1
Oil price shocks and volatility on East-Asian industries from 1994 to 2014.
Student number: s2038943 Name: Johan van Tienen Study program: DD MSc IFM Supervisor: Mr. Scholtens
Abstract:
2
I Introduction:
This research will examine the impact of oil price returns and oil price volatility on stock returns. Chen et al (1986) performed research into the effect of different macroeconomic variables on the stock markets and find that oil prices do not have a significant effect on the stock markets. Huang et al (1996) research the oil and stock markets and how both of these markets influence each other. They do this on a industry level and market level and find that oil prices don’t have a relationship with the market but only a positive relationship with the oil and gas companies. However the research by Jones and Kaul (1996) find that an oil price increase causes a negative impact on stock market returns.
Oil prices influence industries differently based on the fact if oil is used as an input or an output. Other industry specific factors like competition and price elasticity may influence the ability of passing on higher fuel costs to customers and thus reduce the negative impact of oil price shocks (Nandha and Faff, 2008). Most of the previous oil price research is focused on the US and Europe, however the research into East-Asian countries is still limited. Furthermore the majority of the research is focused on one specific country and the number of industries that is being researched is very small. The research on the level of countries (indices) in previous research might hide the effects of oil price returns on industry level. Another point in which the current literature is limited is on the effect of oil price volatility on stock returns and stock return volatility. This relationship of oil price volatility on stock return volatility might especially have an impact on a sector level. This is because the standard deviation of sector stock returns generally exceeds the standard deviation of the aggregate market returns. In recent research into the relationship between oil prices and stock returns it is found that the financial crisis is believed to have influenced oil prices as well as other markets such as stock and exchange rate markets (Zhu et al. 2013).
This research will try to contribute to the current gaps in the literature by giving an insight into the following research questions: What is the response of East-Asian stock markets to changes in the oil price? How does the response to a change in the oil price differ between industries? How does the response to a change in the oil price differ between countries? What is the response of East-Asian stock markets to oil price volatility? How does the response to a change in oil price volatility differ between industries? How does the response to a change in oil price volatility differ between countries? Is there a difference in response to oil price changes before the crisis (26-8-1994 to 30-06-2008) compared to during/after the crisis (01-08-2008 to 30-06-2014)?
3 returns. This research will contribute to the current literature by looking at the effect of oil price returns on a sector and national level, while previous research mostly focused on national level. This will give us a clear view on how the impact differs between national and industry level for the East-Asian countries. If previous research did research on a sector level they usually used a very small amount of industries and used to focus on the oil & gas industries. This research will discuss 39 different industries using the Industrial Benchmark Classification (ICB). By using a larger number of industries this research
contributes to the current literature by researching industries that have been ignored in previous research and giving a clear overview. This research will also contribute by examining the relationship between oil price volatility and stock returns and stock price volatility. The findings have implications for investors and for companies . As different sectors in East-Asian countries might have different stock return responses to oil price returns and oil price volatility, this paper might provide information regarding the best way to diversify your portfolio to minimize the risk. If investors or companies expect an increase in oil price return or oil price volatility they can adjust their portfolios accordingly. This paper also has implications regarding international financial management in the following ways. First of all if the company has shares in other firms these might be influenced by oil price changes and oil price volatility changes. This can be crucial if the company is planning to sell these stocks in the near future when there is an increase in oil price and/or volatility. Second, oil prices and/or volatility might influence the stock price of the company itself, making it more prone to hostile takeovers.
4
II Literature review:
According to the paper by Scholtens and Yurtsever (2012), there has been surprisingly little research regarding the exact relationship between financial markets and oil price changes. Findings on the relationship between oil price shocks and returns on stocks are mixed and differ in methodology and data periods. Chen et al. (1986) researched the risk factors on stock markets in the US between 1953 and 1978 and find no effect of oil price risk on the US stock markets. Sadorsky (1999) also performed research on the US stock market in the period 1-1947 to 4-1996 using a VAR model and finds that oil price movements do explain movements of the US stock markets. Jones and Kaul (1996) looked at the UK, US Canada and Japan for the postwar period using an asset pricing model and find that changes in oil price have a negative impact on real stock returns in several countries. The oil prices could directly influence cash flows of the company if oil is an input in the production and also because oil price changes can influence the demand for certain industries (Ratti and Hasan, 2013). Kilian (2008) mentions that an increase in oil price reduces consumer income and raises savings. He mentions that the effects of oil price shocks on stock prices is through the demand effects. Kilian (2009) also identifies the different kind of oil price shocks and the different effects on the stock prices. In total they find that oil shocks account for 22% of the long-run variations in the US real stock returns.
5 mention that the impact of the oil price shocks is different for each Asian country according to the industries that country has, however it fails to provide the data to support these assumptions.
Therefore it might be important to look at a industry level. Scholtens and Yurtsever (2012) mention that the effect of oil price shocks on stock prices differs on industry level as oil might be used as an input or as an output and Nandha and Brooks (2007) add that factors like price elasticity and the degree of competition in a sector influence the ability of passing on the higher prices of oil to the customers and minimizing the impact of higher oil prices. Huang et al. (1996) use a VAR model for the relationship between US stock returns and oil-future returns between 9-10-1979 and 16-03-1990. They find that the oil prices have a positive relationship with the stock returns for the oil and petroleum companies, but they do not find a relationship for the other industries or for the total index. Nandha and Faff (2008) use a standard market model for 35 global industry indices to determine the impact of oil price on the stock returns of different industries in the period 4-1983 to 9-2005 and find that all the sectors have a
negative relationship between stock returns and oil prices except for mining and oil-gas industries. Reboredo et al. (2014) did research into different sectors from the US and the EU in the period 2000 to 2011 using wavelet multi-resolution analysis and found that pre-crisis the oil price changes had no influence on the stock returns except for the oil and gas companies. Arouri (2011) has studied the effect of oil price changes on 12 EU sector stock markets from 01-01-1998 to 06-30-2010 using a multifactor GARCH model. Arouri et al. (2011) find strong relationships for the EU on sector level which are mostly negative except for the oil & gas sector. They mention that the different sectors have
different sensitivities to oil price changes and that because of this it is important to invest your portfolio across sectors to achieve risk diversification. Scholtens and Yurtsever (2012) use a SVAR model to determine the impact of oil price shocks on European industries between 1983 and 2007. They use the Industry Benchmark Classification (ICB) classifying the companies into 41 different sectors. They find mostly negative responses of stock returns to oil price increases, except for the oil and gas producers, oil equipment and the mining industries which have positive stock returns in case of an oil price increase. Lee et al. (2012)use VEC and VAR models to determine the impact of oil price changes on different sector indices in the G7 countries in the period 1991-01 to 2009-05. They do find a relationship between oil price changes and sector stock returns in Germany, the US and France, however they do not find a relationship for any of the sectors in Japan. The sectors that were impacted most by oil price shocks are information technology and consumer staples, followed by financial, utilities and transportation sectors. Besides just the oil price returns, oil price volatility might also influence the stock market returns.
6 higher oil price volatility causes uncertainty for firms, which may firms to postpone irreversible
investment decisions to the future. Park and Ratti (2008)used a VAR model to determine the impact of oil price volatility on the US and 13 European countries between 1-1986 and 12:2005 and find for many EU countries that increased oil price volatility decrease stock returns, however they do not find this relationship for the US. Elyasiani et al. (2011) find a positive relationship between oil price volatility and stock returns for seven out of 13 US industries between 11-12-1998 and 29-12-2006 using a GARCH model. The relationship is positive oil & gas extraction, building, chemicals, transport equipment, air transportation, depository institutions and insurance. A similar model is used by Ratti and Hasan (2013) that also use a GARCH model for ten Australian industries between 31-3-2001 and 31-12-2010. They also find positive relationships for five out of the ten industries (energy, materials, financial, information technology and utility) and a negative relationship for one industry (industrials). Oil price volatility might not only influence the stock market returns but also the volatility of the stock market returns.
