| Summary: | 博士 === 國立中興大學 === 應用經濟學系所 === 99 === World crude oil prices have undergone a rapid increase since 2004, and several important factors have contributed to the current level of high prices, including a rapid growth in demand by Asian countries, declining spare production and refinery capacity, supply disruptions due to political events, relatively low levels of crude oil and petroleum inventories in the U.S., a weak dollar, and increasing speculative activities in the oil market. When oil price first touched $100 per barrel in February 20, 2008 and soon reached a new historic peak at $145.31 per barrel in July 03, 2008, almost everyone was concerned about how long the high oil price would sustain? Until now, there is still lack of studies on solving this issue. Therefore, the main purpose of this study is to investigate the behavior of oil price and its effects on energy substitution and stock market.
(1) The first purpose of this study is to answer how long the high oil price would sustain? This study adopts panel SURADF unit root test and half-life based on ARMA (1, 1) to investigate the behavior of oil price (Section II). Evidences show that oil prices robust exist mean reversion and the speed of oil price adjustment is about 4~5 years.
(2) Next, I would like to estimate the effect of substitution between energy and non-energy. This study employs panel threshold regression to investigate the substitution relationship between energy and non-energy inputs (Section III). The results of the threshold test find that the relationship between energy and non-energy inputs exist structural change but no threshold effect among non-energy inputs in Taiwan’s case.
(3) Finally, I will devote some space to the discussion of the impact of oil price change on stock market. This study uses so-called interactive double dummy variables approach to re-investigate the relationship between oil price changes and stock returns (Section IV). The empirical results indicate that when the stock market is exhibiting a bull trend, a rise in the oil price will not affect stock returns in all selected OECD countries.
(4) Furthermore, this study provides several important contributions in the study of energy and stock market as follows: (i) this study is the first research that adopts panel unit root test and the half-life formula based on ARMA(1,1) model using high-frequency data to investigate the behavior of oil price. (ii) this study is the first research that employs panel threshold regression to investigate the substitution relationship between energy and non-energy inputs. (iii) this study is the first research that incorporates the stock market trend into the discussion of the relationship between oil price changes and stock market returns. (iv) Overall, because oil price exhibits slow mean reversion process, authority should not only take the impact of oil price change into account in forming its short-run policy but also develop a long-run plan in industrial policy, energy policy and investment strategy.
|
|---|
