical focus and scope of the paper
- Objective: Investigate empirically the determinants of local currency bond markets in sub-Saharan Africa.
- Rationale for focusing on local currency markets:
  - Importance relative to international sovereign bonds.
  - Focus on countries’ ability to overcome “original sin,” i.e., the inability to issue debt in local currency.
- Data and sample:
  - Local currency government securities market capitalization for 36 countries, over the years 1980–2010.
  - A newly developed database for corporate bond market capitalization.
  - The sample is described as the largest of its kind in terms of both number of countries included and number of years covered.
- Methodology:
  - Econometric approach draws upon Eichengreen and Luengnaruemitchai (2004), Claessens, Klingebiel, and Schmukler (2007), and Adelegan and Radzewicz-Bak (2009), among others.
  - Uses generalized method of moments estimation to address possible endogeneity among variables relevant to bond market development.

### Contextual statistics cited in the introduction
- Gross official developmental assistance to sub-Saharan Africa amounted to US$49 billion in 2010, accounting for 32 percent of total government consumption expenditure, with 83 percent grants and 17 percent concessional loans, according to World Bank and Organization for Economic Cooperation and Development data.
- Historical reference: Asian financial crisis in 1997.

### Research aims and paper structure
- Three stated aims of the research:
  - Outline the current situation of local currency bond markets including both government securities and corporate bond markets in sub-Saharan African countries.
  - Discuss and estimate key determinants of bond market development in sub-Saharan Africa.
  - Offer policy advice for enhancing bond market development in sub-Saharan Africa.
- Paper structure: Section II reviews the relevant literature (following the Introduction).

*Source: _wp1312 - 1. Countries in the Sample, Definition of Variables and Sources ......................................25*

### Section III provides an overview of the government securities and corporate bond markets in

### _wp1312 - Section III provides an overview of the government securities and corporate bond markets in

### II. LITERATURE REVIEW — key findings and methodological notes
- Research focus to date has concentrated on banking sector and stock market financial development (examples cited: Detragiache et al., 2005; McDonald and Schumacher, 2007; Yartey and Adjasi, 2007; Andrianaivo and Yartey, 2009; Anayiotos and Toroyan, 2009; Kablan, 2010; and Beck et al., 2011).
- Relatively little attention has focused on development of public and private bond markets.
- International sovereign bonds issued by sub-Saharan countries were estimated at US$5 billion (according to Bloomberg International) compared to local currency government securities markets of US$135 billion (2010 estimate).
- Prior cross-country empirical approaches and results summarized:
  - Eichengreen and Luengnaruemitchai (2004): panel data 1990–2001 for 41 countries; generalized least squares with heteroskedasticity and panel-specific autocorrelation correction. Determinants for public bonds: positive and significant — GDP at purchasing power parity, exports, English origin, distance from the equator, a positive investment profile, and an open capital account; negative and significant — GDP per capita at purchasing power parity, banking sector concentration, bureaucracy quality, the interest rate spread, exchange rate volatility, and the fiscal balance. For private bonds: positive and significant — GDP at purchasing power parity, exports, Asia dummy, distance from the equator, corruption, accounting standards, domestic credit, and bureaucracy quality; negative and significant — English origin, the interest rate spread, and exchange rate volatility. Conclusions: market size matters; poor accounting standards, corruption, and low bureaucratic quality hinder private debt markets; well-capitalized bank systems promote bond markets; exchange rate stability encourages bond market development; absence of need for public financing discourages public bond markets; capital controls discourage bond market development.
  - Eichengreen, Panizza, and Borensztein (2008): panel on developing and developed countries (focus on Latin America). Findings: country size positive and significant with concavity; GDP per capita positive and concave; trade openness positive and significant. Domestic interest rate negative and significant only for government bonds. Interest rate volatility positively correlated with private bond market and negatively with government bond market. Domestic credit positively and concavely related to financial bonds. Interest rate spread positively correlated with corporate but not public bond market. Stricter capital controls correlated with large public bond markets but do not influence private bond markets. Larger public debt linked to large public bond markets but not significant for private bond markets.
  - Other studies (Burger and Warnock, 2006; Claessens, Klingebiel, and Schmukler, 2007; Jeanne and Guscina, 2006; Bae, 2012) highlight roles for inflation performance, rule of law, fiscal deficits, economy size, domestic investor base, exchange regime flexibility, inflation history, GDP per capita, banking sector size, and government bond market development across public, private, and financial bond markets.
- Counterintuitive institutional findings reported in some studies: stronger creditor rights linked to smaller private bond markets; legal origin effects (French, German, Scandinavian) and latitude/region dummies also matter.
- Adelegan and Radzewicz-Bak (2009): focused on corporate debt in sub-Saharan Africa for 23 African countries, 1990–2008. For public debt stock: positive and significant — investment profile, exchange rate variability, no capital controls, and the fiscal balance; negative and significant — quality of the bureaucracy and the interest rate spread. For private debt stock: positive and significant — domestic bank credit, exchange rate variability, no capital controls, and the fiscal balance; negative and significant — GDP per capita and interest rate variability. Note: limited sample size, especially for corporate bond market estimation, possibly affecting robustness.
- Endogeneity is a concern across the literature: fiscal balance, interest rate volatility, and spread may be endogenous to debt stock; many studies have not sufficiently addressed endogeneity.
- This study’s empirical contributions relative to the literature:
  - Distinguishes marketable from non-marketable central bank debt in measuring government securities market capitalization.
  - Provides more complete coverage of sub-Saharan African domestic government securities and corporate debt: 36 countries covered over the period 1980 to 2010 for government securities; corporate debt database includes 24 countries active over 1980 to 2010.
  - Employs a wide array of variables drawn from existing literature.
  - Confronts endogeneity by using generalized method of moments estimation.

