Abstract
This paper aims to examine the effect of financial development on economic growth volatility for a sample of 63 countries during the period 1996-2016. Previous studies have reported mixed and inconclusive results regarding such an effect. I ascribe these controversial findings to the limitations of the ordinary least squares (OLS) regression and adopt the method of quantile regression with panel data as developed in Canay (2011). This methodological contribution allows us to test whether the effect of financial development varies across the full distribution, especially at the extreme quantiles of economic growth volatility. Unlike OLS regression, quantile regression captures the whole picture of the relationship between financial development and economic growth volatility by estimating the effect at each quantile of the distribution. Overall, our empirical results show that the effect of financial development on economic growth volatility is negative. However, this effect appears to not be uniform across the quantiles of the economic volatility distribution. This paper sheds more light on the association between financial development and economic growth volatility.
Keywords: Financial development, Economic growth volatility, Panel quantile regression, Developed countries, Developing countries.
JEL Classification: C1; C2; F1.
Received: 2 August 2021 / Revised: 6 April 2022 / Accepted: 22 April 2022/ Published: 9 May 2022
Contribution/ Originality
This study contributes to the existing literature by empirically investigating the potentially heterogeneous patterns of financial development effects across economic growth volatility distribution by applying the quantile regression approach proposed by Koenker and Bassett Jr (1978). Moreover, I adopt fixed-effect models which provide the ability to control unobservable time-invariant country characteristics.
1. INTRODUCTION
The link between financial development and economic growth has been the subject of extensive literature. From a theoretical perspective, it is firmly established that well-developed stock markets and correctly functioning banking systems spur economic growth and are therefore consistent with the proposition of ‘‘more finance, more growth.” From the standpoint of empirical research (Beck, Levine, & Loayza, 2000; Beck & Levine, 2004; Beck, Demirguc-Kunt, Laeven, & Levine, 2008; Goldsmith, 1969; King & Levine, 1993a; King & Levine, 1993b; Levine, 1997; Levine & Zervos, 1998; Levine, Loayza, & Beck, 2000; McCaig & Stengos, 2005; Rajan & Zingales, 1998; and Thumrongvit, Kim, & Pyun, 2013) studies on the finance-growth nexus report strong and robust evidence sustaining the hypothesis that economic growth is positively associated with financial development. In addition to economic growth, policymakers have also considered economic stability as a major objective of macroeconomic policies (Huang, Fang, & Miller, 2014; Mishkin, 2009; Yellen & Akerlof, 2006) . Therefore, academic researchers have focused on examining the relationship between financial development and macroeconomic volatility.
The theory provides competing hypotheses regarding how financial development can affect economic growth volatility. First, it is argued that a well-developed financial system can smooth economic growth volatility by withdrawing or reducing financial constraints that cause the propagation of business cycles and thus lead to a higher fluctuation of economic growth indicators (Bernanke & Gertler, 1989; Kiyotaki & Moore, 1997). A second argument is that the positive or negative effects of financial development on growth volatility are subject to real and monetary shocks, credit supply, and the prevailing stage (early, intermediate, or late) of financial development (Aghion, Bacchetta, & Banerjee, 2004; Bacchetta & Caminal, 2000; Morgan, Rime, & Strahan, 2004) .
Some existing empirical studies on the impact of financial development on economic growth volatility also provide conflicting and inconclusive findings. For example, Beck, Lundberg, and Majnoni (2006), by using a panel of 63 countries during the period 1960–1997, and the volatility of terms of trade and inflation as a proxy for real and monetary volatility, respectively, presented weak evidence that financial intermediaries alleviate economic expansion. Other empirical studies report a negative effect of financial development on macroeconomic volatility (Acharya, Imbs, & Sturgess, 2011; Da Silva, 2002; Dabla-Norris & Srivisal, 2013; Denizer, Iyigun, & Owen, 2000; Larrain, 2006; Mallick, 2014; Manganelli & Popov, 2015; and Raddatz, 2006) . Conversely, further studies report the evidence of a nonlinear relationship, especially U-shaped, between financial development and output volatility (Alatrash, Leff, Minten, Soupre, & Van Schoot, 2014; Arcand, Berkes, & Panizza, 2012; Dabla-Norris & Srivisal, 2013; Easterly, Islam, & Stiglitz, 2002; and Ibrahim & Alagidede, 2017) . That is, financial development dampens economic growth volatility to an optimal point, after which any increase in the level of financial development will increase economic volatility. More recently, Zouaoui, Mazioud, and Ellouz (2018) examine the relationship between financial development and economic growth volatility for a sample of 50 developing countries during the period 1960-2016 by using a semi-parametric, panel-fixed effects regression model and find that the financial development-volatility relationship is not linear and has multiple turning points.
I ascribe these conflicting findings to the regression techniques used by the referenced studies. Most of these studies employ classical regression estimation techniques such as ordinary least squares (OLS) estimation, panel regression, or instrumental variables regression. These estimation techniques only focus on the conditional mean of the dependent variable, which may result in biased results, especially where the distribution heterogeneity of the dependent variable is neglected (Binder & Coad, 2011). In contrast to previous studies, in this paper, I examine the effect of financial development on economic growth volatility by taking into account the distribution heterogeneity and using the panel quantile regression with fixed effects as introduced by Canay (2011).
The advantages of employing the panel quantile regression with fixed effects are threefold. First, the use of the panel regression methodology allows us to get greater efficiency in estimating the effect of financial development on economic growth volatility because employing a cross-country study provides more informative data, variability, and therefore more degree of freedom. Second, the panel data framework allows the consideration of unobserved individual heterogeneity in the estimation process. The omission of such heterogeneity may lead to unbiased estimators. Third, the use of the quantile regression framework helps us to get a complete picture of the relationship between financial development and economic growth volatility because this method allows us to describe the entire distribution of the economic growth volatility by estimating the effect of financial development at each quantile of the economic growth volatility distribution. In addition, the panel quantile regression with fixed effects produces more robust estimators compared with classical approaches, especially when the distribution of the dependent variable is skewed and/or the error term is not normal distributed.
In summary, I employ a fixed effect panel quantile regression to explore the effect of financial development on economic growth volatility for selected developed and developing countries during the period 1996-2016. This paper may contribute to the existing literature in several ways. First, the work gives a more detailed picture than is currently available on the association between financial development and economic growth volatility by constructing point estimates at each quantile, especially in the highest and lowest ones, by using panel quantile regression. Second, I use the principal component analysis (PCA) framework to construct a comprehensive proxy of financial development. As such, I construct three indices: bank system development index, stock market development index, and overall financial development index. Third, I estimate based on both developed and developing countries.
The remaining paper below is structured with section 2 describing the data, variables, and descriptive statistics, section 3 presenting the estimation approach, section 4 providing empirical results and analysis, and section 5 consisting of the conclusions.
