An overview of renewable energy companies in stock exchange: Evidence from minimal spanning tree approach

Renewable Energy 102 (2017) 107e117

Contents lists available at ScienceDirect

Renewable Energy journal homepage: www.elsevier.com/locate/renene

An overview of renewable energy companies in stock exchange: Evidence from minimal spanning tree approach Mansooreh Kazemilari a, Abbas Mardani b, Dalia Streimikiene c, *, Edmundas Kazimieras Zavadskas d a

Department of Mathematical Sciences, Faculty of Science, Universiti Teknologi Malaysia (UTM), Skudai Johor, 81300, Malaysia Faculty of Management, Universiti Teknologi Malaysia (UTM), Skudai Johor, 81300, Malaysia Lithuanian Energy Institute, Breslaujos3, LT-44403, Kaunas, Lithuania d Department of Construction Technology and Management, Vilnius Gediminas Technical University, Sauletekio al. 11, LT-10223, Vilnius, Lithuania b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 19 February 2016 Received in revised form 20 September 2016 Accepted 16 October 2016 Available online 20 October 2016

The renewable energy industry has the important role in overall growth in the worldwide economy in the last two decades. This paper constitutes a first analysis on renewable energy companies in stock exchange by used the minimum spanning trees (MSTs) approach. In this study, the daily closure prices data of 70 stocks of renewable energy companies during the time period from 13th October 2010 to 4th March 2015 are investigated. In stock market analysis, the interrelations among stocks are represented in terms of Pearson correlation coefficient (PCC). This is the standard practice to construct the stock network. With the precondition of time series are synchronous, the similarity among the stocks is quantified by the PCC between the logarithmic exchange rate returns of i and j stocks. In order to analyze the topological properties of MST, three major centrality measures are used, namely; degree, closeness and betweenness centralities. Result of this paper indicated that; First Solar Inc., General Cable Corporation and Trina Solar are more important within the network, moreover; we found that; these stocks play a significant role in renewable energy development in terms of market capitals. © 2016 Elsevier Ltd. All rights reserved.

Keywords: Renewable energy Stock exchange Minimum spanning trees (MSTs) Financial market

1. Introduction The worldwide financial system has several kinds of markets which are located in different industries such as renewable energy and in that a broad several financial goods and products are traded. Despite the variety of marketplaces, index activities frequently respond to the similar financial statements or marketplace news [1e3] which suggests that monetary time sequence may show related features and be associated. Renewable energy is important in terms of strengthening the local improvements. That is why there is a growing interest in renewable energy both in the developed and developing countries. Though, there are several problems facing [4,5] the activities to increase use of renewable energy, that should be understood and correctly interpreted into an inclusive controlling framework. The initial and notable problem for greater renewable penetration into energy organizations refers to the high-up front charges and associated inadequate cost

* Corresponding author. E-mail address: [email protected] (D. Streimikiene). http://dx.doi.org/10.1016/j.renene.2016.10.029 0960-1481/© 2016 Elsevier Ltd. All rights reserved.

usefulness. There are several market barriers and failures preventing penetration of renewables, such as commercialization barriers faced by new technologies competing with mature technologies; price distortions from existing subsidies and unequal tax burdens between renewables and other energy sources; failure of the market to value the public benefits of renewables and other, such market barriers such as inadequate information, lack of access to capital, “split incentives” between building owners and tenants, and high transaction costs for making small purchases etc. Consequently, it is significant to present financial support mechanism and satisfactory advancement arrangements, particularly ones which will interest isolated funding into energy segment and in such way decrease the monetary load on the public budget [6,7]. Numerous of previous studies have attempted to use the networks theories for analysis of exchange market in the international level. Mizuno, Takayasu and Takayasu [8], used minimum spanning tree (MST) and hierarchical taxonomy for analysis of foreign exchange market data. McDonald, Suleman, Williams, Howison and Johnson [9], applied MST for examine of the currency correlations among foreign exchange markets, finding of this analysis indicated that; dependent and dominant currency

108

M. Kazemilari et al. / Renewable Energy 102 (2017) 107e117

structures can be as some global foreign exchange dynamics in the MST network. Therefore; there are very useful correlation based networks among financial markets for investing and building the optimal strategy in renewable energy markets. Additionally, knowing about how evolving of stock markets is useful and important over the time. Consequently, in the past years explaining the patterns among stock markets networks have been popular study agenda. Development of stock market network theory allowed to introduce these patterns in an elucidative and simple way [10e14]. Using of statistical physics to financial exchange market data has been very interesting issue in last decade. Network analysis is assumed as the most important technique among other techniques, because, the introducing of financial exchange market as network topology represents optimal ways for understanding of the financial market structural properties [15e17]. While; some evaluations performed on various stock exchange markets in the social science, in several industries, due to the important of renewable energy stock market in the financial world, unfortunately; no previous study paid attention on this issue. First of all, few of previous scholars have conducted research on understanding of behaviour of network properties of renewable energy stocks of the market and interpretation of the network topology in financial terms of those stocks. Second, previous research focuses solely on econometric techniques, and does not try to explain their findings upon financial theory. Therefore, this paper attempted to use the network theory to the analysis of the renewable energy exchange market. Moreover; this study utilized MST to finding and filtering significant information related to those complex networks. As a real example, 72 renewable energy companies’ data from 13th October 2010 to 4th March 2015 were used. The rest of the paper is organized as follows; in Section 2, the researchers present literature review on stock market networks, renewable industry in stock exchange market, MST and graph theory. Section 3 describes the research methodology of this paper; in following section presents the results based on MST approach, market network correlation and graph theory. Finally, Section 5 provides concluding remarks on this paper. 2. Literature review 2.1. Stock market networks In the stock market, all corporations are related to each other and therefore their stock charges are correlated. This correlation, recognized as the possible of deep internal influence, forms the stock market network. Network theory was extended into an extensive
si, Albert and Jeong [22], prevariety of themes [18e21]. Baraba sented the scale-free network that is made by the development rule and the preferential attachment rule. In monetary marketplaces, the topology network analysis is an extensively applied practical instrument and delivers well-organized approaches for describing their market and structural possessions [23]. In the stock markets, Mantegna [10] first offered topology network analysis technical instrument of minimal spanning tree (MST) for analyzing the similarities among stock prices and found important results for portfolio optimization. After that, the network-based approach was applied extensively for exploring market belongings in monetary marketplaces, like equity marketplaces [16], the Korean stock marketplace [13], the Athens stock marketplace [24], European equity marketplaces [25], the NYSE stock marketplace [26], the Brazil stock marketplace [27], and commodity marketplaces [28]. Additionally, the topology network model was has been presented for assessing the FX market properties [28,29]. Ortega and Matesanz [30], evaluated some actual FX rates of 28 currencies during 1990e2002 for

