DOES ENERGY PRICES AFFECT THE INVESTOR SENTIMENT?: SHORT- AND
LONG-TERM ANALYSIS IN EQUITY MARKET OF ISTANBUL STOCK EXCHANGE
Emrah Keleş1, Burç Ülengin2, Sibel Yılmaz Türkmen3 and Ömer Faruk Tan4
Abstract
In this study, we investigate the short- and long-run effects of energy prices on investor sentiment. Using some market-based variables and consumer confidence index among widespread sentiment measures, we construct two sentiment indexes by using principle component analysis. We employ Johansen Cointegration, Granger Causality and Vector Autoregressive (VAR) models and find that there is long-run equilibrium between sentiment index with raw data and energy prices. However, we do not find any evidence for short-term effect of change in energy price on investor sentiment. Developing a local sentiment index and finding evidence of long-term relationship between energy prices and sentiment for an energy-dependent emerging country like Turkey, we contribute to the related literature.
Keywords: Energy prices, investor sentiment, vector autoregressive, cointegration, causality
ENERJİ FİYATLARI YATIRIMCI DUYARLILIĞINI ETKİLER Mİ?: BORSA İSTANBUL PAY SENEDİ PİYASASINDA KISA VE UZUN DÖNEM ANALİZİ
Özet
Bu çalışmada, enerji fiyatlarının yatırımcı duyarlılığına kısa ve uzun dönemli etkileri araştırılmaktadır. Çok sayıdaki duyarlılık değişkeni arasından birtakım piyasa tabanlı değişkenler ve tüketici güven endeksi kullanılarak temel bileşen analizi aracılığıyla iki adet duyarlılık endeksi oluşturulmuştur. Johansen Eşbütünleşme, Granger Nedensellik ve Vektör Otoregresif (VAR) modelleri kurulmuş ve düzey değerlerden oluşan duyarlılık endeksi ile enerji fiyatları arasında uzun dönem ilişkisi tespit edilmiştir. Bununla birlikte enerji fiyatlarındaki değişimin yatırımcı duyarlılığı üzerinde kısa dönemli bir etkisine yönelik bir kanıt elde edilememiştir. Yerel bir duyarlılık endeksi oluşturulması ve Türkiye gibi enerji bağımlısı gelişmekte olan bir ülke için enerji fiyatları ile duyarlılık arasında uzun dönemli ilişkiye kanıt sağlaması bakımından ilgili literatüre katkı sağlanmaktadır.
Anahtar Kelimeler: Enerji fiyatları, yatırımcı duyarlılığı, vektör otoregresif, eşbütünleşme, nedensellik
1. INTRODUCTION
Energy, which is one of the most basic elements of the universe and which is responsible for almost all of the things around us, can be described as the ability to act (Schwarz and Seaton: 1986, 158). Energy and energy resources, which are extremely important for the economy, are an indispensable part of the industrialization infrastructure and everyday life. World countries are competing with each other to obtain energy resources. The cheap, clean and adequate energy is an important part of both economic and social life because of energy consumption, environmental factors and external dependence. In parallel with the increase in population and economy in our growing country, our energy demand is constantly increasing.
1 Asst. Prof, Department of Accounting and Finance, Marmara University, Istanbul, Turkey, emrah.keles@marmara.edu.tr
2 Prof., Department of Management Engineering, Istanbul Technical University, Istanbul, Turkey, ulengin@itu.edu.tr 3 Assoc. Prof., Department of Accounting and Finance, Marmara University, Istanbul, Turkey, sibelyilmaz@marmara.edu.tr
4 Res. Asst., Department of Accounting and Finance, Marmara University, Istanbul, Turkey, omer.tan@marmara.edu.tr
The continuous increase in energy prices and the fact that the energy production is very expensive reveals the significance of alternative energy sources.
The need for energy in the world is increasing. According to BP Statistical Review of World Energy (2017), the total primary energy consumption in the world in 2016 is 13,276.3 mtoe5. US (2,272.7
mtoe), Russian Federation (673.9 mtoe), China (3,053.0 mtoe) and India (723.9 mtoe) are the most energy-consuming countries. The energy consumption of Turkey is 137.9 mtoe.
Given its proximity to the world’s largest proven oil and gas reserves, Turkey is of strategic importance and has the long-standing ambition to become a major hub for the energy trade. Turkey is well positioned to reap the benefits of diverse supply sources and routes for its development and to support the diversification and security of the supply of the European Union and the region. The country’s geographical conditions and climate bring about substantial potential for renewable energy, particularly hydro, wind, solar and geothermal energy (IEA: 2016, 21).
Nevertheless, Turkey’s domestic energy production only covered 32.2 mtoe in 2015. Growing energy needs have been met by a higher use of natural gas and oil. Given their limited domestic availability, oil and natural gas supply is almost entirely covered by imports. Driven by economic growth, final consumption of energy has also increased in all sectors, by 35.8% since 2004 (IEA: 2016, 22-25).
Even though subdued domestic demand and depreciation in real exchange rates restrained an increase in imports, imports increased in nominal and real terms in the final quarter of 2016 and the first quarter of 2017 due to the gradual increase in oil prices coupled with the sizable gold imports (TCMB: 2016, 6; TCMB: 2017, 5).
In year 2016, current account deficit was 32.6 billion dollars (www.tcmb.gov.tr). The current account deficit is increasing, mainly because of the increase in oil prices. As oil prices continue to increase, this deficit will grow even bigger. The increase in energy prices negatively affects the current account deficit in energy importing countries like Turkey.
Current account deficit has increased fast in Turkey after 2000’s because of the overvaluation of TL, fast economic growth and the foreign source dependency at energy (Eşiyok, 2012: 63). In this period, Turkey was able to import energy at cheaper prices (Yüksel, 2016: 104). The commodity trade balance declined in 2014, and the current account deficit decreased considerably due to favorable conditions in global raw material and energy prices after 2015. Nevertheless, the current account deficit in world standards is high (Kaya, 2016: 57).
Due to Turkey's dependency on foreign energy, giving priority to domestic resources that will reduce this dependency will affect the current deficit positively.
The relationship between energy consumption and economic growth has been intensively analyzed in recent years. Mucuk and Uysal (2009), Akan et al. (2010), Özata (2010), Aydın (2010), Yanar and Kerimoğlu (2011), Karhan et al. (2012), Çetin and Seker (2012), Korkmaz and Develi (2012), Uzunöz and Akçay (2012), Akpolat and Altıntaş (2013), Altıntaş (2013), Çağıl et al. (2013), Bayar (2014), Erdoğan and Gürbüz (2014), Topallı and Alagöz (2014), Uçak and Usupbeyli (2015) have done researches and shown that the economic growth in Turkey is the cause of energy consumption or the consumption of energy is the stimulus of economic growth. They mostly discovered a causality relationship between economic growth and energy consumption.
