Chapter 14
Empirical Assessment of the Finance–Growth Nexus in Frontier Markets
G. Giorgioni*
Abstract
This chapter contributes to the literature on the relationship between financial development and economic growth by including variables related to both the banking sector and the development of stock markets and by distinguishing among developed, emerging, and “frontier” stock markets. The results provide evidence that, in line with recent empirical literature, the banking sector has an overall negative impact, once the effect of stock markets is taken into account, although the results do not provide evidence of a threshold leading to instances of “too much finance.” However, in countries with emerging stock markets and particularly in countries with frontier markets, the impact of credit depends on the nature of the provider. Finally, the impact of stock markets depends on their own level of development, with a positive impact only for developed stock markets, but an insignificant impact for both emerging and frontier markets.
Keywords
financial development
stock exchanges
least developed countries
emerging markets
frontier markets
economic growth
JEL classification
F39 International Finance: Other
G39 Corporate Finance and Governance: Other
O1 – Economic Development
O16 - Financial Markets, Saving and Capital Investment P43 Other Economic Systems: Public Economics
Financial Economics
1. Introduction
This chapter contributes to the literature on the nexus between finance and economic growth by focusing on the impact of frontier stock markets. The literature on finance and economic growth has established empirically that although both banks and stock markets may be important for economic growth (Chakraborty and Ray, 2006; Deidda and Fattouh, 2008; Levine, 2002; Levine and Zervos, 1998), the impact of stock markets appears to be stronger, while the impact of the banking sector could even be detrimental. For instance, Atje and Jovanovic (1993), in one of the very first cross-country growth analyses involving banks and stock markets, found that stock markets had a positive and significant influence on economic growth, whereas banks had a negative impact.
We have found a large effect of stock markets on subsequent development. We have failed to find a similar effect of bank lending. That this differential effect should exist is in itself surprising. But if it is true, then it is even more surprising that more countries are not developing their stock markets as quickly as they can as a means of speeding up their economic development. (Atje and Jovanovic, 1993, p. 663)
Also, Beck and Levine (2004), Saci et al. (2009), and Shen and Lee (2006) have provided evidence of positive effects of stock market development on economic growth, while Minier (2009) has shown that the establishment of a stock exchange can boost economic growth even in the poorest regions of the world. Minier (2009) finds that countries experience higher economic growth during the first 5 years of existence of a stock exchange. However, the longer-term results are more ambiguous.
More recently, a metaanalysis conducted by Valickova et al. (2014) on the empirical literature on the nexus between financial development and economic growth has provided evidence indicating that the literature has managed to identify an “authentic” (ie, without publication selection bias) positive link between financial development and economic growth. However, it has to be noted that this result depended on region, time period, and structure of the model. In general, the positive effect is stronger in more developed nations but over more recent times (ie, from the 1980s onward) has demonstrated a weakening propensity. The analysts also find that stock market–oriented systems tend to be more conducive to growth than bank-oriented systems.
In recent papers, Arcand et al. (2012), Law et al. (2013), and Cecchetti and Kharroubi (2012) show that when banking credit reaches a certain threshold, typically 90% of the aggregate gross domestic product (GDP) of a country, it has an insignificant if not negative impact on economic growth, leading to the conclusion that there is such a thing as “too much finance,” although the conclusion of “too much banking credit” would be more accurate.
Therefore this study aims at empirically testing the relationship between finance and growth by specifically assessing the impact of stock market and banking sector development on economic growth for a broad sample of countries whose stock exchanges are classified as either developed, emerging, or frontier. In total there are 68 countries consisting of 23 developed markets, 20 emerging markets, and 25 frontier markets. We also test the relationship for the subsamples (developed, emerging, and frontier). In testing the nexus between finance provision and economic growth, a one-step generalized method of moments (GMM) estimation will be utilized on a relatively balanced data set of developed, emerging, and frontier market nations. To the best of the authors’ knowledge, this is the first attempt to incorporate frontier markets into such a study, although existing research such as Saci et al. (2009) and Rioja and Valev (2004, 2014) have analyzed emerging markets or, in broader definition, markets in less developed countries.
The chapter is organized as follows: in Section 2 the theoretical and empirical literature on the nexus between financial development and economic growth is briefly reviewed. Section 3 focuses on the extant literature on frontier markets, while Section 4 explains the model and describes the variables. Section 5 comments on the results. Finally, the concluding remarks are in Section 6.
2. Review of the Theoretical and Empirical Literature on the Finance and Growth Relationship Involving Banks and Stock Markets
In this section we briefly review the literature on finance and growth. Unavoidably, choices had to be made, and it was decided to focus on a selection of papers, in particular those that emphasize the role of stock markets.
