Master's Thesis, 2012
47 Pages, Grade: 80
2 Literature review
3 Theoretical foundations
6 Empirical results
A3 Autocorrelation tests
A4 Newey-West variance-covariance estimator
A5 Ensuring strict exogeneity in DOLS
Figure 1: Real oil prices and value of US dollar.
Figure 2: Summary demand and supply framework.
Figure 3: Global benchmarks since 2006.
Figure 4: Combined plot of all series.
Figure 5: Granger causality results.
Table 1: Results Augmented Dickey-Fuller test.
Table 2: Results Johansen LM test.
Table 3: Johansen rank test.
Table 4: DOLS estimates.
Table 5: Results Granger causality testing.
As an invoice currency the US dollar is frequently cited to cause changes in real oil prices through a demand and supply channel. The aim of this paper is to shed more light on whether the US dollar is driving oil prices. In order to do so, we incorporate the bivariate relationship in a broader demand and supply framework in which real oil prices are determined by demand and supply factors. We find cointegration among the variables involved and that the value of the US dollar is a significant part in the long-run relationship. The causal links within the long-run relationship are examined using the procedure suggested by Toda & Yamamoto (1995) to test for Granger causality in the presence of I(1) variables. The results show that no causal effect of the US dollar on oil prices can be found. This contradicts the view that the US dollar is driving real oil prices. There is even evidence that none of the fundamentals is causing real oil prices. Moreover, real oil prices are found to have an indirect causal effect on the US dollar. This contradicts standard models such as the Krugman (1983) model which suggest a direct link.
Crude oil is one of the most important sources of energy and therefore a key element of our worldwide economy. The constant increases in oil prices in recent years led to an intense public debate about the drivers of oil prices. In this debate, the value of the US dollar is frequently cited as a significant contributing factor to changes in oil prices. This conclusion is often drawn by comparing the series of oil prices and the value of the dollar which indicates an inverse comovement (see Figure 1 in appendix A1). Clearly, this is only a casual observation and more profound statistical evidence is needed. The rationale behind the argument that the US dollar is driving oil prices relies on the fact that crude oil is priced in US dollars. Through this property as an invoice currency, the US dollar is supposed to be a demand and supply factor of crude oil. Thus, the US dollar is supposed to cause changes in oil prices through a demand and supply channel. However, this direction of causality is not obvious as there is theoretical and empirical evidence which suggests that oil prices are driving the value of the US dollar and not vice versa. The aim of this paper is to shed more light on whether the US dollar is driving oil prices through a demand and supply channel. In order to do so, we incorporate this bivariate relationship in a broader demand and supply framework. Hence, we include other variables which are considered as demand and supply factors besides the US dollar. Based on this framework, we examine whether a causality running from the value of the US dollar to oil prices is supported by the data. In order to do so, we first test whether a long-run relationship between oil prices and the fundamentals exists using a cointegration test. Second, we estimate the long-run parameters and assess whether the value of the US dollar is a significant part of it. Third, we examine the causal links within this long-run relationship using a Granger causality test. This is a necessary step because a cointegrating relationship only indicates long-run comovements of the variables. It does not tell us which variables cause the other in the system to create the long-run equilibrium.
The paper is structured as follows: In section 2, a review of the existing literature about the relationship between the US dollar and oil prices is presented. In section 3, the economic theory for the specification of the demand and supply model is presented. In addition, explanations for causal effects of oil prices are outlined. The concrete measurement of the demand and supply factors and the choice of the sample period are discussed in the data part in section 4. In section 5, the econometric methods to test for the existence of the long-run relationship, to estimate its parameters and to conduct inference on them are outlined. Moreover, a Granger causality test is presented. In section 6 the empirical results are discussed and section 7 concludes.
The 3 works of Krichene (2006), Cheng (2008) and He et al. (2010) incorporate the US dollar as a demand and supply factor in a structural demand and supply model. A reduced form expression for the oil price serves as the theoretical basis for an expected long-run relationship. They test for this long-run relationship by using cointegration tests. Because of the use of a demand and supply model, these works are coined “demand and supply literature” in this paper.
Demand and supply literature Krichene (2006) supposes that crude oil prices are cointegrated with the value of the US dollar and world interest rates. The value of the US dollar is measured as the nominal effective exchange rate. The world interest rates are measured by the US three-month treasury bill rate. The IMF crude oil price index is used as a measure for oil prices. The interest rates and the oil price index are in nominal terms. The IMF oil price index and the nominal effective exchange rate are transformed into logs. The sample period ranges from 1974 until 2004. Using monthly data, Krichene finds a cointegrating relationship among the variables with the oil price as dependent variable. The coefficient estimate of the value of the dollar in the cointegrating relationship is -7.76 and significant. So an appreciation of the US dollar is accompanied by a decrease in oil prices. The estimate of the coefficient of world interest rates is positive (0.28). However, Krichene provides no causality tests to assess whether the fundamentals are causing oil prices.