2. DATA DESCRIPTION, VARIABLES, AND DESCRIPTIVE STATISTICS
2.1. Data
The aim of this paper is to examine the effect of financial development on economic growth volatility by using data from 64 developing and developed countries over the period 1996-2016. Countries are selected so that there are enough available data on variables used in the regression analysis. The list of countries considered in our study is given in Table 1.
Table 1. List of countries.
| Developed countries (30 countries) |
Australia, Austria, Belgium, Canada, Chile, Cyprus, Denmark, Finland, France, Germany, Greece, Hong Kong, Iceland, Ireland, Israel, Italy, Japan, Korea Rep., Luxembourg, Malta, Netherlands, New Zealand, Norway, Portugal, Singapore, Spain, Sweden, Switzerland, United Kingdom, United States. |
| Developing countries (34 countries) |
Argentina, Bahrain, Brazil, Bulgaria, China, Colombia, Croatia, Czech Republic, Egypt, Estonia, Hungary, India, Indonesia, Jordan, Kuwait, Lithuania, Malaysia, Mexico, Morocco, Nigeria, Pakistan, Peru, Philippines, Poland, Qatar, Romania, Russian Federation, Saudi Arabia, South Africa, Sri Lanka, Thailand, Tunisia, Turkey, United Arab Emirates. |
Note: The countries are classified into developed and developing countries according to the World Bank’s classification of countries.
Table 2. Principal component analysis.
| Panel A: Bank development subindex (FD_BANK) | |||||||
PC_1 |
PC_2 |
PC_3 |
PC_4 |
||||
| Eigen values | 3.240 |
0.640 |
0.094 |
0.023 |
|||
| Percentage of variance | 0.810 |
0.160 |
0.024 |
0.006 |
|||
| Cumulative percentage | 0.810 |
0.970 |
0.994 |
1.000 |
|||
| Variables | Comp_1 |
Comp_2 |
Comp_3 |
Comp_4 |
|||
| DMBA | 0.489 |
0.527 |
0.694 |
-0.026 |
|||
| FSD | 0.507 |
-0.491 |
-0.011 |
-0.708 |
|||
| LL | 0.506 |
-0.494 |
0.045 |
0.705 |
|||
| PRIVATE | 0.496 |
0.485 |
-0.718 |
0.029 |
|||
| Panel B: Market development subindex (FD_MARK) | |||||||
PC_1 |
PC_2 |
PC_3 |
|||||
| Eigen values | 2.037 |
0.850 |
0.112 |
||||
| Percentage of variance | 0.679 |
0.283 |
0.038 |
||||
| Cumulative percentage | 0.679 |
0.962 |
1.000 |
||||
| Variables | Comp_1 |
Comp_2 |
Comp_3 |
||||
| SMKT | 0.507 |
0.728 |
0.459 |
||||
| SMTV | 0.679 |
-0.009 |
-0.734 |
||||
| TURNOVER | 0.530 |
-0.684 |
0.499 |
||||
| Panel C: Overall financial development index (FD_OVERALL) | |||||||
PC_1 |
PC_2 |
||||||
| Eigen values | 1.414 |
0.585 |
|||||
| Percentage of variance | 0.707 |
0.293 |
|||||
| Cumulative percentage | 0.707 |
1.000 |
|||||
| Variables | Comp_1 |
Comp_2 |
|||||
| FD_BANK | 0.707 |
0.707 |
|||||
| FD_MARK | 0.707 |
-0.707 |
|||||
2.2. Variable Measures
2.2.1. Financial Development Measures: Principal Component Analysis (PCA)
The review of previous studies reveals that financial development is measured through a variety of proxies. However, there is no consensus evidence showing the appropriate proxy to measure the extent of financial development (Tang & Tan, 2014). To overcome this problem, I employ the method of principal component analysis to construct a comprehensive indicator for financial development. To do so, I follow (Ang & McKibbin, 2007; Ang, 2009; Coban & Topcu, 2013; Gries, Kraft, & Meierrieks, 2009; Huang, 2010; Saci & Holden, 2008; Shahbaz, Shahzad, Ahmad, & Alam, 2016; Tang & Tan, 2014; and Topcu & Payne, 2017) together with three sub-indices and the overall financial development index. The first financial development subindex (FD_BANK) is constructed using banking sector development proxies, including deposit money bank assets to GDP (DMBA), financial system deposit to GDP (FSD), liquid liabilities to GDP (LL), and private credit by deposit money banks to GDP (PRIVATE). The second financial development subindex (FD_MARK) is obtained based on the stock market development index, which covers the stock market capitalization to GDP (MKT), stock market turnover ratio (TURNOVER), and stock market total value traded to GDP (SMVT). The third financial development subindex is an overall index computed as the principal component of the seven typical measures of financial development as mentioned above.
Table 2 shows the result of the principal component analysis of each financial development indicator with the factor scores for each variable.
2.2.2. The Dependent Variable and Control Variables
Our main dependent variable is economic growth volatility. Following the studies of Easterly, Islam, and Stiglitz (2000); Acemoglu, Johnson, Robinson, and Thaicharoen (2003); Ahamada and Coulibaly (2011) and Ma and Song (2018) among others, the volatility of economic growth (VOLAT) is measured by the standard deviation of real gross domestic product (GDP). I use a five-year rolling window to compute the standard deviation. The real GDP per capita is drawn from the World Development Indicator (WDI) published by the World Bank.
In accordance with existing literature, I include a set of control variables that are known to influence economic growth volatility. The set of control variables includes GDP growth rate (GROWTH), inflation rate (INFL), financial openness (FO), government expenditure (EXPEND), remittances and compensation of employees as a percentage of GDP (REMITTANCE), the logarithm of real per capita gross domestic product in 1996 (Initial GDP), and the trade dependency ratio being the sum of exports and imports of goods and services measured as a share of gross domestic product (TO). The definition and data sources of our variables are reported in Table 3.
2.3. Descriptive Statistics
Table 4 reports descriptive statistics for the variable of interest used in our study, where I give the number of observations (N), the mean, the standard deviation, and the minimum and the maximum for both the full sample countries and the developing and developed countries. The economic volatility of our sample countries is, on average, equal to 2.3% and ranges between 0.6% and 6.3%. In addition, economic growth volatility is about 2.9% in developing countries and 2% in developed countries over the full period.
The mean difference of economic volatility between developed and developing countries is statistically significant at 1% level. Furthermore, I find that developed countries are characterized by higher levels of financial development compared with developing countries. Indeed, the T-test for the mean difference is statically significant for all proxies of financial development. Regarding control variables, the inflation rate average is 4.3%, constituting about 6.2% in developing countries and 2.1% in developed countries. The economic growth is equal to 3.5% on average and is significantly higher in developing countries when compared with developed countries.