On the other side, many reserachers focus on energy-stock market relation. Findings of these studies are mixed. However, the results show that there is a co-movement (causality, long-run relation, ..etc) between energy prices and stock markets (Maghyereh, 2004; El-Sharif, et al. 2007; Li, et al. 2012; Ünlü and Topçu, 2012; Şener et al. 2013). Common feature of these studies are that they use classical view and do not consider psychology of investors much. Contrary to mainstream finance, behavioral finance is interested in deviation from strict rationality assumption. Researchers in this field of finance assert that mispricing not only stems from earning opportunities but also from psychological foundations (Barberis, Shleifer, and Vishny, 1998; Daniel, Hirshleifer, and Subrahmanyam, 1998). Behavioral finance is based on mainly two constructions: real world arbitrage is limited because it is risky and costly (Shleifer and Vishny, 1997) and investors are affected by sentiment (DeLong et al., 1990). Contrary to arbitrageurs, irrational investors, namely noise traders, form their beliefs with some psychological biases, whereas they make their preferences in accordance with prospect theory (Kahneman and Tversky, 1979). These cognitively-based beliefs about future cash flows that are not explained with information at hand are commonly known as investor sentiment (Baker and Wurgler, 2007). Sentiment effects attract many scholars for that reason. Predictability tests are ideal for measuring these effects. If sentiment causes over-valuation, according to these tests sentiment-sensitive stocks would make lower returns when sentiment changes (Baker and Wurgler, 2007).
There is strong evidence that sentiment is a significant predictor of stock returns. An important finding of sentiment studies in aggregate levels is that the market has lower returns following high sentiment periods. This negative effect of sentiment is captured for companies that are difficult to value and arbitrage (Lemmon and Portniaguina; Baker and Wurgler, 2006). Strong evidence shows that when sentiment is high, subsequent returns of small, young, highly volatile, unprofitable, non-dividend-paying, extreme growth, and distressed stocks are lower (Baker and Wurgler, 2006). When sentiment is low, the opposite is true.
In this study, we explore the impact of changes in the energy prices on investor sentiment. As dependent variables, we develop two sentiment indexes from market-based data and confidence data by using principle component analysis. First index is constructed from raw data whereas second is the common principle component of the residuals obtained from regressing these variables on macro variables. Then, we document long-term relationship between crude oil price and sentiment levels, and between fluctuations in natural gas prices and change in sentiment levels by using cointegration, causality tests and error correction models. We do not find any short-term impact of our energy measures. The study contributes the literature in some ways. i) Although there are many studies related to impact of energy prices on returns or volatility in stock markets, very few papers investigate the relationship between energy prices and investor sentiment. Taking sentiment as dependent variable this paper documents evidence from an energy-importer country. ii) This study constructs a sentiment index for Turkey from the combination of some market indicators and confidence survey as the first time as far as we know. iii) The study also investigates the short- and long-run relationship between energy prices and sentiment.
Rest of the paper is organized as follows: The next section gives detail about the related literature. We present the data and the variables in Section 3. In Section 4, we design the research and discuss the results. We conclude in Section 5.
2. LITERATURE REVIEW
2.1. Energy Prices and Stock Market
El-Sharif, et al. (2007) and Li, et al. (2012) analyzed the movement between oil prices and stock returns and found positive relationship. Lake et al, (2009) attempted to explore of the effects of oil prices
returns on the stock market returns of Germany, USA, Greece and the UK. According to results, Greece and USA stock market returns are sensitive to oil price returns while the returns of UK and German stock returns are not affected. Ono (2011) explored the relationship of oil prices and stock returns in BRIC countries and revealed positive relations for China, Russia, and India while Brazil does not show any responses. Maghyereh (2004), Dhaoui and Khraief (2014) examined the relationship oil prices and stock returns and found no relationship between each other.
Ünlü and Topçu (2012), Eryiğit (2012), Güler and Nalın (2013), Akgün et al. (2013), Acaravcı and Reynaoğlu (2013), Özmerdivanlı (2014), Eyüboğlu and Eyüboğlu (2016) found Borsa Istanbul has positive relationship with oil prices in the long run. Kapusuzoğlu (2011), observed that there was one-way causality relationship between all indexes in the stock exchange market and oil prices, but oil prices were not causal for any of the three indexes as the result of Granger causality analysis. Şener et al. (2013), studied the period 2002-2012 with daily frequencies of Istanbul Stock Exchange and realized that future increases or decreases in oil prices will affect the formation of stock prices. Öztürk et al. (2013), indicated cointegration between oil prices and indices of ISE manufacturing industry and chemical-petroleum-plastic industry for the period 1997 and 2009. Abdioğlu and Değirmenci (2014), Avcı (2015) found causality relationship between oil prices and stock returns. Kılıç et al. (2014) found long run relationship between crude oil and industrial price index. Yıldırım (2016), observed that increase in oil prices has no effect on BIST 100 index but decrease in oil prices affects the BIST 100 index positively both short and long run. Sarı and Soytaş (2006) and İşcan (2010) did not find a relationship between oil prices and stock prices in the long run.
2.2. Investor Sentiment and Its Effects on Stock Market
Psychological evidence of Kahneman and Tversky studies reveals that people do not deviate randomly from rationality randomly, instead they do mostly in the same way. Demand formation of investors leads to highly correlated buying and selling among investors, and buying or selling of the same security at the same time. When irrational investors act as a social living being and follow others’ mistakes, this co-movement strengthens. Investor sentiment explains these common judgment mistakes many investors make rather than uncorrelated common mistakes as classical theories suggest (Shleifer, 2000). Investor sentiment is also used to describe beliefs and preferences that conform the psychological evidences. These heuristics-based beliefs of investors about future cash flows are known as investor sentiment (Shleifer, 2000; Baker and Wurgler, 2007). Baker and Wurgler (2006) define sentiment as speculation tendency of investors.
There are many sentiment measures and proxies in the literature. Differently from Brown and Cliff (2004) who use two main groups of direct and indirect sentiment measures, one can also divide them four main categories as shown in Table 1 (Keleş and Arat, 2016). Among others, Baker and Wurgler (shortly BW) composite index and confidence indexes are famous. Some recent studies compare these two proxies (Chan et al, 2016; Li et al., 2017) whereas some combines to construct different index (Finter, Ruenzi and Ruenzi, 2012). Canbaş and Kandır (2009) test some sentiment proxies such as market turnover ratio, mutual fund flows, closed-end fund discount, odd-lot sales-to-purchases ratio, share of the equity issuance and repo holdings of mutual funds for the period between July 1997 to June 2005. Using six-month observations, they did not find forecasting power of sentiment on stock returns in ISE except market turnover ratio.
