This study examines the impact of stock market development on energy consumption (electricity and oil) in a sample of G20 countries for the period from 2000 to 2023. Several different measures of the development of the stock market are examined, in particular, the market capitalization; stock traded total value (% of GDP), the turnover ratio and the number of listed companies. As an estimation method, we use the autoregressive distributed lag model (ARDL panel), more precisely the two estimators MG and PMG. The results provide dynamic relationships (short- and long-term) between stock market development and electricity consumption and only a long-term relationship for oil consumption. The results found are important for energy policies, the development of the stock market must be taken into account when estimating energy demand.
Research Article
A Multiple threshold panel ARDL Approach for studying the impact of stock market development on energy consumption
https://doi.org/10.21203/rs.3.rs-3241428/v1
This work is licensed under a CC BY 4.0 License
Version 1
posted
You are reading this latest preprint version
stock market development
energy consumption
ARDL panel
Energy has so far been considered not only as a production input but also as a strategic product that forms the basis of international relations and shapes the global economy and politics. The conditions under which energy is purchased and the problems encountered during the procurement process directly affect competition at the national and international levels; these conditions also shape countries' production structures and are one of the main indicators of basic economic variables. For all these reasons, energy is one of the most important issues in today's world (Esen and Bayrak, 2017).
Global energy consumption increased by 2.1% in 2017 compared to 2016, an increase of 253 million tons of oil equivalent. All sources of energy were consumed more in the world last year, including coal (1% due to the increase in demand in India for the first time since 2013). Fossil fuels would have accounted for 85.2% of the world's primary energy consumption in 2017. Oil remains by far the main source of energy in the global energy mix (34.2%) with an increase in consumption of 1.7 million barrels per day in 2017 (International Energy Agency 2018).
The study of the determinants of energy demand has been one of the main research themes in emerging and developing countries for several reasons. First, since energy is a key factor in production, its resources can affect a country's economic growth and development (Stern, 2010), Second, emerging and developing countries have experienced rapid growth. This increases energy demand in these countries. As indicated in the report presented by the bp statistical review of world energy (2018), China is the world's largest consumer of energy, followed by America. China and the United States alone absorbed about 40% of global energy production, followed by India, Russia, and Japan. The Kingdom's energy consumption represented about 2% of the world's total energy in 2017.
In previous studies, most empirical studies focus on the relationship between energy consumption and economic growth. As Chang (2015) points out, previous studies have cited mixed results on the links between energy use and income. This may be explained by the omission of some important variables, such as stock market indicators, as these variables have a significant influence on energy demand estimates Ulusoy and Demiralay (2017).
Although the impact of stock markets on energy consumption is evident, studies on this subject are relatively rare. The preceding studies mainly use "domestic credits to the private sector" as a general indicator of financial development. However, this indicator does not assess the effect of stock market activity and only measures the impact of the accessible credits provided on the level of energy consumption.
This document contributes to the literature review regarding the relationship between energy demand (oil and electricity) and stock market developments in G20 member countries. To do so, we study the impact of stock market development, measured as a function of the market capitalization of listed companies (% of GDP), turnover ratio and stock traded total value (% of GDP, on oil and electricity consumption, using a dynamic panel approach and annual data for 2000 to 2017.
To our knowledge, no empirical studies have examined the impact of stock market development on oil and electricity consumption in G20 member countries, while oil and electricity are important energy sources in the production process and stock market developments are one of the main determinants of energy consumption. Moreover, another important contribution is to estimate the relationship between stock market development and energy consumption in the short and long term, whereas most previous articles in the related field have estimated only short-term parameters.
In this paper, the G20 countries are chosen for a number of reasons: G20 countries represent 85% of the world economy, 75% of world trade and two-thirds of the world's population. The energy composition of G20 economies varies widely, but most countries rely on a significant share of fossil fuels in their total energy supply. At the same time, together, the G20 economies consume 80% of the world's energy among themselves and represent 81% of the world's renewable energy capacity, 82% of global energy-related CO2 emissions and 77% of global energy consumption International Energy Agency IEA (2018).
The rest of the paper is as follows: Section 2 presents a literature review; Section 3 explains methodology and data. Section 4 presents the results and discusses. Section6 concludes and implication.
The link between financial development and energy consumption has been the subject of considerable academic scrutiny in recent years. Financial development is generally cited as a very important driver of economic growth in emerging economies and it is therefore likely that financial development will affect energy demand Sadorsky (2010).
Financial development, referring to a country's decision to allow and promote activities such as increasing foreign direct investment, increased banking activity and increased stock market activity, is a possible path to economic growth, which impacts energy demand (Sadorsky (2010). It also suggests that financial development is important because it can increase the economic efficiency of a country's financial system. Financial development encourages a number of changes in a country, including a reduction in financial risk and borrowing costs, greater transparency between lenders and borrowers, access to financial capital and greater investment flows in advanced technology, all of which can influence energy demand by increasing business consumption and fixed investment.
According to Sadorsky. (2011a), financial development can influence energy consumption in several ways. Directly, one way in which financial development can influence energy demand is to facilitate and reduce the borrowing costs of consumers to buy large items such as cars, houses, refrigerators, air conditioners and washing machines. These large consumer items consume a lot of energy, which can affect a country's overall energy demand. Businesses also benefit from improved financial development because they facilitate and reduce access to financial capital that can be used to develop existing activities (purchase or construction of plants, recruitment of workers and purchase of machinery and equipment) or to create new ones.