Table 3. Definitions and sources of variables.
| Variable | Definition | Source |
| VGDP | Macroeconomic volatility computed as the five-year rolling window standard deviations of real GDP growth rate. | Authors’ calculation |
| VINFL | Aggregate price volatility computed as 5-year rolling window standard deviations of annual inflation rate. | Authors’ calculation |
| DMBA | Deposit money bank assets to GDP (%). | WDI* |
| FSD | Financial system deposits to GDP (%). | WDI |
| LL | Liquid liabilities to GDP (%). | WDI |
| PRIVATE | Private credit by deposit money banks to GDP (%). | WDI |
| SMKT | Stock market capitalization to GDP (%). | WDI |
| SMTV | Stock market total value traded to GDP (%). | WDI |
| TURNOVER | Stock market turnover ratio (%). | WDI |
| GROWTH | Annual growth rate of real GDP. | WDI |
| FO | Financial openness index constructed by Chinn and Ito. | Chinn-Ito website** |
| REMITTANCE | Migrants’ remittances are given by the sum of workers’ remittances and compensation of employees as a percentage of GDP. | WDI |
| EXPENDITURE | Government consumption divided by GDP. | WDI |
| INFLATION | Inflation, GDP deflator (annual %). | WDI |
| GDP_INITIAL | Logarithm of real per capita GDP in 1996. | Authors’ calculation |
| TRADE | The trade dependency ratio is the sum of exports and imports of goods and services measured as a share of GDP. | WDI |
| SCHOOL | Log of one plus the average years of secondary schooling of the adult population. | Authors’ calculation |
| REER | Real effective exchange rate. | WDI |
Note: *WDI denotes the World Bank’s World Development Indicators database; ** Data are available at the Chinn-Ito index website
In addition to descriptive statistics, I conduct further analysis concerning the normality of the variables used in this study. Table 5 reports the Skewness and the Kurtosis statistics for all the variables as well as the test for normality proposed by Doornik and Hansen (2008), which is based on a test by Shenton and Bowman (1977). As can be seen clearly, the distributions of all the variables are skewed, and the Kurtosis values show that the distributions of our variables of interest exhibit tail data exceeding the tails of the normal distribution. The Doornik and Hansen (2008) normality test is statistically significant for all variables, indicating the rejection of the normality hypothesis. This result suggests that the use of the classical OLS regression may not be appropriate for our data analysis, and the quantile approach could be more suitable and more robust for non-normal errors and outliers.
3. PENALIZED PANEL QUANTILE REGRESSION WITH FIXED EFFECTS
To understand how quantile regression works, we start from the classical linear regression model. Consider the following linear model:
Table 4. Descriptive statistics.
Variable |
All countries |
Developing countries |
Developed countries |
T-test for mean difference |
||||||||||||
N |
Mean |
SD |
Min |
Max |
N |
Mean |
SD |
Min |
Max |
N |
Mean |
SD |
Min |
Max |
||
Economic volatility |
||||||||||||||||
VGDP |
1280 |
0.025 |
0.016 |
0.006 |
0.063 |
680 |
0.029 |
0.017 |
0.006 |
0.063 |
600 |
0.020 |
0.013 |
0.006 |
0.063 |
10.16*** |
VINFL |
1274 |
0.030 |
0.034 |
0.004 |
0.131 |
675 |
0.046 |
0.040 |
0.004 |
0.131 |
599 |
0.013 |
0.011 |
0.004 |
0.059 |
19.22*** |
Financial development |
||||||||||||||||
DMBA |
1309 |
0.828 |
0.427 |
0.218 |
1.683 |
703 |
0.577 |
0.299 |
0.218 |
1.683 |
606 |
1.118 |
0.364 |
0.218 |
1.683 |
-29.53*** |
FSD |
1287 |
0.659 |
0.397 |
0.176 |
1.804 |
704 |
0.460 |
0.241 |
0.176 |
1.264 |
583 |
0.898 |
0.416 |
0.176 |
1.804 |
-23.61*** |
LL |
1293 |
0.757 |
0.431 |
0.237 |
1.903 |
696 |
0.565 |
0.306 |
0.237 |
1.903 |
597 |
0.981 |
0.448 |
0.292 |
1.903 |
-19.75*** |
PCDMB |
1308 |
0.701 |
0.411 |
0.140 |
1.555 |
703 |
0.446 |
0.278 |
0.140 |
1.555 |
605 |
0.996 |
0.335 |
0.278 |
1.555 |
-32.43*** |
SMC |
1261 |
0.617 |
0.465 |
0.105 |
1.815 |
665 |
0.459 |
0.401 |
0.105 |
1.815 |
596 |
0.794 |
0.469 |
0.105 |
1.815 |
-13.65*** |
SMTVT |
1274 |
0.366 |
0.420 |
0.007 |
1.451 |
672 |
0.196 |
0.267 |
0.007 |
1.451 |
602 |
0.555 |
0.476 |
0.007 |
1.451 |
-16.80*** |
SMTO |
1262 |
0.541 |
0.465 |
0.037 |
1.651 |
662 |
0.431 |
0.439 |
0.037 |
1.651 |
600 |
0.663 |
0.462 |
0.037 |
1.651 |
-9.13*** |
Control variables |
||||||||||||||||
FO |
1323 |
1.149 |
1.422 |
-1.202 |
2.360 |
714 |
0.460 |
1.410 |
-1.202 |
2.360 |
609 |
1.956 |
0.925 |
-1.202 |
2.360 |
-22.38*** |
REMITTANCE |
1237 |
0.015 |
0.021 |
0.000 |
0.075 |
634 |
0.024 |
0.025 |
0.000 |
0.075 |
603 |
0.006 |
0.008 |
0.000 |
0.053 |
17.58*** |
TO |
1339 |
0.895 |
0.541 |
0.284 |
2.449 |
709 |
0.809 |
0.413 |
0.284 |
2.204 |
630 |
0.992 |
0.643 |
0.284 |
2.449 |
-6.23*** |
INFL |
1338 |
0.043 |
0.047 |
-0.013 |
0.172 |
709 |
0.062 |
0.054 |
-0.013 |
0.172 |
629 |
0.021 |
0.021 |
-0.013 |
0.154 |
18.14*** |
GROWTH |
1344 |
0.035 |
0.029 |
-0.025 |
0.090 |
714 |
0.042 |
0.031 |
-0.025 |
0.090 |
630 |
0.026 |
0.025 |
-0.025 |
0.090 |
10.55*** |
EXPEND |
1339 |
0.170 |
0.045 |
0.092 |
0.247 |
709 |
0.157 |
0.044 |
0.092 |
0.247 |
630 |
0.185 |
0.042 |
0.092 |
0.247 |
-11.91*** |
GDP_INI |
1323 |
9.385 |
0.795 |
7.844 |
10.447 |
693 |
8.855 |
0.735 |
7.844 |
10.447 |
630 |
9.968 |
0.292 |
8.991 |
10.447 |
-35.55*** |
| SCHOOL | ||||||||||||||||
| REER | ||||||||||||||||
Note: This table reports descriptive statistics of the variables of interest. I report the mean (MEAN), the standard deviation (SD), the minimum (MIN), and the maximum (MAX). The statistics are computed for all countries, and both developing countries and developed countries. I also report the T statistics, in the last column, for differences between developing and developed countries. All variables are defined in Table 3.
