Small Financial Systems: Some Stylized Facts1
The challenges of small financial systems are well known. The provision of financial services is often more expensive and more limited in small financial systems compared to larger ones due to limited economies of scale at both the level of each financial institution and the financial system as a whole. Financial service costs are mostly fixed, which means that the cost of financial intermediation decreases as the number of transactions, customers, and institutions increases. Fewer transactions and fewer customers can sustain fewer financial institutions and less competition. They are also more expensive to operate as they face a number of costs including branch network, IT systems, and corporate governance structures that are largely independent of the number of customers served. Similarly, the costs of regulating and supervising a financial system and the cost of payments systems and other financial infrastructure are also largely independent of the number of institutions that make up the financial system. Moreover, it is often difficult for small financial systems to diversify risks, which can further constrain the size of the financial system in small markets and contribute to financial stability risks.2
A. Countries with Small Financial Systems
1. In 2013, out of 170 reporting countries, 19 countries had financial systems with a total size as measured by M2 (currency outside of banks and demand deposits) of less than US$1 billion (Figure 1, Table 1). As Bossone, Honohan, and Long (2002) point out, this is no larger than a small bank in a developed country.3 Small populations characterize countries with small financial systems. Among the 19 countries, population size ranges from less than 100,000 in Dominica and the Seychelles to 10 million in Burundi and is on average less than 2 million. They are also prevalent in small economies as measured by GDP. In terms of GDP per capita, the group of countries varies widely, from low income countries to upper middle-income countries. About half of the countries are located in sub-Saharan Africa (10), followed by East Asia Pacific (6) and Latin American and the Caribbean (3). (Given that a disproportionate number of countries with small populations and low GDP per capita do not report M2, the true number of countries with small financial systems is higher than 19).
Figure 1.Size of Financial System (M2) in 170 Countries
Source: World Bank.
|Country||Region||Income Group||M2 (USD million)||Population (million)||GDP (USD million)||GDP per capita (USD)|
|Central African Republic||Sub-Saharan Africa||LI||432||4.711||1,545||328|
|Dominica||Latin America & Caribbean||UMI||481||0.072||517||7,175|
|Gambia, The||Sub-Saharan Africa||LI||497||1.867||891||477|
|Grenada||Latin America & Caribbean||UMI||759||0.106||836||7,890|
|Micronesia, Fed. Sts.||East Asia & Pacific||LMI||146||0.104||316||3,049|
|Samoa||East Asia & Pacific||LMI||323||0.190||796||4,181|
|Sierra Leone||Sub-Saharan Africa||LI||862||6.179||4,928||798|
|Solomon Islands||East Asia & Pacific||LMI||472||0.561||1,060||1,890|
|St. Vincent and the Grenadines||Latin America & Caribbean||UMI||522||0.109||719||6,575|
|São Tomé and Príncipe||Sub-Saharan Africa||LI||116||0.182||311||1,703|
|Timor-Leste||East Asia & Pacific||LMI||500||1.180||1,468||1,244|
|Tonga||East Asia & Pacific||UMI||201||0.105||433||4,117|
|Vanuatu||East Asia & Pacific||LMI||569||0.253||802||3,167|
2. It is estimated that US$1 billion is a low threshold: 70 countries out of 170 countries had financial systems of less than US$10 billion in terms on M2. By comparison, 63 banks in the United States had more than US$15 billion in average assets at the end of 2013 (US Federal Reserve Bank, FRED database). The average population size for countries with financial systems in terms of M2 between US$1 billion and US$10 billion is 7.5 million, if excluding the Democratic Republic of Congo, which, with a population of 72 million, is twice as large as the next populous country in this category.
3. Compared to 10 years ago, the number of countries with financial systems with M2 smaller than US$10 billion (in 2013 dollars) has decreased from 104 in 2003 to 70 in 2013 (Table 2). As their economies have grown and their financial sectors have expanded, some countries moved from having financial systems with M2 below US$1 billion (in 2013 dollars) to financial systems with M2 between US$1 billion and US$10 billion (in 2013 dollars) while in other countries the financial sector now clears the threshold of US$10 billion. The latter category includes countries with relatively large populations (such as Kenya), which have been able to expand the size of their financial systems as their economies grew and their financial sectors were liberalized. However, many countries with small financial systems are too small in terms of population to develop large financial systems serving their domestic economy.4
|Countries with M2 less than USD 10 billion (in 2013 dollars)||104||70|
|of which M2 less than USD 1 billion (in 2013 dollars)||49||19|
|of which M2 between USD 1 billion to USD 10 billion (in 2013 dollars)||55||51|
4. Smallness and the issues facing small financial systems are of course not limited to countries falling below some thresholds. Rather, as some of the analysis below shows, the issues facing small financial systems are also relevant to relatively small systems beyond those thresholds along the continuum of size, albeit to a lesser extent.