Table 1: Investor Sentiment Measures and Proxies
Measure
Group Measure Type Author(s) Proxies
In di re ct M ea su re s Closed-End Funds
Lee et al. (1991) CEFD
Chen et al. (1993) CEFD
Neal&Wheatley (1998) CEFD Net mutual fund redemptions
Qiu&Welch (2005) CEFD
Composit
Sentiment Index Baker ve Wurgler (2006, 2007, 2012) B&W index Market
Performance Brown&Cliff (2004)
# of bullish stocks / # of bearish # of new high / # of new low
Micro Trade
Kumar&Lee (2006) Brokerage data Barber et al. (2009)
Order imbalance of small trades from TAQ/ISSM data
Trades of individual investor from brokerage data
Derivates
Wang (2001)
Simons&Wiggins (2001)
Bandopadhyaya and Jones (2011) Put/call ratio Brown&Cliff (2004)
Neal&Wheatley (1998) Odd-lot sales/purchases
Lemmon and Ni (2010) Synthetic demand index for put/call options Other Brown&Cliff (2004) IPOs returns IPOs numbers
Co ns um er /I nv es to rs Co nf id en ce a nd Se nt im en t Su rv ey s Consumer Confidence Survey and Indexes
Fisher&Statman (2000, 2003) AAII Investor SS6
Otoo (1999) UM CSI7
Lemmon&Portniaguina (2006) UM CSI and CB CCI8 Qiu& Welch (2006) UBS/Gallup index Menkhoff& Rebitzky (2008) UM CSI Schmeling (2009) DG ECFIN, 18 EU UM CSI, US Lemmon&Ni June (2010)
UM CSI B&W
Investors’ Intelligence Bull-Bear spread
Kandır et al. (2013) CNBC-e CCI
Direct Sentiment
Indexes Brown and Cliff (2004)
Other Index Baron’s CI
Ne ws a nd So ci al M ed
ia Tetlock (2007) Tetlock et al.(2008) Pessimism in Wall Street content Negative words about firms’ earnings/returns in financial media
Han (2008) Investors’ Intelligence Bull-Bear spread Chen et al. (2012) Articles for investors in http://seekingalpha.com
St oc k B oa rd C om m en
ts Tumarkin&Whitelaw (2001) 181,633 messages for 73 stocks in www.ragingbull.com Antweiler& Frank (2004) 1.5 million messages for 45 firms
www.finance.yahoo.com www.ragingbull.com
Das&Chen (2007) 145,110 messages for 24 tech stocks in www.finance.yahoo.com
Kim&Kim (2014) 32 million messages for 91 firms www.finance.yahoo.com 2.3. Investor Sentiment and Energy Prices
6 American Association of Individual Investors 7 University of Michigan Consumer Sentiment Index 8 Conference Board Consumer Confidence Index
Contrary to vast literature on energy and stock market, there are fewer studies about the relationship between sentiment and energy prices. Although there are some recent papers related to the effect of sentiment on energy prices, there are few studies investigating the impact of energy prices on investor sentiment. Johnson and Lamdin (2012) analyze the relationship between changes in gasoline prices and its impact on consumer sentiment. The authors document negative relationship between gasoline prices and consumer sentiment. Narayan and Narayan (2014) investigate the psychological barrier effect when the oil price reaches US$100 or more per barrel on stock returns. They find evidence for the negative effect of the US$100 per barrel on listed firms in American stock exchanges, both foreign and domestic firms. Ding et al. (2017) report contagion effect of international crude oil fluctuations on Chinese stock market investor sentiment. The results indicate that in the long term, crude oil price fluctuations significantly Granger cause of Chinese stock market investor sentiment. On the other hand, Panapoulou and Pantelidis (2015) develop regime switching models and test for their forecasting ability for oil prices. Their findings suggest that the regime-switching models are more accurate than the Random Walk model in terms of both statistical and economic evaluation criteria for oil price forecasts. Deeney et al. (2014) find evidence that sentiment is good predictor of oil prices. Narayan and Narayan (2017) investigate the impact of oil price news on stock returns. Their findings inform that oil price news predict market returns of some sectors.
3. DATA AND VARIABLES 3.1. Data and Sentiment Proxies
In this study, our sample covers monthly observations in the period between January 2005 and December 2016. As shown in Table 3, we obtained sentiment data from monthly bulletins of Capital Market Board of Turkey (CMB) whereas oil and macro data come from Federal Reserve Bank of St. Louis and The Organization for Economic Co-operation and Development (OECD), respectively. We can use longer sample period due to lack of monthly data for some sentiment variables and macro variables.
Our dependent variable is investor sentiment. To obtain a sentiment measure, we first derive some market-related data for 144 months from CMB monthly bulletins. Using Principal Component Analysis (PCA) we construct a sentiment index for Turkey. As energy variables prices of Crude Oil, West Texas Intermediate oil (WTI) and Henry Hub Natural Gas are our independent variables. Variables list, some descriptive statistics and data sources are shown in Table 2. As an emerging country, 12 years are such a long period for Turkey where some economic, political and social turbulences observed lead to shape micro and macro effects. Not interestingly, according to the data, energy prices move in a wide range as sentiment and macro indicators do. Two of most used oil price indicators, BRENT and WTI fluctuate between nearly 30$ to 140$ levels. EQUITY is 1 when there is no debt issuance and 0 when lack of equity issuance. In some months (i.e., between 2010 and 2013) higher level of NIPO decrease sharply in especially difficult economic prospects as depicted in Figure 1.
In the study, first, we construct sentiment index by PCA. Second, we investigate long and short-run effect of world energy prices on sentiment on Turkish Equity Markets. We also check for causality between sentiment and energy price.
3.2. Construction of Sentiment Index
We follow Baker and Wurgler (2006) to construct sentiment index for Turkey by PCA. Although Baker and Wurgler use only market-based components some subsequent studies (Finter, Niessen-Ruenzi and Ruenzi, 2012; Ding, et al., 2017) also use consumer confidence index as component. Due to lack of data we remove or change some variables during index construction. Differently from US markets, there is no closed-end fund in Turkey. Although Kandır, Çerçi and Uzkaralar (2009) use real estate investment trust (REIT) discounts as proxy for closed-end fund discount, we choose another variable, mutual fund
flow as Finter, Niessen-Ruenzi and Ruenzi (2012). We also could not use put/call as derivative market indicator since we only obtain data for couple years9. Finally we use CCI, EQUITY, NIPO, RIPO, FUND and TRADE for sentiment index.