Hammoudeh, Dibooglu and Alesia (2004) use multivariate GARCH models for different US oil industry equity indices between 17-7-1995 and 10-10-2001 and find that increase in oil price volatility actually decreases stock return volatility of certain oil sectors, like oil and gas refining, and finds that it increases stock volatility in companies that engage in oil production and exploration. Hammoudeh, Yuan and Nandha (2010) use GARCH models on 27 different US sector stock indices classified by the ICB on the sample period 02-01-1989 to 3-10-2006. They find that oil price volatility has a positive relationship with stock return volatility in sectors that use oil intensively , and a negative relationship with stock return volatility for all the other sectors (including oil-related sectors). Arouri, Jounini, Khuong, and Nguyen (2011)find that there are volatility spillover effects from oil prices to sector stock returns for the EU, and they find a bidirectional relationship between oil prices and stock returns for the US using a VAR-GARCH model on the period 1998 to 2009. The literature of oil price volatility and the effect on stock returns and stock return volatility find both positive and negative relationships and are mostly focused on the US and Europe.
7 report that the dependence structure between crude oil prices and stock returns increased significantly for the US and Chinese stock markets after the recent world financial crisis, finding increased tail dependence in all paired markets.
This research will use a GARCH model similar to Elyasiani et al. (2011) and Ratti and Hasan (2013) that allows to measure the impact of oil price returns and oil price volatility on stock returns and volatility in one model. It will cover quite a large time period of almost 20 year of daily data, including the Asian financial crisis of 1997 and the most recent financial crisis of 2008. This while previous research that uses daily data usually uses a smaller timeframe (Huang et al. 1996, Hammoudeh et al. 2004, Elyasiani et al. 2011, Ratti and Hasan 2013, Reboredo et al. 2014). This research will cover a larger number of industries with 39 different industries using the ICB classification that is also used by Hammoudeh et al. (2010) and Scholtens and Yurtsever (2012). While previous research focused on the US or EU this research will try to explore the relationship between oil prices and stock markets in East-Asia.
III Methodology:
The short term relationship between oil prices and the East-Asian stock markets will be examined in this research. First the data and descriptive statistics of the data will be discussed. Afterwards the GARCH-M model will be discussed.
Data
8 derived from the following stock indexes: China Shanghai Stock Exchange, Hong Kong Hang Seng Index, Japan Nikkei 225, South Korea Kospi 200 and the Taiwan MSCI Index. In total this gives 758 companies of which 753 had suitable data available for this research, as 5 firms did not have a FTSE Industrial
Classification Benchmark (ICB) industry classification. The market return is determined as the returns on the MSCI World Index. The stock returns of the firms and the market returns are calculated in the following way:
R = LN(Pt/Pt-1)
Also the oil price return will be calculated in similar way as described above. The oil price that will be used is the return on one-month future prices of the West Texas Intermediate (WTI) crude oil for a couple of reasons. First of all, the futures on the WTI are according to Sadorsky (2012) the most widely traded oil future contract and serve as a worldwide standard in the oil market. Secondly, spot prices are more affected by temporary random noise than the future price of crude oil are, as mentioned by Sadorsky (2001). Third, Elyasiani et al. (2013) mention that if a firm engages in hedging that the
effectiveness of such hedging activities is judged by future prices and their variability. The WTI oil price is converted into the local oil price using the exchange rates for each country. This is because Nandha and Hammoudeh (2007) only find that countries in the Asia-Pacific region are sensitive to oil price changes when the oil price is expressed in their local currency. For classifying the companies into different industries the FTSE Industrial Classification Benchmark (ICB) is used. The ICB divides the companies into 10 industries, 19 supersectors, 41 sectors and 114 subsectors. This research will use the 41 sectors to classify the different industries. However in our sample there two sectors that do not have any companies that fall under these two sectors. The two sectors are Real Estate Investment Trusts (8760) and Equity Investment Instruments (8980). Excluding these two sectors means that there are 39 sectors remaining in our sample. The risk free rate that will be used is the three-month treasury bill rate for the individual countries, however some countries do not have a liquid treasury bill market. For those
9 ri,t = Ri,t - Rf,t
The excess stock return of the industry is calculated by taking the average excess stock return of all the companies in a specific industry. The excess market return is calculated by deducting the daily
compounded 3 month US treasury bill rate from the market return. All the data used in this research is extracted from Datastream.
The excess daily return data from the industries is described in Table 1. The returns have a skewness that is not close to zero and have a kurtosis that is above three. This gives a Jarque-Bera (J-B) value that rejects the null hypothesis of normality for all the industries. All of the industry returns are negatively skewed, except for Alternative Energy, Household Goods & Home Construction, Gas Water &
Multiutilities and Nonequity Investment Instruments, which are positively skewed. This is shown in Appendix 1. All of the industries are positively correlated with the market and oil price, except for Gas, Water & Multiutilities which has a small negative correlation with the oil price. To test whether the excess returns have autocorrelation this paper has performed a Ljung-Box Test (LB-Q) which is
significant for most of the industries. Because the zero hypothesis of no autocorrelation is rejected the excess returns have autocorrelation in most industries. Also this paper performed an Augmented Dickey-Fuller test (ADF) on the excess daily returns. The ADF test is rejects for all the industries the null hypothesis that there is a unit root. The Kwiatkowski-Philips-Schmidt-Shintest (KPSS) on the daily excess returns is not rejected which means that the series is stationary. The results of these test for the
respective industries can be found in Appendix 1.
10
Table 1: Daily excess stock return descriptive statistics for ICB industries in East-Asian countries between 26-8-1994 and 30-06-2014.
Notes: Descriptive statistics for the period 26-8-1994 to 30-6-2014 of the daily excess stock return. Market and oil price are the correlation of the excess stock return with the excess market return and oil price return in USD. *,**,*** means significant on the 0,10, 0,05 and 0,01 level respectively.
Industry Mean Median Max Min SD J-B Market Oil Price
Oil & Gas Producers 0.000 0.000 0.083 -0.082 0.013 *** 0.352 0.078
Oil Equipment, 0.000 0.000 0.180 -0.146 0.021 *** 0.226 0.035
Alternative Energy -0.001 0.000 0.161 -0.198 0.037 *** 0.144 0.022
Chemicals 0.000 0.000 0.069 -0.083 0.011 *** 0.381 0.095
Forestry & Paper 0.000 0.000 0.115 -0.155 0.016 *** 0.335 0.063 Industrial Metal & Mining 0.000 0.001 0.097 -0.100 0.012 *** 0.382 0.092
Mining 0.000 0.000 0.089 -0.098 0.014 *** 0.354 0.096
Construction & Materials 0.000 0.000 0.075 -0.098 0.012 *** 0.346 0.089 Aerospace & Defence 0.000 0.000 0.130 -0.153 0.016 *** 0.245 0.052 General Industrials 0.000 0.000 0.095 -0.085 0.014 *** 0.281 0.053 Electronic & Electrical Equipment 0.000 0.001 0.088 -0.095 0.012 *** 0.365 0.079 Industrial Engineering 0.000 0.000 0.089 -0.103 0.012 *** 0.344 0.050 Industrial Transportation 0.000 0.000 0.078 -0.095 0.014 *** 0.308 0.096
Support Services 0.000 0.001 0.109 -0.116 0.016 *** 0.310 0.044
Automobiles & Parts 0.000 0.001 0.082 -0.096 0.012 *** 0.355 0.062
Beverages 0.000 0.000 0.070 -0.102 0.013 *** 0.319 0.053
Food Producers 0.000 0.000 0.086 -0.074 0.013 *** 0.251 0.056
Household Goods & Home Construction 0.000 0.000 0.068 -0.065 0.015 *** 0.214 0.031
Leisure Goods 0.000 0.001 0.094 -0.101 0.015 *** 0.266 0.015
Personal Goods 0.000 0.000 0.064 -0.067 0.012 *** 0.283 0.057
Tobacco 0.000 0.000 0.101 -0.104 0.019 *** 0.081 0.025
Healthcare Equipment & Services 0.000 0.000 0.121 -0.122 0.018 *** 0.186 0.023 Pharmaceuticals & Biotechnology 0.000 0.000 0.077 -0.088 0.011 *** 0.372 0.095 Food & Drug Retailers 0.000 0.000 0.087 -0.087 0.012 *** 0.318 0.081
Media 0.000 0.000 0.084 -0.078 0.014 *** 0.242 0.070
Travel & Leasure 0.000 0.000 0.086 -0.078 0.013 *** 0.321 0.077 Fixed Line Telecommunications 0.000 0.000 0.095 -0.109 0.016 *** 0.260 0.070 Mobile Telecommunications 0.000 0.000 0.071 -0.079 0.013 *** 0.220 0.054
Electricity 0.000 0.000 0.090 -0.079 0.013 *** 0.250 0.056
Gas, Water & Multiutilities 0.000 0.000 0.184 -0.108 0.020 *** 0.131 -0.002
Banks 0.000 0.001 0.072 -0.075 0.011 *** 0.339 0.083
Nonlife Insurance 0.000 0.000 0.092 -0.124 0.020 *** 0.181 0.031
Life Insurance 0.000 0.000 0.074 -0.070 0.012 *** 0.339 0.076
Real Estate Investment & Services 0.000 0.000 0.074 -0.074 0.011 *** 0.391 0.074
Financial Services 0.000 0.001 0.086 -0.085 0.011 *** 0.341 0.064
11
Table 2: Daily excess stock return descriptive statistics for East-Asian countries between 26-8-1994 and 30-06-2014.
