For developing countries, most financial systems would be characterised as bank-based rather than market-based, though this is not universally so. More generally, financial systems are mostly smaller in developing countries, with less activity – and less efficiency – in terms of banking, non-bank financial institutions and capital markets.
These differences are illustrated in Table 4.2. As we can see, the scale of the financial sector in high-income countries, relative to GDP, is vastly greater than is the case in low-income countries. For example, while total bank deposits equate to 22% of GDP in low-income countries, the figure for high- income countries is 95%. As a result, while high-income country banks pro- vide credit at a level almost equal to national GDP, the respective figure for low-income countries is just 15%.
When we consider capital markets we see a similar disparity. Average stock market capitalisation represents 12% of GDP in low-income countries, but 99% in high-income countries, while the figures for private bond market cap- italisation to GDP are 1% and 45% respectively. For public bond market capitalisation the difference is less marked, which reflects the central role that the domestic financial system often plays in financing government activity in developing countries. Despite this, the ratio remains higher in high-income countries than is the case in their low-income counterparts.
The only component of the financial system that is relatively larger in low- income countries – in Table 4.2 – is the ratio of central bank assets to GDP, Table 4.2 Average financial structures of high-income vs. low-income countries, 2005*
Variables Low income High income
Central bank assets / GDP 0.10 0.03
Bank assets / GDP 0.19 1.14
Other financial institutions assets / GDP 0.08 0.54
Private credit by banks / GDP 0.15 1.08
Bank deposits / GDP 0.22 0.95
Bank overhead costs / total assets 0.07 0.03
Net interest margina 0.08 0.02
Bank concentrationb 0.83 0.76
Stock market capitalisation / GDP 0.12 0.99
Stock market turnover ratioc 0.64 0.80
Private bond market capitalisation / GDP 0.01 0.45 Public bond market capitalisation / GDP 0.33 0.47 Source: World Bank Financial Structure Dataset, 2006.
* Sample of 55 high-income and 55 low-income countries.
a: Banks’ net interest revenue as a share of interest-bearing assets (0.08 = 8%).
b: Assets of three largest banks as a share of assets of all commercial banks.
c: Ratio of the value of total shares traded to average real market capitalisation.
which in turn is a reflection of the more dominant role that the public sector has played, and continues to play, in these economies. Although we saw in the previous chapter that the process of financial liberalisation has swept away many of the structures associated with financial repression, the far greater importance of central banks in the financial systems of low-income countries clearly illustrates that this process is far from complete.
The three other shaded rows in the table also represent variables where the ratios are higher for low-income countries. In broad terms these meas- ure the inefficiency of the financial sector: bank overheads amount to 7% of total assets in low-income countries, compared to 3% in high-income coun- tries, while the figures for banks’ net interest margin are 8% and 2%
respectively.
For bank concentration, the situation is broadly similar in the two categor- ies, where the three largest banks account for a little over 80% of commercial banking assets in low-income countries, and a little under this figure in high-income countries.
In combination, these statistics suggests that, although one would expect the relative size and efficiency of the financial sector to move towards levels seen in high-income countries as countries become wealthier, there is no clear trend in terms of bank concentration: the tendency is for a small group of banks to dominate regardless of the level of national wealth.
Another key difference between developing and (most) developed coun- tries’ banking systems is the extent of foreign ownership. Particularly in the last decade, there has been an increasing trend towards foreign banks from developed countries buying up banking assets in the developing world. In part this is supply-led – the reduction or removal of restrictions on foreign ownership as a component of financial liberalisation has made foreign own- ership possible where this was previously not the case. The realisation of this potential, however, has been demand-driven – developed country banks have increasingly wished to ‘cross the border’ and conduct their business from within the country using local currency.
Much of the impetus here relates to the exchange rate risks associated with lending in international currencies. As we shall see in later chapters, a country with a large quantity of foreign-denominated debt faces extreme difficulties if the value of this debt increases following a currency crisis. Indeed, it may be that the presence of such a debt burden can itself precipitate the onset of such a crisis.
Furthermore, as well as increasing ‘demand’ for domestic banking assets in developing countries, financial crises can also directly affect the ‘supply’. As illustrated in Table 4.3, one effect of the Asian crisis was to accelerate the trend towards foreign ownership in all emerging regions. The process was made possible in Asia – where before the crisis foreign ownership of the domestic banking system was largely prohibited – as the conditions attached to IMF rescue packages obliged countries to remove these restrictions.
Foreign banking groups were thus able to avoid currency risk by lending
in local currency through their newly acquired domestic banking assets, the attractiveness of which was further enhanced by the fact that the crisis had left these assets extremely undervalued.3
However, while the logic behind these trends from both a supply and demand perspective is understandable, the fact that a large proportion of a country’s domestic banking system may be foreign-owned raises important issues, which will be covered in subsequent chapters in depth.