Cheng's (2008) demand and supply model suggests a long-run relationship between oil prices and four demand and supply factors. The first is the value of the US dollar which is measured by the IMF nominal effective exchange rate. World industrial production is included to capture the effects of world economic activity. World real interest rates are included as a supply factor to capture the effect of interest rates on intertemporal supply decisions and inventory holdings. Interest rates are measured by the federal funds rate which is deflated by the US CPI. As a last factor, inventory holdings are included and measured by the inventories of the OECD countries provided by the OECD. All variables are in logs except for the interest rates. The data is on a monthly basis and the sample period is January 1985 until February 2008. Upon finding cointegration, Cheng uses the Dynamic OLS estimator of Stock & Watson (1993) to estimate the long-run parameters. He finds that a 1% increase in the value of the US dollar is associated with a decrease in oil prices of 1.43%. However, again no causality tests are conducted.
He et al. (2010) derive a long-run relationship in which real oil prices are determined by global real economic activity and the value of the US dollar. Real oil prices are measured using WTI future prices with one month maturity sourced from the US Energy Information Administration (EIA). The nominal prices are deflated by the US CPI. The value of the US dollar is captured by the trade weighted nominal effective exchange rate of the US dollar including major currencies provided by the US FED. The real global economic activity is measured by the Kilian index. The sample period is January 1988 until December 2007. Using Johansen's (1988) rank test, they find one cointegrating vector among the variables. However, the result from the trace statistic is only significant on the 10% level. A second cointegrating vector is clearly rejected. Using the Johansen maximum likelihood estimator of Johansen (1988), they estimate the long-run relationship. They cannot reject the null that the coefficient of the value of the dollar is insignificant at the 5% significance level. The long-run coefficient estimate of the real economic activity is 1.7. As they find real economic activity and the value of the US dollar to be weakly exogenous, they also estimate a general autoregressive distributed lag (ADL) model. After stepwise simplification of this ADL model, they obtain the embedded long-run solution using OLS estimation. In this long-run solution, the estimate of the real economic is 0.98 and the one of the value of the dollar is -0.89. Because of the results of the weak exogeneity tests, He et al. conclude that the two fundamentals are Granger causing oil prices but are not caused by oil prices. However, the fact that real economic activity and the US dollar are not responding to disequilibrium does not imply that no Granger causality is running from oil prices to these to variables. In a vector error correction model the disequilibrium error is only one of the two channels through which a variable can be Granger caused by another variable. If a variable is weakly exogenous, it can still be caused by lagged differences of the other variables (Granger 1988, p. 203). Thus, these causality results have to be taken with a grain of salt.
Real exchange-rate literature There is another strand of literature that investigates the relationship between the real exchange rate of the US dollar and oil prices (see Amano & Norden (1995), Amano & Norden (1998), Benassy-Quere et al. (2007), Chen & Chen (2007) and Coudert et al. (2008)). The real exchange rate is usually measured as the real effective exchange rate which is a weighted average of bilateral real exchange rates using CPIs of the respective countries. Those papers are mainly interested in assessing whether real oil prices are the single long-run determinant of real exchange rates and in particular of the US dollar real exchange rate. Therefore these papers are summarised under the name “real exchange rate literature”. Several different theoretical foundations are given in this literature. For example, Amano & Norden (1995) use a model of a small open economy to show that oil prices might influence the real exchange rate through causing changes in the terms of trade. A similar terms of trade model is used in Chen & Chen (2007). Overall, they find that real oil prices are cointegrated with the real effective exchange rate of the US dollar. Except for Chen & Chen (2007), who use a panel cointegration test, standard Johansen rank tests are used to test for cointegration. Amano & Norden (1995), Amano & Norden (1998), Benassy-Quere et al. (2007) and Coudert et al. (2008) find a positive long-run relationship and that the causality is running from real oil prices to the US dollar real exchange rate. Chen & Chen (2007) are not conducting causality tests. In this paper, we use the concept of the nominal exchange rate as this is the natural way to capture the value of the US dollar. However, the real exchange rate is not simply the nominal exchange rate in real terms but a different economic concept. The nominal exchange rate refers to the value of one currency in terms of other currencies. On the contrary, the real exchange rate refers to the relative price of a representative basket of goods of one country in terms of a foreign basket making it to a measure of international competitiveness. Therefore, the theoretical explanations for a link between oil prices and the value of the US dollar of these papers are only of limited connection. However, their empirical results provide evidence against the result that the US dollar is a demand and supply factor driving oil prices. On the contrary, these results favour the view that the opposite causality is the case.
The contribution of this paper to the existing literature consists of several elements. First, a recent data set is used which incorporates the time period from 2008 until 2011 that is not covered so far. Second, the paper combines the choice of variables, estimation techniques, and theoretical reasoning of different works. Third, a solid (Granger) causality test procedure is employed to shed more light on the causal links involved. So far, no (convincing) causality testing has been conducted in the demand and supply literature.