building the FX and MST networks in that the worldwide FX markets are segmented into subdivisions including nations from the similar geographical areas. Then, a related assumption was presented by Ref. [8] investigated the network of FX market of 26 currencies based on three kinds of metals duration of 1999 and 2003 for classification of currencies based on correlation coefficient and MST network. Results of this study indicated that; USD has played as an important role in the FX markets. Subsequently, Naylor, Rose and Moyle [31] used USD and NZD as two currencies for examine of two different currency networks based on topology evolution of 44 currencies from 1995 to 2001. This study constructed two MST networks and demonstrated that the structure of network has changes. Also, Kwapien, Gworek and Drozdz [29] assessed a set of FX market rates in 46 currencies based on metals between 1998 and 2008, finding of this study found that; in the different FX markets currencies had the different MST networks. They examined the temporal and topology evolution of FX markets and determined that the USD node gradually loses its centrality, but the EUR node has developed slightly more dominant than before in all the network activities. More lately, Keskin, Deviren and Kocakaplan [32] investigated the correlation topology of networks in 34 currencies by applying the hierarchical tree and MST tools from 2007 to 2008. They applied USD and TRY as numeraires to create the networks and presented that the groups coordinated well with corresponding countries from the similar geographical areas of Asia and Europe. Jang, Lee and Chang [23], examined the time series properties of FX market since 1990 to 2008 about the history of the currency crises according to the MST technique. Their consequences showed that EUR as well as USD had a strong negative correlation afterward the Southeast. Numerous renewable corporations were listed on stock relations in the retro after 2000. The early 21st century was considered as a high dynamic period for the renewable energy manufacturing, as several administrations set long term renewable energy development aims. Some selected to directly fund the renewables by feed-in tariffs as well as related short-term procedures for bridging the gap for fulling cost accounting which would correctly reward these tools for their short releases and absence of interference with ecosystem amenities, and likewise for ensuring some volume and incentive for installing conservation-focused smart grid machineries. Monetary markets around the globe might be viewed as a compound scheme. This forces us for focusing on a global-level explanation for analyzing the communication assembly among marketplaces that might be obtained over signifying the scheme as a network. Association among stock markets has a significant character in asset theory and risk management, and likewise is important basics for the optimization problem in the Markowitz [33] portfolio theory. For the optimization of portfolio in the stock exchange market, there are some trading activities which show the similarity among stocks [14,34]. Compound networks deliver a high general framework, according to the notions of statistical physics, for examining the systems with large amounts of interrelating properties. These networks could positively define the topological possessions and features of several real-life schemes like multi-locus arrangement typing to analyze the clonality [35], scientific association in the European outline programmes [24], global hotel business in Spain [36], classification of correlations of wind speed [37], Brazilian term structure awareness rates [38] and legislative election consequences [39]. MST method is not merely applied inside only one stock market but likewise for networks made from interest rates [38], currencies [40], economic segments [41] and product charges [42]. Though, possibly the most promising method according to network modeling was applied to examine the associations among stock marketplaces as the consequences confirm a growth in the interdependence amongst markets over the past two decades [25,43,44].

M. Kazemilari et al. / Renewable Energy 102 (2017) 107e117

2.2. Renewable energy industry in stock exchange Energy creates one of the main responses for economic and societal improvement [45e47]. In line with the developing populace, growth, automation, dispersal of technology and growing of wealth, energy use has been increased. Energy use [48] at minimum quantities and prices is the main goal of sustainable energy development that support financial and societal improvement which have the lowest negative influence on the atmosphere at the lowest level. The renewable energy sector has accomplished substantial overall growth in the global economy during the last decade. Estimates by the International Energy Agency (IEA) suggest that renewable energy will be the fastest growing component of global energy demand with an annual growth rate of more than 7% within the next two decades [49]. Some of this development may be attributable to the conjunction of government policies, rising oil prices and evolving stock market liquidity for investments in renewable energy companies. Table 1 represented renewable energy global indicators for the 21st century. Several renewable, sustainable or clean energy stock indices have been created. There has also been an increased interest in examining returns of renewable energy companies, as well as in identifying potential drivers of these returns, see, e.g. [50e54]. These studies typically focus on the relationship between renewable energy stocks, changes in the oil price, other equity indices and carbon prices. The authors typically find evidence for the impact of several of these variables on renewable energy stock markets. In particular, returns of high technology and renewable energy stocks seem to be highly correlated. On the other hand, results are not that clear-cut for the influence of changes in the oil price. While Henriques and Sadorsky [50] suggested that changes in oil prices have only limited impact on returns from investment in renewable energy stocks, Kumar, Managi and Matsuda [51], Sadorsky [52] and Managi and Okimoto [54] find some evidence for a significant relationship between these variables. In recent years, knowledge of increased pollution and climate change has caused an enhanced focus on sustainable energy development, atmospheric emissions reduction and the usage of clean energy. These issues have been addressed through various agreements such as the Kyoto protocol, which is a treaty that sets binding obligations on industrialized countries in order to reduce the emission of greenhouse gasses [55]. Recently, the main sources of energy were fossil fuels, i.e. oil and coal, both of which are large contributors to the increase GHG emissions into atmosphere. In order to achieve sustainable energy development goals the dependence of fossil fuels has to decrease, consequently the importance of renewable energy sources is bound to increase. Investment bank Deutsche Bank is predicting that solar systems will be at grid parity in up to 80% of the global market within 2 years, and they mentioned that the collapse in the oil price will do little to slow down the solar juggernaut [56]. Consequently renewable energy is rarely adopted in the market without a subsidy, and the available supply is highly dependent on government funding and

109

tax deduction schemes [57]. Consequently the supply of renewable energy is dependent of economic development. During recessions, governments may scale back these subsidies making the alternative energy less attractive. On the other hand the subsidies may increase when economy is booming. Grøm [58] extends the research of Henriques and Sadorsky [50] by utilizing a dynamic multi-factor setting based on a state-space model with time-varying coefficients. In another article, Schmitz [59] applied a multi-factor market model for investigating the association between oil charges and substitute energy portions. The investigation is done by applying CAPMGARCH (General Autoregressive Conditional Heteroscedasticity) multi-factor market model to examine the association between oil charges and renewable energy guides. Outcomes display that a growth in oil charges and the broad market have a statistically important and positive influence on renewable energy stock revenues. Remarkably the oil charge beta for MAC is closely twice that of substitute energy representing that solar segment revenue is more sensitive to variations in oil charges than the comprehensive substitute energy marketplace. Based on the report this is most probable based on the vast amount of volatile small cap frameworks and solar corporations that include of the solar index applied by Schmitz. Huang, Cheng, Chen and Hu [60], applied a Vector Error Correction Model (VECM) to investigate the relationship between crude oil prices and stock performances of renewable energy companies (from 2001 to 2010). They have divided the sample period into three sub-periods with the two Middle East wars (Iraq and Lebanon) as natural divisions. The research indicates that the oil prices behaved differently during these sub-periods, but no significant relationship between oil prices and renewable energy stocks were detected in the first two periods. However in the last period, post 2006, when oil prices reach historical high and crash back with volatile dynamics, oil price behaviour has a significant effect on performances of alternative energy companies. As suggested by Ref. [61], in particular experienced investors consider supportive policy environments as an important way to encourage investment in clean energy technologies. The proposed model and set of variables are motivated by a number of previous studies investigating the effects of energy and stock market prices on the renewables sector [62]. Faff and Brailsford [63], inspected the association between oil charges and stock marketplace indices of numerous businesses in Australia. They discovered important effects of oil charges on equity revenues, particularly for the oil, gas, supply and construction manufacturing frameworks. Sadorsky [64], found the positive relationship in the growth oil price on Canadian gas and oil stocks. The results are confirmed by Boyer and Filion [62] who find evidence of a significant relationship among oil, stock returns and natural gas prices of Canadian gas and oil companies. Henriques and Sadorsky [50], considered the relationship between returns in equity investments and oil prices. The authors use a four variable vector-autoregressive model including returns on renewable energy stocks, technology stocks, crude oil price and interest rates. Conducting Granger causality tests, they find that movements

