Journal Issue

Financial Inclusion and Stability In Africa’s Middle-Income Countries Including Namibia

International Monetary Fund. African Dept.
Published Date:
February 2014
  • ShareShare
Show Summary Details

Financial Inclusion and Stability In Africa’s Middle-Income Countries Including Namibia1

The financial sectors of middle-Income countries (MICs) in sub-Saharan Africa (SSA) continue to evolve and recent global regulatory reforms will likely steer funding structures further towards equity and deposits. This chapter examines two questions. What is the financial landscape (stability and inclusion) in SSA MICs relative to emerging market economies (EMs)? How do financial inclusion and shifts in funding structure of banks affect financial stability? Our findings suggest that: (i) financial stability and access to finance for households in SSA is comparable to EMs, while access to finance for small and medium-sized enterprise (SMEs) lags behind those of EMs; (ii) SME access to finance and/or savings-oriented financial inclusion enhances financial stability, while overly consumption-oriented financial inclusion likely undermines financial stability; and (iii) a more equity and deposit-based funding structure enhances financial stability.

A. Introduction

1. The rapidly expanding financial sectors in MICs in SSA, including Namibia, plays a critical role in the general economy of these countries. Given that this role is also expected to expand over time, the financial environment itself must be understood in greater detail. First, as we know, the financial sector is becoming more and more important in IMF surveillance. Thus, it is important to have a better knowledge of how the financial landscapes of SSA MICs compares with those of EMs.

2. While financial stability has been central to the IMF’s work, over the recent years, financial inclusion/access to finance has become increasingly important. Financial stability and financial inclusion are often considered to be conflicting goals, just as fighting inflation and maximizing growth are often at odds in a central bank’s objective function. For example, financial stability focuses on asset quality. This means that credit institutions (including banks) need to be more selective about whom they lend and provide financial services. In contrast, financial inclusion focuses on quantity such as size of assets and the number of customers (in percent of population). This often means that credit institutions should lower their standards for lending and reduce the minimum balance requirement or fees associated with opening or maintaining a bank account. A natural question therefore is how different types of financial inclusion tend to affect financial stability.

3. The current global regulatory reforms will likely encourage banks’ funding structures to shift towards greater holdings of equity and deposits. This raises the question will this funding structure shift impact financial stability in SSA MICs?

4. In this chapter, we address each of the above issues by exploring the following questions: How has the financial landscape in SSA MICs evolved? How does the financial landscape (stability and inclusion) in SSA MICs compare to that of EMs? How do financial inclusion and funding structure shifts affect financial stability?

5. Our findings suggest the following: (i) financial stability and financial inclusion for households in SSA MICs are comparable to a group of EMs while financial inclusion for SMEs lags behind that in EMs;2 (ii) aspects of financial inclusion that focuses on expanding SMEs access to finance and individuals’ access to savings accounts enhances financial stability, while financial inclusion that focuses purely on expanding the percentage of individuals with credit undermines financial stability; and (iii) a more equity and deposit based funding structure enhances financial stability.

B. Literature Review

6. This chapter explores the nexus between financial inclusion and stability. Much work has been done in the literature on financial sector stability and a growing body of ongoing work is focusing on financial access as well. Here, we present an overview of the literature, as it relates to this chapter, and give a sense of the limited work available on exploring the interactions between financial stability and access.

7. The literature on financial stability is vast and diverse, with most of it focused on measures and indicators for financial stability (FSIs).Demirgüç-Kunt and Detragiache (1998), Kaminsky (1998), and Bordo and Schwartz (2000), among others, pioneered early warning indicators on macro-financial stability based on risk spreads, and market liquidity. The primary FSIs cover a few key banking areas: capital adequacy, asset quality, management effectiveness, earnings, liquidity, and sensitivity to market risks.

8. Work has also been done to look at predictors of banking crises.Demirgüç-Kunt and Detragiache (1997) use a multivariate logit estimation to identify determinants of banking crises in a panel of developing and industrialized countries. They found that weak macroeconomic environments with low growth, high inflation, high real interest rates, explicit deposit insurance schemes and weak law enforcement were particularly vulnerable to economy-wide banking crises. Demirgüç-Kunt and Detragiache (2005) survey the body of work done on crisis prediction and identify two main methodologies of the cross-country empirical work in this field based on the signals approach and the multivariate probability model.