***Statistical significance at the 1% level.
Table 5. Normality test.
| Variable | Skewness |
Kurtosis |
D-H |
P-Value |
| VGDP | 1.011 |
3.014 |
496.96*** |
0.000 |
| DMBA | 0.413 |
2.145 |
59.59*** |
0.000 |
| FSD | 1.290 |
4.505 |
335.63*** |
0.000 |
| LL | 1.130 |
3.729 |
318.96*** |
0.000 |
| PCDMB | 0.504 |
2.267 |
60.532*** |
0.000 |
| SMC | 1.093 |
3.392 |
323.11*** |
0.000 |
| SMTVT | 1.314 |
3.661 |
382.07*** |
0.000 |
| SMTO | 0.983 |
3.020 |
167.34*** |
0.000 |
| FO | -0.649 |
1.767 |
305.93*** |
0.000 |
| REMITTANCE | 1.747 |
5.054 |
2856.95*** |
0.000 |
| TRADE | 1.442 |
4.612 |
393.23*** |
0.000 |
| INFLATION | 1.408 |
4.386 |
559.47*** |
0.000 |
| GROWTH | -0.042 |
2.633 |
1.16 |
0.559 |
| EXPENDITURE | -0.134 |
2.017 |
48.18*** |
0.000 |
| GDP_INITIAL | -0.428 |
1.939 |
147.08*** |
0.000 |
| SCHOOL | -0.422 |
3.293 |
15.73*** |
0.000 |
| REER | 0.280 |
2.945 |
29.49*** |
0.000 |
Note: this table reports Skewness and Kurtosis for the variables of interest.
***Statistical significance at the 1% level.
OLS and LAD regression, as given by Equation 2 and Equation 3, respectively, may suffer from some limitations, which, in turn, could alter the quality of the estimates. First, OLS or LAD produce estimators that are best unbiased linear estimators if errors are independent and identically distributed as a normal distribution. However, these assumptions may not hold, especially when the mean distribution of the dependent variable is commonly affected by extreme values. Therefore, it is more appropriate to use non-mean-based models. Second, OLS and LAD methods provide only one estimate based on the central distribution of the dependent variable. Therefore, OLS and LAS provide only a partial view of the relationship between the dependent variable and the independent variables and the covariates. In other words, neither the OLS nor the LAD methods consider the full distribution of the dependent variable, liquidity creation, especially in the tail regions. To overcome these drawbacks, other regression models have emerged, like the QR approach.
Koenker and Bassett Jr (1978) developed a new approach named the quantile regression (QR), which is a natural extension of the OLS and LAD models. However, unlike the OLS and LAD regression, the QR approach enables the estimation of a set of models for conditional quantile functions by allowing the consideration of the effect of the independent variables on the entire distribution of the response variable, not merely its conditional mean. Furthermore, the QR approach is proved to be more robust in the face of non-normal errors and extreme values and does not need strict assumptions as for classical linear regression like normality, homoscedasticity, or absence of outliers (Johnston & Dinardo, 1997).
To minimize the objective function in Equation 4, the Barrodale and Roberts (1974) approach can be used which is mainly based on the simplex algorithm. Koenker and Machado (1999) establish that the minimization of Equation 4 is related to likelihood-based inference using independently distributed asymmetric Laplace densities (ALD). That is, minimizing Equation 4 is equivalent to maximizing an ALD-based likelihood function. is consistent and asymptotically normal under certain regularity conditions (Canay, 2011). The quantile approach has received increased attention in academic circles and has been widely employed in different research fields, especially in economic studies (Binder & Coad, 2015; Rosendo, Simões, & Andrade, 2018; Wang, Zhu, Guo, & Peng, 2018; Zhu, Duan, Guo, & Yu, 2016; Zhu, Guo, You, & Xu, 2016; Zhu, Xia, and Guo, & Peng, 2018) .
However, the classical QR method does not consider the unobserved individual heterogeneity. To overcome this shortcoming, some works, including those by Koenker (2004); Lamarche (2010); Galvao Jr (2011), and Canay (2011), provide econometric theory to deal with unobserved individual heterogeneity and propose applying quantile regression to the case of panel data. Therefore, the proposed fixed effect panel quantile regression model is given as follows:
As discussed in Lancaster (2000) and Neyman and Scott (1948) the incorporation of a substantial number of fixed effect may lead to incidental parameters problems that render estimators inconsistent. Besides, in the quantile regression framework, the removal of unobserved fixed effects cannot be achieved by classical approaches that rely on the hypothesis that expectations are linear operators (Canay, 2011). To overcome this problem of eliminating individual fixed effects in the QR framework, Koenker (2004); Lamarche (2010); Galvao Jr (2011), and Canay (2011) propose the penalized quantile regression for panel data by introducing a penalty term in the minimization problem. Formally, the estimation of parameters in Equation 5 is conducted by solving the following minimization problem:
Where the countries are indexed by i and time by t. VOLAT is the economic growth volatility. The description of other variables is provided in Table 3.
4. ESTIMATION RESULTS AND DISCUSSION
4.1. Panel Unit Root Test Results
Before running our fixed effects panel quantile regression models, it is important to test whether the variables of interest are stationary. I conduct three panel unit root tests, namely the Fisher-type unit root test based on an augmented Dickey-Fuller (Fisher-ADF) (Dickey & Fuller, 1979), the Fisher-type unit root test based on the Phillips-Perron test (Fisher-PP) Phillips and Perron (1988), and the Im, Pesaran, and Shin (2003) test (IPS). Table 6 presents the results of the panel unit root tests. The findings suggest that the null hypothesis that all the panels contain a unit root could not be rejected for all the variables at level. However, at the first difference of all variables, the null hypothesis could almost be rejected at the 1% level. These results imply that the use of the first difference for all variables is necessary.
Table 6. Panel unit root test.