Bagehot (1906), Schumpeter (1934), Gurley and Shaw (1955), Goldsmith (1969), McKinnon (1973), and Shaw (1973) have suggested a positive impact of financial development over economic growth, although Goldsmith states that it might be difficult to establish the direction of causality. The issue of the direction of causality is also mentioned by Driffill (2003) and Trew (2006). However, earlier authors like Robinson (1952) and Lucas (1988) did not identify finance as an important determinant for economic growth.
The literature on the role of finance in economic growth was given a new lease on life by King and Levine (1993a,b), Levine (1997, 2005), Levine and Zervos (1998), Levine et al. (2000), and Beck and Levine (2004). Conceptually, a developed financial system should enhance economic growth by enhancing both productivity growth and physical capital accumulation in five different ways: (1) by mobilizing savings, (2) by allocating funds to capital investment, (3) by improving corporate governance and management, (4) by redistributing risk, and (5) by facilitating trade.
The crucial role of the financial sector is to minimize the problems brought on by information frictions (information asymmetry) and transaction frictions in performing the aforementioned five functions.
In addition to the aforementioned functions, Rousseau and Wachtel (2000) explain the role of stock markets as providers of (1) an exit option for start-up entrepreneurs and venture capitalists, (2) capital inflows (domestic and foreign), (3) liquidity, and (4) information on the quality of potential investments.
There has been no shortage of papers attempting to empirically establish the existence and the type of relationship between finance and economic growth. In terms of the findings, the existing research has provided evidence that both banks and stock markets are important for growth (among others, see Levine and Zervos, 1998; Levine, 2002; Chakraborty and Ray, 2006). Deidda and Fattouh (2008) find that the role of banks is reduced as stock markets develop. However, Atje and Jovanovic (1993), Beck and Levine (2004), Loayza and Ranciere (2006), Shen and Lee (2006), and Saci et al. (2009) have shown that while stock markets have a positive impact upon growth, banks tend to have a negative impact. For instance, De Gregorio and Guidotti (1995) observe that the relationship between banks and growth in Latin America was in fact negative, and explained this finding by citing cycles of credit booms and busts; that is, an overexpansion of credit was followed by a credit crunch. The role of banking and financial crises was reiterated more recently by Jordà et al. (2013) who provide evidence on the consequences of financial crises. They find that financial crises have been costlier (than the nonfinancially induced recessions) in terms of lost GDP over the period 1870–2008, and more credit-intensive expansions tend to be followed by deeper recessions (in financial crises or otherwise) and slower recoveries, lending support to how financial factors play an important role in the modern business cycle.
The interest in the role of stock exchanges is reinforced by Atje and Jovanovic (1993), who find that a 1% point rise in the stock market-related variable has had a positive impact on economic growth, whereas they cannot find a significant effect of banks. Rioja and Valev (2014) focus on economic growth, productivity, and capital accumulation of high- and low-income countries. They find that in both low- and high-income countries banks are relatively important for capital accumulation; stock markets do not have an impact on capital accumulation and productivity in low-income countries, but have a strong impact in high-income countries.
Three important conclusions are observed: banks’ apparently negative relationship with economic growth; the relative importance of the role of stock markets in economic growth (initially excluded from the analysis); and the nonlinearity of the relationship between stock markets and economic growth, with different impacts at different stages of development.
3. Empirical Literature on Frontier Markets
The definition of frontier markets arose in the early 1990s, where on the back of substantial investment performance by emerging markets throughout the 1970s and 1980s, there was a desire to provide move visibility to less known investment markets in less developed nations. Along the lines of the favorable “emerging” moniker, markets that were fundamentally smaller and less developed were classed as “frontier” (Speidell and Krohne, 2007). Given the relatively young age of these “frontier” classifications, the empirical academic literature still possesses much room for development. Predominantly, the existing literature on frontier markets can be divided into two fronts: investment finance and liquidity risk measurement. Much of the empirical work draws from the meta-description by Speidell and Krohne (2007), who provide a seminal overview of the frontier markets. In suggesting that emerging markets are quickly becoming part of the developed world and that they are the next source of global capital, there have been numerous studies looking into the diversification benefits and risks of frontier markets.
Although frontier markets are still an underresearched area of study, recent publications have focused on the potential for portfolio diversification afforded by them (Berger et al., 2011; Kiviaho et al., 2012), with particular emphasis on how closely correlated these markets are with developed and emerging ones and more recently on liquidity, both at firm level and at market level.