Table 1 Source: global status report of the renewable energy policy network (2014). Selected some global indicators of renewable energy 9

Ethanol production (annual) (10 L) Hydropower capacity (existing) (GWe) Renewables power capacity (existing) (GWe) Solar PV capacity (grid-connected) (GWe) Investment in new renewable capacity (annual) (109 USD) Solar hot water capacity (existing) (GWth) Biodiesel production (annual) (109 L) Countries with policy targets for renewable energy use Wind power capacity (existing) (GWe)

2014

2013

2012

2011

2010

2009

2008

94 1055 1712 177 270 406 29.7 164 370

87 1018 1578 138 232 373 26 144 319

83 990 1470 100 256 255 22.5 138 283

86 970 1360 70 279 232 21.4 118 238

86 945 1320 40 237 185 18.5 98 198

76 915 1230 23 178 160 17.8 89 159

67 885 1140 16 182 130 12 79 121

110

M. Kazemilari et al. / Renewable Energy 102 (2017) 107e117

in technology stock prices, oil prices and interest rates each have some impact on the movements of the stock price of renewable energy companies. However, their study also shows that while shocks to technology stock prices have a large impact on prices of renewable energy companies, shocks to the oil price seem to have only little and no significant impact on renewables. Kumar, Managi and Matsuda [51], examined the relationship between alternate energy prices, oil prices, technology stocks and interest rates, but extend the analysis by also including carbon prices. Similar to Sieczka and Hołyst [28], they also applied a vector-autoregressive model to study the relationship between the variables. Their results suggest that both the oil price and technology stock prices separately affect stock prices of clean energy firms. However, the authors do not find a significant relationship between carbon prices and renewable energy stocks. Sadorsky [52], applied multivariate GARCH and dynamic conditional correlation models to examine volatility spillover effects between oil prices, technology stocks and clean energy companies. The results of this study suggest that returns from clean energy companies correlate highly with returns from technology stocks and less so with changes in the oil price. The results also have a significant impact on hedging the risk of price movements of renewable energy companies: technology stocks cannot be considered a good hedge, while due to significantly lower correlations oil provides a more useful hedge for clean energy stocks. Managi and Okimoto [54], analyzed the relationship between oil prices, technology stock prices and clean energy stocks. They extend previous work by analyzing data up to 2010 and apply Markovswitching vector autoregressive models to detect possible structural changes in the examined relationship. In contrast to Henriques and Sadorsky [50], the authors find a relationship between clean energy prices and oil prices, also suggest that; there is a strong similarity between the market response of clean energy and technology stock prices to oil price shocks, which can be explained by both types of companies benefiting from government policies. 2.3. Graph theory In computer science and mathematics lexicon, graph theory refers to the examination of graphs. Pairwise associations amongst objects from a confident gathering are modeled through some mathematical assemblies. Most regularly, in modern texts on graph theory, except stated otherwise, a graph means an undirected simple finite graph. Consequently, a graph is typically defined G: ¼ (V, E). About this description, we can simply get that the vertex (V) and edge (E) are the two important basics of a graph. A vertex is merely drawn as a node or a dot. The order of a graph is the amount of its vertices. An edge joins pairs of vertices. The graph dimension is the amount of its edges. Fig. 1 displays a simple sample of a tree

Fig. 1. A tree graph.

graph. Vertex or node 1 is a origin of this graph. Once manipulating a graph, a data assembly is applied to store graphs in a computer scheme. There are two types of data assemblies, the list construction and the matrix construction. Practically, a mixture of both is the construction most extensively applied in tangible uses. 2.4. Minimum spanning tree (MST) In the literature, various properties of graphs have been well studied and a great number of practical applications of graph theory have been considered [65,66]. The use of developed methods in graph theory enables us to improve our understanding of the economic phenomena in which the embeddedness of individuals in their social interactions cannot be neglected [67]. The application of graph theory in finance is relatively recent but has explored in the last decades. Vast literature were reviewed on the application of network theory to financial markets. Several methods are reviewed to extract the information from a network related to a similarity matrix with a special emphasis to correlation based network, i.e., network where the similarity measure is the linear cross correlation. These methods include the threshold networks, the MSTs, and the planer maximally filtered graphs (PMFG) [68]. In financial studies, MST is used for reducing the complete network to a simpler correlation based network that helps in getting better visual insights about network. The method of constructing a MST connecting a set of n elements is direct and it is well-known approach in multivariate analysis as the nearest neighbor single linkage cluster analysis. The concept of MST was originally developed in the field of graph theory. In the last two decades we see its widespread use in many disciplines such as biology, social science, economy, anthropometry and general taxonomy, data analysis, regression analysis, computers science, networking, and multivariate and clustering analysis. Since MST in order to conduct a more effective examination of the information flow among vertices from the perspective of the overall network can be applied to reduce the number of edges among vertices in network, has been widely applied to various problems such as in cell-phone network design problem [69], traveling salesman problem [70], and etc. MST is also used for filtering networks of financial market that resulting in simpler forms of graphs that can facilitate the analysis of stock markets. It has been applied into financial market studies, for instance, the work of Mantegna and Stanley [15], Onnela, Chakraborti, Kaski, Kertesz and Kanto [14] and Onnela, Kaski
sz [71]. MST is a kind of currency map and is helpful for and Kerte establishing a stable portfolio on the financial market. In financial market, since the last decade, MST is considered as one of the chief streams in econophysics to filter the significant data enclosed in monetary networks. For finding an MST econophysicists are typically applying Kruskal’s algorithm [10,15,72]. In practice, the MST which is applied for analyzing a network is typically gained from the two most recommended algorithms of the previous studies, i.e., Kruskal’s algorithm and Prim’s algorithm [10,15]. See likewise Zhang, Lee, Wong, Kok, Prusty and Cheong [41] who use these algorithms in economy analysis as well as Ulusoy, Keskin, Shirvani, Deviren, €nmez [73] in transaction. The significant character of Kantar and Do these algorithms in the assembly of filtered network topology and economic taxonomy interested Huang, Gao and Wang [74] to associate them to each other about the computational difficulty. These two algorithms used for finding a minimum spanning tree form a network. According to the comparison of Prim’s and Kruskal’s algorithm in the study of Huang, Gao and Wang [74], Kruskal’s algorithm is more superior to use in constructing network if the number of nodes is not more than 100 [74]. Otherwise, Prim’s algorithm is more superior. Though, several of previous scholars agree that the prominent role is performed by Kruskal’s algorithm