Levels |
First difference |
|||||
Fisher-ADF |
Fisher-PP |
IPS |
Fisher-ADF |
Fisher-PP |
IPS |
|
| VGDP | 3.184 |
-3.872*** |
1.502 |
-4.387*** |
-23.487*** |
-8.947*** |
| SDINFL | -1.954*** |
-3.453*** |
-2.741*** |
-9.579*** |
-19.474*** |
-8.021*** |
| FD_BANK | 3.515 |
7.019 |
3.211 |
-9.591*** |
-12.208*** |
-8.030*** |
| FD_MARK | -5.508*** |
-0.838 |
-4.558*** |
-13.868*** |
-12.001*** |
-11.486*** |
| FDBOND | 5.226 |
3.968 |
-4.864*** |
-14.343*** |
||
| FD_OVERALL | 4.560 |
3.510 |
-5.087*** |
-9.697*** |
||
| DMBA | 6.620 |
7.111 |
4.739 |
-8.937*** |
-9.961*** |
-7.511*** |
| FSD | 4.952 |
2.815 |
3.477 |
-13.359*** |
-13.249*** |
-11.494*** |
| LL | 3.396 |
3.769 |
2.516 |
-13.590*** |
-11.973*** |
-11.552*** |
| PCDMB | 6.974 |
7.202 |
4.924 |
-7.709*** |
-9.484*** |
-6.402*** |
| SMC | -0.267 |
-0.046 |
-2.439*** |
-14.591*** |
-10.594*** |
-12.254*** |
| SMTVT | 2.350 |
2.888 |
0.554 |
-13.929*** |
-6.923*** |
-11.492*** |
| SMTO | 1.080 |
-6.932*** |
-0.359 |
-17.902*** |
-27.922*** |
-15.242*** |
| FO | -2.743*** |
2.288 |
-3.972 |
-4.974*** |
-17.355*** |
-4.437*** |
| REMITTANCE | 0.157 |
2.249 |
-10.727*** |
-17.724*** |
||
| TO | -0.299 |
-0.904 |
-1.297 |
-13.381*** |
-18.702*** |
-12.290*** |
| INFL | -2.159*** |
-15.378*** |
-2.971*** |
-26.789*** |
-40.848*** |
-24.203*** |
| GROWTH | 0.9496 |
-10.956 |
-0.031 |
-25.330*** |
-33.105*** |
-22.761*** |
| EXPEND | 3.3371 |
3.771 |
2.179 |
-12.062*** |
-20.060*** |
-10.102*** |
| SCHOOL | 1.271 |
0.984 |
-2.340*** |
-13.944*** |
||
| REER | 2.715 |
3.012 |
1.178 |
-2.302** |
-18.575*** |
-2.978*** |
Note: Fisher-ADF and Fisher-PP represent the Maddala and Wu (1999) Fisher-ADF and Fisher-PP panel unit root tests, respectively. IPS represents the panel unit root tests of Im et al. (2003), respectively. The maximum number of lags is set to one. The Schwarz information criterion (SIC) is used to select the lag length. The bandwidth is selected using the Newey–West method. Bartlett is used as the spectral estimation method. The exogenous variables are the individual effects and individual linear trends. n.a. refers to not available.
*** Statistical significance at the 1% level.
** Statistical significance at the 5% level.
4.2. Panel Quantile Regression Results
Before estimating the panel quantile regression model and to facilitate comparison, the model in Equation 7 is first estimated by the pooled OLS and panel data regression technique. The results are reported in Table 7, where financial development is measured using the aforementioned indices FD_BANK, FD_MARK and FD_OVERALL. For each proxy, I report the pooled OLS regression estimates and panel data estimations with both fixed and random effects. Overall, the results reported in Table 7 show that all the measures of financial development have a negative and statically significant effect on economic growth volatility. Specifically, an increase in the overall financial development level will lead to the reduction of economic volatility. In addition, the negative effect of financial development on economic volatility comes from both banking sector and financial market development. However, as discussed above, the classical regression approaches provide only a description of the average relationship between financial development and economic growth volatility and do not take into consideration the distributional heterogeneity.
Table 7. Financial development and economic growth volatility: OLS and panel regression.
Notes: This table reports the result of the impact of financial development on economic growth volatility using OLS and panel regression with fixed (FE) and random (RE) effects. FD_BANK is bank system development index. FD_MARK is financial market development index. FD_OVERALL is overall financial development index. The other variables are defined in Table 3. Figures in parentheses are t-values. is the first difference.
***Statistical significance at the 1% level.
** Statistical significance at the 5% level.
* Statistical significance at the 10% level.
Table 8. The effect of banking sector development on economic growth volatility: A panel quantile regression.
Notes: This table reports the result of the panel quantile regression model with Bond development index (FD_BANK) as a measure of financial development. All variables are defined in Table 3. Figures in parentheses are t-values. is the first difference
*** Statistical significance at the 1% level.
** Statistical significance at the 5% level.
* Statistical significance at the 10% level.
To control for the distributional heterogeneity and to have a more complete picture of the association between financial development and economic growth volatility, I employ the panel quantile regression with fixed effects as introduced by Canay (2011). Table 8, Table 9, and Table 10 report the result of the fixed effects panel quantile regression for the three measures of financial development, namely FD_BANK, FD_MARK, and FD_OVERALL, respectively. The results are reported for the 5th, 10th, 20th, 30th, 50th, 60th, 70th, 80th, 90th, and 95th percentiles of the conditional economic volatility distribution. The higher quantiles, such as the 60th, 70th, 80th, 90th, and 95th refer to the countries with higher economic growth volatility. However, the lower quantiles, such as the 10th, 20th, 30th, 40th, and 50th refer to the countries with lower economic growth volatility. Overall, the findings indicate that the impacts of various measures of financial development on economic growth volatility are heterogeneous.
Regarding the effect of the banking sector development (FD_BANK) on economic growth volatility, as reported in Table 8, the impact is heterogeneous. At the 95th quantile, the coefficient of FD_BANK is negative but is insignificant at the 10% level. For the other quantiles, the coefficient is negative and statistically significant at conventional level, implying that the effect of banking sector development on economic growth volatility is negative and that the effect is more significant in lower volatility countries. However, the coefficient of
FD_BANK is greater at higher quantiles compared with lower quantiles. That is, in countries with higher economic growth volatility, the impact of bank sector development is more pronounced compared with countries with lower economic growth volatility. The corresponding panel quantile regression diagram of the coefficients is provided in Figure 1. As can be seen in Figure 1, there is a monotonic increase in the impact of bank sector development over the quantiles of economic growth volatility distribution.
Figure 1. Quantile regression estimates with 95% confidence intervals for the impact of bank sector development (FD_BANK) on economic growth volatility. The vertical axes show the coefficient estimates of the variables over the economic growth volatility’ distribution. The horizontal axes depict the quantile levels. The red horizontal dashed lines represent the corresponding OLS estimations with their 95% confidence interval.
Table 9. The effect of financial market development on economic growth volatility: A panel quantile regression.
Notes: This table reports the result of the panel quantile regression model with Bond development index (FD_MARK) as a measure of financial development. All variables are defined in Table 3. Figures in parentheses are t-values. is the first difference.
***Statistical significance at the 1% level.
** Statistical significance at the 5% level.
* Statistical significance at the 10% level.
Figure 2. Quantile regression estimates with 95% confidence intervals for the impact of financial market development (FD_MARK) on economic growth volatility. The vertical axes show the coefficient estimates of the variables over the economic growth volatility’s distribution. The horizontal axes depict the quantile levels. The red horizontal dashed lines represent the corresponding OLS estimations with their 95% confidence interval.