There are several iterations on the study of the diversification benefits of frontier markets, and these range from an analysis of correlation to a cointegration analysis of frontier countries with both emerging and developed nations. Along the lines of correlation, studies such as Quisenberry and Griffith (2010) suggest that frontier markets possess low cross-correlations with emerging and developed markets, hence providing substantial diversification opportunities. More recently, Amin and Orlowski (2014) further develop this correlation analysis to include an analysis of relative volatility of frontier, emerging, and developed markets over the period 1999–2011. The sample consists of three frontier markets, one emerging market, and one developed market. The findings suggest that frontier markets are more resilient to market shocks with volatility clustering, which could provide a strong case for diversification. Moreover, the study also suggests that there is a strong influence on the returns of the frontier markets from developed markets during “normal” economic periods, suggesting that diversification opportunities would be present only during “abnormal” economic periods. However, we should caution against complete generalization of these characteristics to all frontier markets; the frontier sample within the Amin and Orlowski (2014) study focuses primarily on the relationship between South Asian and US markets. Nevertheless, these studies on correlation provide arguments toward the diversification benefits of frontier markets.
With the focus still on diversification, the second main group of empirical research adopts a cointegration analysis to examine the diversification benefits of frontier markets. Studies such as Berger et al. (2011) adopt this method of analysis and show that frontier markets lack cointegrating vectors with both emerging and developed markets, thus providing, once again, an argument for diversification benefits. More recently, Su and Yip (2014) adopt this methodology for their study of 32 nations—15 emerging, 16 frontier, and 1 developed, based on the MSCI classifications. The study evaluates monthly data over a 13-year period 2000–13. The results suggest that frontier markets exhibit relatively weak levels of cointegration with emerging and developed markets; however, this level of cointegration has risen marginally during periods of shocks, providing some support for the relative integration of global markets.
Research on the relative liquidity of frontier markets predominantly takes the form of asset pricing studies. Research such as Milunovich and Minović (2014) attempts to price this liquidity element through the use of established multifactor models such as the Fama and French (1992) three-factor model with the inclusion of a liquidity proxy. A detailed analysis of multifactor asset pricing on frontier markets can be found in Hearn (2012). Conducting this analysis on a sample of sub-Saharan equity markets, including the Bourse Régionale des Valeurs Mobiliéres SA (BRVM), Hearn (2012) attempts to determine if liquidity is a significant factor within the pricing of equity assets. Using a liquidity measure defined by O’Hara (2003) and Amihud (2002), this study discovers that liquidity is a significant factor in influencing equity returns within frontier markets. The findings for liquidity being a significant factor within frontier equity markets concur with that of Lesmond (2005). Moreover, Hearn (2012) also suggests information asymmetry issues and questions the underlying macroeconomic stability of frontier markets (Hearn, 2014; Hearn and Piesse, 2010; Marshall et al., 2013). However, it should be noted that while the study does identify liquidity as one of the influencing factors on equity asset pricing in frontier markets, we should caution against suggesting direction of causality.
In terms of the relationship between frontier stock exchanges and development, Hearn and Piesse (2010) analyze barriers to development of stock markets, but also limits to the contributions of stock exchanges to development. Hearn and Piesse focus on the cases of Swaziland and Mozambique and very briefly on Côte d’Ivoire. They identify political reasons for the failure to sustain development (state-owned assets effectively transferred from the state to a small postcolonial elite) and practical reasons like lack of market depth, liquidity, and transparency. These two markets are not part of the sample of countries studied in this chapter.
4. Model and Variables
The study uses a rich and up-to-date database of financial variables constructed from the World Development Indicators database established by the World Bank. We have opted not to use financial data from the Global Financial Development Database (GFDD), which was developed by Cihak et al. (2012) for the World Bank because of its ad hoc nature and too many missing observations. We are also aware of the availability of a database on educational attainment produced by Barro and Lee (2013). But we decided that using the World Development Indicators database would allow for continuity and consistency in the quality of data.
To align this chapter with the existing literature and allow comparability and generalizability with results from previous empirical work, the following variables have been selected. Economic growth, as the dependent variable, is represented by annual GDP per capita, which is the consistent measure within the existing literature (Beck et al., 2014). The independent variables can be broken down into two subsets capturing (1) banking development and (2) stock market development effects on sovereign economic growth. These variables are widely used in the literature to capture the level of financial development and are further defined in the following paragraphs.
4.1. Measuring Banking Sector Development on Economic Growth
PRICRE or private credit is defined as the domestic credit provided to the private sector by banks, and it refers to financial resources provided by other depository corporations (deposit-taking corporations except central banks), such as through loans, purchases of nonequity securities, and trade credits and other accounts receivable, that establish a claim (source: World Bank World Development Indicators).