Notes: Descriptive statistics for the period 26-8-1994 to 30-6-2014 of the daily excess stock return. Market and oil price are the correlation of the excess stock return with the excess market return and oil price return in USD. *,**,*** means significant on the 0,10, 0,05 and 0,01 level respectively.
Model
This paper will use an asset pricing theory to test whether oil price returns and oil price volatility influence the stock returns. A GARCH-M approach will be used similar to the models used by Ratti and Hasan (2013) and Elyasiani et al. (2011) to model stock returns and the conditional variance of the stock returns. As the data is non-normal as is suggested by the J-B test, using an ARCH or GARCH model will be preferred as these type of models can deal with the non-normality and thick tails that are in the data. This research will use the GARCH-M model for estimating the stock price returns as it deals with the heteroskedasticity as found by the LB-Q. The GARCH-M adds an extra heteroskedasticity term into the model. The use of GARCH-M enables to test the effect of oil price returns and oil price volatility on the stock prices. The GARCH-M model estimates the conditional volatility of the stocks and allows it to change over time and with a GARCH-M model the effects of oil price return and oil price return volatility can be tested in one model. The model is the following and is similar to the model used by Ratti and Hasan (2013), except for the removal of the exchange rate as a variable. Because the portfolios of industries are composed from different countries with different exchange rates it is not possible to have one exchange rate variable in this model as used by Ratti and Hasan (2013). Instead of one exchange rate variable the oil price return per industry is calculated per individual industry. As every country has a different oil price return after it has been adjusted for the exchange rate, the individual industry oil price returns are as the weighted average of the oil price returns in the countries according to where the individual companies are located. For example if an industry has three companies from China and two from Hong Kong the oil price return for that day is ( 3 * China oil price return + 2 * Hong Kong oil price return ) / 5 = rio,t. The following equations are also used by Ratti and Hasan (2013) to capture the effects of oil price returns and oil price volatility on the stock returns:
Country Mean Median Max Min SD J-B Market Oil Price
12 ri,t = ci + 1ri,t-1 + 2rm,t + 3rio,t-1 + 42o,t-1 + iln(i,t) + i,t (1)
i = 1,2....,J
2 i,t = ωi + αiε2t−1 + βiσ2t−1 + 2io,t-1 (2)
i,t i,t x i,t (3)
Where i is the average of the industry in the GARCH model, ri,t is the excess return of industry i at time t, ri,t-1 is the excess return of the industry stock at time t-1 and rm,t-1 is the excess market return on time t. The conditional variance
i,t is based on the residual ε2t−1 on time t-1 and an autoregressive term σ2t−1. The residual i,t is based on the conditional variance t multiplied by the standardized residual i,t. The variable 2io,t-1 is the conditional oil price volatility based on the information on day t-1 using a GARCH (1,1) model as described under (4). This is calculated by using a GARCH (1,1) model on the oil price returns. Sadorsky (1999) find that oil return volatility calculated from a GARCH (1,1) is well suited to study the relationship between stock returns and oil price shocks as they do not find evidence of serial correlation in their residuals and thus appears adequate. In later research from Sadorsky (2006) they compare different GARCH models in forecasting oil return volatility and mentions that the GARCH (1,1) model is most suitable to sample forecast and recommends this class of model in estimating the oil return volatility. Elyasiani et al (2011) and Ratti and Hasan (2013) also use a GARCH (1,1) model in forecasting the oil return volatility. The conditional oil price volatility will be calculated for every individual industry.
Rio,t = + irio,t-1 + t (4)
2o,t =
t-1 + 2o,t-1 (5)
13 South Korea. Zhu et al. (2013) describe that the economic downturn caused a sharp decline in energy and oil demand, and positive expectations for economic growth made the oil prices rise throughout 2009. Besides this breakpoint in 2008 the graphs indicate a large breakpoint around 1998, so a second breakpoint test is performed on the period 26-08-1994 to 16-07-2008 for China, Hong Kong, Japan and Taiwan, and on the period 26-08-1994 to 07-07-2008 for South Korea. The second breakpoint is found on 24-12-1998 for China, Hong Kong, Japan and Taiwan and on 25-12-1997 for South Korea. Elyasiani et al. (2011) mention that the second breakpoint in December 1998 reflects the collapse of oil prices due to a significant decline in oil consumption in the Asian-Pacific economies in the aftermath of the Asian currency crisis of 1997. This was because of a combination of lower consumption and higher OPEC production had a significant negative effect on the oil prices.
To test whether the most current crisis had an impact on the model this paper will include a dummy variable in (1) that becomes 1 after 01-08-2008. Although the breakpoints take place in July 2008, the dummy variable becomes one on the first of August 2008. This is because the whole month of July breaks from the trend as is shown in Appendix 3. The month of July 2008 is removed to have a clear pre- and post-crisis period when testing equation (6) to check if the dummy variable influenced the other variables. With the added dummy variable the formula becomes the following:
ri,t = i + 1ri,t-1 + 2rm,t + 3ro,t-1 + 42o,t-1 + DF + iln(i,t) + i,t (6)
Where DF is the financial crisis dummy variable with a value of 1 between 01-08-2008 and 30-06-2014 and a value of 0 between 26-08-1994 and 30-06-2008.
This paper will briefly discuss another measure of volatility: “half-life” (HL) that is explained by Engle and Bollerslev (1986) as the length of time it takes for the volatility to move half way back to its
unconditional volatility following a deviation from it. So it measures the persistence of variance of the stocks. If there is a break from the trend this is the amount of time it will take to move back to its unconditional volatility that is calculated under equation (2). This method is also used by Elyasiani et al. (2011) for the industries. The “half-life” is calculated in the following way:
HL = LOG(0.5)/LOG(αi+βi) (7)
14 characteristics of the volatility of East-Asian industries and East-Asian countries which could be used for future research.
IV Results:
In this part the results will be discussed of the GARCH-M model regarding East-Asian industries and their respective countries. First the results regarding East-Asian industries will be discussed in the following order: the effect of oil price on stock returns, the effect of oil price volatility on stock returns, the effect of oil price volatility on stock volatility. After that the results of the countries will be discussed in the same order. At the end the impact of the crisis dummy variable will be discussed for the industries and the countries.