### III. CURRENT SITUATION OF GOVERNMENT SECURITIES AND CORPORATE BOND MARKETS IN SUB-SAHARAN AFRICA — stylized facts and recent evolution
- Overall characterization:
  - Local currency bond markets in sub-Saharan African countries are described as still at a nascent stage of development.
  - Market capitalization of both government securities and corporate bonds is typically much lower than other developing, emerging, and advanced economies (referenced Tables and Figures in source).
- Aggregate capitalization metrics (2010):
  - Government securities market capitalization as a percent of GDP: 14.8 percent in 2010 for sub-Saharan Africa.
  - Corporate bonds capitalization as a percent of GDP: 1.8 percent in 2010 for these countries.
  - Share of government securities in total local currency bond market capitalization: 89.2 percent in 2010.
  - Share of corporate bonds in total bonds: 10.8 percent in 2010.
  - Comparative corporate bond capitalization examples cited: Canada 26.5 percent of GDP; United States 98.6 percent of GDP.
- Temporal trends:
  - Government securities market capitalization: fell from 18.7 percent of GDP in 2006 to 14.1 percent in 2009; rose to 14.8 percent in 2010 amid widening fiscal deficits.
  - Corporate bond market capitalization: grew as a share of GDP from 1 percent in 2006 to 1.8 percent in 2010 (growth trajectory noted in source).
- Cross-regional and subgroup heterogeneity:
  - Regionally, Asian and Central European countries surpass sub-Saharan Africa’s government securities capitalization measure; most Latin American countries also generally surpass it, excluding Argentina and Chile.
  - Aside from Poland, the share of corporate bonds in total bonds in sub-Saharan Africa is smaller than in other developing and emerging economies.
  - Sample subgrouping used for analysis: South Africa; all sub-Saharan African countries in the sample excluding South Africa; Central African (CEMAC) member states; West African (WAEMU) member states; oil exporters; fragile countries recently emerged from conflict; low-income countries; and middle-income countries.
  - With the exception of South Africa, subgroups display relatively low percentage capitalizations and relatively less developed corporate bond markets compared to government securities.
  - Middle-income countries have the most developed corporate bond market capitalization among subgroups, followed by WAEMU countries.
- Data coverage and limitations:
  - Analysis restricted to 36 countries between 1980–2010 due to missing, insufficient, inconsistent, or unreliable data for some countries.
  - Countries dropped for various reasons: Benin, Burkina Faso, and Equatorial Guinea (no government debt data); São Tomé (insufficient government debt data); Liberia, Mozambique, and Niger (inconsistencies with reported central bank lending); and Zimbabwe (unreliable data).
  - Corporate debt database covers 24 countries that had active corporate bond markets over the period 1980 to 2010; intention is for this database to provide groundwork for future expansion.

*Source: _wp1312 - Section III provides an overview of the government securities and corporate bond markets in (PDF chapter/section).*

### 1.8 percent in 2010 (Figure 3). When taken together, the share of corporate bonds in total

### _wp1312 - 1.8 percent in 2010 (Figure 3). When taken together, the share of corporate bonds in total

### Market trends and evolution
- Corporate bond market capitalization in sub-Saharan Africa climbed from approximately zero in 1989 to over 1.3 percent of GDP in 2010.
- The share of corporate bonds in total bonds increased from 5.1 percent in 2006 to 10.8 percent in 2010.
- South Africa is a clear outlier:
  - Corporate bond market capitalization declined from around 25 percent of GDP in 1989 to less than 15 percent by 2000, then recovered to almost 20 percent of GDP by 2010.
  - In 2010 South Africa’s government securities market capitalization was 31.2 percent of GDP and corporate 20.0 percent of GDP (Table 4).
- Number of active corporate bond markets increased (Figure 5), and both aggregate size and the number of countries with active markets rose during 1980–2010.