It should be remembered, however, that there is no point in having a large and efficient financial sector simply for its own sake. Efforts to move towards these ends are stimulated by the belief that improvements will bring meaning- ful benefits in terms of both economic growth and, ultimately, poverty allevi- ation. What is the evidence in this respect?
As we can see from Table 4.4 there is a clear positive correlation between the size of the financial sector – both in terms of banking and stock markets – and GDP per capita. Furthermore, there is a clear negative correlation between the inefficiency of the financial sector and GDP per capita, as well as between the relative size of the public sector in the financial system and GDP per capita.
Table 4.3 Foreign bank ownership in emerging markets before and after the Asian crisis
Foreign control (a) December 1994
Foreign control (a) December 1999
Foreign control (b) December 1999
Central Europe 7.8% 52.3% 56.9%
Latin America 7.5% 25.0% 25.5%
Asia 1.6% 6.0% 13.2%
Source: IMF, 2000.
(a) Where foreigners own more than 50% of total equity.
(b) Where foreigners own more than 20% of total equity.
Table 4.4 Correlations of financial sector development and GDP per capita
Measure of financial sector development Correlation
Liquid liabilities/GDP 0.465
Bank assets/GDP 0.663
Claims of banks on private sector/GDP 0.639
Central bank assets/GDP −0.442
Overhead costs −0.353
Bank net interest margin −0.443 Public share in total bank assets −0.462 Stock market capitalisation/GDP 0.282
Turnover ratio 0.409
Source: Demirgỹỗ-Kunt and Levine, 1999.
As we have seen, in terms of positive correlations, all of these indicators are significantly higher in richer than in poorer groups of countries, and have a tendency to rise as economies become more developed. Similarly, the effi- ciency of the financial system is higher in richer countries and correlated with higher GDP per capita levels.
However, although stock markets also tend to become larger, more efficient and more important to the economy as incomes rise, this does not lead to a uniform outcome, particularly as different measures are often used to meas- ure the degree of stock market development. There are some countries that show up as well developed by all measures (Australia, Great Britain, Hong Kong, Malaysia, the Netherlands, Singapore, Sweden, Switzerland, Thailand, and the United States, for example). Other markets – such as Chile and South Africa – are large but relatively illiquid, whilst others – such as Korea and Germany – have stock markets that are relatively small but are also quite active and liquid.
When we look at non-bank financial institutions (NBFIs), the same pat- tern emerges. Insurance companies, pension funds, mutual funds and other NBFIs are all larger as a share of GDP in richer countries. They are therefore not just larger in absolute terms, but also more important relative to other components of the financial sector, particularly banks and development banks.
It is therefore clear that the different components of financial sector devel- opment are strongly associated with higher levels of GDP per capita. Fur- thermore, as we saw in Chapter 2, this relationship is, at least in part, a causal one: growth may lead to financial sector development, but financial sector development does itself lead to growth.
4.1.1 The formal vs. the informal sector
As well as the specific structural differences between developed and develop- ing countries described above, a key distinction between economic systems in general and the financial sector in particular is the relative sizes of the ‘for- mal’ and ‘informal’ sectors.
As discussed later in this chapter, the tax take (as a percentage of GDP) in developing countries is generally much lower than that in developed coun- tries. Although there are numerous reasons for this – and therefore numerous remedies proposed – a key factor is the size of the (untaxed) informal sector.
In this regard, it is estimated that within developing countries informal firms account for around 30% of total production and between 50% and 75% of jobs in the non-agricultural labour force.4 That is, in many developing coun- tries, the informal sector’s importance is comparable to that of the formal sector.
Table 4.5 gives comparative estimates of the total size of the informal sector in developed countries compared to three developing regions, and makes this importance abundantly clear. Clearly, therefore, no national
development strategy can afford to ignore a sector which is equivalent to between a third and one half of total GNP. The issues in this regard will be explored fully in the following chapter on reform of the domestic system.
A corollary to the size of the informal sector is the extent of financial exclu- sion in developing countries. That is, to a far greater extent than is the case in developed countries, a high proportion of the population of many developed countries have no contact with and no access to the formal financial system.
Table 4.6 highlights the lack of access to formal finance starkly: in none of the countries, regions or cities considered does more than 50% of the popula- tion have a bank account, whereas in the USA 91% of people have bank accounts. For many this is a clear example of market failure: it must be the case that a large proportion of those excluded from the financial sector could effectively use its services. Furthermore, any country that excludes more than 50% of its population from the formal financial system cannot be optimally employing its human resources. Consequently, economic growth must be lower than would be the case if the financially excluded could be drawn into the formal financial system.