Effect of oil price on sector stock return:
The coefficient tables of equation (1) will be shown in this section (table 3), while the results of equation (2) are posted in the appendix 4. The coefficient tables of equation (2) are shown in the appendix to reduce the amount of tables in this results section. The model diagnostics of the model are based on the standardized residuals of the estimated GARCH-M model and are shown in appendix 5. First the model diagnostics will be briefly discussed. Under the null hypothesis of the Jarque-Bera the assumption is that the residuals are normally distributed, however all the Jarque-Bera values of the residuals reject this assumption. This means that residuals are not normally distributed and show that there are extreme values in all of the industries in this sample. The residuals of equation (1) have a skewness closer to zero compared to the real stock returns without the model. The residuals also have a kurtosis closer to three compared to real stock returns which means that the residuals are less skewed and have less leptokurtosis than the return values. Ratti and Hasan (2013) also find non-normality in the residuals.In appendix 6 the ARCH-LM values of the residuals can be found, which is insignificant for most industries. This means that there is no autocorrelation between the residuals and that the GARCH(1,1) model has reduced the autocorrelation in the residuals compared to the real stock returns.
15
Table 3: Results of GARCH(1,1)-M model on the daily excess stock returns of ICB industries in East-Asian countries between 26-08-1994 and 30-06-2014. (equation(1)).
Industry Intercept Return t-1 Market return Oil return Oil
volatility Con. Var.
Oil & Gas Producers -0.001 0.070*** 0.445*** 0.030*** 0.380 0.000
Oil Equipment, -0.002 -0.012 0.417*** 0.020 0.560 0.000
Alternative Energy -0.006 0.068*** 0.445*** 0.026 2.546** -0.001
Chemicals -0.002 0.042*** 0.372*** 0.021*** 0.332 0.000
Forestry & Paper 0 0.038*** 0.462*** 0.035*** 0.300 0.000 Industrial Metal & Mining -0.005 0.033** 0.474*** 0.020*** 0.719 0.000
Mining 0 0.059*** 0.517*** 0.032*** 0.501 0.000
Construction & Materials -0.001 0.065*** 0.416*** 0.026*** 0.436 0.000 Aerospace & Defence 0.003 -0.024 0.412*** 0.023*** 0.581 0.000 General Industrials -0.002 0.042*** 0.357*** 0.031*** 0.480 0.000 Electronic & Electrical Equipment -0.001 0.049*** 0.450*** 0.026*** 0.721 0.000 Industrial Engineering -0.003 0.081*** 0.420*** 0.019*** 0.759 0.000 Industrial Transportation -0.003 0.018 0.437*** 0.041*** 0.450 0.000 Support Services -0.001 0.077*** 0.481*** 0.028*** 0.917 0.000 Automobiles & Parts -0.004 0.054*** 0.432*** 0.026*** 0.845* 0.000
Beverages 0 0.051*** 0.412*** 0.019*** 0.483 0.000
Food Producers 0 0.038*** 0.296*** 0.029*** 0.9267* 0.000
Household Goods & Home Construction 0.004 0.034** 0.282*** 0.022*** 0.328 0.000
Leisure Goods -0.001 0.050*** 0.365*** 0.018** 0.787 0.000
Personal Goods 0.004 0.027* 0.335*** 0.023*** 0.349 0.000
Tobacco 0.004 -0.017 0.152*** 0.018 0.496 0.000
Healthcare Equipment & Services -0.001 -0.019 0.297*** 0.012 0.174 0.000 Pharmaceuticals & Biotechnology -0.002 0.027 0.406*** 0.027*** 0.528 0.000 Food & Drug Retailers -0.002 0.055*** 0.364*** 0.040*** 0.911 0.000
Media -0.001 0.078*** 0.349*** 0.034*** 0.473 0.000
Travel & Leasure 0.002 0.049*** 0.430*** 0.028*** 0.686 0.000 Fixed Line Telecommunications 0.004 0.044*** 0.372*** 0.020** 0.205 0.000 Mobile Telecommunications -0.001 -0.001 0.281*** 0.021*** 0.553 0.000
Electricity -0.004 0.053*** 0.343*** 0.025*** 0.347 0.000
Gas, Water & Multiutilities -0.001 0.042*** 0.266*** 0.001 0.532 0.000
Banks -0.006 0.031** 0.368*** 0.026*** 0.835* -0.001
Nonlife Insurance 0.002 0.071*** 0.366*** 0.029*** 0.867 0.000 Life Insurance -0.005 0.043*** 0.387*** 0.026*** 0.909** -0.001 Real Estate Investment & Services 0 0.045*** 0.425*** 0.020*** 0.286 0.000 Financial Services 0.001 0.069*** 0.385*** 0.025*** 0.592 0.000 Nonequity Investment Instruments 0.005 -0.011 0.714*** 0.049*** -0.785 0.001 Software & Computer Services -0.002 0.067*** 0.420*** 0.031*** 0.613 0.000 Technology Hardware & Equipment 0.001 0.035** 0.484*** 0.034*** 0.145 0.000
Notes: Return t-1 is the industry excess return on the previous day. Market return is the MSCI World Index on t=0. Oil volatility is the conditional variance of a GARCH (1,1) model on the oil price return which is described by
16 Ratti and Hasan (2013) mention that the GARCH model requires to have a stationary position in the persistence of volatility. This means that the ARCH (αi) and GARCH (βi) combined should have a value that is smaller than 1. This is satisfied for all the industries in the model. This means that it is appropriate to use the GARCH-M model and that it is preferred over the ARCH-M and ARCH models (Ratti and Hasan, 2013).
The half-life values are shown in appendix 4, all of the ARCH (αi) and GARCH (βi) combined are below zero, which means that the volatility will return to its mean for all the different sectors. The half-life values are also described in appendix 4 under the column “HL”. Elyasiani et al (2011) describe that return volatilities have a long memory if the HL is higher than 30 days, and that return volatilities have a relatively short memory when it is under 30 days. The industries with the shortest return volatility memories are Oil Equipment and Industrial Transportation with 13 and 14 days respectively. The one industry with the longest return volatility memory is Gas, Water & Multiutilities which has a HL value of almost 693 days.
As the results in table 3 indicate, all of the East-Asian industries have a positive relationship with the MSCI World Index, as all of the coefficients are significant and positive. The coefficient ranges between 0.266 for Gas, Water & Multiutilities and 0.714 for Nonequity Investment Instruments. Also it finds that the returns on t-1 have a positive and significant influence on the returns of t=0 for 31 out of 39
industries, and no relationship for 8 industries. The most important part of this research are the effect of the oil price return on the different industries. This paper finds that 34 out of 39 industries have a positive and significant relationship between oil price return and daily excess stock return, while for the other 5 there is no relationship found between the oil price returns and the stock returns. It is
remarkable and different from previous research as this paper does not find any negative relationships between stock price returns and oil price returns, while previous research does for certain industries. It is difficult to compare these findings directly with other research, as other research use different industry classifications and different models for determining the influence of oil prices on stock returns which might contribute to the difference in findings.
The industry that responds strongest to oil price returns is Nonequity Investment Instruments. The other financial sectors also have positive relationships between the oil price return and the stock price
17 Services, Financial Services and Nonequity Investment Instruments all have a positive relationship with the oil price return in this research. The industry with the second strongest impact of oil price returns on stock returns is Industrial Transportation. And this is quite remarkable as the literature finds a negative relationship between the transport sector and oil price returns (McSweeney and Worthington,
2008)(Nandha and Brooks, 2009). Nandha and Brooks (2009) used 38 different countries in their research and found a negative response between oil price returns and transport stock returns. The sectors that have been most researched are the oil and gas sectors. This paper finds that the Oil & Gas Producers stock returns have a positive and significant relationship with the oil price returns. These sectors have been researched extensively and they find a positive relationship in line with the findings in this paper (Mohanty and Nandha, 2011)(Ramos and Veiga, 2011)(Arouri et al., 2011).