DOMCRE or domestic credit is defined as the domestic credit provided to the private sector, and it refers to the financial resources provided by financial corporations, such as through loans, purchases of nonequity securities, and trade credits and other accounts receivable, that establish a claim for repayment. For some countries these claims include credit to public enterprises. The financial corporations include monetary authorities and deposit money banks, as well as other financial corporations where data are available, including corporations that do not accept transferable deposits but do incur such liabilities as time and savings deposits (source: World Bank World Development Indicators).
LL or liquid liabilities are also known as M3. They are the sum of currency and deposits in the central bank (M0), plus transferable deposits and electronic currency (M1), plus time and savings deposits, foreign currency transferable deposits, certificates of deposit, and securities repurchase agreements (M2), plus traveler’s checks (source: World Bank World Development Indicators).
LL can be seen as the ratio of broad money to GDP, where broad money is defined as currency held outside the banking system plus interest-bearing total deposit liabilities of banks and other financial institutions. However, care should be taken in the interpretation of this variable, as it could potentially lead to some ambiguity. A low ratio could be the result of either an underdeveloped banking sector or, equally, a highly sophisticated financial sector that allows economic agents to reduce money balances held with the banking system.
4.2. Measuring Stock Market Development on Economic Growth
SECAP is the market capitalization of listed domestic companies (% of GDP) where market capitalization (or market value) is calculated as the share price times the number of shares outstanding. They are the domestically incorporated companies listed on the country’s stock exchanges at the end of the year. Listed companies do not include investment companies, mutual funds, or other collective investment vehicles.
While market capitalization is widely used within the literature, there is evidence to suggest that it is not a good predictor of economic growth. Since markets are forward-looking, they will anticipate higher economic growth through an appreciation of share prices. As value traded is the product of quantity and price, this indicator can rise even without an increase in the number of transactions.
SEVAL is the ratio of total value of stocks traded (% of GDP) during the period. This indicator complements the market capitalization ratio by showing whether market size is matched by trading activity.
The ratio of value of stocks traded to GDP is used to measure stock market activity, and it measures trading volume relative to the size of the economy. Being the product of market price and the number of shares traded, it includes elements of both liquidity and size.
SETURN is the turnover ratio of stock traded (%). The turnover ratio is defined as the total value of shares traded during the period divided by the average market capitalization for the period, where average market capitalization is calculated as the average of the end-of-period values for the current period and the previous period.
The turnover ratio is the trading volume of the stock market relative to the average market capitalization, and it measures stock market liquidity, showing the importance and the credibility of available information. In emerging markets, an increase in liquidity is a good indication of financial development. Furthermore, it indicates low transaction costs, which facilitate fund transfers and increase the number of firms and traded shares, hence potentially promoting growth (Rousseau and Wachtel, 2000).
Computationally, turnover equals the value of the trading of shares on domestic exchanges divided by total value of listed shares, and indicates the trading volume of the stock market relative to its size. Since both numerator and denominator contain the share price, the turnover ratio is not plagued by the weaknesses found in the market capitalization ratio.
4.3. Instrumental Variables
Undeniably, there could be an endogeneity problem between the dependent variable, GDP per capita growth, and the explanatory variables related to the banking and financial sectors. To account for this problem, instrumental variables have been included in the model. The instrumental variables are enrollment in secondary schools, import and export to GDP (trade openness), general government consumption to GDP (government consumption), inflation measured by the consumer price index (CPI), gross capital formation (% of GDP), initial GDP per capita, and dummy variables for the years.
ENROLLSECONDARY or the gross enrollment ratio is the ratio of total enrollment, regardless of age, to the population of the age group that officially corresponds to the level of education shown. Secondary education completes the provision of basic education that began at the primary level, and aims at laying the foundations for lifelong learning and human development, by offering more subject- or skill-oriented instruction using more specialized teachers.a The use of enrollment numbers in secondary schools as control presents further challenges in assuming a strong causal relationship between secondary school education and human capital development, which may not necessarily be the case. However, to be consistent with previous empirical research and because we feel that the variable could still be interpreted as an overall indicator of the commitment toward investments in human capital, the variable is included as another control variable in the regression.
OPEN is the exports and imports of goods and services over GDP (%). It represents the value of all goods and other market services provided to and coming from the rest of the world. They include the value of merchandise, freight, insurance, transport, travel, royalties, license fees, and other services, such as communication, construction, financial, information, business, personal, and government services. They exclude compensation of employees and investment income (formerly called factor services) and transfer payments.
GOVCONS represents general government final consumption expenditures (formerly general government consumption) and includes all government current expenditures for purchases of goods and services (including compensation of employees). It also includes most expenditure on national defense and security, but excludes government military expenditures that are part of government capital formation.