Regarding the financial market development (FD_MARK), the results of its impact on economic growth volatility is reported in Table 9. The corresponding panel quantile regression diagram of the coefficients is given in Figure 2. I can observe from Table 9 and Figure 2 that the effect of financial market development (FD_MARK) on economic growth volatility is negative and statistically significant for all quantiles and the effect is more pronounced at higher quantiles. These results indicate that an increase in the development level of financial markets can impede economic output volatility and therefore foster economic stability.
As regards the impact of the overall financial development level, the results are reported in Table 10. As for the other measures of financial development, I find that the coefficient of FD_OVERALL is also negative and statistically significant for all quantiles except for the 95th quantile where the coefficient is negative but not significant at 10% level. The corresponding panel quantile regression diagram of the coefficients is given in Figure 3. As can be seen from Figure 3, there is a monotonic increase in the impact of overall financial development over the quantiles of economic growth volatility distribution. This implies that well-developed financial systems, including both banking sector and financial market development, can dampen output volatility by removing or reducing financial constraints and information asymmetry.
To verify the heterogeneity of the estimates, I conduct inter-quantile tests. These tests allow us to check whether the differences among the estimated coefficients of financial development indices are significant across quantiles. Following Koenker and Bassett (1982), I conduct Wald tests to check the equality of financial development coefficients across quantiles. To save space, I present only the results of the Wald test concerning whether the lower quantile (the 5th quantile) is equal to the middle quantile (the 50th quantile) and the higher quantile (the 95th quantile). The result of the Wald test of slope equality across quantiles is reported in Table 11.
Table 10. The effect of overall financial system development on economic growth volatility: A panel quantile regression.
Notes: This table reports the result of the panel quantile regression model with overall financial development index (FD_OVERALL) as a measure of financial development. All variables are defined in Table 3. Figures in parentheses are t-values. is the first difference.
*** Statistical significance at the 1% level.
** Statistical significance at the 5% level.
* Statistical significance at the 10% level.
Figure 3. Quantile regression estimates with 95% confidence intervals for the impact of overall financial development (FD_OVERALL) on economic growth volatility. The vertical axes show the coefficient estimates of the variables over the economic growth volatility’s distribution. The horizontal axes depict the quantile levels. The red horizontal dashed lines represent the corresponding OLS estimations with their 95% confidence interval.
As can be seen from Table 11, the null hypothesis of equality of slopes is rejected for all cases implying that the parameter estimates of financial development are heterogeneous across quantiles. In conclusion, our results show that it is important to consider the distributional heterogeneity when examining the effect of financial development on economic growth volatility. Besides, compared with OLS regression, the panel quantile regression with fixed effects provides a more complete picture of such associations. Next, our findings suggest that the impact of financial development, including both banking system and financial market development, on economic growth volatility is clearly heterogeneous. The results demonstrate that an increase in the level of bank system development and/or financial market development can lead to a more stable economic growth rate.
Table 11. Wald tests for the equality of slopes (5th against 50th and 95th quantiles).
Against the 50th quantile |
Against the 95th quantile |
|||
Test statistic |
p-value |
Test statistic |
p-value |
|
19.29⁎⁎⁎ |
0.000 |
2.72⁎ |
0.099 |
|
16.18⁎⁎⁎ |
0.000 |
5.73⁎⁎ |
0.016 |
|
12.45⁎⁎⁎ |
0.057 |
3.63⁎ |
0.057 |
|
⁎⁎⁎ Statistical significance at the 1% level.
⁎⁎Statistical significance at the 5% level.
⁎ Statistical significance at the 10% level.
5. CONCLUSION
The main objective of this study is to explore the impact of financial development on economic growth volatility. To do so, I employ quantile regression under the panel data framework. Unlike classical estimation methods, including OLS and panel regressions, the proposed method considers the distributional heterogeneity and unobserved individual heterogeneity. Besides, the panel quantile regression is preferred to OLS because it allows us to obtain a more complete image of the impact of financial development on economic volatility. For implementation, I use a sample of 63 developed and developing countries during the period 1996-2016. To measure the level of financial development, I use the methodology of principal component analysis and construct three indices based on commonly used variables that describe the level of banking sector development and financial market development. The third proxy used is an overall financial development index.
For the case of OLS and panel regression, I find that banking sector development, financial market development, and overall financial development impact economic growth volatility both negatively and significantly. However, our empirical results based on the panel quantile regression show that the impact of various proxies of financial development on economic growth volatility is found to be heterogeneous across quantiles. This implies that it is important to take into consideration the distributional heterogeneity when testing the effect of financial development on economic growth volatility. All in all, looking at the full distribution of economic growth volatility instead of focusing on the average effects allows us to shed more light on the association between financial development and economic output volatility and reconcile previous empirical results.
Funding: This study received no specific financial support. |
Competing Interests: The author declares that there are no conflicts of interests regarding the publication of this paper. |
REFERENCES
Acemoglu, D., Johnson, S., Robinson, J., & Thaicharoen, Y. (2003). Institutional causes, macroeconomic symptoms: Volatility, crises and growth. Journal of Monetary Economics, 50(1), 49-123. Available at: https://doi.org/10.1016/s0304-3932(02)00207-6.
Acharya, V., Imbs, J., & Sturgess, J. (2011). Finance and efficiency: Do bank branching regulations matter? European Finance Association, 15(1), 135- 172. Available at: https://doi.org/10.1093/rof/rfq009.
Aghion, P., Bacchetta, P., & Banerjee, A. (2004). Financial development and the instability of open economies. Journal of monetary Economics, 51(6), 1077-1106. Available at: https://doi.org/10.1016/j.jmoneco.2003.12.001.
Ahamada, I., & Coulibaly, D. (2011). How does financial development influence the impact of remittances on growth volatility? Economic Modelling, 28(6), 2748-2760. Available at: https://doi.org/10.1016/j.econmod.2011.08.019.
Alatrash, Y., Leff, D., Minten, T., Soupre, M., & Van Schoot, D. (2014). Financial development and macroeconomic volatility. Working Paper. Utrecht University School of Economics. Retrieved from: https://www.uu.nl/sites/default/files/rebo_use_dp_2017_1713.pdf.
Ang, J. B., & McKibbin, W. J. (2007). Financial liberalization, financial sector development and growth: Evidence from Malaysia. Journal of Development Economics, 84(1), 215-233. Available at: https://doi.org/10.1016/j.jdeveco.2006.11.006.
Ang., J. B. (2009). Financial development and the FDI-growth nexus: The Malaysian experience. Applied Economics, 41(13), 1595-1601. Available at: https://doi.org/10.1080/00036840701222553.
Arcand, J., Berkes, E., & Panizza, U. (2012). Too much finance? Working paper WP/12/161, International Monetary Fund. Retrieved from: https://www.imf.org/external/pubs/ft/wp/2012/wp12161.pdf.