Basher et al. (2012) performed research into emerging countries (including China) between 01-1988 and 12-2008. They find that an increase in stock market prices lead to an increase in oil prices. They mention that the economic growth in most developed economies is flat or in decline. Because of this the
emerging markets economic growth (as measured by the stock prices) is likely to be an important source of the demand side pressuring the oil market. They also find that oil prices respond negatively to an unexpected increase in oil supply and respond positively to an increase in demand.
As stock prices are a proxy for economic growth because of more positive future cash flows, the higher economic growth in East-Asia might trigger the higher oil prices. As explained by Basher et al. (2012) is that most of the EU and the US are developed markets and have flat economic growth or are in decline, and the demand for oil will remain stable. This is in contrast with East-Asia where there is more rapidly economic growth and is likely to be important in the demand side of pressuring the oil market. This could be the explanation in the difference of findings between the reactions of industries in the EU and US (which most previous literature focuses on) and East-Asia.
Effect of oil price volatility on sector stock return:
18 often related with oil price changes that move stocks in a particular direction. This is in line with what this paper finds for the four industries.However for the other 35 industries no relationship has been found, which means the oil price volatility does not influence their stock price returns.
Oil return volatility and industry stock return volatility:
The effect of oil return volatility on the industry stock volatility is measured by coefficient in equation (2). This is relationship between the conditional variance of the oil price returns and the conditional variance of the stock returns. The results this can be seen in appendix 4. It is found that the oil return volatility influences the stock volatility of the 7 out of 39 industries, while there is no relationship in 32 industries. The relationship is positive meaning that an increase in oil price volatility increases the volatility of the industry daily excess stock return. The industries are: Oil Equipment, Alternative Energy, Forestry & Paper, Industrial Transportation, Healthcare Equipment & Services, Travel & Leasure and Fixed Line Telecommunications. One point of interest is that for Alternative Energy the oil price return volatility has an influence on the stock return and the stock return volatility. However the oil price return itself does not have a direct influence on the stock returns. For Oil Equipment, Alternative Energy and Healthcare Equipment & Services the oil price return volatility influences the volatility of the stock returns, but the oil price returns do not influence the stock returns itself, only the volatility. Ratti and Hasan (2013) find that an increase in oil price volatility decreases the stock return volatility for 5 out of 10 Australian industries, and only find that a positive relationship between oil price volatility and stock price volatility for the financial sector.Elyasiani et al. (2011) find both positive and negative relationships regarding oil price volatility on stock price volatility for US firms. For East-Asian industries only a
relatively small amount of industries have a relationship between oil price volatility and stock return volatility, and the relationships that have been found are positive. When oil price returns increase in volatility, the future cash flows of firms that use oil as an input or output will become more volatile, and the stock price returns will become more volatile. This is the case when the impact of the oil price on the future cash flows cannot be hedge or is not hedged.
Effect of oil price on country stock return:
19 residuals have a leptokurtic distribution. This differs from the return distributions, which were
platykurtic. The ARCH-LM of the standardized residuals show that the GARCH-M has taken into account most of the autocorrelation, except for Taiwan for which the ARCH-LM is significant. Without the GARCH-M four out of five countries had autocorrelation. For equation (2) we find that both the αi and βi are found to be significant for all the different countries, which means there are significant ARCH and GARCH effects. Also the constant (ωi), ARCH (αi) and GARCH (βi) are non-negative. And as (ARCH (αi) + GARCH (βi)) <1 for all the countries it has a stationary position in the persistence of volatility. So for the countries both of these conditions are satisfied, and it means that the use of the GARCH-M model is preferred over the ARCH-M and ARCH models for countries.
Also the “half-life” values are calculated for the different East-Asian countries, which are shown in appendix 6. The values are on average higher than the HL values for the industries. The country with the lowest mean reversion is Japan with a HL value of around 33 days, and the highest is South Korea with a HL value of 173 days. As ARCH (αi) + GARCH (βi) is smaller than 1 all of the countries have a mean reversion.
Table 4:Results of GARCH(1,1)-M model on the daily excess stock returns of East-Asian countries between 26-08-1994 and 30-06-2014. (equation(1)).
Notes: Return t-1 is the industry excess return on the previous day. Market return is the MSCI World Index on t=0. Oil volatility is the conditional variance of a GARCH (1,1) model on the oil price return which is described by
equation (5). Con. Var. Is the conditional variance and result of equation(2). *,**,*** means significant on the 0,05, 0,02 and 0,01 level respectively.
The result of equation (1) for the individual countries in this sample can be seen in table 4. All of the countries have a relationship between the world market index returns and their own domestic markets returns. As can be seen, the relationship between China and the world market is a lot lower than the other East-Asian countries. This could be due to the fact that China still has a large and relatively closed domestic market. Chinn and Ito (2008) created a measure of financial openness for the economy, called the KAOPEN. Here they created an index which ranked the countries according to financial openness.
Country Intercept Return t-1 Market return Oil return Oil volatility Con. Var.
China 0.009*** 0.02 0.076*** 0.028*** 0.703 0.001*** Hong Kong 0.002 0.074*** 0.511*** 0.036*** 0.090 0.000
Japan 0.003 0.024 0.518*** 0.039*** -0.091 0.000
20 This shows that China was ranked as number 162 out of the 181 countries in 2010, which shows their financial market is still relatively closed. Japan and Hong Kong are number 14 and 38 respectively which shows that their financial markets are more open and these countries also have a lot stronger
relationship between the world market returns and their domestic stock market returns. This could be because they more open and influenced by the world markets. All of the countries have a significant and positive relationship between the oil price returns and the stock market returns. Japan has the strongest relationship between the oil price return and stock price return and South Korea and Taiwan the
weakest relationship. However Alom et al. (2013) find that South Korea and Taiwan have the strongest responses to oil price shocks. They mention that they have large manufacturing and information technology while being resource poor countries. This can actually be seen from the weaker positive relationship between oil price return and stock price return. This means that these countries do not benefit as much from an oil price increase as other countries in this sample, because the higher cost of oil prices reduce the stock returns more than other countries. Alom et al. (2013) mention that Hong Kong is dependent on financial services, and not as much on oil, and thus profits more from an oil price and thus demand increase. Abhyankar et al. (2013) did research on Japan and found that the stock price have a positive relationship with an aggregate demand shock in oil prices. They find that a oil supply shock that cause an increase in oil prices don’t affect the Japanese stock markets much. They mention that this could be because of the large strategic oil reserves that are held in public and private domain. This could mean that the negative oil price shocks don’t influence Japanese markets and thus they are only influenced by demand shocks, which increase stock prices. This could be the reason for the largest positive relationship between oil price returns and stock returns. Lin et al. (2010) mention that China responds low to global demand shocks, because it is relatively isolated. This could be the reason why the relationship between oil price returns and stock returns is not as strong as more financial open
economies like Japan and Hong Kong (see Chinn and Ito (2008)).
Effect of oil price volatility on country stock return:
21 relationship between oil price volatility and the Australian stock returns. Although the oil price volatility has a positive relationship with the oil price return on some industries, this is not the case for the country stock indices. This means that the impact of an increase in oil price volatility does not significantly affect the future cash flows of the East-Asian countries.
Effect of oil price volatility on country stock return volatility:
The effect of oil price return volatility on the country stock volatility is measured by in equation (2). The results this can be seen in appendix 6. It finds that for Hong Kong and Taiwan the oil price return volatility has a significant positive impact on the stock price return volatility.This is a point of interest as Hong Kong has no large oil consuming companies and is mostly focused on financial services. Ratti and Hasan (2013) say that an increase in volatility of the oil price return is associated with the oil price moving in one direction, which could influence the country index stock returns positively.
Effect of crisis on sector stock return:
In table 5 a dummy crisis variable is added to the model. This crisis dummy coefficient is zero and insignificant meaning there is a no change in stock returns after and during the financial crisis. Including the dummy variable did not change any of the relationships in this model, meaning that the financial crisis did not account for changes in the relationships.