M. Kazemilari et al. / Renewable Energy 102 (2017) 107e117

and mathematically is very attractive [75], easiest to comprehend and may be simply explained manually. This is maybe the motive why most of previous researchers would like to employ Kruskal’s algorithm more than Prim’s algorithm such as [76,77]. Therefore, Kruskal’s algorithm is most widely used than Prim’s algorithm. Kruskal’s algorithm was invented by Kruskal [70]. In the processes of Kruskal’s algorithm, the edge with least weight (or distance) will be chooses at each stage. In the final of the processes, it will construct a spanning tree which contains every vertex, and the sum of weights of all the edges in that tree is minimal. The steps of Kruskal’s algorithm are described as follows: 1. Sort the edges in the order of increasing weights and identify the edge with minimum weight in the graph. If there is more than one edge with minimum weight, then arbitrary choose one edge from them. 2. Choose the edge with minimum weight from those edges remained which do not form a cycle. (If more than one chose any); 3. Repeat step 2 until V1 edges have been chosen where the number of vertices is V. 3. Research methodology Stock market contains a large number of interactions among stocks and produces a rich data in every trading day. In general, complex system of stocks is represented in the form of a network [2], or equivalently an undirected weighted complete graph with finite number of nodes (stocks) and links (interactions between stocks). The complex system composed of larger stocks leads the system excessive carrying the redundant and complex information. That complexity makes the system difficult to analyze. In order to reduce the complexity of system, minimal spanning tree (MST), a graph theoretical approach has been used to simplify the system into a sub graph with the most relevant interactions of each stock. It is facilitates the stock network analysis. The use of MST in stock network analysis can trace to the work of Mantegna in 1999 [10]. He was first used the MST to study the topological structures of the stocks. Since then, the MST becomes an indispensable tool to filter the economic information contained in the network. In Mantegna’s work, the MST is constructed based on the distance matrix transformed from Pearson correction matrix of stocks. The relationships between a pair of stocks are customarily represented by correlation of their logarithmic closed price returns. 3.1. Method of constructing network In stock market analysis, the interrelations among stocks are represented in terms of Pearson correlation coefficient. This is the

111

standard practice to construct the stock network. Let pi(t) be the stock price of stock i (resp. j) at time t. With the precondition of time series are synchronous, the similarity among the stocks is quantified by the Pearson correlation coefficient (PCC) between the logarithmic exchange rate returns of i and j stocks is given as [10,15,78].



 ri rj  hri i rj ffi r ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi cij ¼ rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi D E D E
ffi 2 r 2i  hri i2 r 2j  rj

(1)

P where hri i ¼ Tt¼1 ri ðtÞ=T, is the statistical average of the time series over all the trading time T and,

ri ðtÞ ¼ ln pi ðtÞ  ln pi ðt­1Þ;

(2)

is the daily logarithmic exchange rate returns; andPi(t) is the exchange rate of i stock at time t. Suppose exchange rates of n stocks are studied, the PCCs of all possible pairs of stocks form a n  n symmetric matrix C, called correlation matrix. All diagonal elements of C are equal to 1. The off diagonal elements of C are varying between 1 to þ1, where value 1 (þ1) indicates that a pair of stocks are completely anticorrelated (correlated). The correlation coefficient is equal to 0 when two stocks are uncorrelated. That matrix C has a significant character in econophysics as the key basis of economic data. Analyzing the complicated arrangement of C is not humble. The more the amount of stocks, the greater the difficulty of that assembly [15]. As mentioned, correlation matrix is a numerical summary of the complex system among the n (n  1)/2 pairs of components. Therefore, the information space of foreign exchange market composed by n stocks is n (n  1)/2. To analyze the correlation structure of stocks in a network form, a distance metric is need to define. Based on PCC, the distance between i and j stocks is defined by Refs. [10,15,78].

dij ¼

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  ffi 2 1  cij :

(3)

This function fulfills the axioms of Euclidean distance [10,15,78]. (i) dij ¼ 0 if and only if i ¼ j, (ii) dij ¼ dji,and (iii) dij  dim þ dmj. The distance matrix D corresponding to C is then used to determine the topology network by using the technical tool, minimal spanning tree (MST). MST is used to reduce the information space of C from n (n  1)/2 into n  1. It constructs a topology network of connecting n stocks with n-1 most significant links which are of

Table 2 The list of websites for popular companies in the stock markets. No.

Resource

1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

http://www.renewable-energy-industry.com/stocks/ http://economictimes.indiatimes.com/topic/List-of-renewable-energy-companies-by-stock-exchange http://en.wikipedia.org/wiki/List_of_renewable_energy_companies_by_stock_exchange http://www.fool.com/investing/general/2014/11/02/5-best-stocks-to-buy-in-renewable-energy.aspx http://www.equities.com/news/3-renewable-energy-stocks-for-solar-investors-to-watch-closely http://www.investopedia.com/articles/active-trading/041415/top-10-alternative-energy-stocks-2015.asp http://www.investorideas.com/Companies/RenewableEnergy/ http://www.renewableenergyworld.com/articles/2015/01/ten-clean-energy-stocks-for-2015.html http://www.marketwatch.com/story/the-four-solar-stocks-all-investors-must-own-2014-10-29 http://digitalmindsoft.org/list-of-renewable-energy-companies-by-stock-exchange.html

112

M. Kazemilari et al. / Renewable Energy 102 (2017) 107e117

Table 3 The list of some renewable energy companies presented by stock exchange market. Symbol

Stock

Sector

Country

X1 X2 X3 X4 X5 X6 X7 X8 X9 X10 X11 X12 X13 X14 X15 X16 X17 X18 X19 X20 X21 X22 X23 X24 X25 X26 X27 X28 X29 X30 X31 X32 X33 X34 X35 X36 X37 X38 X39 X40 X41 X42 X43 X44 X45 X46 X47 X48 X49 X50 X51 X52 X53 X54 X55 X56 X57 X58

Brookfield Renewable Energy CANADIAN SOLAR ALEO SOLAR Americas Wind Energy Anwell Tech Applied Solar Ascent Solar Technologies EDP RENOVAVEIS Ballard Power Systems Carnegie Wave Energy CERAMIC FUEL CELLS China Sunergy Co DayStar Technologies Dongfang Electronics ENLIGHT ENERGY E-TON SOLAR TECH EnviroMission Limited Finavera Renewables First American Scientific First Solar, Inc
n Tecnolo
gica GamesaCorporacio Gintech Energy Corporation GREEN PLAINS INC. Innergex Renewable Energy JA Solar Holdings LDK Solar Co China Longyuan Power Group Mass Megawatts Wind Power MOTECH INDUSTRIES Neo Solar Power Corporation Ocean Power Technologies PHOENIX SOLAR PV Crystalox Solar plc ReneSola Ltd Crosswind Renewable Energy Renewable Energy Generation Renewable Energy Holdings PremierPower Renewable Energy S.A.G. SOLARSTROM SMA Solar Technology AG SOLAR-FABRIK SOLARWORLD Solco Ltd SunPower Corporation Suntech Power Holdings SUZLON ENERGY LTD Vestas Wind Systems MEYER BURGER Ormat Technologies Inc REC Silicon ASA (RECFUT.OL) Trina Solar General Cable Corporation Capstone Infrastructure Nordex Power REIT Ameresco, Inc NextEra Energy Plug Power Inc