### Analytical framework and data
- Model: Two-phase estimation under fixed effects extended from Eichengreen and Luengnaruemitchai (2004); generalized method of moments (system GMM) by Arellano and Bover (1995) and Blundell and Bond (1998) used to address endogeneity.
- Dependent variables:
  - Gdebt: government domestically issued and marketable securities (percent of GDP).
  - Cdebt: corporate bonds outstanding (percent of GDP).
- Endogenous variables accounted for in GMM: fiscal balance (fiscal), interest rate variability (intvol), bank interest rate spread (spread), and the dependent variables.
- Independent variables include: econsize, tradeopen, credit, gdpcap, spread, intvol, xrvol, capopen (Chinn-Ito Index), fiscal, laworder, corruption, invprofile, bureaucracy, comprisk, legalorigin (English dummy), and area.
- Data sources: IMF WEO and IFS; World Bank ADI; ICRG for institutional variables; multiple primary national sources for corporate bond data (Table 2).

### Key empirical findings — Government securities markets (GMM and FE evidence)
- Variables with consistent negative correlation with government securities market development:
  - Spread (bank interest rate spread): negative and significant.
  - Fiscal (three-year moving average of overall fiscal balance): negative and significant.
  - Exchange rate variability (xrvol): negative and significant in many specifications.
  - Trade openness and capital account openness: negative and significant in several specifications.
  - Area (land area): negative and significant in GMM and some fixed effects specifications.
- Variables with positive correlation:
  - Legalorigin (English origin): positive and significant in many specifications.
  - Comprisk (composite lower risk / better institutions) and Laworder: positive and significant.
  - Interest rate volatility (intvol): positive in some GMM specifications.
- Notable contrasts and robustness:
  - GMM B-specification finds positive effect of economic size (contrast to some FE estimates where econsize was negative).
  - GDP per capita (gdpcap) becomes negative and significant in GMM after controlling for risk in some specifications.
  - Diagnostics: Hansen and Arellano-Bond AR tests typically do not reject instrument validity and estimation consistency (Tables 8b).

### Key empirical findings — Corporate bond markets (GMM and RE evidence)
- Robust positive correlates:
  - Econsize (GDP, PPP) and Gdpcap (GDP per capita): positive and significant across specifications.
  - Credit (domestic credit to private sector): positive and strongly significant in GMM estimates, suggesting complementarity between credit market development and corporate bond market development.
  - Area (land area) often positive and significant in some specifications.
- Mixed or context-dependent effects:
  - Tradeopen: negative in several GMM specifications.
  - Capopen (Chinn-Ito Index): positive in regressions that control for risk, negative otherwise.
  - Spread: negative in regressions that control for risk (GMM), mixed elsewhere.
  - Intvol (interest rate volatility): mixed effects across specifications.
  - Fiscal: improving fiscal balance tends to be associated with lagging corporate bond market development in some GMM specifications (negative, not always significant).
- Institutional variables:
  - Comprisk tends to be positively related to corporate bond development in GMM (lower risk improves development), reversing some random effects results that showed negative relations.
  - Laworder and lower corruption are positively related in GMM results.
- Sample size: corporate bond sample smaller (24 active countries) versus 36 for government securities, affecting precision and significance patterns.

### Sensitivity and robustness
- Sensitivity analysis (Table 10) uses GMM specification B2 with the four endogenous variables and the composite risk variable across sample variants:
  - Variants: baseline, 1993–2010, excluding South Africa, excluding fragile states, excluding WAEMU, excluding CEMAC.
  - Results show a fairly high degree of similarity across different estimations, bolstering robustness of main findings.

### Concluding findings (summary)
- Government securities:
  - Development is influenced by a combination of structure, policy, and institutions.
  - Negative correlates: spread, fiscal balance, exchange rate volatility, trade openness, capital openness, land area.
  - Positive correlates: English legal origin, lower composite risk, law and order, interest rate volatility in some specifications.
- Corporate bonds:
  - Key determinants are size measures (econsize, gdpcap, area) and credit market development.
  - Institutional improvements (lower comprisk, better law and order, lower corruption) generally support corporate market development in GMM estimates.
  - Policy variables show mixed effects; corporate markets appear more sensitive to scale and private credit depth than to a consistent set of policy levers.
- Market outlook:
  - Corporate bonds remain nascent in most sub-Saharan African countries (except South Africa) but have been expanding consistently and could become an important future source of finance.