Stock market development, being an important element in financial development, is particularly attractive for companies because it allows them to access an additional source of financing, equity financing that can be used to develop their business in addition to debt financing. The increased stock market activity can increase risk diversification for both consumers and businesses, which is an important element in generating wealth in an economy (sadorsky 2010). The stock market is often considered a leading economic indicator and the increase in stock market activity can be seen as a sign of economic growth and prosperity which, in turn, strengthens consumer and business confidence. Increased economic confidence increases demand for energy-intensive goods.
In this context, Zhang et al (2011) suggest that the stock market attracts more savings and that the share of savings transformed into investment increases, leading to economic growth and increasing energy consumption. The healthy stock market contributes to strengthening the economic strength and efficiency of listed companies, thus affecting energy consumption.
Many empirical studies examine the determinants of energy consumption. Some authors have studied energy consumption using data from several countries, such as Ulusoy and dimaralay (2017); Mahmood and Saleem (2016); Chang (2015); Sadorsky (2010); Çoban and Topcu (2013); Topcu and Payne (2017). Others, such as Zhang et al (2011); Xu (2012); Shahbaz et al, (2013); Komal and Abbas (2015); Sbia et al (2014); Sbia et al (2017); Shahbaz and Lean (2012); Islam et al, (2013); Kaker et al (2011); Shahbaz (2015); Destek (2015); Rafindadi and Ozturk (2016); Doaa and Eyad (2016), have focused on specific countries. These authors find different or even contradictory results. This is due in particular to the different data they use, for different periods and territories.
Ulusoy and dimaralay (2017) examine the impact of stock market developments on OECD member countries' oil and electricity demand over the period 1996 to 2011. Overall results confirm that income, real prices, stock market size and, liquidity are important determinants of oil and electricity demand.
Mahmood and Saleem (2016) examined the relationship between energy consumption and financial development, utilizing secondary data for 22 emerging economies over the period 1999–2012. The moderating role of oil prices was also highlighted. The results of the research reveal that oil prices negatively moderate the relationship between energy consumption and financial development.
Chang (2015), studies the effect of financial development and income on energy consumption. This study is based on a sample of 53 countries (high and low income) for the period 1999–2008, indicating a single threshold effect on energy consumption when private credit, domestic credit, the value of shares traded and stock market turnover are used as indicators of financial development. In high income countries, energy consumption increases with private and domestic credit.
Sadorsky (2010) studied the impact of financial development on energy consumption in 22 emerging economies using several financial development indicators over the period 1990–2006. The results of the GMM estimation indicate that financial development positively affects energy consumption when financial development is measured using stock market variables.
For the EU countries (European Union), Çoban and Topcu (2013) using the GMM (Generalized Moment Methodology) approach to examine the relationship between energy consumption and financial development over the period 1990–2011. The results show that for former members, higher levels of financial development lead to higher energy consumption, whether financial development comes from the banking sector or the stock market. While for new members, they conclude that the impact of financial development depends on how financial development is measured.
Topcu and Payne (2017) examine the impact of financial development on energy consumption for a panel of 32 high-income countries over the period 1990–2014. The results of this study indicate that there is no statistical relationship between the overall financial development index and energy consumption.
To study the influence of the development of Chinese stock markets on energy consumption, Zhang et al (2011) use the grey relational analysis and Granger causality test approaches. The empirical results indicate that first; the grey relational level appears to be relatively high for both the stock market scale and for efficiency and energy consumption between 1992 and 2009, at 0.84 and 0.73 respectively. Second, the expansion of the stock market in China is becoming an obvious driver of increased energy consumption, while the influence of stock market efficiency is negligible. In the same context, Xu (2012) studied the relationship between financial development and energy consumption in 29 provinces during the period 1999–2009 in China. The results of the System-GMM (general methods of the moment) indicates the presence of a significant positive relationship between financial development and energy consumption when financial development is measured by the ratio of loans from financial institutions to GDP (gross domestic product) as well as by the ratio of FDI (foreign direct investment) to GDP.
Also, Shahbaz et al, (2013) have studied the production function by integrating financial development and energy consumption in the case of China. They have applied the ARDL approach to cointegration and VECM Granger causality to examine the long-term and causal relationship between the series. Their results show that financial development has a positive impact on energy consumption. They also noted that financial development is a cause of Granger's energy consumption.
Using the GMM system estimation technique, the study by Komal and Abbas (2015) analyses the impact of financial development on energy consumption through the economic growth channel for Pakistan over the period 1972–2012 and includes energy prices and urbanization in the structural model. The study reveals a positive and significant impact of economic growth and urbanization on energy consumption, while the effect of energy prices on energy consumption is significant but negative. Financial development has a positive and significant impact on energy consumption through the economic growth channel.
Sbia et al (2014) examined the relationship between FDI, clean energy, trade openness, carbon emissions and economic growth in the case of the UAE over the period 1975Q1-2011Q4. They tested the unit properties of the variables in the presence of structural breaks. The ARDL approach was applied to investigate cointegration taking into account the structural breaks resulting from the series. Their empirical results confirm the existence of co-integration between the series. They note that FDI, trade openness and carbon emissions reduce energy demand. Economic growth and clean energy have a positive impact on energy consumption.
Sbia et al (2017) discuss the relationship between financial development and electricity consumption in the United Arab Emirates for the period 1975–2011. The results show that financial development increases electricity consumption, there is a two-way causality between financial development and electricity consumption.