Canada Canada Germany United States China United States United States Spain Canada Australia Australia China Canada China Israel Taiwan Australia Ireland Canada United States Spain Taiwan United States Canada China China China United States Taiwan Taiwan Australia Germany United Kingdom United States United States United Kingdom Wales United States Germany Germany Germany Germany Australia United States China India Denmark Taiwan United States Norway China United States Canada Germany United States United States Canada United States

X59 X60 X61 X62 X63 X64 X65 X66 X67 X68 X69 X70

SunEdison Alterra Power Corp Amyris Enersis S.A. Iberdrola, S.A SW Umwelttechnik Stoiser ZBB Energy Corporation Real Goods Solar, Inc. Münchmeyer Petersen Capital ACCIONA Solartech Energy Corp AREVA SA

Hydroelectric Solar Solar Wind Solar Solar Solar Wind Fuel Cell Wave Fuel Cell Solar Solar Electronics Solar and Wind Solar Solar Solar General Solar Wind Solar Fuel General Solar Solar wind wind Solar Solar wave Solar Solar Solar General General General Solar Solar Solar Solar Solar Wave Solar Solar Wind Wind Solar Geothermal solar and electronic Solar General General Wind General General Solar Hydrogen and fuel cell Solar General Fuel cell Electric power wind General General Solar General General Solar Nuclear & Renewable energy

United States Canada United States Chile Spain Austria United States United States Germany Spain Taiwan France

M. Kazemilari et al. / Renewable Energy 102 (2017) 107e117

shortest distance. In MST, each stock that represented by exchange rate time series is treated as a node with strongest connections (links) between stocks. For constructing the MST, Kruskal’s algorithm is the most powerful and widely used method [15]. 3.2. Data There are several renewable companies in the world; we attempted to collect our data from those companies were popular in the stock markets (Table 2). In this study, the daily closure prices data of 70 stocks of renewable energy during the time period from 13th October 2010 to 4th March 2015 are investigated. The following results of this paper are based on the historical data of the stock price downloaded from http://finance. yahoo.com. The list of stocks and the corresponding of stocks are provided in Table 2: The 70 stocks and their respective sectors, countries and symbols are presented in Table 3. 4. Results 4.1. Distribution of renewable energy companies based on sector Renewable energy capitals happen over wide geographical zones, compared to other energy foundations, that are focused in a partial amount of regions [79]. Rapid development of renewable energy and increase in energy efficiency is resultant in important energy safety, climate change mitigation, and economic profits [80]. In international public belief studies there is robust support to improve renewable bases such as wind power and solar power [81]. Renewable energy markets in the national level are likely to endure to improve powerfully in the coming period and beyond [82]. Several of renewable energy companies developed on stock relations in the retro afterward 2000. The initial 21st century was a very dynamic period for the renewable energy manufacturing, as many administrations set long term renewable energy goals. Some selected to straight fund the renewables with feed-in tariffs and extra temporary actions for bridging the gap to full cost accounting that would suitably reward these skills for their low releases and absence of interference with ecosystem facilities, and likewise to guarantee some volume and incentive to connect conservation-focused smart grid machineries. Numerous public corporations associated in the improvement of this

113

manufacturing and responsible for great marketplace portion do not contribute completely in renewable energy and have been absent for the list of the current study. In addition; due to lack of historical data of the stock price in http://finance.yahoo.com, our study only focused on 70 renewable energy companies by stock exchange. We attempted to choose the important renewable energy companies by stock exchange based on some information provided in Table 2. Totally; 70 renewable energy companies were included of this study. Fig. 2 provided the percentage of each sector. Results of this figure demonstrated that; solar companies were the first rank with 46% and general companies with 26% had the second rank.

4.2. Distribution of renewable energy companies based on country Renewable energy capitals and important chances for energy efficacy occur over extensive geographical parts, compared to related energy foundations that are focused in a limited amount of nations. Fast development of technological change of energy foundations, energy efficacy and renewable energy would end in important energy security and financial advantageous [80]. It would likewise decrease environmental pollution like air pollution and climate change produced by burning of fossil oils and develop public well-being, decrease premature deaths because of the pollution and keep related well-being costs which amount to different 100 billion dollars yearly merely over the US [83]. Renewable energy bases, which derive their energy out of the sunlight, both straight or indirect, like the hydro and wind, are predicted to be accomplished of providing humankind energy for nearly other 1 billion years, in that the forecasted upsurge in hotness from the sun is anticipated to make the surface of the soil excessively warm for liquid water to occur [84]. Given the favourable policy framework, the renewable energy sector has been experiencing fast growth and has seen a significant increase in investment in recent years [85]. Renewable energy sources are an option worthy of serious consideration for governments [86,87]. Some developing and developed countries extended their investments in international level by stock exchange market. According to results of Fig. 3, 70 renewable energy companies were from 17 different countries. Finding of this figure showed that; United State with 19 companies (27%) had the first rank and Canada with nine companies (13%) had the second rank.

1%

Solar

1%

Wind

26%

Fuel Cell

46%

Wave Electronics

3% 4% 6%

General Geothermal

13%

Fig. 2. Distribution of companies based on sector (Source: study calculation).

other

114

M. Kazemilari et al. / Renewable Energy 102 (2017) 107e117

1% 1% 3% 1% 1%

1% 1% 1%

Canada Germany

13%

United States

9% 11%

China Chile

7%

Spain 1%

6%

Austria 27% 11%

1%

Australia Israel Taiwan

Fig. 3. Distribution of companies based on country (Source: study calculation).

4.3. Analysis of data In order to analyze the topological properties of MST, three major centrality measures are used, namely; degree, closeness and betweenness centralities. The centrality measure of stock in a network represents the role or degree of influence and importance of the stock in that network. These measures will allow us to make decision about market behaviour. To elaborate the findings more clearly, based on the MST, we present their network topology. Table 4 presents the score of each centrality measures of MST for top 38 stocks. Table 4 Centrality measures based on MST. #

Degree Centrality

Betweens Centrality

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38

X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X

X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X

52 62 51 23 40 69 6 7 8 9 12 20 22 25 34 54 60 61 68 10 21 26 29 30 31 43 44 45 47 48 50 55 59 66 31 43 44 45

0.116 0.101 0.087 0.058 0.058 0.058 0.043 0.043 0.043 0.043 0.043 0.043 0.043 0.043 0.043 0.043 0.043 0.043 0.043 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029

52 20 51 8 68 54 23 47 21 29 22 62 34 44 40 61 69 25 6 7 9 12 60 45 10 26 30 31 43 48 50 55 59 66 10 26 30 31

Closeness Centrality 0.655 0.579 0.539 0.454 0.450 0.315 0.309 0.272 0.251 0.230 0.214 0.194 0.165 0.161 0.140 0.139 0.113 0.113 0.058 0.058 0.058 0.058 0.058 0.057 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029 0.029

X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X

52 20 51 68 8 62 44 1 23 24 49 56 67 25 45 59 27 42 54 40 61 10 5 46 53 57 70 34 9 31 47 55 66 26 53 57 70 34

0.239 0.235 0.220 0.220 0.202 0.201 0.198 0.193 0.193 0.193 0.193 0.193 0.193 0.185 0.183 0.182 0.181 0.181 0.180 0.173 0.170 0.169 0.168 0.168 0.168 0.168 0.168 0.167 0.164 0.163 0.161 0.157 0.157 0.156 0.168 0.168 0.168 0.167