### Policy implications and recommendations
- Foster economic growth and structural expansion:
  - Growth and larger economic size (econsize, gdpcap) promote corporate bond market development.
- Strengthen macroeconomic policy and institutions:
  - Improve fiscal positions and reduce exchange rate volatility to support government securities markets.
  - Enhance law and order, reduce corruption, and improve institutional quality (comprisk, laworder, bureaucracy) to encourage both government and corporate bond markets.
- Develop credit markets:
  - Deepening domestic credit (credit) complements corporate bond market development—support policies that strengthen banking intermediation while encouraging market-based financing.
- Consider capital account and trade policies carefully:
  - Trade openness and capital account openness exhibit mixed effects; policy design should weigh external access benefits against incentives for domestic market development.
- Support market infrastructure and data collection:
  - Continued efforts to compile reliable corporate bond market data (from stock exchanges, securities commissions, central banks) are necessary to monitor and facilitate market growth.

- Key empirical diagnostics:
  - GMM estimations generally pass Hansen and Arellano-Bond AR tests (instrument validity and consistency) across specifications.
  - Hausman tests favor fixed effects for government securities in many cases and random effects for corporate bonds in several specifications; GMM helps address endogeneity concerns.

- Representative exact numeric highlights:
  - Corporate bond capitalization: 1.3 percent of GDP in 2010 (aggregate, Figure 4 statement).
  - Share of corporate bonds in total bonds: 5.1 percent in 2006 and 10.8 percent in 2010.
  - South Africa corporate bond capitalization: almost 20 percent of GDP as of 2010.
  - Sub-Saharan Africa aggregates (2010, Table 4): Government 14.8 percent of GDP; Corporate 1.8 percent of GDP; Contribution of Corporate Bonds to Total Domestic Debt 9.2 percent.
  - Descriptive means (1980–2010, Table 3): Gdebt mean 17.36 (percent of GDP); Cdebt mean 2.31 (percent of GDP); Econsize mean 22.24 (US$ billions); Tradeopen mean 31.88 (percent of GDP); Credit mean 18.73 (percent of GDP); Gdpcap mean 2.21 (US$ thousands); Spread mean 12.83.

*Source: IMF staff compilations.*

### 11. Corporate: Credit and Capitalization

### 11. Corporate: Credit and Capitalization

### Overview of Figures and Variables
- The section presents a series of figures relating corporate bond market capitalization (percent of GDP) and government bond market capitalization (percent of GDP) to a range of macroeconomic, institutional, and market variables.
- Figures compare capitalization to variables for both government and corporate bond markets where available, with sources including National Sources in Table 2, IMF, IFS, WEO, World Bank, ADI, ICRG, Chinn and Ito (2008), BBA, and World Factbook as indicated in each figure caption.

### Capitalization versus Credit and Interest Spreads
- Figure 11 (Corporate): Corporate Bond Market Capitalization (percent of GDP) and Private Sector Credit (percent of GDP)
  - Axes shown: Corporate Bond Market Capitalization (percent of GDP) with range "0 5 10 15 20 25"
  - Private Sector Credit (percent of GDP) with range "0 50 100 150"
  - Sources: National Sources in Table 2, World Bank, ADI
- Figure 13 (Corporate): Corporate Bond Market Capitalization (percent of GDP) and Bank Interest Rate Spread (Lending Rate Minus LIBOR)
  - Axes shown: Corporate Bond Market Capitalization (percent of GDP) with range "0 5 10 15 20 25"
  - Bank Interest Rate Spread (Lending Rate Minus LIBOR) with range "0 10 20 30 40"
  - Sources: National Sources in Table 2, World Bank, ADI

### Capitalization and Interest Rate Volatility
- Figure 15 (Corporate): Corporate Bond Market Capitalization (percent of GDP) and Standard Deviation of Log Interbank Rate
  - Axes shown: Corporate Bond Market Capitalization (percent of GDP) with range "0 5 10 15 20 25"
  - Standard Deviation of Log Interbank Rate with range ".2 .4 .6 .8 1"
  - Sources: National Sources in Table 2, IMF, IFS