Shahbaz and Lean (2012) investigated the impact of financial development on energy consumption by applying the energy demand function in the case of Tunisia from 1971 to 2008. They concluded that financial development increases energy demand by stimulating stock market development and stimulating real economic activity. The results show that financial development and Granger-cause each other's energy consumption. Shahbaz and Lean also suggest that a financially developed system is an appropriate way to reallocate financial resources to high-return investment projects. Consequently, investment stimulates economic growth, which in turn increases energy demand. In addition, a strong and developed financial system that can attract investors, stimulate the stock market and improve the efficiency of economic activities should be encouraged in the country.
Islam et al, (2013) using the Vector Error Correction Model (VECM) to test the causality between energy consumption, financial development, economic growth and population in Malaysia from 1971 to 2008. Have shown that financial development and economic growth have a positive impact on energy consumption. They have found bidirectional causality between financial development and long-term energy consumption. In the short term, Granger's financial development leads to energy consumption.
Kaker et al (2011) discuss the association between economic growth, financial development and, energy consumption in Pakistan over the period 1980–2009. From the results of co-integration and error correction, it can be seen that financial development affects energy consumption in the long term, while a statistically significant association does not exist in the short term. Shahbaz (2015) found a bi-directional causality between electricity consumption and financial development in Pakistan.
Destek (2015) has shown that financial development in Turkey has a negative impact on long-term energy consumption. Rafindadi and Ozturk (2016) show that, in the long term, a 1% increase in financial development in Japan increases electricity consumption to 0.2429% by using ARDL.
Doaa and Eyad (2016) study the impact of financial development and economic growth on energy consumption in three North African countries during the period 1980 to 2010. The empirical results provide a significant positive link between financial development and energy consumption in Algeria and Tunisia. On the other hand, Egypt's results show a significant negative relationship between energy consumption and financial development.
3.1. Data description
This study is based on panel data covering G203 member countries over the period 2000–2017, to examine the dynamic relationship between stock market development and energy consumption. Appendix 1 provides all the detailed definitions and sources of the variables in the empirical analysis.
a) The dependent variable and the control variables
The dependent variable is the consumption of oil and electricity, measured per kg of oil equivalent per capita and kWh per capita, respectively. As far as the control variables are concerned, we initially considered a broad set of control variables typically used in the growth literature: foreign direct investment as measured by net inflows (% of GDP), economic growth, Income, and energy prices.
b) Measures of stock market development
We consider three stock market indicators: market capitalization of listed companies (% of GDP), turnover ratio, and stock traded total value (% of GDP). All the financial variables selected reflect the deepening of the stock markets. The market capitalization variable refers to the size of the company; investors often rank companies according to their size. The turnover ratio is essentially used to assess liquidity in order to measure the ease with which investors can buy and sell financial assets.
The stock traded total value (% of GDP) is the total number of shares traded, including domestic and foreign assets, and therefore reflects the size of the market. The number of listed companies is another measure of market size; the total number of listed companies in developed countries is generally higher than in emerging countries. All variables are transformed into natural logarithms.
3.2. Model specification
According to Johansen (1995), a long-term relationship can only exist between variables with the same order of integration. Pesaran and Shin (1999) show that the ARDL panel can be used even with variables with different integration orders, either I (0) or I (1) or a mixture of the two.
Mean Group (MG)
The average group (MG) introduced by Pesaran and Smith (1995) asked to estimate separate regressions for each country and to calculate the coefficients as unweighted averages of the estimated coefficients. This imposes no restrictions. This allows all coefficients to vary and be heterogeneous in the long term and in the short term. However, the necessary condition for the consistency and validity of this approach is that the data have a sufficiently large time series dimension. The transnational dimension should also be important.
Pooled Mean Group (PMG)
The second technique (PMG) proposed by Pesaran and Shin (1999), the main feature of PMG is that it allows short-term coefficients, including intercepts, the adjustment speed of long-term equilibrium values and error variances to be heterogeneous country by country, while long-term slope coefficients are homogeneous between them. This is particularly useful when there is the reason to believe that the long-term equilibrium relationship between the variables is similar across countries or, at least, across a subset of them. Short-term adjustment is allowed to be country-specific, due to the very different impact of vulnerability to financial crises and external shocks, stabilization policies, monetary policy, etc.
The model estimated has the form of an ARDL (p,q,q,…,q) model (1):
$${EC}_{it}=\sum _{j=1}^{p}{\delta }_{i}{EC}_{i,t-j}+\sum _{j=0}^{q}{\beta {\prime }}_{ij}{X}_{i,t-j}+{\phi }_{i}+{e}_{it}$$
1
Where X is the vector of the explanatory variables. the model turns into:
Where
\(\beta\) are our vectors of interest, which measures the long-run impact of the explanatory 𝛽𝑖 variables on the share of energy consumption and\({\theta }\) is the error corrector mechanism impact. The remaining 𝜑𝑖 parameters are the short run coefficients. The disturbances \({\text{e}}_{\text{i}\text{t}}\) are independently distributed across 𝜀 𝑖𝑡 time and units, with zero mean and variance constant within each unit.
