Using bank survey data, this article compares the scale and type of bank lending to SMEs in Africa to that available outside the region. The data shows that bank financing to SMEs in Africa is less significant, more short-term, and more expensive than in other developing countries.
Policy-makers all over the world show a keen interest in the subject of SME financing. This can be explained by both the contribution that SMEs make to the private sector and the common perception that SMEs are credit constrained. These are significant factors for African economies where SMEs account for close to 50 percent of employment2. Furthermore, enterprise surveys conducted by the World Bank reveal that close to 40 percent of small firms and 30 percent of medium-sized firms perceive access to finance as a major stumbling block to the growth of their operations3.
African banking survey
Using newly gathered data collected through a survey of banks conducted in 2007/8, this article quantifies the scale and type of bank lending to SMEs in Africa and compares it to lending available to SMEs in the rest of the developing world (see also Beck et alii, 2008). We specifically analyze information obtained from 16 banks in 8 African countries (Ethiopia, Kenya, Malawi, Sierra Leone, South Africa, Swaziland, Zambia, and Zimbabwe) and from 64 banks operating in 30 developing countries outside of the region. The banks included in the survey are among the largest 5 in each country.
We compare bank financing to SMEs in Africa vis-à-vis financing available in other developing countries from various standpoints. First, we examine the scale of financing by comparing the share of loans directed to SMEs (as a percentage of total lending) and the percentage of applications approved (out of total SME loan applications received). Second, we consider the type of lending by comparing the percentage of SME loans devoted to investment and the share of SME loans that are secured. Third, we examine the cost of SME lending by comparing the fees and interest rates charged on these loans. Finally, we analyze the riskiness of SME loans by comparing the share of non-performing SME loans in Africa with the rest of the developing world. The table displays the results.
Low levels of financing in Africa
Our data shows that bank financing to SMEs in Africa is less significant and more short-term than in other developing countries. This is particularly the case for small firm financing. In the large group of developing countries, bank loans devoted to financing small firms average 13.1 percent, while only 5.4 percent of loans are allocated to such firms by banks in Africa. Similarly, bank approvals for loan applications by small firms in non-African developing countries average 81.4 percent, whereas only 68.7 percent of such applications are approved by banks in Africa. Furthermore, while an average of 47 percent of small business loans are used to finance investments (as opposed to working capital) in non-African developing countries, the figure only reaches 28 percent in African countries. Importantly, all these differences are statistically significant.
Bank lending to SMEs in Africa is also more costly than in other developing countries. Fees charged on SME loans in Africa – an average of 1.97 percent of the loan value for small firms and 1.79 percent for medium-sized firms – are generally almost twice as high as in other developing economies. Interest rates on SME clients are also 5 to 6 percentage points higher on average in Africa than elsewhere in the developing world. For instance, banks in Africa charge on average close to 15.6 percent for loans to their best small firm borrowers, whereas interest rates in other developing countries hardly exceed 11 percent for these clients.
At the same time, SME loans in the region appear to be riskier than those in other developing countries. This may be at the root of the higher interest rates observed in Africa. Indeed, the share of non-performing loans (NPLs) among small firm loans in Africa averages 14.5 percent compared to 5.5 percent in other developing countries. The NPL ratio for medium-sized firms is also higher in Africa (6.8 percent) than in other countries (5.1 percent), but the difference in means is not statistically significant.
Why African banks are reluctant to finance SMEs
How can we explain the differences found in the scale, cost and riskiness of bank financing to SMEs in Africa compared to the rest of the developing world? While a full analysis of this issue is not possible here, banks’ responses to questions on the factors that drive and impede SME financing across countries can prove revealing (see figures opposite). In non-African developing countries, more than three quarters of banks respond that the perceived profitability of the SME segment is a key driver for their involvement with SMEs, whereas only two thirds of African banks point to this factor as being important. At the same time, less than 40 percent of banks in non-African developing countries indicate that macroeconomic factors are a significant obstacle to their involvement with SMEs, yet 60 percent of banks in Africa point to the state of the economy as constraining their involvement with these businesses. These responses suggest that policy-makers in Africa have a role to play – more so than elsewhere in the developing world – in helping to promote SME finance by adopting sound macroeconomic policies that reduce risks and increase profitability when doing business in Africa and financing firms in the region.
¹ The views expressed in this article are my own and do not represent the opinions of the World Bank or its Executive Directors.
² IFC Micro, Small and Medium Enterprises Database. http://rru.worldbank.org/Documents/other/MSMEdatabase/msme_database.htm
³ Calculations made using data from enterprise surveys available at http://www.enterprisesurveys.org.
References / Beck, T., Demirguc-Kunt, A.N Martinez Peria, M.S., 2008. Bank Financing for SMEs around the World: Drivers, Obstacles,Business Models and Lending Practices, World Bank Policy Research Working Paper 4785.