9. Other stands of the literature on financial stability have looked at the central bank’s role in financial sector stability (Nier 2009), bank competition and stability (Berger et al. 2009).3 Other work includes financial liberalization and crisis (Caprio and Summers 1993), external shocks and crisis (Eichengreen and Rose 1998), bank ownership and structure as related to crisis (La Porta, Lopez-de-Silanes, and Shleifer 2002) and the role of institutions and the political system in causing and preventing crises (Beck, Demirgüç-Kunt, and Levine 2004).

10. The literature on financial access focuses on the effect of individual access on income inequality, poverty, and GDP growth. The Global Financial Index (Findex) surveys on “how adults in 148 economies save, borrow, make payments, and manage risk” finds that high cost, physical distance, and lack of proper documentation are the most common barriers to household access to finance (WB GFDR 2013, Demirgüç-Kunt & Klapper 2012). “Policies and Pitfalls in Expanding Access” - Demirgüç-Kunt, Beck and Honohan. (2008) illustrates that financial access is quite limited around the world and identifies barriers that may be preventing small firms and poor households from using financial services. Based on this research, the report derives principles for effective government policy on broadening access. Beck, Demirgüç-Kunt, and Maksimovic (2003) explores the effects of firm-level financial access and find that financial constraints are strongest for small firms and weakening these constraints disproportionately benefits smaller firms.

11. The literature, is however, limited on the interactions between financial stability and access.Hannig and Jansen (2010) acknowledge that financial inclusion poses risks in terms of reputation and quality, low-income savers and borrowers maintained “solid financial behaviour” through crises periods and that the presence of “vulnerable clients” in the financial system has negligible risks. They show various graphical correlations such as a positive correlation between GDP per capita and inclusion, but stop short of establishing a causal quantitative relationship. Similarly, Khan (2011) provides a graphical correlation showing the positive relationship between financial inclusion, as indicated by, and development as measured by World Bank development levels (High Income, Low Income, etc.). This, however, does not provide a causal link or direction of causation. Aduda and Kalunda (2012) explore financial inclusion and stability with reference to Kenya and postulate that it is very likely that banking performance, and the likelihood of crises, may depend on the structure and degree of development of the financial systems, which is one of the focal point in financial inclusion. However, there is no quantitative analysis performed in that paper and the observation is purely speculative.

12. The chapter addresses a different question and provides quantitative evidence on the impact of financial inclusion and funding structure shifts on financial stability. Rather than focusing on a qualitative or descriptive assessment based on correlations, we attempt to establish quantitatively whether a causal link exists among financial inclusion, funding structure and stability. We also look at whether and which types of financial inclusion exist that are stability friendly at the bank-level. This approach is taken in the context of SSA for several reasons. First, the banking sector in SSA is relatively small; with 227 banks in the sample, solvency and stability of each individual bank can have great ramifications for MICs in SSA. Second, financial access levels can have differing impacts on heterogeneous banks and a bank-level analysis can account for fundamental differences between individual banks that may be affecting stability, which helps to isolate the effect of each individual explanatory variable without use of aggregation and its resulting distortions.

C. Data

13. The dataset includes bank-level data for 227 banks in SSA, over the period 1998-2013. Eleven SSA MIC countries are covered: Botswana, Cape Verde, Ghana, Lesotho, Mauritius, Namibia, Senegal, Seychelles, South Africa, Swaziland, and Zambia. Country-level financial stability and access data for a group of EM countries is also included for 2011 for benchmarking purposes. The EM countries are: Argentina, Brazil, Chile, China, Colombia, Hungary, Indonesia, India, Republic of Korea, Mexico, Malaysia, Peru, Philippines, Poland, Romania, Russia Federation, Thailand, Turkey and Ukraine. The selection criteria of the EM group are based on the peer group of South Africa (South Africa 2013 Article IV Consultation Staff Report 13/303).