The model of Eq. 1 allows for the parameters to vary between units and as Pesaran and Smith (1995) and Pesaran et al. (1999) showed, this can be consistently estimated using the Mean Group (MG) estimator, which estimates the parameters for each country and then average for the group. However, they also showed that if the long-run coefficients are homogeneous across groups we can use a more efficient estimator, the Pooled Mean Group (PMG) which allows the short run parameters to vary from country to country but forces the long run ones to be homogenous. Finally, to apply these methods the variables have to be a mixture of I(1) and I(0) and for the model to be read as an error correction mechanism, the variables have to be cointegrated.
In unit root panel tests, one of the important issues is whether or not the cross-sectional units that make up the panel are treated independently. Two types of generation, the first generation tests, are used on the assumption that the cross-sectional units are independent of each other. Also second generation tests that take into account the hypothesis of dependence between sections. For this reason, it is necessary to apply cross-sectional dependence tests before the unit root test. In this study, the independence of the cross-section is tested using Pesaran's (2004) test.
Cross sectional dependence: Pesaran Tests 2004 (CD)
The Peasran (CD) 2004 (cross sectional dependency) test has been developed to check whether variables or residues are correlated between groups in a panel. The test is based on the following statistics:
CD=\(\sqrt{\frac{2T}{N(N-1)}}(\sum _{i=1}^{N-1}\sum _{j=i+1}^{N}{\rho }_{\widehat{ij}})\)
Where N is the number of individuals, T the period and \({\rho }_{\widehat{ij}}\)is the pairwise correlation coefficient of the residues. The null hypothesis for this test assumes that there is no dependence between individuals.
|
Variables |
CD-test |
P-value |
Decision |
|---|---|---|---|
|
Ln_elec |
6.882 |
0.000*** |
Dependence |
|
Ln_ener |
1.668 |
0.095* |
Dependence |
|
Ln_mc |
20.436 |
0.000*** |
Dependence |
|
Ln_st |
17.536 |
0.000*** |
Dependence |
|
Ln_tr |
14.981 |
0.000*** |
Dependence |
|
ln_rev |
41.219 |
0.000*** |
Dependence |
|
ln_gdp |
48.093 |
0.000*** |
Dependence |
|
Ln_enerp |
46.967 |
0.000*** |
Dependence |
|
Fdi |
5.379 |
0.000*** |
Dependence |
NB: * and *** refers to significance at the 10% and 1% threshold respectively.
From this table, it is clear that the hypothesis of no cross-sectional dependence is rejected at the 1% significance level for all the variables considered, which implies the existence of cross-sectional dependence in the variables across the countries considered. This implies that a shock on one G20 capital market can affect the other G20 capital markets.
Therefore, from the evidence provided by the Pesaran test (2004), we find that the first generation unit root tests are not applicable. So, we apply the second generation Pesaran unit root test (2003), based on the assumption of a cross-sectional dependence, to identify the order of integration of the variables.
Unit root test: Pesaran (2003)
Pesaran (2003) proposed a test to take into account dependencies between individuals. The test is based on the average of the individual t statistics DF (or ADF) of each unit in the panel. He considered an increased DF model in the interindividual dimension or CADF model. This model is written in the absence of autocorrelation of terms:
|
Level |
First difference |
||||
|---|---|---|---|---|---|
|
Variable |
Constant and no trend |
Avec constant et avec trend |
Avec constant et sans trend |
Avec constant et avec trend |
Décision |
|
Ln_elec |
2.864 (0.998) |
0.623 (0.733) |
-5.013 (0.000)*** |
-2.079 (0.006)*** |
I(1) |
|
Ln_ener |
1.308 (0.904) |
0.610 (0.729) |
-4.014 (0.000)*** |
-2.409 (0.002)*** |
I(1) |
|
Ln_mc |
-0.437 (0.331) |
0.248 (0.598) |
-3.681 (0.000)*** |
-1.423 (0.077)* |
I(1) |
|
Ln_st |
-0.067 (0.437) |
-0.165 (0.434) |
-4.289 (0.000)*** |
-2.551 (0.005)*** |
I(1) |
|
Ln_tr |
-0.314 (0.377) |
-2.580 (0.005)*** |
-7.069 (0.000)*** |
-5.638 (0.000)*** |
I(1) |
|
ln_rev |
-0.568 (0.285) |
-0.451 (0.326) |
-3.251 (0.001)*** |
-1.585 (0.057)* |
I(1) |
|
Ln_gdp |
-2.740 (0.003)*** |
-2.233 (0.013)** |
-4.511 (0.000)*** |
-1.329 (0.092)* |
I(0) |
|
Ln_enerp |
-3.101 (0.001)*** |
-2.459 (0.007)*** |
-5.328 (0.000)*** |
-2.325 (0.010)*** |
I(0) |
|
Fdi |
-2.438 (0.007)*** |
-1.795 (0.036)** |
-7.786 (0.000)*** |
-6.520 (0.000)*** |
I(0) |
| Notes ***, ** and * show significance at the 1%, 5% and 10% levels, respectively | |||||
Table 2 presents the results of the unit root tests which indicate that some variables are stationary in level (ln_gdp, ln_enerp and fdi) and others in first difference (ln_ener, ln_mc, ln_st, ln_tr and ln_rev). Due to the existence of mixed levels of integration between series, we use the ARDL Panel approach rather than the static cointegration test. ARDL panel approach is characterized by a multiplicity of advantages that it highlights and allows to estimate different variables with a different order of stationarity, as in our case in Table 2. We also found that our data suffer from I(0) and I(1). In more, these estimators allow us to estimate short- and long-run relationships and error correction coefficients.