Each proxy is likely to include both sentiment and idiosyncratic components that are unrelated to sentiment (Baker and Wurgler, 2006). PCA is good statistical way to extract common component and to obtain a new variable from information contained in multiple variables.
We use six sentiment proxies and their lagged to obtain principle component. Eigenvalues of first four components are above 1.00 and explain the 70% of total variation. There is also significant break in scree plot after fourth component. Therefore, we use weighted average of 1-4 components to form an initial index and as shown in Table 3, we use correlation matrix to determine our first sentiment index, SENT_RAW. We choose positive and higher correlation with the initial index. Then we obtain our first sentiment index, SENT_RAW, as follow:
Table 2: List of Variables and Descriptive Statistics
Variable Proxy Database Mean Std Max Min Obs
Panel A: Energy Variables
BRENT Oil price ($) Brent Crude FED St. Louis 80.58 26.40 132.72 30.70 140
WTI Oil price ($) WTI FED St. Louis 76.95 22.35 139.96 32.74 140
NGAS Natural gas price ($) Henry Hub FED St. Louis 4.93 2.47 13.42 1.73 140
Panel B: Sentiment Variables
CCI Consumer confidence index TSI/CBRT consumer confidence index Turkish Statistical Institute (TUIK) 74.58 7.73 92.20 55.66 144
EQUITY Equity issuance Equity/ (Equity plus Debt) CMB monthly bulletins 0.51 0.44 1.00 0.00 144
NIPO Number of IPOs Number of IPOs CMB monthly bulletins 1.19 1.75 10.00 0.00 144
RIPO IPO returns Average of the differences between offer prices and IPO prices
CMB monthly
bulletins 0.21 1.82 21.20* -2.94 144
FUND Net fund flows (thousand TL) Change in the market value of mutual funds CMB monthly bulletins 346.87 7,803 21,397 -70,990 144
TRADE Trading volume (thousand TL) Market turnover or trading volume CMB monthly bulletins 171.48 37.67 255.58 95.84 144 Panel C: Macro Variables
IND Growth in industrial production Monthly change in the industrial production index OECD 0.41 2.52 16.70 -6.80 144
PPI Industrial producer price index
Increase in the industrial producer price level comparing to the same month of previous year
OECD 110.65 26.17 167.00 71.00 144
CPI Consumer price index Increase in the consumer price level according to the same month of previous year
OECD 8.26 1.64 12.10 4.00 144
UNEMP Unemployment rate Unemployment rate OECD 9.98 1.50 12.10 4.00 144
NBER NBER recession indicator If there is recession 1; 0 otherwise OECD 0.36 0.48 1.00 0.00 144
SHORT Short-term interest rate Call money or immediate rate OECD 9.40 5.06 17.50 1.50 144
DIV Dividend yield BIST 100 dividend yield (value-weighted) Datastream 2.26 0.67 5.10 1.24 144
𝑆𝐸𝑁𝑇_𝑅𝐴𝑊! = 0.647𝐶𝐶𝐼!+ 0.538𝐸𝑄𝑈𝐼𝑇𝑌!!!+ 0.540𝑅𝐼𝑃𝑂! (1)
First principal component explains 45% of total sample variation, which is similar as in previous studies (49% in Baker and Wurgler, 2006; 0.37 in Finter, Niessen-Ruenzi and Ruenzi, 2012). This shows that one factor explains much variation.
We also construct second sentiment index. To eliminate the effects of macro variables we regress our individual sentiment proxies on macroeconomic and financial indicators; industrial production, IND, producer and consumer price indexes, PPI and CPI, unemployment rate, UNEMP, recession dummy, NBER, short-term interest rate, SHORT and dividend yield, DIV. We use residuals of these regression as components. 1 to 6 components are weighted to derive initial index. Then considering the correlations in Table 3 again, we determine high and positive correlations. Then, we derive second sentiment index, SENT_RES as follow:
Figure 1: Number and Average First Day Returns of IPOs in Turkey
Source: https://site.warrington.ufl.edu/ritter/ipo-data/
𝑆𝐸𝑁𝑇_𝑅𝐸𝑆! = 0,463𝐶𝐶𝐼!!!+ 0,538𝐸𝑄𝑈𝐼𝑇𝑌!+ 0,704𝐹𝑈𝑁𝐷! (2)
For the second sentiment index, first principal component captures 42% of total variation. Hence, we use only first component to obtain second sentiment index, which is adjusted for macro changes.
Table 3: Correlation Between Initial Indexes and Sentiment Proxies
Raw Data (Controlling for Macro Variables) Residual Data
Variable Level Lagged Level Lagged
CCI 0.55 0.55 0.67 0.71 EQUITY 0.81 0.82 0.15 0.11 NIPO 0.06 0.07 0.05 0.09 RIPO 0.21 0.18 0.16 0.19 FUND -0.15 -0.16 0.60 0.54 TRADE -0.80 -0.80 -0.57 -0.57 -5% 0% 5% 10% 15% 20% 25% 30% 35% 40% -5 0 5 10 15 20 25 30 35 40 Av er ag e Fi rs t-da y Re tu rn s Num be r of IP O s
The correlation between the 12-term first-stage index and the SENT _RAW is 0.78 whereas it is 0.75 between the 12-term first-stage index and the SENT_RES.
Table 4 shows cross-correlations of sentiment components, and correlations between components and final sentiment indexes. The higher coefficients of correlation between variables and final indexes are shown in first column.
Table 4: Correlation of Sentiment Components
Correlation with
Sentiment Indexes Correlation with Sentiment Components
Variable Coefficient CCI EQUITY RIPO FUND
Panel A: Raw Data
CCI 0.75 1.00
EQUITY(-1) 0.63 0.21 1.00
RIPO 0.63 0.21 0.10 1.00
Panel B: Residual Data (Controlling for Macro Variables)
CCI(-1) 0.51 1.00
EQUITY 0.60 0.01 1.00
FUND 0.78 0.16 0.19 1.00
As shown in Figure 2 SENT_RAW and SENT_RES, one can not easily conclude, there is strict co-movement between two indexes.
Figure 2: Sentiment Index of Turkey, 05/2005 to 12/2016
4. RESEARCH DESIGN
We investigate the dynamic short- and long-run relation between sentiment and energy prices. Hence we test the order of integration of the variables by Augmented Dickey Fuller (ADF) Test (Dickey and Fuller (1979), first. Engle and Granger (1987) and Johansen and Juselius (1990) methods enable us to search for the existence of any long run relationship between sentiment and energy prices for the series that are integrated of the same order. For the long-run relationship or cointegration, as another prequisity, residuals from equation of non-stationary variables are also required to be stationary.