14. Our data sources include: bank-level financial data from Bankscope, macroeconomic indicators data from the World Bank Indicators and Bloomberg, individual financial access data from WB Development Indicators and Financial Access Survey and SME financial access data from the IFC SME Financial Access Survey.

D. Analytical Framework

15. The framework centers around two econometric models of the determinants of financial stability. The first is a baseline Probit model with the probability of bank distress because the dependent variable. The second is a standard OLS estimation with percentage deviation from bank insolvency as the dependent variable.

Baseline probit model

16. In this section, we attempt to explore the effects of financial access on financial stability and, particularly, whether households or SME financial access affects distress probability differently. Traditional measures of bank distress in the literature include the book-price ratio, analyst ratings and the Z-score. We focus on the Z-score measure of bank distress as it has become the most frequently used indicator, in addition to having greater data availability. The Z-score measure of bank stability equals the return on assets (ROA) plus the capital asset ratio (CAR) of each bank divided by the banks’ standard deviation of return on assets. It proxies the risk of bank insolvency as it is the inverse of the probability that losses exceed equity; that is, a higher Z-score implies lower risk of insolvency (see Box 1 for details of its derivation).

Box 1.Derivation of Z-Score

Banks’ probability of distress is defined as the probability that it defaults, i.e. consolidated profit is less than consolidated equity:

Let the following notation hold: p(π˜<-E)=p(π˜A<-EA)

Then, assuming that the distribution of ROA satisfies r˜π˜A~N(μ,σ)r˜-μσ(r˜)~N(0,1)

Then, probability of bank distress can be written as:

Thus the definition of the bank Z-score, Z(EA+ROA)σ(AOA), is a direct inverse measure of the likelihood of bank distress.

17. The traditional factors affecting bank distress can be categorized by: (1) funding structure (e.g., Herfindahl funding diversity index, loans to customer deposits, short-term funding to assets, equity to assets, term deposits to assets), (2) profitability and asset quality (return on average assets, return on average equity, loan loss provisions to gross loans net interest margin), (3) size (total assets, asset growth) and (4) macroeconomic factors (e.g. inflation, output growth). Thus, in tackling our question, we must control for the effects of these other factors.

18. We proceed by first estimating a Probit model of probability of financial distress:

where P{} is the probability that bank yi from country j will be in distress at time t, conditional on bank-specific and country-level characteristics Xijt-1 and Wjt. P{} is based on the Z-score and is a decreasing function of the Z-score as higher levels of Z imply lower probability of distress. F() is the standard normal distribution function that transforms a linear combination of the explanatory variables into the [0,1] interval. The estimations use lagged bank-level explanatory variables in order to reduce endogeneity concerns and report robust standard errors.

19. Distressijt is measured by bank-level Z-scores, with a threshold at the 10th percentile of Z-scores within the sample4, which is equivalent to being above the 10th percentile in probability of default (Box 1). The presence of bank-specific lagged explanatory variables, Xijt1, is primarily to reduce endogeneity concerns and report robust standard errors, but also to control for bank-level characteristics (e.g. size) that may make an individual bank particularly sensitive or insensitive to country-wide macroeconomic conditions. Finally country-specific explanatory variables, Wjt, must also be included to control for macroeconomic conditions that can obviously affect a given bank’s default probability.

Baseline Logarithmic Model

20. We look at the relationship in levels and ensure that our Probit results are not sensitive to the choice of binary cutoff threshold by examining the impact of the explanatory variables on the percentage change in Z-score, that is, ln(Z-score). We estimate a standard linear regression of bank-level explanatory variables, Xijt-1, and macro-level stability variables, Wjt, on ln(Z-score):

21. In order to include observations corresponding to negative values of the Z-score, which cannot be log-transformed, we adjust additively relative to the minimum Z-score. This technique uses ln(Z-score+min(Z-score)+1) in order to keep all observations under the previous binary dependent variable estimation in this analysis as well. All other explanatory variables are kept the same. The estimation again uses lagged bank-level explanatory variables in order to reduce endogeneity concerns and report robust standard errors.