Effect of crisis on country stock return:
22
Table 5:Results of GARCH(1,1)-M model on the daily excess stock returns of ICB industries in East-Asian countries between 26-08-1994 and 30-06-2014. (equation (6)).
Industry Intercept Return t-1 Market return Oil
return Oil vol. Con.
Var. Crisis
Oil & Gas Producers -0.001 0.070*** 0.445*** 0.030*** 0.385 0.000 0.000
Oil Equipment, -0.002 -0.012 0.417*** 0.021 0.572 0.000 0.000
Alternative Energy -0.006 0.068*** 0.445*** 0.026 2.564** -0.001 0.000
Chemicals -0.002 0.042*** 0.372*** 0.021*** 0.341 0.000 0.000
Forestry & Paper 0.000 0.038*** 0.462*** 0.035*** 0.277 0.000 0.000 Industrial Metal & Mining -0.005 0.033** 0.474*** 0.020*** 0.763 -0.001 0.000
Mining 0.000 0.059*** 0.517*** 0.032*** 0.480 0.000 0.000
Construction & Materials -0.001 0.065*** 0.416*** 0.027*** 0.451 0.000 0.000 Aerospace & Defence 0.003 -0.024 0.412*** 0.023*** 0.581 0.000 0.000 General Industrials -0.002 0.042*** 0.357*** 0.031*** 0.503 0.000 0.000 Electronic & Electrical Equipment -0.001 0.049*** 0.450*** 0.026*** 0.715 0.000 0.000 Industrial Engineering -0.003 0.081*** 0.420*** 0.019*** 0.757 0.000 0.000 Industrial Transportation -0.003 0.018 0.437*** 0.041*** 0.452 0.000 0.000 Support Services -0.001 0.077*** 0.481*** 0.028*** 0.917 0.000 0.000 Automobiles & Parts -0.004 0.054*** 0.433*** 0.026*** 0.829* 0.000 0.000
Beverages 0.000 0.051*** 0.412*** 0.019*** 0.460 0.000 0.000
Food Producers 0.000 0.038*** 0.296*** 0.029*** 0.933 0.000 0.000 Household Goods & Home Construction 0.004 0.034** 0.282*** 0.022*** 0.330 0.000 0.000 Leisure Goods -0.001 0.050*** 0.366*** 0.018** 0.849 0.000 0.000 Personal Goods 0.004 0.027* 0.335*** 0.023*** 0.339 0.000 0.000
Tobacco 0.004 -0.017 0.152*** 0.018 0.521 0.000 0.000
Healthcare Equipment & Services -0.001 -0.019 0.297*** 0.012 0.188 0.000 0.000 Pharmaceuticals & Biotechnology -0.002 0.027 0.406*** 0.027*** 0.521 0.000 0.000 Food & Drug Retailers -0.002 0.055*** 0.364*** 0.040*** 0.911* 0.000 0.000
Media -0.001 0.078*** 0.349*** 0.033*** 0.490 0.000 0.000
Travel & Leasure 0.002 0.049*** 0.430*** 0.028*** 0.687 0.000 0.000 Fixed Line Telecommunications 0.004 0.044*** 0.372*** 0.020** 0.222 0.000 0.000 Mobile Telecommunications -0.001 -0.001 0.281*** 0.021*** 0.556 0.000 0.000 Electricity -0.004 0.053*** 0.344*** 0.025*** 0.375 0.000 0.000 Gas, Water & Multiutilities -0.001 0.042*** 0.267*** 0.001 0.544 0.000 0.000
Banks -0.006 0.031** 0.368*** 0.026*** 0.814* -0.001 0.000
Nonlife Insurance 0.002 0.071*** 0.366*** 0.029*** 0.853 0.000 0.000 Life Insurance -0.005 0.043*** 0.387*** 0.026*** 0.907** -0.001 0.000 Real Estate Investment & Services 0.000 0.045*** 0.425*** 0.020*** 0.272 0.000 0.000 Financial Services 0.001 0.069*** 0.385*** 0.025*** 0.593 0.000 0.000 Nonequity Investment Instruments 0.005 -0.011 0.714*** 0.049*** -0.773 0.001 0.000 Software & Computer Services -0.002 0.067*** 0.421*** 0.031*** 0.625 0.000 0.000 Technology Hardware & Equipment 0.001 0.035** 0.484*** 0.034*** 0.100 0.000 0.000
23
Table 6:Results of GARCH(1,1)-M model on the daily excess stock returns of East-Asian countries between 26-08-1994 and 30-06-2014. (equation(6)).
Notes: Return t-1 is the industry excess return on the previous day. Market return is the MSCI World Index on t=0. Oil volatility is the conditional variance of a GARCH (1,1) model on the oil price return which is described by equation (5). Con. Var. Is the conditional variance and result of equation(2). Crisis is a dummy variable that has a value of 0 before 30-06-2008 and a value of 1 on and after 01-08-2008. *,**,*** means significant on the 0,05, 0,02 and 0,01 level respectively.
V Conclusion:
The goal of this study is to measure the effect of the oil price returns and oil return volatility on the return and volatility of East-Asian industries and the countries. This research uses a GARCH-M model to measure the relationship between oil price returns and oil return volatility on the one hand and the real stock returns and stock volatility of 39 different East-Asian industries using the ICB classification on the other hand. The GARCH-M model is appropriate to deal with the non-normal distribution that the real stock returns have and allows to measure the influence of oil price return and oil price volatility in one model.
This paper finds that for 34 out of the 39 industries that oil price return has a positive relationship with the real stock returns in East-Asia, and do not find a relationship for 5 out of the 39 industries. This is remarkable that even the industries that use oil as an input and (for example Industrial Transportation) have a positive relationship with oil price returns, and is different with previous sector research that mostly focused on the US and Europe which found mostly negative relationships in their sector research. The same result can be found for the East-Asian equity indices as all of them have a positive relationship between oil price returns and stock returns which is also different from previous research that finds especially strong negative relationships for manufacturing heavy countries like Korea and Taiwan. It is found for 4 out of 39 industries that oil price volatility has a positive relationship with stock returns, and no relationship has been found for 35 out of 39 industries. There is no relationship between oil price volatility on countries real stock returns in East-Asia for the individual countries. This paper also
examined the relationship between the oil price volatility and the volatility of the stock returns and
Country Intercept Return t-1 Market return Oil return Oil vol. Con. Var. Crisis
24 found that there is a positive relationship between the volatility of oil prices and stock returns for 7 out of 39 industries, and no relationship for the other industries. This relationship has also been found for two countries in East-Asia, which are Hong Kong and Taiwan in which an increase in the oil price volatility increases the stock return volatility. Finally this paper examined the impact of the recent financial crisis on the other variables in the GARCH-M model by including a dummy variable for the crisis, and does not find an impact on industry or country level.
The results of this paper add value to the current literature by researching the impact of oil price returns on real stock returns of East-Asian industries, while previous research is focused on the EU and the US. It also adds value by researching the impact of oil price volatility on both stock returns and stock volatility. The results contribute to the current discussion in the literature regarding the impact of oil price returns on stock returns as most of the previous literature finds negative relationships, while the results find only positive relationships. Furthermore it contributes to the limited amount of literature regarding oil price volatility and the impact on the different industries. In the previous literature the findings are mixed, however in this research it is found that oil price volatility has a positive relationship on both stock returns and stock return volatility.
The results from this study are important to investors and companies. As the different sectors in East-Asian countries mostly have positive relationships between oil price returns and real stock returns, the risk diversifying against oil price changes by investing in different industries might be limited for East-Asian countries. The results also provide information regarding international financial management of companies listed on East-Asian stock markets or companies that own shares of East-Asian companies. The results can be used to provide information when a oil price increase or decrease is expected and how the respective industries stock returns will respond in the short-term. It is assumed that oil prices directly influence the cash flows or indirectly by affecting the interest rates, and this paper gives more information how significant this impact is. The results also provide information about the response of the sector stock returns to oil price volatility which can be used in strategies to hedge oil price risks in portfolios.