Fig. 4 is the correlationebased MST issued from Kruskal’s algorithm for the 70 stocks of renewable energy. Each stock is represented by a labeled node and it is colored by its regions classification. The MST exhibits the topological structure of 70 stocks. This MST complies with the actual situation where the General Cable Corporation, First Solar. Inc and Trina Solar play the most important stocks in renewable energy in terms of market capitals. In comparison between these stocks, it is worth noticing that the range of evolution is very high for First Solar, Inc stock price. It is also important to note that it was detected as a hub. This is understandable in economical terms since First Solar. Inc was commonly considered as the American economy. However, in this paper, First Solar. Inc shows up systematically at the periphery and, instead, some much smaller capitalization companies from hubs. Some potential reasons of this include; the change in the US stock influences as compared to the past. As can be seen in Fig. 5, during March 2012 and March 2013, the performance of First Solar. Inc in terms of its closing price was dominated by General Cable Corporation. First Solar Inc was at the worst performance on June 2012 before it rises back with positive trend. However, the trend and its magnitude is more than General Cable Corporation and Trina Solar. The MST is attractive because it shows that the X52 (General Cable Corporation) is located at the central of network. It is directly connected by 8 stocks. They are X20 (First Solar, Inc), X62 (Enersis S.A.), X68 (ACCIONA), X24 (Innergex Renewable Energy), X67 (Münchmeyer Petersen Capital), X56 (Ameresco, Inc), X49 (Ormat Technologies Inc) and X1 (Brookfield Renewable Energy). X52 is the dominant stock in terms of centrality measures (see Table 1) and most of the information is mainly controlled by this stock. The most influential stocks after General Cable Corporation are First Solar.

Fig. 4. Minimal spanning tree of 70 renewable energy stocks.

M. Kazemilari et al. / Renewable Energy 102 (2017) 107e117

115

General G eneral Ca C Cab Cable able Corpora Corpora o on n FFirst irst rst Solar, rs Solar, Inc Inc TTrina rina Solar Solaarr

Fig. 5. Price evolution of First Solar, Inc (red), Trina Solar (green) and General Cable Corporation (blue) stocks.

Inc, Trina Solar, ACCIONA, EDP RENOVAVEIS and Enersis S.A, respectively. 5. Concluding remarks 1 It is a worldwide demand and mutual responsibility to do activities to cope with problems posed through atmospheric pollution and use of fossil fuels and to reach sustainable development by strengthening worldwide co-operation. Along with the worldwide economic development, energy deficiency and environmental pollution have become a progressively big difficulty. If it is not resolved in a good and effective way, not merely will humanoid society not obtain the aim of sustainable improvement, but it will likewise make a serious effect on the living setting and quality of human society. Consequently, renewable energy supply is rich, fresh and sustainable having positive impact on many social issues including creation of new jobs. Thus, it is a crucial measure for achieving sustainable improvement of human society through accelerating the usage and improvement of renewable energy. 2. The system of stocks market is extremely complex and it is continuously evolving through various heterogeneous interactions between them. Thus, to capture the mechanism of stocks market, it is important to study and figure out the interactions. In this regard, the correlation matrix has long been used to quantify the interactions, and the information generated can be very helpful if enough data has been provided beforehand. MST is an indispensable tool for analyzing the structure of stock markets. Many studies have been focused on analyzing stock markets based on MST-based networks however there is no study dealing with the network of renewable energy stocks. 3. In this paper, PCC coefficient is applied to measure the similarity among stocks. Based on this coefficient, we analyzed the topological properties as usually constructed in topological analysis of networks among stocks. This approach is used to analyze 70 major renewable energy stocks. 4. Results of analysis indicated that; solar companies were the first rank with 46% and general companies with 26% had the second rank. 70 listed renewable energy companies were from 17 different countries: US with 19 companies (27%) had the first

rank and Canada with nine companies (13%) had the second rank. 5. The results of MST analysis indicated that the General Cable Corporation, First Solar. Inc and Trina Solar are the significant stocks in the network. These results comply with the real situation. 6. The present paper has also some limitations and recommendations for future studies. First of all, this study attempted to select the most important renewable energy companies around the world, maybe, some other renewable energy companies were not included on this paper, therefore; for this bias future studies can focus on other companies. This study has chosen renewable energy companies in various regions based on country, in this regard future studies can classify and choose these companies in specific regions or specific country, for example, Asia, Europe, America etc. In addition; this study selected different kinds of renewable energy companies such as solar, wind, geothermal etc. future studies can only focus on one type of these renewable energy companies, for example only solar or wind. For collection of data related to renewable energy companies this study based on the historical data of the stock price downloaded from Yahoo finance website, future papers can collect data from other financial websites. In MST, each stock that represented by exchange rate time series is treated as a node with strongest connections (links) between stocks. For constructing the MST, this study used Kruskal’s algorithm, future papers can use other algorithms such as Prim’s algorithm. This study in order to analyze the topological properties of MST, three major centrality measures are used, namely; degree, closeness and betweenness centralities, future studies can use other centrality measures such as Eigenvector centrality and PageRank centrality. References [1] L.H. Ederington, J.H. Lee, How markets process information: news releases and volatility, J. Finance 48 (4) (1993) 1161e1191. [2] T.G. Andersen, T. Bollerslev, F.X. Diebold, C. Vega, Real-time price discovery in global stock, bond and foreign exchange markets, J. Int. Econ. 73 (2007) 251e277. [3] P. Balduzzi, E.J. Elton, T.C. Green, Economic news and bond prices: evidence