  • P{Z< 10th percentile} is a binary dependent variable representing the probability that a bank’s Z-score is below the 10th percentile of Z-scores of regional banks, putting it at risk of default relative to other banks in the sample. As the Z-score represents the adequacy of a bank’s capital to cover potential equity losses and thus is directly and inversely related to the probability of default, higher values of the Z-score correspond to greater solvency. This cutoff is equivalent to a bank’s probability of a default being amongst the top 10 percentile of sample banks (Box 2). Thus positive coefficients on explanatory variables would indicate a negative contribution of that variable towards bank-level stability. The 10th percentile is chosen as a measure in line with the convention in related literature (IMF’s GFSR October 2013); other nearby cutoffs had very similar results for robustness.

Box 2.Derivation of Z-Score Threshold in Binary Estimation

Since Z-score is inversely related to p(distress), if we look at the case when the p(distress)>p10(distress), the 10th percentile of default probabilities, then we have that:

p(distress) > Ф(p10)⇒-p(distress) < Ф(p10)


And since Z = –Ф–1[p(distress)] ⇒ Z < p10, exploring the case of a 10 percentile threshold on distress probability, is equivalent to exploring the case when Z < p10(Z), i.e. Z<2.435


  • In(Z-score) is a continuous dependent variable representing the percentage increase relative to the lowest bank Z-score present in the sample. Under this dependent variable, positive coefficients on explanatory variables correspond to greater solvency and a decreased probability of bank default. This variable is secondary in our analysis because the true relationship between explanatory variables and bank stability and is likely piece-wise linear. Thus, beyond a certain point of high Z-scores, observed explanatory variables may have increasingly weak marginal impact on stability and unobserved variables may carry more weight. This can appear in an estimate as a weakening of the causal link between independent variables (summarized in Table 1) and stability.
Table 1.List of Independent Variables
Bank-level Variables, XijtCountry-level Variables, Wjt
Funding structureAccess/financial inclusion
  • Herfindahl index of funding diversity
  • Adults saving to total adults
  • Loans to customer deposits ratio
  • Adults borrowing to adults
  • Short-Term funding to assets ratio
  • Percent of SMEs identifying access to finance as a major constraint
  • Equity to assets
  • Percent of small firms with a credit line
  • Term deposits to assets
Profitability and asset quality
Profitability and asset quality 
  • GDP per capita
  • Return on Average Assets
  • GDP Growth
  • Return on Average Equity
  • GDP Growth-Bank Size interaction
  • Loan loss provisions to gross loans
  • Interest rate spread
  • Net interest margin
  • Inflation, GDP deflator
  • Banking regulatory quality and disclosure
  • Volatility of stock price index, 360-day Standard Deviation
  • Total Assets
  • Asset growth
  • Human Development Indicator

E. Empirical Findings

22. Tables 2 and 3 present the key results of the estimates. All estimates distinguish between household financial savings versus borrowing as indicators for household financial access. Table 2 gives the Probit estimation results using the SME Access Constraint as self-reported by SMEs within each country in the IFC Enterprise Survey. Table 3 gives Probit results using percentage of small firms with a credit Line among total small firms as the measure of SME financial access. Our analysis also provides marginal effects on the respective Probit estimations. We provide estimation results using ln (Z-score) as the dependent variable as a robustness check on whether our results are sensitive to chosen thresholds for our Probit analysis. However, this has limited utility because marginal differences in Z-score for already very high Z-score levels are unlikely to be significantly affected by changes in explanatory variables.

Table 2.Marginal Effects: Adults Saving versus SME Access ConstraintMarginal Effects: Adult Saving (%) vs. SME Access Constraint (%)
P(z<10th percentile)dy/dxdy/dxdy/dx
Adults Saving %-0.26690**-0.26299**-0.27211**
Adults Borrowing %0.40460**0.39869**0.44858**
SME Actets Constraint0.02655**0.02601**0.03713*
* p<0.10, ** p<0.05, *** p<0.01
* p<0.10, ** p<0.05, *** p<0.01
Table 3.Individual Access versus SME AccessIndividual Access vs. SME Access
Adults Saving%0.07722***0.07735***0.04155**0.04127**
Adults Borrowing%-0.04223*-0.04237*-0.06211**-0.06245**
SME access constraint%-0.05992***-0.06000***  
Small Firm Credit %  0.01814**0.01873**
* p<0.10, ** p<0.05, *** p<0.01
* p<0.10, ** p<0.05, *** p<0.01