INFLATION as measured by the consumer price index (CPI) reflects the annual percentage change in the cost to the average consumer of acquiring a basket of goods and services that may be fixed or changed at specified intervals, such as yearly. The Laspeyres formula is generally used.
GCF or gross capital formation (formerly gross domestic investment) consists of outlays on additions to the fixed assets of the economy plus net changes in the level of inventories. Fixed assets include land improvements (fences, ditches, drains, and so on); plant, machinery, and equipment purchases; and the construction of roads, railways, and the like, including schools, offices, hospitals, private residential dwellings, and commercial and industrial buildings. Inventories are stocks of goods held by firms to meet temporary or unexpected fluctuations in production or sales, and “work in progress.” According to the 1993 System of National Accounts (SNA), net acquisitions of valuables are also considered capital formation.
IIP or initial real GDP per capita is defined as the start-year current GDP per capita US dollar deflated by the US CPI index of each year. This variable is assumed to capture the catch-up effect.
A summary table of the included variables in the model is given in Table 14.1. All variables are measured as logarithms of the corresponding numbers or, for growth and inflation, as logarithmic rates of change.
Table 14.1
List of Variables
| Economic growth variable (dependent) | |
| GDPPCGRO | Growth of real GDP per capita (%) |
| Financial development variables (independent) | |
| PRICRE | Private credit by deposit money banks to GDP (% of GDP) |
| DOMCRE | Domestic credit to private sector (% of GDP) |
| LL | Liquid liabilities to GDP (%) |
| SECAP | Stock market capitalization to GDP (%) |
| SEVAL | Stock market total value traded to GDP (%) |
| SETURN | Stock market turnover ratio (%) (ie, total value of shares traded/average market capitalization) |
| Instrumental variables | |
| ENROLLSEC | Lower secondary completion rate, total (% of relevant age group) |
| OPEN | (Exports + imports)/GDP (% of GDP) |
| GOVCONS | Government consumption to GDP (% of GDP) |
| INFLATION | Inflation rate via CPI (%) |
| GCF | Gross capital formation to GDP (% of GDP) |
| IIP | Initial real GDP per capita in USD |
Source: World Development Indicators, the World Bank.
The model is thus presented as:
We have a 68-nation sampling frame comprising 23 developed market nations, 20 emerging market nations, and 25 frontier market nations. The MSCI definitions for developed, emerging, and frontier markets were adopted for this study and are given in detail within the Appendix. We have a relatively balanced sample consisting of nations across the three measures of market development. The selection process of nations was determined by data availability over the sample period. The unit of measure is annual data for each of the variables taken over a 22-year sampling period (1990–2011). In order to assess both the combined and the individual characteristics of each of these markets, the sample has been broken down into further subgroupings given as (1) all markets, (2) developed markets only, (3) emerging markets only, and (4) frontier markets only.
The choice of using annual data stems from the observation that different outcomes can be obtained if data are arbitrarily averaged as is done in the literature and to fully exploit the database. For instance, the usual averaging over a 5-year period could lead to a significant loss of important information. Moreover, it would be based on the substantial assumption that the economic cycle lasts 5 years and is synchronous across countries at varying stages of development.
Table 14.2 provides the descriptive statistics (mean) of the variables. A quick base analysis of the rate of growth of GDP indicates that this is highest within the frontier markets, although the difference with both emerging and developed markets is not substantial. Traditionally, it is believed that more developed equity exchanges belong to countries with a higher income per capita and higher initial real GDP per capita. This is evidenced from an examination of the stock market variables (SECAP, SEVAL, SETURN) against initial real GDP per capita (IIP). The variables measuring the development of stock markets reveal that frontier nations possess significantly smaller equity markets in terms of capitalization, value traded, and turnover. This is a potential confirmation of a significant issue of frontier (and to a certain extent emerging) markets: the lack of liquidity. The banking sector variables indicate that bank credit in developed countries appears to have surpassed the threshold of 90% of GDP identified in the literature as an important ratio for a detrimental effect. This, too, clearly indicates one of the issues of frontier (and emerging) markets: the lack of liquidity.