-5 -4 -3 -2 -1 0 1 2 3 4 05 06 07 08 09 10 11 12 13 14 15 16 SENT_RAW SENT_RES
4.1. Unit Root Tests
ADF Test shows test statistics and p values for three test equations as in Table 5. According to test results, individual sentiment proxies are stationary except LTRADE while energy price variables are I(1) series.
Table 4: Augmented Dickey-Fuller Unit Root Tests
Intercept and Trend
Intercept None
Integration Order
Test st. p value Test st. p value Test st. p value Panel A: Investor Sentiment
SENT_RAW I(0) -4.95 0.00*** -3.30 0.02*** -3.32 0.00*** SENT_RES I(0) -6.34 0.00*** -6.37 0.00*** -6.40 0.00*** CCI I(0) -3.24 0.08* -2.95 0.04** -1.11 0.24 EQUITY I(0) -5.72 0.00*** -0.39 0.91 -2.42 0.02** NIPO I(0) -8.63 0.00*** -8.65 0.00*** -4.23 0.00*** RIPO I(0) -12.42 0.00*** -12.11 0.00*** -11.99 0.00*** FUND I(0) -3.38 0.00* -3.44*** 0.00*** -10.56 0.00*** LTRADE I(1) -2.56 0.29 -2.25 0.19 -0.20 0.61
Panel B: Energy Prices
LBRENT I(1) -2.33 0.41 -2.32 0.17 -0.17 0.62
LWTI I(1) -2.60 0.28 -2.47 0.13 -0.21 0.61
LNGAS I(1) -2.79 0.20 -1.75 0.40 -0.88 0.33
The lag length selection is based on SIC and the test equations include intercept and trend, only intercept, and neither intercept or trend. Test statistics are reported here with critical values of -3.44, -2.88 and -1.94, respectively. *, ** and *** denotes 10%, 5% and 1% significance level, respectively. L denotes natural logarithm of the related variables.
4.2. Long-Term Analysis
Johansen’s method allowing one or more cointegrating vector test and estimates in the multivariate system is based on vector autoregressive (VAR) analysis and utilizes the maximum likelihood estimates (Aydoğan, Vardar and Tunç, 2014).
Since SENT_RES is stationary, we do not investigate cointegration relation with energy prices. However, correlogram of SENT_RAW shows slow decreases in autocorrelations and series is trending. Since it is a series with a trend, we detrend the SENT_RAW. 𝑅! is supporting enough to show trend in the series with significant trend coefficient as shown in Table 6.
Table 5: Detrend Equation
Dependent variable: SENT_RAW
Coefficient St. errors
Trend -0.020 (0.002)***
Constant 1.426 (0.143)***
Adj. 𝑅! 0.49
F Statistics 132.92***
Table 6 shows the Johansen Test results. As shown in the table, energy price and investor sentiment index that is not controlled by macro variables are cointegrated, and there is a long-run relationship between them.
Table 6: Johansen Cointegration Test
Trace Eigenvalue
SENT_RAW Cointegrating
vector Test
Critical Values
Test Critical Values
10% 5% 1% 10% 5% 1% LBRENT 0 27.70 13.43 15.49 19.94 24.42 12.30 14.26 18.52 1 3.28 2.71 3.84 6.63 3.28 2.71 3.84 6.63 LWTI 0 27.92 13.43 15.49 19.94 23.91 12.30 14.26 18.52 1 4.010 2.71 3.84 6.63 4.010 2.71 3.84 6.63 LNGAS 0 22.67 13.43 15.49 19.94 27.56 12.30 14.26 18.52 1 4.69 2.71 3.84 6.63 6.54 2.71 3.84 6.63
Number of cointegrations are 2 for LWTI and LNGAS while 1 for LBRENT. L denotes natural logarithm of the related variables.
We demonstrate Granger Causality Test results in Table 7. As in table, we document strong causality from LBRENT, LWTI and LNGAS to SENT_RAW as in Test 1 and Test 3. Test 3 shows the long run causality between change in natural gas price and sentiment change while Test 1 displays causality with oil prices and sentiment levels. There is also reverse causality between LWTI and sentiment but not strongly. We do not find long-run causality between SENT_RES and energy prices.
Table 7: Granger Causality Test
SENT_RAW SENT_RES
Test 1 Test 2 Test 3 Test 4 Test 1 Test 2
LBRENT 7.724 (0.005)*** 1.299 (0.256) - - 0.035 (0.852) 0.054 (0.817) LWTI 6.935 (0.031)** 5.179 (0.075)* - - 0.834 (0.659) 1.147 (0.537) LNGAS 3.315 (0.191) 7.364 (0.025) 8.566*** (0.014) 2.301 (0.316) 0.058 (0.810) 0.475 (0.491) Test 1 shows chi-square test statistics for the causality from oil prices to sentiment variables and Test 2 displays reverse causalities from non-stationary data while Test 3 reveals granger causality through vector error correction estimates with differenced variables from energy to sentiment while Test 4 is opposite. p values are shown in parentheses. *, ** and *** denotes 10%, 5% and 1% significance level, respectively. L denotes natural logarithm of the related variables.
4.3. Short-Term Analysis
As a next step, we determine whether there is short-term relation between energy prices and investor sentiment. In Table 8, we document Vector Error Correction estimates. Before VAR equations we use detrended sentiment index. Results show that there is no statistically significant short-term effect of change in the energy prices on Turkish investor sentiment. On the right side of the table, we report negative and significant error correction coefficients. These coefficients are speed of adjustment from deviations from long-term equilibrium. Vector error correction estimates for prices of three energy indicators are as follow:
D(SENT_RAW) = - 0.695*( SENT_RAW(-1) - 0.435*LBRENT(-1) + 1.914) - 0.157*D(SENT_RAW(-1)) + 0.009*D(SENT_RAW(-2)) - 0.577*D(LBRENT(-0.157*D(SENT_RAW(-1)) + 0.950*D(LBRENT(-2)) + 0.020
D(SENT_RAW) = - 0.677*( SENT_RAW(-1) - 0.466*LWTI(-1) + 2.021) - 0.180*D(SENT_RAW(-1)) + 0.006*D(DET_SENT_RAW(-2)) - 0.072*D(LWTI(-1)) + 0.654*D(LWTI(-2)) + 0.020
D(SENT_RAW) = -0.556*( SENT_RAW(-1) - 0.219*LNGAS(-1) + 0.292) - 0.242*D(SENT_RAW(-1)) - 0.926*D(LNGAS(-1)) + 0.010
According to equations, between 56%-70% of disequilibrium is corrected in 1 month. With other words, i.e., system corrects its previous period disequilibrium at a speed of 69.5% between investor sentiment and change in energy prices for Brent oil.