23. The Probit estimations under either measure of SME access are broadly similar. Under SME Access Constraint as reported in the World Bank Enterprise Survey, individual bank size, term deposits to assets, equity to assets, country legal index and both individual access as well as SME access variables are significant determinants for bank-level financial stability. The signs of these coefficients give a sense of their contribution to financial stability. Note that in the Probit analysis, our dependent variable is a measure of the likelihood of bank distress, that is, the inverse of financial stability. Bank size, measured as total assets, has a positive and highly significant effect on the probability of distress. The proportion of assets that are term deposits increases financial stability significantly.

24. Our results give the directional impact of each explanatory variable on bank distress, and the statistical significance. Negative coefficients indicate the variable is associated with greater bank stability while positive coefficients indicate that the variable is associated with greater bank distress. In general, we find that higher equity to assets, term deposits to assets, and percentage of adults saving lead to an increase in financial stability. We also find that higher ratios of loans to customer deposits, percentage of adults borrowing, and percentage of SMEs facing financial access constraints and less diverse funding sources lead to increased probability of banking sector distress.

25. The lagged equity to assets and ROA are used in the main empirical analysis to address potential endogeneity issues. However, our analysis also provides robustness checks to test the validity of our findings by taking out equity to assets and ROA. All our previous results are substantively unaffected.

26. Note that the signs of the Probit estimation coefficients give the direction of the effects, but the coefficients themselves do not give a sense of the magnitude of the effect. This is because the coefficient magnitudes are in the units of the standard errors. For a sense of true magnitude, we separately calculate the average marginal effects for the explanatory variables to obtain the discrete change in the probability of bank distress averaging across the sample values of the other predictor variables. For example, to calculate the average predicted probability of distress for a given percentage of adults saving, the predicted probability was calculated for each bank-year, using that bank’s value explanatory variables for that year, and the average was taken across all these predicted probabilities.

27. The marginal effect on the households saving percentage tells us that the derivative of the mean expected probability of bank distress with respect to adults saving is -0.267. This suggests that if we had four banks and then increased the percentage of adults saving by 1 percent this would cause one bank to switch from being likely to default to being unlikely to. The 0.02655 coefficient on SME access constraint suggests that if we had 37 banks and then lowered the SME access constraint by 1 percent, one bank would switch from being likely to default to being unlikely to. The 0.4046 coefficient on percentage of adults borrowing suggests that, of just 2.5 high default probability banks, one would become more solvent and unlikely to default if adult borrowing decreased by 1 percent. The coefficient magnitudes where equity to assets and ROA are removed are nearly identical.

28. The analysis using the percentage small firm credit line variable as a measure of SME financial access shows the same directional effects on all variables but the magnitudes of the effects differ. Here, the marginal effect of the households saving percentage and small firm credit line percentage on the probability of bank distress are both equal to -0.039. This suggests that if we had 26 banks and either increased the percentage of adults saving or increased the percentage of small firms with a credit line by 1 percent this would cause one bank to switch from being likely to default to being unlikely to default. The 0.046 coefficient on adults borrowing suggests that, of the 21 high default probability banks, one would become more solvent and unlikely to default if adult borrowing decreased by 1 percent. The -0.0376 coefficient on percentage of small firms with a credit line suggests that a 1 percent increase in the percentage of small firms with a credit line leads to a 0.038 percent decrease in probability of distress. Removing equity to assets and ROA, column 4, gives nearly identical magnitudes.

29. Our analysis also shows the results obtained using the dependent variable ln(Z-score), renormalized so that negative Z-score observations are also reflected. The first two columns represent the estimation using the SME Access Constraint of firm-level financial access, while the last two columns utilize the small firm credit line percentage measure. The interpretation of these coefficients support the probit estimation results in both the directional effect and the magnitude effect of variables.