Table 14.2
Descriptive Statistics
| Developed (mean) | Emerging (mean) | Frontier (mean) | |
| GDPPCGRO (%) | 1.87 | 2.55 | 2.61 |
| Financial and banking sector development variables | |||
| PRICRE (% of GDP) | 99.76 | 45.55 | 36.43 |
| DOMCRE (% of GDP) | 114.14 | 52.73 | 41.12 |
| LL (% of GDP) | 94.08 | 56.25 | 49.64 |
| SECAP (% of GDP) | 83.98 | 52.22 | 29.88 |
| SEVAL (% of GDP) | 72.28 | 24.51 | 8.58 |
| SETURN (%) | 83.50 | 54.86 | 27.73 |
| Control variables | |||
| ENROLLSEC (%) | 107.72 | 78.16 | 79.81 |
| OPEN (% of GDP) | 91.60 | 68.39 | 84.23 |
| GOVCON (% of GDP) | 18.17 | 14.63 | 15.66 |
| INFLATION (%) | 2.49 | 50.69 | 44.47 |
| GCF (% of GDP) | 23.33 | 24.57 | 22.86 |
| IIP (USD) | 17,691 | 3,826 | 2,542 |
With regard to the control variables, while gross capital formation, government consumption, and to a certain extent openness are very similar across the three groups of countries, the record in inflation clearly distinguishes developed countries from emerging and frontier ones by a factor of more than 20.
Tables 14.3A–D present the simple correlations of the independent variables for each of the subgroupings. There are several notable results within the correlation matrices that are common across the entire sample and the three subgroupings. The variables PRICRE and DOMCRE possess a substantially large correlation coefficient; however, this is to be expected given the similar nature of the two variables. Other noteworthy characteristics of the correlation matrices include the coefficients of LL to DOMCRE and SECAP to SEVAL. In terms of the former, the high correlation coefficient could be a result of the classification of credit instruments within the liquid liabilities measure. With regard to SECAP and SEVAL, the variable SEVAL captures the relative change in the total value of stocks traded, which could potentially be captured within the relative change in market capitalization, SECAP; hence the reasonably large correlation coefficient.
Table 14.3a
Correlations of the Variables All Countries (Obs. = 1238)
| PRICRE | LL | DOMCRE | SECAP | SEVAL | SETURN | |
| PRICRE | ********* | |||||
| LL | 0.7862 | ********* | ||||
| DOMCRE | 0.9125 | 0.7418 | ********* | |||
| SECAP | 0.4716 | 0.5879 | 0.5176 | ********* | ||
| SEVAL | 0.4646 | 0.5470 | 0.5510 | 0.7535 | ********* | |
| SETURN | 0.2734 | 0.2248 | 0.3452 | 0.1686 | 0.5542 | ********* |
Values in bold are not significant, even at 10% level of significance.
Table 14.3b
Correlations of the Variables Developed (Obs. = 492)
| PRICRE | LL | DOMCRE | SECAP | SEVAL | SETURN | |
| PRICRE | ********* | |||||
| LL | 0.6579 | ********* | ||||
| DOMCRE | 0.7891 | 0.6237 | ********* | |||
| SECAP | 0.2898 | 0.6050 | 0.3147 | ********* | ||
| SEVAL | 0.2556 | 0.4854 | 0.4101 | 0.7497 | ********* | |
| SETURN | 0.0700 | 0.0426 | 0.2793 | 0.1023 | 0.5742 | ********* |
Values in bold are not significant, even at 10% level of significance.
Table 14.3C
Correlations of the Variables Emerging (Obs. = 383)
| PRICRE | LL | DOMCRE | SECAP | SEVAL | SETURN | |
| PRICRE | ********* | |||||
| LL | 0.8542 | ********* | ||||
| DOMCRE | 0.9149 | 0.6937 | ********* | |||
| SECAP | 0.4572 | 0.3137 | 0.6146 | ********* | ||
| SEVAL | 0.5033 | 0.4595 | 0.5694 | 0.6989 | ********* | |
| SETURN | 0.1800 | 0.2408 | 0.1418 | −0.0674 | 0.4381 | ********* |
Values in bold are not significant, even at 10% level of significance.
Table 14.3D
Correlations of the Variables Frontier (Obs. = 363)
| PRICRE | LL | DOMCRE | SECAP | SEVAL | SETURN | |
| PRICRE | ********* | |||||
| LL | 0.7691 | ********* | ||||
| DOMCRE | 0.9793 | 0.7566 | ********* | |||
| SECAP | 0.5028 | 0.5185 | 0.5022 | ********* | ||
| SEVAL | 0.2499 | 0.2828 | 0.2550 | 0.7123 | ********* | |
| SETURN | −0.0428 | 0.0182 | −0.0396 | 0.0642 | 0.5498 | ********* |
Values in bold are not significant, even at 10% level of significance.