Bacchetta, P., & Caminal, R. (2000). Do capital market imperfections exacerbate output fluctuations? European Economic Review, 44(3), 449-468. Available at: https://doi.org/10.1016/s0014-2921(98)00083-x.
Barrodale, I., & Roberts, F. (1974). Solution of an overdetermined system of equations in the l 1 norm [F4]. Communications of the ACM, 17(6), 319-320. Available at: https://doi.org/10.1145/355616.361024.
Beck, T., Levine, R., & Loayza, N. (2000). Financial intermediation and growth: Causality and causes. Journal of Financial Economics, 58(1-2), 261–300.
Beck, T., & Levine, R. (2004). Stock markets, banks, and growth: Panel evidence. Journal of Banking & Finance, 28(3), 423-442. Available at: https://doi.org/10.1016/s0378-4266(02)00408-9.
Beck, T., Lundberg, M., & Majnoni, G. (2006). Financial intermediary development and growth volatility: Do intermediaries dampen or magnify shocks? Journal of International Money and Finance, 25(7), 1146-1167. Available at: https://doi.org/10.1016/j.jimonfin.2006.08.004.
Beck, T., Demirguc-Kunt, A., Laeven, L., & Levine, R. (2008). Finance, firm size, and growth. Journal of Money, Credit and Banking, 40(7), 1379-1405.
Bernanke, B., & Gertler, M. (1989). Agency costs, net worth, and business fluctuations. The American Economic Review, 79(1), 14-31.
Binder, M., & Coad, A. (2011). From average joe’s happiness to miserable jane and cheerful john: using quantile regressions to analyze the full subjective well-being distribution. Journal of Economic Behavior & Organization, 79(3), 275-290. Available at: https://doi.org/10.1016/j.jebo.2011.02.005.
Binder, M., & Coad, A. (2015). Heterogeneity in the relationship between unemployment and subjective wellbeing: A quantile approach. Economica, 82(328), 865-891. Available at: https://doi.org/10.1111/ecca.12150.
Canay, I. A. (2011). A simple approach to quantile regression for panel data. The Econometrics Journal, 14(3), 368-386. Available at: https://doi.org/10.1111/j.1368-423x.2011.00349.x.
Coban, S., & Topcu, M. (2013). The nexus between financial development and energy consumption in the EU: A dynamic panel data analysis. Energy Economics, 39, 81-88. Available at: https://doi.org/10.1016/j.eneco.2013.04.001.
Da Silva, G. F. (2002). The impact of financial system development on business cycles volatility: Cross-country evidence. Journal of Macroeconomics, 24(2), 233-253. Available at: https://doi.org/10.1016/s0164-0704(02)00021-6.
Dabla-Norris, N., & Srivisal. (2013). Revisiting the link between finance and macroeconomic volatility. IMF Working Paper No. 13/29.
Denizer, C., Iyigun, M., & Owen, A. (2000). Finance and macroeconomic volatility. Board of Governors of the Federal Reserve System. International Finance Discussion Papers No.670.
Dickey, D. A., & Fuller, W. A. (1979). Distribution of the estimators for autoregressive time series with a unit root. Journal of the American Statistical Association, 74(366a), 427-431. Available at: https://doi.org/10.2307/2286348.
Doornik, J. A., & Hansen, H. (2008). An omnibus test for univariate and multivariate normality. Oxford Bulletin of Economics and statistics, 70, 927-939. Available at: https://doi.org/10.1111/j.1468-0084.2008.00537.x.
Easterly, W., Islam, R., & Stiglitz, J. E. (2000). Shaken and stirred: Explaining growth volatility. In B. Pleskovic And J. E. Stiglitz (Eds). Paper presented at the Annual World Bank Conference On Development Economics 2000. Washington Dc: World Bank.
Easterly, W., Islam, R., & Stiglitz, J. (2002). Shaken and stirred: Explaining growth volatility, in: B. Pleskovic, N. Stern (Eds.). Paper presented at the Annual World Bank Conference on Development Economics, World Bank and Oxford University Press, Washington, D.C.
Galvao Jr, A. F. (2011). Quantile regression for dynamic panel data with fixed effects. Journal of Econometrics, 164(1), 142-157. Available at: https://doi.org/10.1016/j.jeconom.2011.02.016.
Goldsmith, R. W. (1969). Financial structure and development. New Haven: Yale University Press.
Gries, T., Kraft, M., & Meierrieks, D. (2009). Linkages between financial deepening, trade openness, and economic development: Causality evidence from Sub-Saharan Africa. World Development, 37(12), 1849-1860. Available at: https://doi.org/10.1016/j.worlddev.2009.05.008.
Huang, Y. (2010). Determinants of financial development. USA: Palgrave Macmillan.
Huang, H.-C., Fang, W., & Miller, S. M. (2014). Does financial development volatility affect industrial growth volatility? International Review of Economics & Finance, 29(1), 307-320. Available at: https://doi.org/10.1016/j.iref.2013.06.006.
Ibrahim, M., & Alagidede, P. (2017). Financial sector development, economic volatility and shocks in sub-Saharan Africa. Physica A: Statistical Mechanics and its Applications, 484, 66-81. Available at: https://doi.org/10.1016/j.physa.2017.04.142.
Im, K. S., Pesaran, M. H., & Shin, Y. (2003). Testing for unit roots in heterogeneous panels. Journal of Econometrics, 115(1), 53-74. Available at: https://doi.org/10.1016/s0304-4076(03)00092-7.
Johnston, J., & Dinardo, J. (1997). Econometrics methods (4th ed.). New York: Mcgrawhill.
King, R., & Levine, R. (1993a). Finance and growth: Schumpeter might be right. The Quarterly Journal Of Economics, 108(3), 717-737. Available at: https://doi.org/10.2307/2118406.
King, R. G., & Levine, R. (1993b). Finance, entrepreneurship and growth. Journal of Monetary Economics, 32(3), 513-542.
Kiyotaki, N., & Moore, J. (1997). Credit cycles. Journal of Political Economy, 105(2), 211-248.
Koenker, R., & Bassett, G. (1982). Robust tests for heteroscedasticity based on regression quantiles. Econometrica, 50(1), 43-61. Available at: https://doi.org/10.2307/1912528.
Koenker, R., & Bassett Jr, G. (1978). Regression quantiles. Econometrica: Journal of the Econometric Society, 46(1), 33-50. Available at: https://doi.org/10.2307/1913643.
Koenker, R., & Machado, J. A. (1999). Goodness of fit and related inference processes for quantile regression. Journal of the American Statistical Association, 94(448), 1296-1310. Available at: https://doi.org/10.1080/01621459.1999.10473882.
Koenker, R. (2004). Quantile regression for longitudinal data. Journal of Multivariate Analysis, 91(1), 74-89.