116

M. Kazemilari et al. / Renewable Energy 102 (2017) 107e117

from the US Treasury market, J. Financial Quant. Anal. 36 (2001) 523e543. ~o, Overview of insular power systems [4] O. Erdinc, N.G. Paterakis, J.P.S. Catala under increasing penetration of renewable energy sources: opportunities and challenges, Renew. Sustain. Energy Rev. 52 (2015) 333e346. [5] S.R. Sivarasu, E. Chandira Sekaran, P. Karthik, Development of renewable energy based microgrid project implementations for residential consumers in India: scope, challenges and possibilities, Renew. Sustain. Energy Rev. 50 (2015) 256e269. [6] V. Bukarica, M.B. Vrhov cak, Renewable Energy Policy in Croatia: Current Status and Future Activities, in: World Renewable Energy Congress (9; 2006), 2006. Hrvatska znanstvena bibliografija i MZOS-Svibor. € Murat, O. € Nuri, Renewable energy market conditions and bar[7] Ç.B. Nalan, O. riers in Turkey, Renew. Sustain. Energy Rev. 13 (2009) 1428e1436. [8] T. Mizuno, H. Takayasu, M. Takayasu, Correlation networks among currencies, Phys. A Stat. Mech. its Appl. 364 (2006) 336e342. [9] M. McDonald, O. Suleman, S. Williams, S. Howison, N.F. Johnson, Detecting a currency’s dominance or dependence using foreign exchange network trees, Phys. Rev. E 72 (2005) 046106. [10] R.N. Mantegna, Hierarchical structure in financial markets, Eur. Phys. J. B Condens. Matter Complex Syst. 11 (1999) 193e197. [11] G. Bonanno, G. Caldarelli, F. Lillo, R.N. Mantegna, Topology of correlationbased minimal spanning trees in real and model markets, Phys. Rev. E 68 (2003) 046130. [12] G. Bonanno, N. Vandewalle, R.N. Mantegna, Taxonomy of stock market indices, Phys. Rev. E 62 (2000) R7615. [13] W.-S. Jung, S. Chae, J.-S. Yang, H.-T. Moon, Characteristics of the Korean stock market correlations, Phys. A Stat. Mech. its Appl. 361 (2006) 263e271. [14] J.-P. Onnela, A. Chakraborti, K. Kaski, J. Kertesz, A. Kanto, Dynamics of market correlations: taxonomy and portfolio analysis, Phys. Rev. E 68 (2003) 056110. [15] R.N. Mantegna, H.E. Stanley, Introduction to Econophysics: Correlations and Complexity in Finance, Cambridge University Press, 1999. , N. Vandewalle, R.N. Mantegna, [16] G. Bonanno, G. Caldarelli, F. Lillo, S. Micciche Networks of equities in financial markets, Eur. Phys. J. B Condens. Matter Complex Syst. 38 (2004) 363e371. [17] R. Coelho, C.G. Gilmore, B. Lucey, P. Richmond, S. Hutzler, The evolution of interdependence in world equity marketsdEvidence from minimum spanning trees, Phys. A Stat. Mech. its Appl. 376 (2007) 455e466. [18] M. Faloutsos, P. Faloutsos, C. Faloutsos, ACM SIGCOMM’99, Comput. Commun. Rev. 29 (1999) 16. [19] R. Albert, H. Jeong, A.-L. Barab
asi, Internet: diameter of the world-wide web, Nature 401 (1999) 130e131. [20] H. Jeong, B. Tombor, R. Albert, Z.N. Oltvai, A.-L. Barab
asi, The large-scale organization of metabolic networks,, Nature 407 (2000) 651e654. [21] J. Jung, J. Lee, H.T. Moon, Topological determinants of protein unfolding rates, PROTEINS Struct. Funct. Bioinforma. 58 (2005) 389e395.
si, R. Albert, H. Jeong, Mean-field theory for scale-free random [22] A.-L. Baraba networks, Phys. A Stat. Mech. its Appl. 272 (1999) 173e187. [23] W. Jang, J. Lee, W. Chang, Currency crises and the evolution of foreign exchange market: evidence from minimum spanning tree, Phys. A Stat. Mech. its Appl. 390 (2011) 707e718. [24] A. Garas, P. Argyrakis, Correlation study of the Athens stock exchange, Phys. A Stat. Mech. its Appl. 380 (2007) 399e410. [25] C.G. Gilmore, B.M. Lucey, M. Boscia, An ever-closer union? Examining the evolution of linkages of European equity markets via minimum spanning trees, Phys. A Stat. Mech. its Appl. 387 (2008) 6319e6329. [26] F. Pozzi, T. di Matteo, T. Aste, Centrality and peripherality in filtered graphs from dynamical financial correlations, Adv. Complex Syst. 11 (2008) 927e950. [27] B.M. Tabak, T.R. Serra, D.O. Cajueiro, Topological properties of stock market networks: the case of Brazil, Phys. A Stat. Mech. its Appl. 389 (2010) 3240e3249. [28] P. Sieczka, J.A. Hołyst, Correlations in commodity markets, Phys. A Stat. Mech. its Appl. 388 (2009) 1621e1630. [29] J. Kwapien, S. Gworek, S. Drozdz, Structure and Evolution of the Foreign Exchange Networks, ArXiv Preprint arXiv:0901.4793, 2009. [30] G.J. Ortega, D. Matesanz, Cross-country hierarchical structure and currency crises, Int. J. Mod. Phys. C 17 (2006) 333e341. [31] M.J. Naylor, L.C. Rose, B.J. Moyle, Topology of foreign exchange markets using hierarchical structure methods, Phys. A Stat. Mech. its Appl. 382 (2007) 199e208. [32] M. Keskin, B. Deviren, Y. Kocakaplan, Topology of the correlation networks among major currencies using hierarchical structure methods, Phys. A Stat. Mech. its Appl. 390 (2011) 719e730. [33] H. Markowitz, Portfolio selection*, J. Finance 7 (1952) 77e91. [34] V. Tola, F. Lillo, M. Gallegati, R.N. Mantegna, Cluster analysis for portfolio optimization, J. Econ. Dyn. Control 32 (2008) 235e258. [35] K.-W. Chen, Y.-C. Chen, H.-J. Lo, F.C. Odds, T.-H. Wang, C.-Y. Lin, S.-Y. Li, Multilocus sequence typing for analyses of clonality of Candida albicans strains in Taiwan, J. Clin. Microbiol. 44 (2006) 2172e2178. [36] J.G. Brida, L.P. Esteban, W.A. Risso, M.J.S. Devesa, The international hotel industry in Spain: its hierarchical structure, Tour. Manag. 31 (2010) 57e73. [37] S. Bivona, G. Bonanno, R. Burlon, D. Gurrera, C. Leone, Taxonomy of correlations of wind velocitydan application to the Sicilian area, Phys. A Stat. Mech. its Appl. 387 (2008) 5910e5915. [38] B.M. Tabak, T.R. Serra, D.O. Cajueiro, The expectation hypothesis of interest rates and network theory: the case of Brazil, Phys. A Stat. Mech. its Appl. 388