### Capitalization and Exchange Rate Volatility
- Figure 17 (Corporate): Corporate Bond Market Capitalization (percent of GDP) and Standard Deviation of Change in Log Nominal Exchange Rate (LCU per US$)
  - Axes shown: Corporate Bond Market Capitalization (percent of GDP) with range "0 5 10 15 20 25"
  - Standard Deviation of Change in Log Nominal Exchange Rate (LCU per US$) with range ".0 5 .1 1.5 2.2 5" (as printed)
  - Sources: National Sources in Table 2, IMF, IFS

### Capitalization and Capital Account Openness
- Figure 19 (Corporate): Corporate Bond Market Capitalization (percent of GDP) and Capital Account Openness (Chinn-Ito Index)
  - Axes shown: Corporate Bond Market Capitalization (percent of GDP) with range "0 5 10 15 20 25"
  - Capital Account Openness (Chinn-Ito Index) with range "-2 -1 0 1 2 3"
  - Sources: National Sources in Table 2, Chinn and Ito (2008)

### Capitalization and Fiscal Balance
- Figure 21 (Corporate): Corporate Bond Market Capitalization (percent of GDP) and Moving Average of Overall Fiscal Balance (percent of GDP)
  - Axes shown: Corporate Bond Market Capitalization (percent of GDP) with range "0 5 10 15 20 25"
  - Moving Average of Overall Fiscal Balance (percent of GDP) with range "-20 -10 0 10 20"
  - Sources: National Sources in Table 2, World Bank, ADI

### Capitalization and GDP per Capita
- Figure 23 (Corporate): Corporate Bond Market Capitalization (percent of GDP) and GDP per Capita, PPP (US$1,000s)
  - Axes shown: Corporate Bond Market Capitalization (percent of GDP) with range "0 5 10 15 20 25"
  - GDP per Capita, PPP (US$1,000s) with range "0 5 10 15"
  - Sources: National Sources in Table 2, World Bank, ADI

### Capitalization and Institutional Quality Indicators
- Figure 25 (Corporate): Corporate Bond Market Capitalization (percent of GDP) and Law and Order (Index)
  - Axes shown: Corporate Bond Market Capitalization (percent of GDP) with range "0 5 10 15 20 25"
  - Law and Order (Index) with range "1 2 3 4 5 6"
  - Sources: National Sources in Table 2, ICRG
- Figure 27 (Corporate): Corporate Bond Market Capitalization (percent of GDP) and Corruption (Index)
  - Axes shown: Corporate Bond Market Capitalization (percent of GDP) with range "0 10 20 30"
  - Corruption (Index) with range "0 1 2 3 4 5"
  - Sources: National Sources in Table 2, ICRG
- Figure 29 (Corporate): Corporate Bond Market Capitalization (percent of GDP) and Legal Origin (Dummy Variable)
  - Axes shown: Corporate Bond Market Capitalization (percent of GDP) with range "0 5 10 15 20 25"
  - Legal Origin (Dummy Variable) with range "0 .2 .4 .6 .8 1"
  - Sources: National Sources in Table 2, World Factbook
- Figure 31 (Corporate): Corporate Bond Market Capitalization (percent of GDP) and Investment Profile (Index)
  - Axes shown: Corporate Bond Market Capitalization (percent of GDP) with range "0 5 10 15 20 25"
  - Investment Profile (Index) with range "2 4 6 8 10 12"
  - Sources: National Sources in Table 2, ICRG
- Figure 33 (Corporate): Corporate Bond Market Capitalization (percent of GDP) and Bureaucracy (Index)
  - Axes shown: Corporate Bond Market Capitalization (percent of GDP) with range "0 5 10 15 20 25"
  - Bureaucracy (Index) with range "0 1 2 3 4"
  - Sources: National Sources in Table 2, ICRG
- Figure 37 (Corporate): Corporate Bond Market Capitalization (percent of GDP) and Composite Risk (Index)
  - Axes shown: Corporate Bond Market Capitalization (percent of GDP) with range "0 5 10 15 20 25"
  - Composite Risk (Index) with range "40 50 60 70 80 90"
  - Sources: National Sources in Table 2, ICRG

### Capitalization and Geography / Size
- Figure 35 (Corporate): Corporate Bond Market Capitalization (percent of GDP) and Country Size (Square kilometers in thousands)
  - Axes shown: Corporate Bond Market Capitalization (percent of GDP) with range "0 5 10 15 20 25"
  - Country Size (Square kilometers in thousands) with range "0 500 1000 1500"
  - Sources: National Sources in Table 2, World Bank, ADI

*Source: _wp1312 - 11. Corporate: Credit and Capitalization (figures and captions as provided in the source PDF).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1312.pdf_