Table 8: Short- and Long-Term Effects of Energy Prices
Short-Term Effect Long-Term Effect (Error Corr.)
SENT_RAW SENT_RES SENT_RAW SENT_RES
LBRENT -0.577 (0.987)a 0.262 (1.406) 0.695 (0.141)*** - LWTI -0.072 (0.926)a -0.138 (0.345) -0.677 (0.140)*** - LHHUB -0.926 (0.753)a -0.100 (0.416) -0.556 (0.125)*** -
Short-term effects are documented for one lag. Standard errors are in parentheses. a denotes cointegration relationship and restricted VAR estimates. Long-term effect reflects the speed of adjustment. L denotes natural logarithm of the related variables.
As we can see from the equations and Table 8, there is long-run relationship between sentiment and energy prices. However, we do not find any short-term effect of energy prices or change in energy prices on investor sentiment. The results show that changes in energy prices should be considered carefuly by policy makers and companies since they affect not only the fundamentals but also psychology of investors. Policy makers and companies should also care energy-saving policies and promotions to decrease the risk stemming from energy price movements.
5. CONCLUSION
The need for energy for near future is growing. Although there is an impressive trend in renewable energy especially in developed markets, fossil energy use of industries and households have still higher share. Hence, how the movement in energy prices attract financial markets is interesting. This study analyzes the long and short-run effect of change in energy prices on sentiment. To do that, we first construct a sentiment index with raw data and with data adjusted by macro indicators. Then with some long run tests, we find that changes in energy prices have impact on one of our sentiment measure. Conversely, we find no evidence of short-run relation. Furthermore, much of deviation from this relation in one period is adjusted next period. The results can attract policy makers in some ways. Movement in energy prices should be observed carefully since it has influence on investor optimism as well as macro and micro indicators. At this point, encouraging for energy-saving consumption and production would be beneficiary for stock market participants. This would reduce the risk related to upward movements in energy prices.
Developing a sentiment index for Turkey that include both market-based measures and consumer confidence and documenting evidence on long-term effect of energy prices on sentiment and causality between two, we contribute to the asset pricing, behavioral finance and energy-related literature. As further studies, sentiment proxies that are robust to macro changes can be added to the index construction, impact of commodity prices on sentiment can be tested for different sub-samples.
6. REFERENCES Books
Shleifer, A. (2000). Inefficient Markets: An Introduction to Behavioral Finance.New York: Oxford University Press Inc.
Journals
Abdioğlu, Z. and Değirmenci, N. (2016). Petrol Fiyatı Şoklarının Hisse Senedi Getirileri Üzerindeki Etkileri. TISK Academy. 11 (22). 330-351.
Acaravcı, S.K. and Reynaoğlu, İ. (2013). Enerji Fiyatları ve Hisse Senedi Getirileri: Türkiye Ekonomisi için Bir Uygulama. Nevşehir Hacı Bektaş Veli Üniversitesi Sosyal Bilimler Enstitüsü Dergisi. 3. 94-110.
Aissia, D. (2016). Home and Foreign Investor Sentiment and the Stock Returns. The Quarterly Review of Economics and Finance. 59, 71-77.
Akan, Y., Doğan, E.M. and Işık, C. (2010). The Causality Relationship between Energy Consumption and Economic Growth: the Case of Turkey. Enerji Piyasa ve Düzenleme. 1 (1), 101-120.
Akgün, A., Şahin, İ.E. and Yılmaz, B. (2013). The Effects of Variations in Gold and Oil Prices on BIST 100 Index. Mediterranean Journal of Social Sciences. 4 (10). 726-730.
Akpolat, A.G. and Altıntaş, N. (2013). Enerji Tüketimi ile Reel GSYİH Arasındaki Eşbütünleşme ve Nedensellik İlişkisi: 1961-2010 Dönemi. Bilgi Ekonomisi ve Yönetimi Dergisi. 8 (2), 115-127.
Altay, E. (2008). Sermaye Piyasasında Sürü Davranışı. İMKB'de Piyasa Yönünde Sürü Davranışının Analizi. BDDK Bankacılık ve Finansal Piyasalar Dergisi, 2 (1), 27-58.
Altıntaş, H. (2013). Türkiye’de Birincil Enerji Tüketimi, Karbondioksit Emisyonu ve Ekonomik Büyüme İlişkisi: Eşbütünleşme ve Nedensellik Analizi. Eskişehir Osmangazi Üniversitesi İİBF Dergisi. 8 (1), 263-294.
Avcı, Ö. B. (2015). Petrol Fiyatlarının Hisse Senedi Piyasasına Etkisi. Adnan Menderes Üniversitesi Sosyal Bilimler Enstitüsü Dergisi. 2 (3). 27-34.
Aydın, F.F. (2010). Enerji Tüketimi ve Ekonomik Büyüme. Erciyes Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi. 35, 317-340.
Aydogan, B., Vardar, G., & Tunç, G. (2014). The Interaction of Mutual Fund Flows and Stock Returns: Evidence from the Turkish Capital Market. Akademik Bakis. 14 (2), 163-173.
Baker, M., and Wurgler, J. (2006). Investor Sentiment and the Cross-Section of Stock Returns. The Journal of Finance. 61(4), 1645-1680.
Baker, M., and Wurgler, J. (2007). Investor Sentiment Effect in Stock Markets: Stock Characteristics or Country-Specific Factors? International Review of Economics and Finance. 6, 103-111.
Barberisa, N., Shleifer, A., and Vishny, R. (1998). A Model of Investor Sentiment. Journal of Financial Economics. 49 (3), 307-343.
Bayar, Y. (2014). Türkiye’de Birincil Enerji Kullanımı ve Ekonomik Büyüme. Atatürk Üniversitesi İktisadi ve İdari Bilimler Dergisi. 28 (2), 253-269.
Cakan, E. and Balagyozyan, A. (2014). Herd Behavior in the Turkish Banking Sector. Applied Economics Letters, 21 (2), 75-79.
Cakan, E. and Balagyozyan, A. (2015). Sectoral Herding: Evidence from an Emerging Market. Retrieved from https://www.researchgate.net/publication/289801944_Sectoral_Herding_Evidence_from_an_Emerging_Market. Canbaş, S., and Kandır, S. Y. (2009). Investor Sentiment and Stock Returns: Evidence from Turkey. Emerging Markets Finance and Trade. 45(4). 36-52.
Chan, F., Durand, R. B., Khuu, J., and Smales, L. A. (2016). The Validity of Investor Sentiment Proxies. International Review of Finance.1-5.
Corredor, P. Ferrer, E. and Santamaria R. (2013). Investor Sentiment Effect in Stock Markets: Stock Characteristics or Country-Specific Factors? International Review of Economics and Finance. 27, 572-591.