(2009) 1137e1149. [39] H. Situngkir, Hierarchical Taxonomy in Multi-party System, ArXiv Preprint Nlin/0405005, 2004.
, S. Gworek, S. Drozd _ z, _ A. Go
rski, Analysis of a network structure of [40] J. Kwapien the foreign currency exchange market, J. Econ. Interact. Coord. 4 (2009) 55e72. [41] Y. Zhang, G.H.T. Lee, J.C. Wong, J.L. Kok, M. Prusty, S.A. Cheong, Will the US economy recover in 2010? A minimal spanning tree study, Phys. A Stat. Mech. its Appl. 390 (2011) 2020e2050. [42] B.M. Tabak, T. Serra, D.O. Cajueiro, Topological properties of commodities networks, Eur. Phys. J. B 74 (2010) 243e249. it, R. Eryig it, Network structure of cross-correlations among the world [43] M. Eryig market indices, Phys. A Stat. Mech. its Appl. 388 (2009) 3551e3562. [44] R.B. Roy, U.K. Sarkar, Identifying influential stock indices from global stock markets: a social network analysis approach, Proced. Comput. Sci. 5 (2011) 442e449. [45] D.A. Katsaprakakis, D.G. Christakis, The exploitation of electricity production projects from Renewable Energy Sources for the social and economic development of remote communities. The case of Greece: an example to avoid, Renew. Sustain. Energy Rev. 54 (2016) 341e349. [46] H.-f. Lu, B.-l. Lin, D.E. Campbell, M. Sagisaka, H. Ren, Interactions among energy consumption, economic development and greenhouse gas emissions in Japan after World War II, Renew. Sustain. Energy Rev. 54 (2016) 1060e1072. [47] A.A. Rafindadi, I. Ozturk, Effects of financial development, economic growth and trade on electricity consumption: evidence from post-Fukushima Japan, Renew. Sustain. Energy Rev. 54 (2016) 1073e1084.
rdova, Homeostatic control, [48] F.F. Yanine, F.I. Caballero, E.E. Sauma, F.M. Co smart metering and efficient energy supply and consumption criteria: a means to building more sustainable hybrid micro-generation systems, Renew. Sustain. Energy Rev. 38 (2014) 235e258. [49] I.I.E. Agency), Trends in Photovoltaic Applications, Survey Report of Selected IEA Countries between 1992 and 2008, IEA (International Energy Agency), 2009. [50] I. Henriques, P. Sadorsky, Oil prices and the stock prices of alternative energy companies, Energy Econ. 30 (2008) 998e1010. [51] S. Kumar, S. Managi, A. Matsuda, Stock prices of clean energy firms, oil and carbon markets: a vector autoregressive analysis, Energy Econ. 34 (2012) 215e226. [52] P. Sadorsky, Correlations and volatility spillovers between oil prices and the stock prices of clean energy and technology companies, Energy Econ. 34 (2012) 248e255. [53] M.T. Bohl, P. Kaufmann, P.M. Stephan, From hero to zero: evidence of performance reversal and speculative bubbles in german renewable energy stocks, Energy Econ. 37 (2013) 40e51. [54] S. Managi, T. Okimoto, Does the price of oil interact with clean energy prices in the stock market? Jpn. World Econ. 27 (2013) 1e9. [55] C. Breidenich, D. Magraw, A. Rowley, J.W. Rubin, The Kyoto protocol to the United Nations framework convention on climate change, Am. J. Int. Law 92 (1998) 315e331. [56] R.A. Ahmad, The Sun Rises on the Solar Sector, The World Bank, 2009. [57] Young Ernst, Renewable Energy Country Attractiveness Indices, EYGM Limited London, 2008. [58] H.A. Grøm, The Relationship between Renewable Energy Assets and Crude Oil Prices, NORWEGIAN SCHOOL of ECONOMICS, 2013. [59] A. Schmitz, Effect of Oil Prices on Returns to Alternative Energy Investments, Georgia Institute of Technology, 2009. [60] A.Y. Huang, C.-M. Cheng, C.-C. Chen, W.-C. Hu, Oil prices and stock prices of alternative energy companies: time varying relationship with recent evidence, J. Econo. Manag. 8 (2) (2012) 221e258. [61] M.J. Bürer, R. Wüstenhagen, Which renewable energy policy is a venture capitalist’s best friend? Empirical evidence from a survey of international cleantech investors, Energy Policy 37 (2009) 4997e5006. [62] M.M. Boyer, D. Filion, Common and fundamental factors in stock returns of Canadian oil and gas companies, Energy Econ. 29 (2007) 428e453. [63] R.W. Faff, T.J. Brailsford, Oil price risk and the Australian stock market, J. Energy Finance Dev. 4 (1999) 69e87. [64] P. Sadorsky, Risk factors in stock returns of Canadian oil and gas companies, Energy Econ. 23 (2001) 17e28. [65] G. Avondo Bodino, Economic Applications of the Theory of Graphs, 1962. [66] N. Deo, Graph Theory with Applications to Engineering and Computer Science, PHI Learning Pvt. Ltd., 2004. [67] D.K. Michael, S. Battiston, From graph theory to models of economic networks. a tutorial, in: Networks, Topology and Dynamics, Springer, 2009, pp. 23e63. [68] S.A. Cheong, Econophysics: an experimental course for advanced undergraduates in the Nanyang technological university, IIM Kozhikode Soc. Manag. Rev. 2 (2013) 79e99. [69] D. Khanafiah, H. Situngkir, 1 Innovation as Evolution Phylomemetic of Cellphone Designs, 2004. [70] J.B. Kruskal, On the shortest spanning subtree of a graph and the traveling salesman problem, Proc. Am. Math. Soc. 7 (1956) 48e50.
sz, Clustering and information in correlation [71] J.-P. Onnela, K. Kaski, J. Kerte based financial networks, Eur. Phys. J. B Condens. Matter Complex Syst. 38 (2004) 353e362. [72] R.L. Graham, P. Hell, On the history of the minimum spanning tree problem, Ann. Hist. Comput. 7 (1985) 43e57.

M. Kazemilari et al. / Renewable Energy 102 (2017) 107e117 €nmez, [73] T. Ulusoy, M. Keskin, A. Shirvani, B. Deviren, E. Kantar, C.Ç. Do Complexity of major UK companies between 2006 and 2010: hierarchical structure method approach, Phys. A Stat. Mech. Appl. 391 (2012) 5121e5131. [74] F. Huang, P. Gao, Y. Wang, Comparison of Prim and Kruskal on Shanghai and Shenzhen 300 index hierarchical structure tree, in: Web Information Systems and Mining, 2009. WISM 2009. International Conference on, IEEE, 2009, pp. 237e241. [75] J. Malkevitch, Trees: a Mathematical Tool for All Seasons, American Mathematical Society, 2012. Feature Column. [76] D. Cieslik, The vertex degrees of minimum spanning trees, Eur. J. Op. Res. 125 (2000) 278e282. [77] J.G. Brida, W.A. Risso, Hierarchical structure of the German stock market, Expert Syst. Appl. 37 (2010) 3846e3852. [78] X.-Y. Wu, Z.-G. Zheng, Hierarchical cluster-tendency analysis of the group structure in the foreign exchange market, Front. Phys. 8 (2013) 451e460. [79] J.L. Sawin, F. Sverrisson, K. Chawla, C. Lins, R. Adib, M. Hullin, S. Leitner, S. Mazzaccaro, H. Murdock, L.E. Williamson, Renewables 2014, Global status report 2014, 2014.

117

[80] I.E. Agency, International Energy Agency International Energy Agency  9 rue
de
ration  75739 Paris Cedex 15, France, 2012. de la Fe [81] C. Greenwood, A. Hohler, V. Sonntag-O’Brien, Global Trends in Sustainable ^ 2007. Energy Investment 2007, in, UNEP, ParyS, [82] E. Martinot, Renewables. Global Futures Report 2013, 2013. [83] M.Z. Jacobson, M. DeLucchi, G. Bazouin, Z.A. Bauer, C.C. Heavey, E. Fisher, S.B. Morris, D.J. Piekutowski, T.A. Vencill, T.W. Yeskoo, 100% clean and renewable wind, water, and sunlight (WWS) all-sector energy roadmaps for the 50 United States, Energy Environ. Sci. 8 (2015) 2093e2117. €der, R.C. Smith, Distant future of the sun and earth revisited, Mon. [84] K.-P. Schro Not. R. Astron. Soc. 386 (2008) 155e163. [85] J.C. Reboredo, Renewable energy contribution to the energy supply: is there convergence across countries? Renew. Sustain. Energy Rev. 45 (2015) 290e295. [86] S. Moosavian, N. Rahim, J. Selvaraj, K. Solangi, Energy policy to promote photovoltaic generation, Renew. Sustain. Energy Rev. 25 (2013) 44e58. [87] M. Pacesila, S.G. Burcea, S.E. Colesca, Analysis of renewable energies in European union, Renew. Sustain. Energy Rev. 56 (2016) 156e170.