Results of PMG and MG Estimators
In order to identify the impact of the variable of interest, an error correction based on the ARDL model (p,q) with autoregressive distributed lag has been used, focusing on the exclusive property of the PMG model compared to other estimates based on the error correction such as the MG estimator.
|
Model 1: Electricity consumption Ln_elec |
Model 2: Oil consumption Ln_ener |
|||
|---|---|---|---|---|
|
PMG |
MG |
PMG |
MG |
|
|
Long-run coefficients (LR) |
||||
|
Ln_mc |
0.0831 (3.47)*** |
-0.0423 (-0.63) |
0.1672 (4.80)*** |
0.0962 (2.18)** |
|
Ln_rev |
0.8154 (6.81)*** |
0.9056 (1.86)* |
0.7597 (6.54)*** |
1.1661 (1.09) |
|
Ln_gdp |
0.1219 (4.03)*** |
-0.0484 (-0.63) |
-0.0636 (-1.47) |
-0.2504 (-1.02) |
|
Ln_enerp |
-0.1563 (-6.04)*** |
-0.0707 (-1.54) |
-0.0947 (-4.45)*** |
-0.0245 (-0.48) |
|
Fdi |
-0.0417 (-4.36)*** |
0.0077 (0.70) |
0.0095 (1.55) |
-0.0142 (-0.58) |
|
Short-run coefficients (SR) |
||||
|
Error-correction coefficient (ec) |
-0.1362 (-2.61)*** |
-0.9605 (-7.43)*** |
-0.1793 (-4.61)*** |
-0.8630 (-6.13)*** |
|
Δln_mc |
0.0021 (0.43) |
0.0022 (0.12) |
-0.0059 (-0.95) |
-0.0392 (-1.45) |
|
Δln_rev |
0.5383 (2.45)** |
0.1280 (0.34) |
0.1547 (065) |
-0.0591 (0.10) |
|
Δln_gdp |
-0.0696 (-1.20) |
-0.1076 (-0.73) |
0.0030 (0.04) |
-0.2634 (-1.20) |
|
Δln_enerp |
0.0354 (2.10)** |
0.0569 (2.05)** |
-0.0000 (-0.00) |
0.0506 (2.07)** |
|
Δfdi |
0.0048 (2.05)** |
0.0045 (1.20) |
0.0007 (0.35) |
0.0034 (0.89) |
|
Constante |
-0.4520 (-2.56)*** |
3.6865 (0.90) |
0.2743 (4.54)*** |
-2.5311 (-0.43) |
|
Tests de Hausman |
(MG PMG) = 3.41 (0.6371) |
(MG PMG) = 0.44 (0.9941) |
||
| Notes: *, ** and *** indicate a significance at 10%, 5% and 1% respectively. The values in brackets represent statistical Zs. For model 1, the offset structure is ARDL (1, 1, 1, 1, 1, 1) and for model 2 is ARDL (1, 1, 1, 0, 1, 1). | ||||
|
Modèle n°3: Electricity consumption |
Modèle n°4: Oil consumption |
|||
|---|---|---|---|---|
|
PMG |
MG |
PMG |
MG |
|
|
Coefficients de long terme (LR) |
||||
|
Ln_st |
-0.0026 (-0.26) |
0.0249 (1.03) |
0.1401 (7.30)*** |
0.1295 (1.26) |
|
Ln_rev |
0.9253 (13.40)*** |
0.2604 (0.75) |
0.9043 (12.33)*** |
-2.3829 (-0.86) |
|
Ln_gdp |
0.0933 (4.03)*** |
0.0416 (0.59) |
-0.1019 (-4.66)*** |
0.5992 (0.97) |
|
Ln_enerp |
-0.0989 (-7.32)*** |
-0.0797 (-1.51) |
-0.0327 (-1.95)* |
-0.1156 (-1.80)* |
|
Fdi |
0.0056 (1.69)* |
-0.0008 (-0.09) |
0.0001 (0.02) |
0.0145 (0.39) |
|
Coefficients de court-terme (SR) |
||||
|
Coefficients de correction d’erreur (ec) |
-1.9184 (-2.29)*** |
-0.7355 (-4.09)*** |
-0.1527 (-3.01)*** |
-0.8031 (-7.73)*** |
|
Δln_st |
0.0069 (1.69)* |
0.0019 (0.17) |
-0.0010 (-0.14) |
.0033 (0.28) |
|
Δln_rev |
0.4948 (2.58)*** |
0.3579 (0.88) |
0.2291 (0.75) |
-0.0594 (-0.15) |
|
Δln_gdp |
-0.0873 (-1.13) |
-0.0576 (-0.61) |
-0.0283 (-0.21) |
-0.0046 (-0.04) |
|
Δln_enerp |
0.0348 (2.20)** |
0.0464 (1.97)** |
-0.0144 (-0.35) |
0.0305 (1.66)* |
|
Δfdi |
0.0006 (0.46) |
0.0078 (1.54) |
0 .0019 (0.93) |
0.0058 (1.22) |
|
Constante |
-0.6505 (-2.16)** |
2.8776 (0.78) |
0.2269 (3.10)*** |
3.8942 (1.57) |
|
Tests de Hausman |
(MG PMG) = 0.59 (0.9884) |
(MG PMG) = 5.22 ( 0.3896) |
||
| Notes: *, ** and *** indicate a significance at 10%, 5% and 1% respectively. The values in brackets represent statistical Zs. For model 1, the offset structure is ARDL (1, 0, 0, 1, 1, 1) and for model 2 is ARDL (1, 1, 0, 0, 1, 1). | ||||
|
Modèle n°5: Electricity consumption |
Modèle n°6: Oil consumption |
|||
|---|---|---|---|---|
|
PMG |
MG |
PMG |
MG |
|
|
Coefficients de long terme (LR) |
||||
|
Ln_tr |
-0.0067 (-0.59) |
-0.0889 (-0.93) |
-0.0208 (-1.02) |
0.0518 (1.16) |
|
Ln_rev |
1.0133 (16.11)*** |
0.3869 (1.04) |
0.4264 (3.85)*** |
0.6164 (1.31) |
|
Ln_gdp |
0.0579 (2.93)*** |
0.0634 (1.02) |
0.0641 (1.39) |
-0.1889 (-0.97) |
|
Ln_enerp |
-0.0828 (-7.37)*** |
-0.0791 (-1.44) |
-0.0926 (-5.64)*** |
-0.0378 (-0.89) |
|
Fdi |
0.0063 (1.75)* |
-0.0240 (-1.17) |
0.0363 (6.53)*** |
-0.0050 (-0.25) |
|
Coefficients de court-terme (SR) |
||||
|
Coefficients de correction d’erreur (ec) |
-0.2309 (-2.70)*** |
-1.1406 (4.95)*** |
-0.1929 (-3.16)*** |
-1.0601 (-4.87)*** |
|
Δln_tr |
-0.0024 (-0.35) |
-0.0241 (-1.24) |
0.0079 (0.69) |
− .00118 (-0.42) |
|
Δln_rev |