Çağıl, G., Yılmaz Türkmen, S. and Çakır, Ö. (2013). Enerji ve Makroekonomik Değişkenler Arasındaki İlişki: Türkiye Açısından Bir Uygulama. Muhasebe ve Finansman Dergisi. 58, 161-174.
Çetin, M. and Seker, F. (2012). Enerji Tüketiminin Ekonomik Büyüme Üzerindeki Etkisi: Türkiye Örneği. Uludağ Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi. 31 (1), 85-106.
Deeney, P., Cummins, M., Dowling, M., and Bermingham, A. (2015). Sentiment in Oil Markets. International Review of Financial Analysis, 39, 179-185.
Daniel, K., Hirshleifer, D., and Subrahmanyam, A. (1998). Investor Pyschology and Security Market Under and Over Reactions. The Journal of Finance. 53(6), 1839-1885.
Delong, J.B., Shleifer, A., and Summers, L. (1990). Noise Trader Risk in Financial Markets. 98(4), 703-738.
Ding, Z., Liu, Z., Zhang, Y., and Long, R. (2017). The Contagion Effect of International Crude Oil Price Fluctuations on Chinese Stock Market Investor Sentiment. Applied Energy. 187, 27-36.
El-Sharif, I., Brown, D., Burton, B., and Russell, A. (2005). Evidence on the nature and extend of the relationship between oil prices and equity values in the UK. Energy Economics, 27 (6), 819-830.
Engle, R. F., and Granger, W. J. (1987). Con-Integration and Error Correction: Representation, Estimation and Testing. Econometrica. 55(2), 251-276.
Erdoğan, S. and Gürbüz, S. (2014). Türkiye’de Enerji Tüketimi ve Ekonomik Büyüme İlişkisi: Yapısal Kırılmalı Zaman Serisi Analizi. Selçuk Üniversitesi Sosyal Bilimler Enstitüsü Dergisi. 32, 79-87.
Eryiğit, M. (2012). The Dynamical Relationship between Oil Price Shocks and Selected Macroeconomic Variables in Turkey. Economic Research. 25(2), 263-276.
Eyüboğlu, K. and Eyüboğlu, S. (2016). Doğal Gaz ve Petrol Fiyatları ile BIST Sanayi Sektörü Endeksleri Arasındaki İlişkinin İncelenmesi. Journal of Yasar University. 11 (42), 150-162.
Eşiyok, B.A. (2012). Türkiye Ekonomisinde Cari Açık ve Nedenleri. Finans Politik ve Ekonomik Yorumlar. 49 (569), 63-86.
Finter, P., Niessen-Ruenzi, A., & Ruenzi, S. (2012). The Impact of Investor Sentiment on the German Stock Market. Zeitschrift für Betriebswirtschaft. 82 (2), 133-163.
Frugier, A. (2016). Returns, Volatility and Investor Sentiment: Evidence from European Stock Markets. Research International Business Finance. 38, 45-55.
Güler, S. and Nalın, H.T. (2013). Petrol Fiyatlarının İMKB Endeksleri Üzerindeki Etkisi. Ekonomik ve Sosyal Araştırmalar Dergisi. 9(2). 79-97.
He, L., and Casey, K. (2015). Forecasting Ability of the Investor Sentiment Endurance Index: The Case of Oil Service Stock Returns and Crude Oil Prices. Energy Economics. 47, 121-128.
İşcan, E. (2010). Petrol Fiyatının Hisse Senedi Piyasası Üzerindeki Etkisi. Maliye Dergisi. 158, 607-617.
Johansen, S., and Juselius, K. (1990). Maximum Likelihood Estimation and Inference on Cointegration – With Applications to the Demand for Money. Oxford Bulletin of Economics and Statistics. 52(2), 169-210.
Johnson, M. A. and Lamdin, D. J. (2012). Changes in Gasoline Prices and Consumer Sentiment. Journal of Applied Business and Economics. 43-51.
Kadilli, A. (2015). Predictability of Stock Returns of Financial Companies and the Role of Investor Sentiment: A Multi-Country Analysis. Journal of Financial Stability. 21, 26-45.
Kahneman, D., and Tversky, A. (1979). Prospect Theory: An Analysis of Decision under Risk. Econometrica. 47(2), 263-292.
Kandır, S. Y., Çerçi, G., & Uzkaralar, Ö. (2013). Yatırımcı Duyarlılığı Temsilcileri: Yatırım Ortaklıkları İskontosu ve Tüketici Güven Endeksi Örneği. BDDK Bankacılık ve Finansal Piyasalar Dergisi, 7-2.
Kapusuzoğlu, A. (2011). Relationships between Oil Price and Stock Market: An Empirical Analysis from Istanbul Stock Exchange (ISE). International Journal of Economics and Finance. 3(6). 99-106.
Kapusuzoğlu, A. (2011). Herding in Istanbul Stock Exchange (ISE): A Case of Behavioral Finance. African Journal of Business Management. 5 (27), 11210-11218.
Karhan, G., Silinir, M., Çayın, M. and Aydeniz, N. (2012). Enerji ve Ekonomik Büyüme İlişkisi: Türkiye Örneği. Batman Üniversitesi Yaşam Bilimleri Dergisi. 2(1), 80-87.
Kaya, M. (2016). Türkiye’de Cari Açık Sorunu ve Nedenleri. Dicle Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi. 6(10), 51-75.
Keleş, E., and Arat, M. E. (2016). Yatırımcı Duyarlılığı Temsilcileri ve Sermaye Getirilerinin Tahmini. Öneri Dergisi. 45(12), 307-326.
Kılıç, C., Bayar, Y. and Özcan, B. (2014). Petrol Fiyatlarının Borsa İstanbul Sanayi Fiyat Endeksi Üzerindeki Etkisi. Kamu-İş. 13 (3), 125-141.
Kim, J., Ryu, D., and Seo S. (2014). Investor Sentiment and Return Predictability of Disagreement. Journal of Banking and Finance. 42, 166-178.
Korkmaz, Ö. and Develi, A. (2012). Türkiye’de Birincil Enerji Kullanımı, Üretimi ve Gayri Safi Yurt İçi Hasıla (GSYİH) Arasındaki İlişki. Dokuz Eylül Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi. 27(2), 1-25. Kumari, J., and Mahakud, J. (2015). Does Investor Sentiment Predict the Asset Volatility? Evidence from Emerging Stock Market India. Journal of Behavioral Experimental Finance. 8, 25-39.
Lake, A., and Katrakilidis, C. (2009). The Effects of Increasing Oil Price Returns and its Volatility on Four Emerged Stock Markets. European Research Studies. 12(1), 149-161.