25 increases while European and US industries have mostly negative response, with the exception of oil and gas companies.
References:
Abhyankar, Abhay, Xu Bing, and Wang Jiayue. "Oil Price Shocks And The Stock Market: Evidence From Japan." Energy Journal 34.2 (2013): 199-222
Alom, Fardous, Bert D. Ward, and Baiding Hu. "Macroeconomic Effects Of World Oil And Food Price Shocks In Asia And Pacific Economies: Application Of SVAR Models." OPEC Energy Review 37.3 (2013): 327-372.
Arouri, Mohamed El Hedi. "Does crude oil move stock markets in Europe? A sector investigation." Economic Modelling 28.4 (2011): 1716-1725.
Arouri, Mohamed El Hedi, Jamel Jouini, and Duc Khuong Nguyen. "Volatility spillovers between oil prices and stock sector returns: implications for portfolio management." Journal of International Money and
Finance 30.7 (2011): 1387-1405.
Basher, Syed Abul, Alfred A. Haug, and Perry Sadorsky. "Oil prices, exchange rates and emerging stock markets." Energy Economics 34.1 (2012): 227-240.
Bernanke, Ben S. "Irreversibility, uncertainty, and cyclical investment." (1980).
Chen, Nai-Fu, Richard Roll, and Stephen A. Ross. "Economic forces and the stock market." Journal of
business (1986): 383-403.
Chinn, Menzie D., and Hiro Ito. "A new measure of financial openness." Journal of comparative policy
analysis 10.3 (2008): 309-322.
Cong, Rong-Gang, Yi-ming Wei, Jian-Jin Jiao, and Ying-Fan. "Relationships between oil price shocks and stock market: An empirical analysis from China." Energy Policy 36.9 (2008): 3544-3553.
26 Elyasiani, Elyas, Iqbal Mansur, and Babatunde Odusami. "Oil price shocks and industry stock
returns." Energy Economics 33.5 (2011): 966-974.
Engle, Robert F., and Tim Bollerslev. "Modelling the persistence of conditional variances." Econometric
reviews 5.1 (1986): 1-50.
Ewing, Bradley T., and Farooq Malik. "Volatility transmission between gold and oil futures under structural breaks." International Review of Economics & Finance 25 (2013): 113-121.
Fang, Chung-Rou, and Shih-Yi You. "The impact of oil price shocks on the large emerging countries' stock prices: Evidence from China, India and Russia."International Review of Economics & Finance 29 (2014): 330-338.
Hammoudeh, Shawkat, Sel Dibooglu, and Eisa Aleisa. "Relationships among US oil prices and oil industry equity indices." International Review of Economics & Finance 13.4 (2004): 427-453.
Hammoudeh, Shawkat, Yuan Yuan, Thomas Chiang, and Mohan Nandha. “Symmetric and asymmetric US sector return volatilities in presence of oil, financial and economic risks.” Energy Policy, 38.8
(2010):3922-3932.
Huang, Roger D., Ronald W. Masulis, and Hans R. Stoll. "Energy shocks and financial markets." Journal of
Futures Markets 16.1 (1996): 1-27.
Jones, Charles M., and Gautam Kaul. "Oil and the stock markets." The Journal of Finance 51.2 (1996): 463-491.
Kilian, Lutz. "Not all oil price shocks are alike: Disentangling demand and supply shocks in the crude oil market." The American Economic Review(2009): 1053-1069.
Kilian, Lutz. "The economic effects of energy price shocks." Journal of Economic Literature (2008): 871-909.
27 Masih, Rumi, Sanjay Peters, and Lurion De Mello. "Oil price volatility and stock price fluctuations in an emerging market: evidence from South Korea." Energy Economics 33.5 (2011): 975-986.
McSweeney, Evan J., and Andrew C. Worthington. "A comparative analysis of oil as a risk factor in Australian industry stock returns, 1980-2006." Studies in economics and finance 25.2 (2008): 131-145. Mohanty, Sunil K., and Mohan Nandha. "Oil risk exposure: The case of the US oil and gas
sector." Financial Review 46.1 (2011): 165-191.
Nandha, Mohan, and Robert Brooks. "Oil prices and transport sector returns: an international analysis." Review of Quantitative Finance and Accounting 33.4 (2009): 393-409.
Nandha, Mohan, and Robert Faff. "Does oil move equity prices? A global view."Energy Economics 30.3 (2008): 986-997.
Nandha, Mohan, and Shawkat Hammoudeh. "Systematic risk, and oil price and exchange rate
sensitivities in Asia-Pacific stock markets." Research in International Business and Finance 21.2 (2007): 326-341.
Park, Jungwook, and Ronald A. Ratti. "Oil price shocks and stock markets in the US and 13 European countries." Energy Economics 30.5 (2008): 2587-2608.
Ramos, Sofia B., and Helena Veiga. "Risk factors in oil and gas industry returns: International evidence." Energy Economics 33.3 (2011): 525-542.
Ratti, Ronald A., and M. Zahid Hasan. "Oil Price Shocks And Volatility In Australian Stock Returns Oil Price Shocks And Volatility In Australian Stock Returns." Economic Record 89.(2013): 67-83
Reboredo, Juan C., and Miguel A. Rivera-Castro. "Wavelet-based evidence of the impact of oil prices on stock returns." International Review of Economics & Finance 29 (2014): 145-176.
Sadorsky, Perry. "Correlations and volatility spillovers between oil prices and the stock prices of clean energy and technology companies." Energy Economics 34.1 (2012): 248-255.
Sadorsky, Perry. "Modeling and forecasting petroleum futures volatility." Energy Economics 28.4 (2006): 467-488.
28 Sadorsky, Perry. "Risk factors in stock returns of Canadian oil and gas companies." Energy
economics 23.1 (2001): 17-28.
Scholtens, Bert, and Cenk Yurtsever. "Oil Price Shocks And European Industries." Energy Economics 34.4 (2012): 1187-1195.
Wang, Yudong, Chongfeng Wu, and Li Yang. "Oil price shocks and stock market activities: Evidence from oil-importing and oil-exporting countries."Journal of Comparative Economics 41.4 (2013): 1220-1239. Wen, Xiaoqian, Yu Wei, and Dengshi Huang. "Measuring contagion between energy market and stock market during financial crisis: A copula approach."Energy Economics 34.5 (2012): 1435-1446.
Zhu, Hui-Ming, Rong Li, and Sufang Li. "Modelling Dynamic Dependence Between Crude Oil Prices And Asia-Pacific Stock Market Returns." International Review Of Economics & Finance 29.(2014): 208-223. Zivot, Eric, and Donald WK Andrews. "Further Evidence on the Great Crash, the Oil-Price Shock, and the Unit-Root." Journal of Business & Economic Statistics 10.3 (1992).
Internet resources:
29
Appendices:
Appendix 1: Daily excess stock return statistics for East-Asian industries between 26-08-1994 and 30-06-2014.