30. If we now turn to the SME access constraint measure, we find that a 1 percentage point increase in the percentage of households saving increases the Z-score, and thus improves bank level stability, by 0.0772 percent. A 1 percent increase in adults borrowing reduces the Z-score by 0.0422 percent and a 1 percent increase in SME access constraint is associated with a 0.0599 percent reduction in the Z-score measure of stability. Using the small firm credit line measure of SME financial access, we find consistent results. A 1 percent increase in percentage of adults saving is associated with a 0.0416 percent increase in Z-score-measured bank stability, a 1 percent increase in small firms with a credit line leads to a 0.018 percent increase in the Z-score, while a 1 percent increase in percentage of adults borrowing leads to a 0.0621 percent decrease in stability. As expected and consistent with the probit analysis, term deposits to assets, equity to assets and return on average equity all have positive and significant causal effects on bank stability, while banks’ total asset growth has a negative impact on stability.

F. Conclusions

31. Main findings:

The results of the empirical analysis in this chapter suggest the following broad findings for selected MICs in SSA, including Namibia:

  • First, the financial landscape in SSA MICs has deepened in past decades, as reflected in the increasing share of the stock of private credit to GDP and deposits as ratio to GDP. Furthermore, their return on assets has stabilized closer to the level of a group of EMs, reflecting a more mature and competitive financial sector in SSA MICs.
  • Second, in many MICs in SSA, one reason for the low level of financial inclusion is SME’s lack of access to finance. Our study shows that SME access to finance has a positive and significant impact on financial stability. Financial usage can have stabilizing effects on the financial sector by helping to increase financial sector depth. If borrowing tends to be used for investment or to finance asset purchases generating returns, overall this would be beneficial to the financial sector and the economy in general. Thus, financial inclusion focusing on enhancing SME access to finance tends to enhance financial stability.
  • Furthermore, another reason for the low level of financial inclusion is that relatively poor households do not have access to bank accounts for various reasons such as the minimum balance requirement, fees for opening or maintaining a bank account with low balance, or low presence of financial institutions in lower-income communities. Therefore, financial inclusion focusing on improving households’ access to bank accounts will likely enhance financial stability. Specifically, policy measures such as reducing or eliminating the minimum balance requirement or fees for opening/maintaining bank accounts with lower balance should increase inclusion. Alternatively, using new technology such as e-banking or mobile banking will facilitate the population’s access to finance.
  • Moreover, financial inclusion that overly relies on increasing households’ access to credit by lowering lending standards encouraging low quality creditors to take out unaffordable loans/mortgages can be destabilizing. In the short run, these policies may mechanically increase financial inclusion, but in the long run, this will likely jeopardize financial stability which will ultimately undermine financial inclusion.
  • Finally, the primary components of global regulatory reforms will likely steer banks’ funding structures further toward deposits and equity with less reliance on short-term wholesale funding. This funding structure will likely have a positive impact on financial stability.

32. Policy implications:

We draw the following key policy messages from our findings.

  • First, policies promoting SME sector development should enhance financial stability. Second, reforms that facilitate households’ access to savings accounts will also promote financial stability. However, mechanically expanding the number of people with credit may lead to over indebtedness. This could undermine financial stability. In a number of countries, the authorities tend to overly promote loans to households in pursuit of greater financial inclusion. This has led to over indebtedness of households. Finally, global regulatory reforms will likely have a positive impact on financial stability by shifting banks’ funding structure more towards capital and deposits.
  • However, our results and messages should be interpreted with some caution. There are no one-size-fits-all approaches to striking an appropriate balance between financial inclusion and financial stability. The policy of enhancing financial stability of each country has to consider its country-specific circumstances.

    AdmatiA. R &HellwigM. F. (2013). The bankers’ new clothes : what’s wrong with banking and what to do about it.Princeton: Princeton University Press.

    • Search Google Scholar
    • Export Citation

    AdudaJ. &KalundaE. (2012). Financial Inclusion and Financial Sector Stability with Reference to Kenya: A Review of Literature. Journal of Applied Finance and Banking 2(6)95120.