Lamarche, C. (2010). Robust penalized quantile regression estimation for panel data. Journal of Econometrics, 157(2), 396-408. Available at: https://doi.org/10.1016/j.jeconom.2010.03.042.
Lancaster, T. (2000). The incidental parameter problem since 1948. Journal of Econometrics, 95(2), 391-413. Available at: https://doi.org/10.1016/s0304-4076(99)00044-5.
Larrain, B. (2006). Do banks affect the level and composition of industrial volatility? The Journal of Finance, 61(4), 1897-1925.
Levine, R. (1997). Financial development and economic growth: Views and agenda. Journal of Economic Literature, 35(2), 688-726.
Levine, R., & Zervos, S. (1998). Stock markets, banks, and economic growth. American Economic Review, 88(3), 537-558.
Levine, R., Loayza, N., & Beck, T. (2000). Financial intermediation and growth: Causality and causes. Journal of monetary Economics, 46(1), 31-77. Available at: https://doi.org/10.1016/s0304-3932(00)00017-9.
Lv, Z. (2017). The effect of democracy on CO2 emissions in emerging countries: Does the level of income matter? Renewable and Sustainable Energy Reviews, 72, 900-906. Available at: https://doi.org/10.1016/j.rser.2017.01.096.
Ma, Y., & Song, K. (2018). Financial development and macroeconomic volatility. Bulletin of Economic Research, 70(3), 205-225. Available at: https://doi.org/10.1111/boer.12123.
Maddala, G. S., & Wu, S. (1999). A comparative study of unit root tests with panel data and a new simple test. Oxford Bulletin of Economics and statistics, 61(S1), 631-652. Available at: https://doi.org/10.1111/1468-0084.61.s1.13.
Mallick, D. (2014). Financial development, shocks, and growth volatility. Macroeconomic Dynamics, 18(3), 651-688.
Manganelli, S., & Popov, A. (2015). Financial development, sectoral reallocation, and volatility: International evidence. Journal of International Economics, 96(2), 323-337. Available at: https://doi.org/10.1016/j.jinteco.2015.03.008.
McCaig, B., & Stengos, T. (2005). Financial intermediation and growth: Some robustness results. Economics Letters, 88(3), 306-312. Available at: https://doi.org/10.1016/j.econlet.2004.12.031.
Mishkin, F. S. (2009). Globalization, macroeconomic performance, and monetary policy. Journal of Money, Credit and Banking, 41, 187-196. Available at: https://doi.org/10.1111/j.1538-4616.2008.00204.x.
Morgan, D. P., Rime, B., & Strahan, P. E. (2004). Bank integration and state business cycles. The Quarterly Journal of Economics, 119(4), 1555-1584.
Neyman, J., & Scott, E. L. (1948). Consistent estimates based on partially consistent observations. Econometrica: Journal of the Econometric Society, 16(1), 1-32. Available at: https://doi.org/10.2307/1914288.
Phillips, P., & Perron, P. (1988). Testing for unit root in time series regression. Biometrika 75, 335–346. Available at: https://doi.org/10.1093/biomet/75.2.335.
Raddatz, C. (2006). Liquidity needs and vulnerability to financial underdevelopment. Journal of Financial Economics, 80(3), 677-722. Available at: https://doi.org/10.1016/j.jfineco.2005.03.012.
Rajan, R., & Zingales, L. (1998). Financial development and growth. American Economic Review, 88(3), 559-586.
Rosendo, S. F., Simões, M., & Andrade, J. S. (2018). Health investments and economic growth: A quantile regression approach. International Journal of Development Issues, 17(2), 220–245. Available at: https://doi.org/10.1108/ijdi-12-2017-0200.
Saci, K., & Holden, K. (2008). Evidence on growth and financial development using principal components. Applied Financial Economics, 18(19), 1549-1560. Available at: https://doi.org/10.1080/09603100701720286.
Shahbaz, M., Shahzad, S. J. H., Ahmad, N., & Alam, S. (2016). Financial development and environmental quality: The way forward. Energy Policy, 98(11), 353-364.
Shenton, L., & Bowman, K. (1977). A bivariate model for the distribution of√ b1 and b2. Journal of the American Statistical Association, 72(357), 206-211. Available at: https://doi.org/10.1080/01621459.1977.10479940.
Tang, C. F., & Tan, B. W. (2014). The linkages among energy consumption, economic growth, relative price, foreign direct investment, and financial development in Malaysia. Quality & Quantity, 48(2), 781-797.
Thumrongvit, P., Kim, Y., & Pyun, C. S. (2013). Linking the missing market: The effect of bond markets on economic growth. International Review of Economics & Finance, 27, 529-541. Available at: https://doi.org/10.1016/j.iref.2013.01.008.
Topcu, M., & Payne, J. E. (2017). The financial development–energy consumption nexus revisited. Energy Sources, Part B: Economics, Planning, and Policy, 12(9), 822-830. Available at: https://doi.org/10.1080/15567249.2017.1300959.
Wang, N., Zhu, H., Guo, Y., & Peng, C. (2018). The heterogeneous effect of democracy, political globalization, and urbanization on PM2. 5 concentrations in G20 countries: Evidence from panel quantile regression. Journal of Cleaner Production, 194, 54-68. Available at: https://doi.org/10.1016/j.jclepro.2018.05.092.
Yellen, J. L., & Akerlof, G. A. (2006). Stabilization policy: A reconsideration. Economic Inquiry, 44(1), 1-22. Available at: https://doi.org/10.1093/ei/cbj002.
Zhu, H., Duan, L., Guo, Y., & Yu, K. (2016). The effects of FDI, economic growth and energy consumption on carbon emissions in ASEAN-5: Evidence from panel quantile regression. Economic Modelling, 58, 237-248. Available at: https://doi.org/10.1016/j.econmod.2016.05.003.
Zhu, H., Guo, Y., You, W., & Xu, Y. (2016). The heterogeneity dependence between crude oil price changes and industry stock market returns in China: Evidence from a quantile regression approach. Energy Economics, 55, 30-41. Available at: https://doi.org/10.1016/j.eneco.2015.12.027.
Zhu, H., Xia, H., Guo, Y., & Peng, C. (2018). The heterogeneous effects of urbanization and income inequality on CO 2 emissions in BRICS economies: evidence from panel quantile regression. Environmental Science and Pollution Research, 25(17), 17176-17193. Available at: https://doi.org/10.1007/s11356-018-1900-y.
Zouaoui, H., Mazioud, M., & Ellouz, N. Z. (2018). A semi-parametric panel data analysis on financial development-economic volatility nexus in developing countries. Economics Letters, 172, 50-55. Available at: https://doi.org/10.1016/j.econlet.2018.08.010.
Views and opinions expressed in this article are the views and opinions of the author(s), The Economics and Finance Letters shall not be responsible or answerable for any loss, damage or liability etc. caused in relation to/arising out of the use of the content. |