Industry Skewness Kurtosis LB-Q KPSS ADF
Oil & Gas Producers -0.268 6.804 35,913*** 0.046 ***
Oil Equipment, -0.065 7.688 28,536*** 0.057 ***
Alternative Energy 0.151 6.724 48,554*** 0.058 ***
Chemicals -0.617 8.014 18,487* 0.068 ***
Forestry & Paper -0.325 10.079 46,426*** 0.044 *** Industrial Metal & Mining -0.481 8.465 8.050 0.050 ***
Mining -0.331 5.952 18,695* 0.035 ***
Construction & Materials -0.582 7.403 22,769** 0.055 *** Aerospace & Defence -0.223 7.748 9.949 0.091 *** General Industrials -0.336 6.081 23,66* 0.039 *** Electronic & Electrical Equipment -0.520 7.750 20,139* 0.070 *** Industrial Engineering -0.712 8.559 61,387*** 0.058 *** Industrial Transportation -0.342 6.582 17.123 0.054 *** Support Services -0.474 7.947 70,877*** 0.035 *** Automobiles & Parts -0.634 8.117 23,785*** 0.065 ***
Beverages -0.539 7.527 25,55*** 0.042 ***
Food Producers -0.331 5.942 35,022*** 0.058 ***
Household Goods & Home Construction 0.016 4.866 30,779*** 0.107 ***
Leisure Goods -0.370 8.022 63,544*** 0.090 ***
Personal Goods -0.298 5.685 18.275 0.049 ***
Tobacco -0.057 5.049 21,074* 0.025 ***
Healthcare Equipment & Services -0.031 6.841 21,599** 0.064 *** Pharmaceuticals & Biotechnology -0.504 8.350 10.723 0.047 *** Food & Drug Retailers -0.464 7.318 35,888*** 0.101 ***
Media -0.252 4.649 33,288*** 0.038 ***
Travel & Leasure -0.249 5.674 43,503*** 0.058 *** Fixed Line Telecommunications -0.062 5.749 53,076*** 0.063 *** Mobile Telecommunications -0.301 5.258 16.068 0.040 ***
Electricity -0.162 5.458 19,835* 0.090 ***
Gas, Water & Multiutilities 0.288 8.225 34,021*** 0.024 ***
Banks -0.465 6.838 15.784 0.051 ***
Nonlife Insurance -0.080 4.701 59,332*** 0.041 ***
Life Insurance -0.254 5.580 33,096*** 0.047 ***
Real Estate Investment & Services -0.378 7.691 16.373 0.077 *** Financial Services -0.517 7.527 52,007*** 0.044 *** Nonequity Investment Instruments 0.176 6.523 10.459 0.047 *** Software & Computer Services -0.302 5.763 50,359*** 0.037 *** Technology Hardware & Equipment -0.535 9.451 14.517 0.082 ***
30 Appendix 2: Daily excess stock return statistics for East-Asian countries between 26-08-1994 and 30-06-2014:
LB-Q is the Ljung-Box test on the 10th lag. KPSS is the KPSS is the Kwiatkowski-Philips-Schmidt-Shin test on the level using the trend and intercept, the number of lags is determined by the Newey-West Bandwith. ADF is the
augmented Dickey Fuller test on the level with trend and intercept, the number of lags is determined by the Akaike Info Criterion. *,**,*** means significant on the 0,05, 0,02 and 0,01 level respectively.
Appendix 3: Zivot-Andrews breakpoint tests on the oil price returns of each country in their respective currency.
China: The Zivot-Andrew Breakpoint test on the trend and intercept for the oil price returns measured in Chinese Yuan between 26-08-1994 and 30-06-2014. The more negative the value of the Dickey-Fuller test the more likely that the zero hypothesis of no united root is rejected.
D ic ke y-Fu lle r te st v al u e Years
Country Skewness Kurtosis LB-Q KPSS ADF
31
Hong Kong : The Zivot-Andrew Breakpoint test on the trend and intercept for the oil price returns measured in Hong Kong Dollar between 26-08-1994 and 30-06-2014. The more negative the value of the Dickey-Fuller test the more likely that the zero hypothesis of no united root is rejected.
D ic ke y-Fu lle r te st v al u e Years
Japan: The Zivot-Andrew Breakpoint test on the trend and intercept for the oil price returns measured in Japanese Yen between 26-08-1994 and 30-06-2014. The more negative the value of the Dickey-Fuller test the more likely that the zero hypothesis of no united root is rejected.
32
South Korea: The Zivot-Andrew Breakpoint test on the trend and intercept for the oil price returns measured in Korean Won between 26-08-1994 and 30-06-2014. The more negative the value of the Dickey-Fuller test the more likely that the zero hypothesis of no united root is rejected.
D ic ke y-Fu lle r te st v al u e Years
Taiwan: The Zivot-Andrew Breakpoint test on the trend and intercept for the oil price returns measured in Taiwanese Dollar between 26-08-1994 and 30-06-2014. The more negative the value of the Dickey-Fuller test the more likely that the zero hypothesis of no united root is rejected.
33
United States: The Zivot-Andrew Breakpoint test on the trend and intercept for the oil price returns measured in US Dollar between 26-08-1994 and 30-06-2014. The more negative the value of the Dickey-Fuller test the more likely that the zero hypothesis of no united root is rejected.
34 Appendix 4: Conditional variance equation of the GARCH(1,1)-M model for industries in East-Asian countries using daily excess return data between 26-08-1994 and 30-06-2014 (equation (2)).
Industry ω α β HL
Oil & Gas Producers 0.000*** 0.062*** 0.923*** 0.001 45.862 Oil Equipment, 0.000*** 0.112*** 0.836*** 0.016*** 13.242 Alternative Energy 0.000*** 0.067*** 0.917*** 0.007** 49.163
Chemicals 0.000*** 0.071*** 0.903*** 0.001 26.311
Forestry & Paper 0.000*** 0.079*** 0.892*** 0.003** 23.553 Industrial Metal & Mining 0.000*** 0.077*** 0.903*** 0.001 34.310
Mining 0.000*** 0.073*** 0.909*** 0.001 38.161
Construction & Materials 0.000*** 0.061*** 0.919*** 0.000 36.134 Aerospace & Defence 0.000*** 0.072*** 0.911*** 0.000 40.426 General Industrials 0.000*** 0.064*** 0.917*** 0.001 36.134 Electronic & Electrical Equipment 0.000*** 0.060*** 0.933*** 0.000 98.674 Industrial Engineering 0.000*** 0.082*** 0.906*** 0.001 57.415 Industrial Transportation 0.000*** 0.082*** 0.871*** 0.007*** 14.398 Support Services 0.000*** 0.072*** 0.918*** 0.002 68.968 Automobiles & Parts 0.000*** 0.079*** 0.896*** 0.001 27.378
Beverages 0.000*** 0.067*** 0.911*** 0.001 31.159
Food Producers 0.000*** 0.048*** 0.935*** 0.000 40.426 Household Goods & Home Construction 0.000*** 0.036*** 0.958*** 0.000 115.178 Leisure Goods 0.000*** 0.082*** 0.909*** 0.001 76.669 Personal Goods 0.000*** 0.052*** 0.933*** 0.000 45.862
Tobacco 0.000*** 0.038*** 0.950*** 0.001 57.415
Healthcare Equipment & Services 0.000*** 0.071*** 0.909*** 0.003*** 34.310 Pharmaceuticals & Biotechnology 0.000*** 0.083*** 0.881*** 0.001 18.905 Food & Drug Retailers 0.000*** 0.062*** 0.920*** 0.001 38.161
Media 0.000* 0.038*** 0.957*** 0.001 138.283
Travel & Leasure 0.000*** 0.051*** 0.935*** 0.001*** 49.163 Fixed Line Telecommunications 0.000*** 0.070*** 0.908*** 0.003** 31.159 Mobile Telecommunications 0.000*** 0.045*** 0.937*** 0.001 38.161
Electricity 0.000*** 0.044*** 0.941*** 0.001 45.862
Gas, Water & Multiutilities 0.000*** 0.052*** 0.947*** 0.000 692.801
Banks 0.000*** 0.063*** 0.915*** 0.001 31.159
Nonlife Insurance 0.000*** 0.052*** 0.943*** 0.001 138.283 Life Insurance 0.000*** 0.065*** 0.919*** 0.000 42.974 Real Estate Investment & Services 0.000*** 0.076*** 0.902*** 0.001 31.159 Financial Services 0.000*** 0.068*** 0.919*** 0.001 52.972 Nonequity Investment Instruments 0.000*** 0.053*** 0.938*** 0.002 76.669 Software & Computer Services 0.000*** 0.058*** 0.933*** 0.001 76.669 Technology Hardware & Equipment 0.000*** 0.089*** 0.888*** 0.001 29.789