Although those working in the informal sector – as well as small-scale entrepreneurs – certainly do require financial services, these differ in certain key ways to those offered by the mainstream financial sector. In particular, the sums involved may be very small and the potential borrowers widely disbursed with no ‘credit history’ and little or no collateral. Given the cost of setting up bank branches, the relatively high fixed costs of monitoring
Table 4.5 Informal sector as percentage of GNP
Developed countries 12%
Africa 44%
Latin America 39%
Asia 35%
Source: Gerxhani, 2004.
Table 4.6 Proportion of the population with bank accounts
Botswana 47%
Brazil (urban) 43%
Colombia (Bogota) 39%
Djibouti 24.8%
Lesotho 17%
Mexico City 21.3%
Namibia 28.4%
South Africa 31.7%
Swaziland 35.3%
Tanzania 6.4%
Source: Emerging Market Economics, 2005.
small-scale loans (i.e. it costs much the same to monitor a loan for £100,000 as for one for £100), and the very low returns earned on small-scale savings, most formal financial institutions have simply not attempted to access this potential market.
For many, the solution to this problem is microfinance. In its modern form, microfinance emerged from the work of the Grameen Bank in Bangladesh, though there have been historical versions as long ago as seventeenth- century Ireland and eighteenth-century Germany.5 Today, there are a number of differing forms, but microfinance institutions generally provide small loans to the poor, with minimal bureaucracy and often require little or no collateral.
The great attention that has been paid to microfinance from the development community and some mainstream economists has been the result of microfi- nance schemes’ seeming ability to reach the financially excluded and to achieve very high repayment rates. This latter point is attributed to many factors, but particularly to the focus on ‘group lending’ (where if one member of the group defaults the remainder become liable for their debt) and the associated ex ante and ex post peer group monitoring that this involves.
By getting borrowers themselves to screen their fellow borrowers and to monitor repayments, microfinance institutions have been able to alter the nature of the financial calculus described above; if the costs of monitoring small loans is dramatically reduced, the economic viability of making such loans is greatly increased.
From small beginnings in the 1970s, these successes led to the replication of the original microfinance model, and to its development into other forms. By 2002 it was estimated that more than 10,000 microfinance institutions had been established, reaching more than 30 million people. However, demand for microfinance was estimated at between 400 and 500 million, leading the United Nations to designate 2005 as the ‘Year of Microcredit’ in an effort to expand its reach significantly.
For supporters the microfinance revolution has the potential to effectively eliminate poverty in developing countries. For others, however, the obstacles that prevented the formal financial sector targeting the poor and excluded in the first place will place inevitable limits on what can be achieved in terms of poverty reduction.
To summarise, observable features of financial structure and development are as follows:
• As per capita incomes rise, the role of the state in the financial system tends to fall.
• As per capita incomes rise, the size and efficiency of the private financial sector tends to rise.
• As per capita incomes rise, stock market activity tends to rise relative to the activity of the banking sector and to that of NBFIs.
• As per capita incomes rise, stock markets tend to become more efficient relative to banks.
• As per capita incomes rise, financial systems tend to become more market-based, though this is not universally the case (e.g. Germany and Japan) and care must be taken in interpreting the data in this regard. For example, Germany and Ecuador are both classified as bank-based sys- tems – due to the relative size of banks vs. capital markets – but in Ecuador banks are not as well developed as German banks, and there- fore less able to efficiently perform their key functions. In the same way, the United States and the Philippines are both classified as market-based systems, but the markets in the Philippines are clearly not as effective at providing financial services as those in the US.
• The informal economy in developing countries can be almost 50% of the size of total GNP, but tends to diminish in relative importance as national income rises.
• Financial exclusion is a major issue in many developing countries: at the extreme more than 93% of the population of Tanzania was without a bank account in 2005.
• For some, the ‘microfinance revolution’ has the potential both to effect- ively eliminate poverty, and to provide a bridge between the informal and formal sectors. For others, its growth is inherently self-limiting.
We have examined the key structural differences between financial systems in developed and developing countries, as well as considering the different pat- terns of financial sector development within these two groups of countries.
Furthermore, we have seen how economic growth exerts a strong influence over the nature of financial sector development (as well as the relative size of the informal sector), and it has been suggested that, at least in part, the causality in this regard may run both ways.
The next section considers factors other than a country’s level of develop- ment that impact the structure of the financial system and the pathway of financial sector development. As we shall see, these factors have much explanatory power when we survey the very different financial systems that countries at similar levels of income have developed.