5. Empirical Results
Diagnostic tests for a check of the overall fit of the regressions (Wald test), the validity of the instrumental variables (Sargan test), and the presence of first-order and second-order serial correlation were conducted. Under conventional circumstances, the validity of the instruments can be evaluated via the Sargan test. However, given that the one-step GMM with robust standard errors is used, the test is not possible. Moreover, a consequence of the first differencing is to introduce first-order autocorrelation, so this is expected on estimation. Moreover, since first-order serial correlation is introduced automatically when the basic equation is differenced, any evidence of first-order serial correlation can be ignored, and instead the results of the tests for second-order serial correlation are considered. All the estimations satisfy the requirements of no second-order correlation. Table 14.4 provides the summarized results of the regression model for the entire sample and its subgroupings—developed, emerging, and frontier markets.
Table 14.4
Growth, Stock Market Development, and Credit Allocation (Dependent Variable GDP per Capita Growth)
| All | Developed | Emerging | Frontier | |
| Banking development variables | ||||
| PRICRE | −0.089 (−2.38) [0.017]** | −0.014 (−0.80) [0.422] | −0.20 (−3.13) [0.002]*** | −0.33 (−3.33) [0.001]*** |
| DOMCRE | 0.038 (1.44) [0.151] | −0.015 (−0.96) [0.336] | 0.10 (1.62) [0.105] | 0.16 (2.41) [0.016]** |
| LL | −0.094 (−3.39) [0.001]*** | −0.014 (−1.19) [0.236] | −0.025 (−0.62) [0.537] | −0.006 (−0.14) [0.887] |
| Stock market development variables | ||||
| SECAP | 0.021 (4.14) [0.000]*** | 0.031 (5.31) [0.000]*** | 0.025 (1.72) [0.085]* | 0.012 (0.85) [0.394] |
| SEVAL | −0.001 (−0.23) [0.818] | −0.008 (−5.27) [0.0000]*** | 0.005 (0.32) [0.745] | 0.024 (1.19) [0.233] |
| SETURN | 0.004 (0.52) [0.607] | 0.013 (3.15) [0.002]*** | −0.009 (−0.71) [0.480] | 0.004 (0.44) [0.661] |
| Number of observations | 930 | 401 | 288 | 241 |
| Wald chi-test (7) | 51.06*** | 156.93*** | 170.83*** | 64.02*** |
| Arellano–Bond test for order 1 | Z = −4.45 [0.0000]*** | Z = −3.68 [0.0002]*** | Z = −3.00 [0.0027]*** | Z = −2.85 [0.0044]*** |
| Arellano–Bond test for order 2 | Z = 1.76 [0.0779] | Z = −1.40 [0.1622] | Z = 1.38 [0.1669] | Z = 1.63 [0.1031] |
1. For the Wald chi-test, the null hypothesis is that none of the variables are worth including, and the alternative is that some variables are needed.
2. For the Arellano–Bond test of order 1, the null hypothesis is that the errors in the first-difference regression exhibit no first-order serial correlation.
3. For the Arellano–Bond test of order 2, the null hypothesis is that the errors in the first-difference regression exhibit no second-order serial correlation.
In parentheses are Z-statistics values, while in square brackets are the corresponding p-values.
*** Denotes statistically significant at 1%, ** statistically significant at 5%, * statistically significant at 10%.
The results indicate that PRICRE or private credit is statistically significant with a negative coefficient for the entire sample and the subgroupings of emerging and frontier markets, while it is not significant in the sample developed markets. While this provides some tentative arguments suggesting that private credit is detrimental toward economic development with both emerging and frontier markets, an even more interesting argument presents itself when PRICRE is compared with DOMCRE, which is statistically significant and yields a positive regression coefficient. Note again that the principal difference between the variables PRICRE and DOMCRE is that the former excludes lending from nondeposit-taking institutions and central banks.
The suggestion, based on the results of the regression, could be that financial intermediation within the context of purely deposit-taking institutions is detrimental toward economic growth and development. There could be a potential argument indicating that capital allocation taken on by deposit-taking institutions in terms of the investment decision might be poorer, at least in countries with emerging and frontier markets, than other private and public sources of credit.
Moreover, given the strong correlation coefficient between the two variables, PRICRE and DOMCRE, the model was also estimated without the variable DOMCRE. The results were markedly similar, with the only exception being that the variable PRICRE was now significant for developed markets. Moreover, the coefficients of the variable PRICRE (−0.027 for developed, −0.086 for emerging, and −0.16 for frontier markets) were very close to the sum of the coefficients of the variables PRICRE and DOMCRE when the two variables were used at the same time, indicating that the impact of the variable DOMCRE was now accounted for within private credit. However, it was decided to keep the two variables in the model to highlight the difference in the behavior of the variables PRICRE and DOMCRE in the context of countries with emerging and frontier markets.