0.6290 (3.12)*** |
-0.1882 (-0.47) |
0.3249 (1.24) |
-0.1459 (-0.28) |
|
Δln_gdp |
-0.0901 (-1.23) |
0.0420 (0.41) |
-0.0166 (-0.18) |
-0.0086 (-0.08) |
|
Δln_enerp |
0.0279 (1.88)* |
0.0875 (2.98)*** |
-0.0056 (-0.17) |
0.0745 (2.33)** |
|
Δfdi |
0.0011 (0.86) |
0.0064 (0.79) |
-0.0016 (-0.62) |
0.0042 (0.53) |
|
Constante |
-0.7464 (-2.52)** |
5.7411 (1.96)** |
0.3474 (3.00)*** |
8.339 (1.56) |
|
Tests de Hausman |
(MG PMG) = 0.69 (0.9834) |
(MG PMG) = 0.69 (0.9835) |
||
| Notes: *, ** and *** indicate a significance at 10%, 5% and 1% respectively. The values in brackets represent statistical Zs. For model 1, the offset structure is ARDL (1, 0, 0, 1, 1, 0) and for model 2 is ARDL (1, 0, 1, 0, 1, 0). | ||||
Tables 3, 4 and 5 presents the estimate of PMG and MG along with the Haussman test to measure efficiency and consistency between them.
Before analyzing the variable coefficients, it is important to mention that the estimated coefficient for the error correction mechanism (ECT) is significant and negative in all cases. The tables (3, 4 and 5) show that the error correction coefficient is negative and significant at a level of 1% for models (1) to (6), which confirms the accuracy of the error. Indeed, there is a statistically significant long-term relationship between dependent and independent variables.
The results of estimation by PMG, as the most efficient estimator compared to the MG estimator, indicate that stock market development indicators have higher values of long-term coefficients than short-term parameters. The findings show that the increase in stock market size measured by market capitalization of listed companies (% of GDP)of 1% increases electricity and oil consumption by 0.083 and 0.167 respectively in the long term (Table 3). However, short-run results show that market capitalization has no effect on energy consumption (electricity and oil).
For indicators measuring stock market liquidity, the results show, in the long term, that the stock traded total value (% of GDP) increases positively and significantly in oil consumption (Table 4). The results show that a 1% increase in the value of the stock traded total value (% of GDP) increases oil consumption by 0.1405. However, for electricity consumption, the stock traded total value has no significant effect.
On the short run, the stock traded total value increases electricity consumption with a slight coefficient. An increase of this indicator by 1% increases electricity consumption by 0.0069.
Table 5 presents the results of the turnover ratio as another indicator that measures stock market liquidity. These results show that the turnover ratio has no significant effect on electricity and oil consumption in either the long or short run. These results are similar to the results of Ulusoy and Demiralay (2017) who found that the turnover ratio does not affect energy consumption in OECD member countries. Contrary to the results of Chang (2015) which find that the turnover ratio has a negative and significant impact on energy consumption in advanced economies while it has a positive and significant impact in emerging and developing countries.
Concerning the control variables, our result show that economic growth (ln_gdp) significantly affects electricity consumption while, overall, it does not significantly affect oil consumption in the long term. However, in the short term, economic growth has no effect on energy consumption (oil and electricity).
The income has a positive and significant statistical effect on electricity consumption in the long and short run. On the other hand, income has a positive and significant impact on oil consumption only in the long run. The results show that electricity demand in G20 countries is much more affected by changes in per capita income than oil demand. As found Ulusoy and Demiralay (2017) for OECD member countries.
Concerning energy price, we have found that it negatively affects electricity consumption in the long term; this is in line with economic theory which suggests that when the price increases, demand decreases. In the short term, the price of energy has a positive impact on electricity consumption. Oil consumption has not been affected by energy prices in either the long or short term.