    • Search Google Scholar
    • Export Citation

    Bank for International Settlements1997. “Financial stability in emerging market economies.” Basel (

    BeckT.H.L. &Demirgüç-KuntA. &MaksimovicV.2004. “Financial and legal constraints to firm growth: Does firm size matter?”Open Access publications from Tilburg University.

    • Search Google Scholar
    • Export Citation

    BeckThorsten &Demirguc-KantAsli &MaksimovicVojislav2003. “Bank competition, financing obstacles, and access to credit,”Policy Research Working Paper Series 2996The World Bank.

    • Search Google Scholar
    • Export Citation

    BeckThorstenAsliDemirgüç-Kunt andMariaMartinez Peria.2009. “Bank Financing for SMEs around the World: Drivers, Obstacles, Business Models, and Lending Practices.”World Bank Policy Research Working Paper 4785World BankWashington, DC.

    • Search Google Scholar
    • Export Citation

    BeckTDemirguc-KuntA andMartinezPeria M S2005. “Reaching out: Access to and use of banking services across countries.

    BergerA.L.Klapper andR.Turk-Ariss.“Bank Competition and Financial Stability.”Journal of Financial Services Research April 2009 Volume 35 Issue 2 pp 99118.

    • Search Google Scholar
    • Export Citation

    BordoMichael D. &SchwartzAnna J.2000. “Measuring real economic effects of bailouts: historical perspectives on how countries in financial distress have fared with and without bailouts,”Carnegie-Rochester Conference Series on Public Policy Elsevier vol. 53(1) pages 81167December.

    • Search Google Scholar
    • Export Citation

    CaprioGerard Jr. &SummersLawrence H.1993. “Finance and its reform : beyond laissez-faire,”Policy Research Working Paper Series 1171The World Bank.

    • Search Google Scholar
    • Export Citation

    Demirgüç-KuntA.T.Beck andP.Honohan2008. Finance for All? Policies and Pitfalls in Expanding Access. World Bank Policy Research Report. Washington: World Bank (

    • Search Google Scholar
    • Export Citation

    Demirgüç-KuntAsli andRossLevine.2008. “Finance, Financial Sector Policies, and Long Run Growth.”M. Spence Growth Commission Background Paper 11World BankWashington, DC.

    • Search Google Scholar
    • Export Citation

    Demirgüç-KuntAsli andDetragiacheEnricaCross-Country Empirical Studies of Systemic Bank Distress: A Survey (September 2005). World Bank Policy Research Working Paper No. 3719.

    • Search Google Scholar
    • Export Citation

    Demirgüç-KuntAsli andDetragiacheEnricaThe Determinants of Banking Crises in Developing and Developed Countries (1998). IMF Staff Paper (Volume 45).

    • Search Google Scholar
    • Export Citation

    Demirgüç-KuntAsli andDetragiacheEnricaThe Determinants of Banking Crises: Evidence from Industrial and Developing Countries (May1997). World Bank Policy Research Working Paper No. 1828.

    • Search Google Scholar
    • Export Citation

    AsliDemirguc-Kunt andLeoraKlapper2012“Measuring Financial Inclusion: The Global Findex Database”World Bank Policy Research Paper 6025.

    • Search Google Scholar
    • Export Citation

    DiamondDouglas W. &RajanRaghuram G.2001. “Banks, Short Term Debt and Financial Crises: Theory, Policy Implications and Applications. Carnegie-Rochester Conference Series on Public Policy.” Elsevier vol. 54(1) pages 3771June.

    • Search Google Scholar
    • Export Citation

    DykF.2010. “The new importance of financial stability’s 2nd dimension: towards Basel III & post the recent credit-liquidity crisis.”Financial Markets Journal.

    • Search Google Scholar
    • Export Citation

    EichengreenBarry &RoseAndrew K1998. “Staying Afloat When the Wind Shifts: External Factors and Emerging-Market Banking Crises,”CEPR Discussion Papers 1828 C.E.P.R. Discussion Papers.

    • Search Google Scholar
    • Export Citation

    FalkenaH B2004. The Access and Stability objectives of regulation. Workshop on tiered banking legislation at the South African Reserve Bank Conference CentrePretoria22November.