At a more aggregate level, looking at the statistically significant coefficients across these three groups, the results provide some argument toward indicating that banking sector development does not contribute to economic growth. This is in line with some of the existing literature suggesting that the banking sector is detrimental toward economic growth and development (Atje and Jovanovic, 1993; Beck and Levine, 2004; Shen and Lee, 2006; Saci et al., 2009). The results, however, to not support the idea of the existence of a specific threshold as identified by Arcand et al. (2012), Law et al. (2013), and Cecchetti and Kharroubi (2012). Therefore, these results do not provide evidence for the notion of “too much finance” (or, more accurately, “too much banking credit”).
Overall, the results provide some evidence of an overall negative (or insignificant) impact of the banking sector on economic growth. However, the banking sector itself appears to have a different impact, depending on the characteristics of the provider of credit.
In contrast, the impact of stock markets is much more clear-cut. Stock markets appear to play an important role in developed markets, with all three variables for stock market development, SECAP, SEVAL, and SETURN, being statistically significant at 1% level of significance. However, the results indicate that only the variable SECAP is statistically significant (and only at 10% level of significance) for emerging markets. However, unlike previous studies into the relationship between stock market development and economic growth in emerging markets, neither value traded nor turnover play a significant role in influencing this growth. Moreover, the results for stock market development in frontier markets are statistically insignificant.
Stock markets within developed markets are far superior to their counterparts within frontier markets in terms of underlying development and size. Given that the MSCI definition of frontier markets dictates that companies wishing to be included within the frontier index must have a market capitalization of USD 630 million and a float size of USD 49 million, this could potentially suggest that stock markets within frontier markets could be dominated by a few large companies with a low volume of trading, as evidenced from the descriptive statistics in Table 14.2. As stock markets are more established within developed markets, it is plausible to suggest, then, that they would have an impact on overall GDP per capita growth.
6. Conclusions
In this chapter the joint contribution of stock markets’ and banks’ development to economic growth has been examined by using annual panel data for 1990–2011 for a sample of 68 countries (23 classified as having a developed stock market, 20 emerging markets, and 25 frontier markets) and utilizing the GMM estimation method.
The results support a number of interesting conclusions; for instance, the role of domestic credit is particularly strong in frontier markets. The study indicates that both private and domestic credit have a relatively significant influence on the countries hosting frontier markets. Interestingly, the impacts of these variables are in opposite directions, with private credit being negative and domestic credit being positive.
The results from the evaluation of stock market variables are less surprising, with significance only for developed markets. This is in line with existing literature concerning the level of development of equity markets within developed markets.
In frontier markets, DOMCRE plays a particularly strong role, but is not able to completely offset the negative impact of PRICRE. The stock market variables, albeit positively signed, are not significant.
In aggregate, the banking sector appears to provide a negative impact on growth in line with previous research, although there does not seem to be particularly strong evidence to support the existence of a threshold leading to “too much finance.” However, the stock markets are able to provide a strong contribution effectively only in the case of developed markets.
Appendix
The MSCI Emerging Markets Index is a free-float-adjusted market capitalization weighted index that is designed to measure the equity market performance of emerging markets. It consists of 23 emerging market country indexes: Brazil, Chile, China, Colombia, Czech Republic, Egypt, Greece, Hungary, India, Indonesia, South Korea, Malaysia, Mexico, Peru, Philippines, Poland, Qatar, Russia, South Africa, Taiwan, Thailand, Turkey, and United Arab Emirates.
The MSCI Frontier Markets Index is a free-float-adjusted market capitalization weighted index that is designed to measure the equity market performance of frontier markets. It consists of 24 frontier market country indexes: Argentina, Bahrain, Bangladesh, Bulgaria, Croatia, Estonia, Jordan, Kenya, Kuwait, Lebanon, Lithuania, Morocco, Kazakhstan, Mauritius, Nigeria, Oman, Pakistan, Romania, Serbia, Slovenia, Sri Lanka, Tunisia, Ukraine, and Vietnam. The MSCI Saudi Arabia Index is currently not included in the MSCI Frontier Markets Index but is part of the MSCI Gulf Cooperation Council (GCC) Countries Index. The MSCI Bosnia Herzegovina Index, the MSCI Botswana Index, the MSCI Ghana Index, the MSCI Jamaica Index, the MSCI Palestine Investable Market Index (IMI), the MSCI Trinidad & Tobago Index, and the MSCI Zimbabwe Index are currently stand-alone country indexes and are not included in the MSCI Frontier Markets Index. The addition of these country indexes to the MSCI Frontier Markets Index is under consideration.
The MSCI World Index is a free-float-adjusted market capitalization weighted index that is designed to measure the equity market performance of developed markets. It consists of 23 developed market country indexes: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Hong Kong, Ireland, Israel, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Singapore, Spain, Sweden, Switzerland, the United Kingdom, and the United States.
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.