Many studies have shown significant impacts of per capita income and real price on energy demand estimates (see, for example, Belke et al (2011), Bernstein and Madlener (2015).
Destek (2018), observed that real income is positively correlated with energy consumption and that energy prices have a negative impact on energy consumption.
The results indicate that overall foreign direct investment inflows do not significantly affect energy consumption. Lee (2013) shows that there is no convincing evidence of a link between FDI and the use of clean energy.
Sadorsky (2010) suggests two reasons why stock markets affect energy demand. The first effect, short-run activity, is the wealth (level) effect. As a leading indicator of future economic prospects, the increase in stock market activity affects consumer and business confidence, which in turn increases energy demand. The successful development of the stock market is accompanied by an increase in financial regulations such as accounting and reporting standards. These increases in financial regulation strengthen investor confidence. This increased investor confidence is particularly important to attract foreign investors.
The second theoretical effect is the efficiency effect, the investment canal that helps companies access the source of financing with equity financing. The development of stock markets increases diversification and liquidity, which increases the amount of investment in higher-yield, higher-risk projects. As the development of stock markets increases the amount of funds available for investment projects, it is expected that the development of stock markets will lead to more investment, economic growth and energy demand.
Our empirical results confirm the existence of these two effects. The size of the stock market (market capitalization) has a positive and significant long-term impact on electricity consumption at the 1% level in G20 countries. For electricity consumption, long-term results indicate that the size of the stock market measured by market capitalization and liquidity (total value of traded shares) have a positive and statistically significant impact on oil demand at the 1% threshold. In the short term, the liquidity of the stock markets increases electricity consumption.
The positive effect of the development of stock markets on energy consumption is consistent with some previous work. For example, Hasnaoui (2014) finds a significant effect of stock market developments on OECD countries' energy consumption by conducting a Principal Component Analysis (PCA).
In the case of EU countries, the study by Çoban and Topçu (2013) shows that stock market developments have no significant effect on energy consumption in general, but that they have a significant influence on energy consumption in the old member countries.
The empirical results presented show that the increased development of stock markets increases energy demand in the economies of G20 member countries. Since the development of stock markets in these economies is expected to increase only in the future, this additional increase in energy demand linked to increased financial development should be taken into account when modeling energy demand in G20 countries.
Recently, several studies have examined the effects of financial development measured by banking variables on energy consumption, but few studies have focused on the relationship between stock market development, as an essential element of financial development, and energy consumption. This paper fills the gaps in the literature by applying ARDL panel techniques to dynamic panels to examine the impact of stock market development on energy consumption.
The empirical analysis was carried out on the basis of a sample of 19 G20 member countries over the period 2000–2017. We conducted the Pesaran 2004 cross-sectional dependency test to examine the dependency of G20 countries. We find that there is a cross-sectional dependence for all variables. To do this, we apply the second generation test, which takes into account the transversal dependence, to test the stationary of the variables.
From these tests we find that the integration order is different such that I(0) and I(1), the mixed integration order prompted us to apply the ARDL panel model and we particularly apply the two estimators PMG and MG to examine the long and short term relationship.
Referring to our estimates, the size of the stock market measured by market capitalization as a % of GDP has a positive and significant impact on electricity consumption, while liquidity, measured by the two indicators (The stock traded total value (% of GDP) as a % of GDP and the turnover ratio as a %), has no significant effect on electricity consumption in the long term. Regarding oil consumption, the long-term results show that market capitalization, as a measure of the size of the stock market, significantly positively affects oil consumption. On the other hand, market liquidity, measured by the value of The stock traded total value, has a positive and significant impact on oil consumption. On the other hand, the short-term results indicate that the size of the stock market has no significant effect on electricity consumption, whereas liquidity (the value of the shares traded) have a positive and significant impact on electricity consumption. As far as oil consumption is concerned, the size and liquidity of the stock market has no significant effect.
The results obtained validate the hypothesis that the development of the stock market has a positive and significant impact on energy consumption.
Energy saving policies based only on energy demand and income relationships would tend to underestimate energy demand when the development of the stock market is excluded. As a result, energy conservation policies that exclude the effects of stock market development on energy demand are probably not sufficient to achieve the desired objectives.
- Belke A, Dobnik F, Dreger C. (2011). Energy consumption and economic growth: new insights into the cointegration relationship. Energy Econ 2011; 33(5):782–9.
- Bernstein R, Madlener R. (2015). Short-and long-run electricity demand elasticities at the subsectoral level: a cointegration analysis for German manufacturing industries. Energy Econ 2015.
- Chang S. C. 2015. Effects of financial developments and income on energy consumption. International Review of Economics & Finance, 35, 28-44.
- Çoban .S, Topcu.M (2013), “The nexus between financial development and energy consumption in the EU: A dynamic panel data analysis” Energy Economics, Volume 39, September 2013, Pages 81-88.
- Destek .M.A,(2018), Financial development and energy consumption nexus in emerging economies, Energy Sources, Part B: Economics, Planning, and Policy, Volume 13, 2018 - Issue 1.
- Destek M.A., 2015. Energy consumption, economic growth, financial development and trade openness in Turkey: Maki cointegration test. Bulletin of Energy Economics 3(4), 162-168
- Doaa M. Salmanai , Eyad M. Atyab , (2016). What is the role of Financial Development and Energy Consumption on Economic Growth? New Evidence from North African Countries, International Journal of Finance & Banking Studies, Vol.3 No.1, 2014 ISSN: 2147-4486.