B. Some Stylized Facts of Small Financial Systems
5. Given the presence of economies of scale, the absolute size of financial systems matters but are small financial systems also small in relative terms? A scatter plot of private credit as a percent of GDP against size and a tabulation of financial depth by size show that small financial systems are not only small in absolute terms as measured by M2, but also in relative terms (Figure 2 and Table 3). In 2013, very small and small financial systems on average reported a private credit to GDP ratio of around 30 percent. This compares to an average ratio of 45 percent in other developing countries and 101 percent in high-income OECD countries. However, as reflected by the scatter plot of private credit as a percent of GDP against M2, factors other than the absolute size influence financial depth and policy can play a potential role in further deepening financial sectors even for a given absolute size.
Figure 2.Private Credit by Banks (Percent of GDP) vs. M2
Source: World Bank.
|M2 (as % of GDP)||Private Credit provided by|
banks (as % of GDP)
|Very small (M2 less than USD 1 billion in 2013)||36.6||48.0||19.1||29.0|
|Small (M2 between USD 1 and 10 billion in 2013)||42.7||50.5||26.3||35.2|
|Other developing (M2 more than USD 10 billion in 2013)||63.7||63.8||39.4||44.8|
|High income: OECD||89.3||108.4||84.7||101.2|
6. Compared to countries with larger financial systems, small financial systems are characterized by fewer financial institutions and a higher share in the number of foreign institutions and the percentage of total bank assets held by foreign banks (Table 4).5 Given the typically smaller market size of countries with small financial systems in terms of overall population and often also in terms of income, the smaller number of banks is expected. On average, very small and small financial systems have 9 banks compared to 32 banks in other developing countries and 52 banks in high-income OECD countries. The share of foreign banks in very small and small financial systems is 62 percent (number) and 75 percent (assets), respectively, about twice as large as in other developing and high-income OECD countries. The higher share of foreign banks in small financial systems might just be a reflection that any given number of foreign banks leads to very different shares of foreign bank ownership depending on the number of domestic institutions. But it can also point to the fact that smaller systems are more open because the presence of foreign financial institutions can help small financial systems address some of the challenges they face such as making the system more competitive and efficient by importing knowledge, risk-management skills, governance, access to capital and economies of scale. However, the presence of foreign banks can also raise questions in terms of foreign banks crowding out the market instead of introducing competition, increasing contagion risk, and overwhelming supervisory capacity.6
|Average number of banks||Average % of foreign banks among total number of banks||Average % of foreign bank assets among total bank assets|
|Very small and small (M2 less than 10 billion in 2013)||9||62%||75%|
|Other developing (M2 more than USD 10 billion in 2013)||32||38%||34%|
|High income: OECD||51||34%||31%|
7. Asset concentration among the three largest banks is highest in very small and small financial systems at 82 percent; it is 55 percent in other developing countries and 70 percent in high-income OECD countries. The scatter plot of asset concentration among the largest three banks and M2 shows there is a large variation in the concentration ratios across financial systems, again pointing to other influencing factors and the potential role of policy in determining the level of concentration (Figure 3). Highly concentrated systems are typically thought to be less competitive; however, in small financial systems there might be a trade-off between competition and size given the prevailing economies of scale in finance, which might make relatively higher levels of concentration desirable.
Figure 3.Largest Three Banks’ Asset Concentration vs. M2
Source: World Bank.
8. Financial systems across all sizes appear to be well-capitalized. Asset quality in small financial systems is on average somewhat lower than in countries with larger systems (Table 5). In 2013, average non-performing loans (NPLs) as a percent of total gross loans stood at 8.7 percent in countries with very small or small financial systems compared to 7 percent in other developing countries and 5.8 percent in high-income OECD countries. While variation across countries along the size continuum does exist, this relationship is also illustrated in a scatter plot (Figure 4). One factor that might explain higher NPL ratios in very small and small financial systems might be relatively scarcer risk management skills both at the level of banks and at the level of regulation and supervision. It might also be a reflection of the generally greater difficulty to diversify risks in small financial systems.
|Bank capital to assets (%)||Regulatory capital to risk (%)||NPLs to total gross loans (%)||Provisions to NPLs (%)|
|Very small and small (M2 less than USD 10 billion in 2013)||12.4||19.3||8.7||58.6|
|Other developing (M2 more than USD 10 billion in 2013)||10.8||16.6||7.0||95.3|
|High income: OECD||7.9||16.0||5.7||48.5|Figure 4.NPL (% total loans) vs. M2
Source: World Bank.