    • Search Google Scholar
    • Export Citation

    Financial Stability Board2012. “Identifying the Effects of Regulatory Reforms on Emerging Market and Developing Economies: A Review of Potential Unintended Consequences.

    • Search Google Scholar
    • Export Citation

    FSB IMF and WB2011. “Financial Stability Issues in Emerging Market and Developing Economies.” IMF:

    • Search Google Scholar
    • Export Citation

    HawkinsPenelope2006. “Financial access and financial stability.”Bank for International Settlements 1997.

    HannigAlfred andJansenStefan2010. “Financial Inclusion and Financial Stability: Current Policy Issues”ADBI Working Paper No. 259 (December212010). Available at SSRN: or

    • Search Google Scholar
    • Export Citation

    JohnstonDon andJonathanMorduch2008. “The Unbanked: Evidence from Indonesia.”World Bank Economic Review World Bank Group vol. 22(3) pages 517537October.

    • Search Google Scholar
    • Export Citation

    KaminskyGraciela L.1998. “Currency and banking crises: the early warnings of distress,”International Finance Discussion Papers 629Board of Governors of the Federal Reserve System (U.S.).

    • Search Google Scholar
    • Export Citation

    KrishnamurthyA. andA.Vissing-JorgensenNov. 2012. “Short-term Debt and Financial Crises: What we can learn from U.S. Treasury Supply.”Kellogg SchoolNorthwestern University.

    • Search Google Scholar
    • Export Citation

    La PortaRafaelFlorencioLopez-De-Silanes andAndreiShleifer2002. “Government Ownership of Banks,”Journal of Finance American Finance Association vol. 57(1) pages 265-301 02.

    • Search Google Scholar
    • Export Citation

    NierErlend2009. “Financial Stability Frameworks and the Role of Central Banks: Lessons from the Crisis,”IMF Working Papers 09/70International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    RochaRoberto R.ZsofiaArvai andSubikaFarazi2011. “Financial Access and Stability: A Road Map for the Middle East and North Africa.”World BankWashington, DC.

    • Search Google Scholar
    • Export Citation

    Ruiz-PorrasA.2009. “Financial Structure, Financial Development and Banking Fragility: International Evidence”Análisis Económico Núm. 24 (56).

    • Search Google Scholar
    • Export Citation

    VazquezFrancisco andPabloFederico.Bank Funding Structures and Risk: Evidence from the Global Financial Crisis. IMF Working Paper WP/12/29.

    • Search Google Scholar
    • Export Citation

    The World Bank Global Financial Development Report (2013). World BankWashington, DC.

1This chapter was prepared by Yibin Mu and Jenny Lin (a summer intern during 2013). The key findings were presented to the Namibian authorities by Andrew Jonelis during the 2013 Article IV consultation mission.
2The definition of SMEs is the same across all countries as the paper’s uses the World Bank Enterprise Survey definition on this.
3Under the traditional “competition-fragility” view, more bank competition erodes market power, decreases profit margins, and results in reduced franchise value that encourages bank risk taking. Under the alternative “competition-stability” view, more market power in the loan market may result in greater bank risk as the higher interest rates charged to loan customers make it more difficult to repay loans and exacerbate moral hazard and adverse selection problems. But even if market power in the loan market results in riskier loan portfolios, the overall risks of banks need not increase if banks protect their franchise values by increasing their equity capital or engaging in other risk-mitigating techniques. The Berger et al test these theories by regressing measures of loan risk, bank risk, and bank equity capital on several measures of market power, as well as indicators of the business environment, using data for 8,235 banks in 23 developed nations. The results suggest that - consistent with the traditional “competition-fragility” view - banks with a greater degree of market power also have less overall risk exposure. The data also provide some support for one element of the “competition-stability” view - that market power increases loan portfolio risk. The Berger et al show that this risk may be offset in part by higher equity capital ratios.
4The 10th percentile threshold was chosen in line with that of a recent estimation in the October 2013 Global Financial Stability Report using a similar model of probability of bank default. In that estimation, the 10th percentile of bank Z-scores for that sample was chosen.

Other Resources Citing This Publication