Lemmon, M., and Portniaguina, E. (2006). Consumer Confidence and Asset Prices: Some Empirical Evidence. The Review of Financial Studies. 19(4), 1499-1529.
Li, S.F., Zhu, H.M., and Yu, K. (2012). Oil Prices and Stock Market in China: A Sector Analysis Using Panel Cointegration with Multiple Breaks. Energy Economics. 34 (6), 1951-1958.
Li, H., Gou, Y., and Park, S. Y. (2017). Asymmetric Relationship between Investors’ Sentiment and Stock Returns: Evidence from Quantile Non-Causality Test. International Review of Finance. 1-10
Liston, D. (2016). Sin Stock Returns and Investor Sentiment. The Quarterly Review of Economics and Finance, 59, 63-70.
Maghyereh, A. (2004). Oil Price Shocks and Emerging Stock Markets: A Generalized VAR Approach. International Journal of Applied Econometrics and Quantitative Studies. 1 (2). 27-40.
Mucuk, M. and Uysal, D. (2009). Türkiye Ekonomisinde Enerji Tüketimi ve Ekonomik Büyüme. Maliye Dergisi. 157, 105-115.
Narayan, S., and Narayan, P. K. (2014). Psychological Oil Price Barrier and Firm Returns. Journal of Behavioral Finance. 15 (4), 318-333.
Narayan, S., and Narayan, P. K. (2017). Are Oil Price News Headlines Statistically and Economically Significant for Investors? Journal of Behavioral Finance. 1-13.
Nie, Z., Wang, D., Xue, W. (2015). Investor Sentiment and its Nonlinear Effect on Stock Returns – New Evidence from the Chinese Stock Market Based on Panel Quantile Regression Model. Economic Modelling, 50, 45-55. Ono, S. (2011). Oil Price Shocks and Stock Markets in BRICs. The European Journal of Comparative Economics. 8 (1), 29-45.
Özata, E. (2010). Türkiye’de Enerji Tüketimi ve Ekonomik Büyüme Arasındaki İlişkilerin Ekonometrik İncelemesi. Dumlupınar Üniversitesi Sosyal Bilimler Dergisi. 26, 101-113.
Özmerdivanlı, A. (2014). Petrol Fiyatları ile BIST 100 Endeksi Kapanış Fiyatları Arasındaki İlişki. Akademik Bakış Dergisi. 43, 1-12.
Öztürk, M.B., Gümüş, G.K., F.D. Taşkın and Çağlı, E.Ç. (2013). Petrol ve Doğalgaz Fiyatları ile İmalat ve Kimya-Petrol-Plastik Sektörlerinin Endeksleri Arasındaki İlişki. Academic Review of Economics & Administrative Sciences. 6(2). 64-74.
Panopoulou, E., and Pantelidis, T. (2015). Speculative Behaviour and Oil Price Predictability. Economic Modelling. 47, 128-136.
Sarı, R. and Soytaş, U. (2006). The Relationship between Stock Returns, Crude Oil Prices, Interest Rates, and Output: Evidence from a Developing Economy. Empirical Economics Letters. 5(4). 205-220.
Shleifer, A., and Vishny, R. (1997). A Survey of Corporate Governance. The Journal of Finance. 52(2), 737-783. Şener, S., Yılancı V. and Tıraşoğlu, M. (2013). Petrol Fiyatları ile Borsa İstanbul’un Kapanış Fiyatları Arasındaki Saklı İlişkinin Analizi. Selçuk Üniversitesi İktisadi ve İdari Bilimler Fakültesi Sosyal ve Ekonomik Araştırmalar Dergisi. 13(26). 231-248.
Tatlıyer, M. (2014). Türkiye’de Cari Açığın Belirleyicileri ve Belirledikleri. Akademik Bakış Dergisi. 42, 1-28. Topallı, N. and Alagöz, M. (2014). Energy Consumption and Economic Growth in Turkey: An Empirical Analysis. Selçuk Üniversitesi Sosyal Bilimler Enstitüsü Dergisi. 32, 151-159.
Uçak, S. and Usupbeyli, A. (2015). Türkiye’de Petrol Tüketimi ve Ekonomik Büyüme Arasındaki Nedensellik İlişkisi. Ankara Üniversitesi SBF Dergisi. 70(3), 769-787.
Uysal, D., Yılmaz, K.Ç. and Taş, T. (2015). Enerji İthalatı ve Cari Açık İlişkisi: Türkiye Örneği. Muş Alparslan ÜniversitesiSosyal Bilimler Dergisi. 3(1), 63-78.
Uzunöz, M. and Akçay, Y. (2012). Türkiye’de Büyüme ve Enerji Tüketimi Arasındaki Nedensellik İlişkisi: 1970-2010. Çankırı Karatekin Üniversitesi Sosyal Bilimler Enstitüsü Dergisi. 3(2), 1-16.
Ünlü, U. and Topçu, M. (2012). Petrol Fiyatları Hisse Senedi Piyasalarını Doğrudan Etkiler mi: İMKB Örneği. İktisat İşletme ve Finans. 27 (319), 75-88.
Vozlyublennaia, N. (2014). Investor Attention, Index Performance and Return Predictability. Journal of Banking and Finance, 41, 17-35.
Yanar, R. and Kerimoğlu, G. (2011). Türkiye’de Enerji Tüketimi, Ekonomik Büyüme ve Cari Açık İlişkisi. Ekonomi Bilimleri Dergisi. 3(2), 191-201.
Yıldırım, E. (2016). Enerji Fiyat Şoklarının Hisse Senedi Piyasasına Etkisi: BIST Örneğinde Asimetrik Nedensellik ve Etki-Tepki Analizi Kanıtları. Uluslararası Yönetim İktisat ve İşletme Dergisi. 16 (Özel Sayı). 187-200.
Yüksel, S. (2016). Türkiye’de Cari İşlemler Açığının Belirleyicileri: Mars Yöntemi ile Bir İnceleme. Bankacılar Dergisi. 96(27), 102-121.
Websites
Balance of Payments. (2016). www.tcmb.org.tr
Reports
BP Statistical Review of World Energy (2017). United Kingdom: Pureprint Group Limited.
Dhaoui, A. and Khraief, N. (2014). Empirical Linkage between Oil Price and Stock Market Returns and Volatility: Evidence from International Developed Markets. Economic Discussion Papers, No 2014-12, Kiel Institute for the World Economy.
IEA – International Energy Agency (2016). Energy Policies of IEA Countries: Turkey 2016 Review, France: IEA Publications.
TCMB (2016). Balance of Payments and International Investment Position Report, 2016-IV. TCMB (2017). Ödemeler Dengesi ve Uluslararası Yatırım Pozisyonu Raporu, 2017-I.
Dictionary