9. The profitability of banks varies by financial sector size. Return on assets and return on equity in very small financial systems are on average higher than in high-income OECD, but about half of those in countries with small financial sectors, which have the highest, average profitability (Table 6). This indicates that contestability remains limited overall especially in small financial systems despite the large share of foreign banks in these countries as documented above.
|Return on Assets (%)||Return on Equity (%)||Private credit to deposits (%)||Net interest margin (%)||Non-interest income to total income (%)||Cost to income||Overhead costs to total assets (%)||Interest rate spread (%)|
|Very small (M2 less than USD 1 billion in 2013)||0.89||8.24||69.09||6.13||38.86||70.62||5.07||9.25|
|Small (M2 between USD 1 and 10 billion in 2013)||1.96||16.66||85.33||5.55||40.47||56.98||4.13||9.67|
|Other developing (M2 more than USD 10 billion in 2013)||1.29||12.29||87.35||4.64||34.09||54.27||3.19||6.13|
|High income: OECD||0.22||3.39||111.91||1.76||39.77||63.53||1.52||2.82|
10. Operational efficiency in countries with small financial systems is typically lower compared to countries with larger financial systems (Table 6). Given the presence of economies of scale, it comes as no surprise that countries with smaller financial systems have higher cost to income and overhead costs to assets ratios (Figure 5). The interest rate spread—the difference in lending rate minus deposit rate—is also higher in smaller financial systems and decreases as the financial system size increases (Figure 6). Higher interest rate spreads might reflect higher lending risk, but also the higher overhead costs typically found in smaller financial systems. Other factors, including statutory reserve ratios and provisioning requirements, also explain differences in interest rate spreads. Higher interest rate spreads in small financial systems might also explain at least in part higher average net interest margins in small financial systems. In contrast, the ratio of noninterest income to total income does not appear to vary systematically with financial system size.
Figure 5.Overheads (% total assets) vs. M2
Source: World Bank.
Figure 6.Interest Rate Spread vs. M2
Source: World Bank.
11. Another measure of efficiency is the rate at which the banks are able to transform savings into loans. The private credit to deposits ratio is lower in countries with very small financial systems compared to the ratio in countries with small financial systems and other developing countries. This might be a reflection that banks perceive lending exposure as relatively more risky in these countries, but it could also be a reflection of more limited supply of projects to invest in or a combination of both.
12. In terms of access to financial services, smaller financial systems typically have lower account ownership rates among adults and fewer bank branches and ATMs per 100,000 adults (Table 7). A simple average of account ownership rates across countries shows that countries with very small or small financial systems show an account ownership rate of 33 percent among adults compared to one of 48 percent in other developing countries. Data on financial access points show that while differences in bank branches by financial system size in developing countries are relatively small, they are more pronounced for ATMs with on average 22 ATMs per 100,000 adults in very small financial systems, 30 in small financial systems, and 38 in other developing countries.7
|Account ownership (% of adults)||Commercial bank branches per 100,000 adults||ATMs per 100,000 adults|
|Very small (M2 less than USD 1 billion in 2013)||16||22|
|Small (M2 between USD 1 and 10 billion in 2013)||14||30|
|Very small and small (M2 less than USD 10 billion in 2013)||33%|
|Other developing (M2 more than USD 10 billion in 2013)||48%||20||38|
|High income: OECD||93%||32||94|
This is a summary of analytical work done during the preparation of the São Tomé and Príncipe Financial Sector Development Implementation Plan by Dorothe Singer and Julian Casal, both World Bank.
Bossone, Honohan, and Long (2002) provide a discussion of how small financial systems differ from larger ones and propose policies that can help small financial systems overcome some of these challenges. In particular, they show how openness can help offset some of the disadvantages of being small. Work by Beck and De La Torre (2007), Barajas et al. (2013), and Beck and Feyen (2013) benchmarks the size, depth, and performance of financial systems relative to structural characteristics of countries such as size in terms of economic activity and population. They find that poorer countries and countries with smaller populations have smaller, shallower and poorer performing financial sectors.
Adjusted for inflation, US$1 billion in 1999 is equivalent to US$1.35 billion in 2013. See also the US Federal Reserve Board’s FRED database for the size of US banks.
Some countries with small populations have developed large financial sectors through the provision of offshore financial services. However, it is not clear that offshore financial services contribute to the quality of depth of onshore financial services in small developing economies (Bossone, Honohan, and Long, 2002).
Data on the characteristics for a large number of countries (136) are available in Claessen and van Horen’s (2014) Bank Ownership Database. The most recent year for which the data are available is 2009.
For an overview of the benefits and risks of foreign banks, see Beck et al (2014) and, for a specific overview as relevant to small financial systems, see Bossone, Honohan, and Long (2002).
Data on firm access to financial services are available from the World Bank’s Enterprise Surveys but country coverage among very small and small financial systems is too poor to calculate meaningful averages.