- Esen .Ö, Bayrak.M, (2017) "Does more energy consumption support economic growth in net energy-importing countries? » Journal of Economics, Finance and Administrative Science, Vol. 22 Issue: 42, pp.75-98, https://doi.org/10.1108/JEFAS-01-2017-0015.
- Hasnaoui H. (2014). The nexus between financial development and energy consumption in the high-income OECD members: a dynamic panel data analysis. Prof Cent Bus Res 2014; 12.
- Islam F., Shahbaz, M., & Alam, M. (2013). Financial development and energy consumption nexus in Malaysia: A multivariate time series analysis. Economic Modelling, 30, 435–441.
- Johansen, S. (1995). Likelihood-based inference in cointegrated vector autoregressive models. Oxford, New York: Oxford University Press.
- Kaker K.K., Khilji, B.A., Khan, M.J., 2011. Financial development and energy consumption: empirical evidence from Pakistan. International Journal of Trade, Economics and Finance 2, 469-471.
- Komal R., & Abbas, F. (2015). Linking financial development, economic growth and energy consumption in Pakistan. Renewable and Sustainable Energy Reviews, 44, 211-220.
- Lee JW. (2013).The contribution of foreign direct investment to clean energy use, carbon emissions and economic growth. Energy Policy 2013; 55:483–9.
- Levine R., Zervos, S. (1998), Capital Control Liberalization and Stock Market Development. World Development, 26, 1169-1184.
- Mahmood .F and Saleem.M, (2016), “An Empirical Analysis of the Impact of Energy Consumption on the Financial Development of the Emerging Economies: A Moderating Role of Oil Prices”, International Journal of Economics and Finance; Vol. 8, No. 9; 2016.
- Pesaran H.M. (2003). ìA Simple Panel Unit Root Test in the Presence of Cross Section Dependenceî, Mimeo, University of Southern California.
- Pesaran M. H. (2004). General Diagnostic Tests for Cross Section Dependence in Panels. https://doi.org/10.17863/CAM.5113.
- Pesaran M. Hashem, Shin.Y et Ron P. (1999). Smith Pooled Mean Group Estimation of Dynamic Heterogeneous Panels, Journal of the American Statistical Association, Volume 94, 1999 - Issue 446.
- Pesaran M.H. and Shin, Y. (1999) An Autoregressive Distributed Lag Modelling Approach to Cointegration Analysis. In: Strom, S., Ed., Chapter 11 in Econometrics and Economic Theory in the 20th Century the Ragnar Frisch Centennial Symposium, Cambridge University Press, Cambridge, 371-413.
- Pesaran, H., & Smith, R. (1995). Estimating long-run relationships from dynamic heterogeneous panels. Journal of Econometrics, 68(1), 79–113.
- Rafindadia.A.A, Ozturkb.I (2016), Effects of financial development, economic growth and trade on electricity consumption: Evidence from post-Fukushima Japan, Renewable and Sustainable Energy Reviews, Volume 54, February 2016, Pages 1073-1084.
- Sadorsky .P, (2010). “The impact of financial development on energy consumption in emerging economies, Energy Policy, Volume 38, Issue 5, May 2010, Pages 2528-2535.
- Sadorsky .P, (2011a). “Financial development and energy consumption in Central and Eastern European frontier economies”, Energy Policy, Volume 39, Issue 2, February 2011, Pages 999-1006.
- Sbia .R, Shahbaz .M et Ozturk.I, (2017). Economic growth, financial development, urbanisation and electricity consumption nexus in UAE, Economic Research, 2017 VOL. 30. 1, 527–549.
- Sbia R., Shahbaz, M., & Hamdi, H. (2014). A contribution of foreign direct investment, clean energy, trade openness, carbon emissions and economic growth to energy demand in UAE. Economic Modelling, 36, 191–197.
- Shahbaz M., & Lean, H. H. (2012). Does financial development increase energy consumption? The role of industrialization and urbanization in Tunisia. Energy Policy, 40, 373–479.
- Shahbaz M., (2015). Electricity consumption, financial development and economic growth nexus in Pakistan: a visit. Bulletin of Energy Economics 3(2), 48-65.
- Shahbaz M., Khan, S., Tahir, M.I., (2013). The dynamic links between energy consumption, economic growth, financial development and trade in China: Fresh evidence from multivariate framework analysis. Energy Econ. 40, 8-21.
- Stern, D.I. (2010), “Energy quality”, Ecological Economics, Vol. 69 No. 7, pp. 1471-1478, available at:http://dx.doi.org/10.1016/j.ecolecon.2010.02.005.
- Topcu .M and James E. Payne (2017), “The financial development–energy consumption nexus revisited”, Energy Sources, Part B: Economics, Planning, and Policy, Volume 12, 201.
- Ulusoy .V, and Demiralay.S, (2017). Energy demand and stock market development in OECD countries: A panel data analysis, Renewable and Sustainable Energy Reviews, Volume 71, May 2017, Pages 141-149.
- Xu S. (2012). The impact of financial development on energy consumption in China: Based on SYSGMM estimation. Advanced Materials Research, 527, 2977–2981.
- Zhang, Y.-j., Fan, J.-l., Chang, H.-r. 2011. Impact of China's stock market development on energy consumption: an empirical analysis. Ener. Proc., 5: 1927-1931.
No competing interests reported.