In this section, we outline the demand and supply model of the crude oil market in which we can incorporate the US dollar as demand and supply factor. It is necessary to consider the US dollar in a broader demand and supply framework in order not to confound any effects of the US dollar with those of other fundamentals. The modelling strategy builds up on the work of He et al. (2010). We first work with general forms and specify the variables of the system more concretely later.
Demand and supply framework The demand equation can be written as:
Abbildung in dieser Leseprobe nicht enthalten
is a column vector compromising factors that determine the demand for crude oil and is the price of crude oil (in real terms) which, clearly, also influences the demand. In a -space, are demand curve shifters. We assume a well behaved downward sloping demand function so is expected to be negative. captures random demand shocks such as precautionary purchases if uncertainties about future oil supply arise. Similarly, one can obtain an expression for the supply of crude oil:
Abbildung in dieser Leseprobe nicht enthalten
is a column vector of the factors that determine the supply of oil, which are supply shifters in -space. Assuming a standard upward sloping supply curve, is positive. incorporates random supply shocks such as production shortfalls because of wars in the middle east. Let us assume that the market clears, , so that one can solve (1) and (2) for which gives the reduced form expression:
Abbildung in dieser Leseprobe nicht enthalten
where captures demand and supply shocks. As is derived by the assumption that the market is cleared, it is the market clearing price. If there are changes in supply or demand factors a new market clearing price settles down. Graphically, the crossing point of the demand and supply curves changes as the two curves shift inwards or outwards. Thus, the demand and supply model suggests a long-run relationship between the oil price and the fundamentals and market clearing is the mechanism that connects changes in the fundamentals with the market price.
Specifying the system In the remainder of this section, we pin down the demand and supply shifters in (3). Which variables to include, depends on the specifications of (1) and (2) based on theoretical reasoning. The emphasis in this section lies on the underlying economics. Hence, a discussion of the concrete measures of the variables will be postponed until the data section. Let us start with the demand equation which is specified as:
Abbildung in dieser Leseprobe nicht enthalten
The variable stands for real worldwide economic activity. The idea is that global economic expansions and slumps influence the demand for crude oil. Thus, the demand for crude oil is influenced by the global business cycle. should be positive as global economic expansions should increase the demand for crude oil.
The value of the US dollar is denoted as . Oil transactions are conducted primarily in US dollar making the dollar to an invoice currency. Through this invoice currency channel, changes in the value of the dollar can have effects on the demand and supply for oil. These demand and supply effects are the theoretical reason to suggest that the value of the US dollar is driving oil prices. Let us first focus on the potential demand side effects. If the dollar appreciates against other currencies, oil becomes more expensive in the domestic currency of oil importing countries outside the US. If the number of these countries is large enough, worldwide demand for oil considerably falls (He et al. 2010, p. 869). Clearly, the real price of oil influences negatively the demand for oil.
Let us now turn to the specification of the supply equation:
Abbildung in dieser Leseprobe nicht enthalten
The first variable are world real interest rates . They have an impact on crude oil supply by influencing the intertemporal supply decisions and the propensity of oil companies to carry inventories. Starting from an optimal intertemporal extraction path, higher real interest rates increase the incentives for extracting more today rather than tomorrow. One can think of the incentive to change the rates at which oil is pumped (Frankel 2006, p. 5). Also, the higher real interest rates the higher the opportunity cost of holding oil inventories (Frankel 2006, p. 5). Thus, firms want to reduce inventories which brings more crude oil on the market. The resulting higher supply in both cases would decrease oil prices.
As mentioned, the value of the US dollar is also a supply factor. The reason for this is again the property of the US dollar as an invoice currency. This is because changes in the value of the US dollar alter the ratio between output prices in US dollar and input prices in domestic currencies (Coudert et al. 2008). If the US dollar appreciates, the oil price producer outside the US obtain is higher in their domestic currency. This gives an incentive to increase production given unchanged input costs.
Figure 2 in appendix A1 summarises graphically the demand and supply framework. If we use as above the market clearing condition, we can combine the specified demand and supply equations and solve for to get:
Abbildung in dieser Leseprobe nicht enthalten
The demand and supply model suggests that there is a long-run relationship of the form (6) between oil prices and demand and supply factors. Hence, we should observe the long-run value of oil prices to be a function of the values of the fundamentals. Moreover, the demand and supply model suggests that the demand and supply factors have causal effects on oil prices. Thus, the fundamentals induce oil prices to change. If the US dollar is one of these causal fundamentals, it would imply that the US dollar is driving oil prices.
Impact of real oil prices However , as mentioned in the introduction, there are also theoretical approaches that support the view that real oil prices have causal effects on the US dollar. We consider the model of Krugman (1983) here. In his model, there are three countries: The US and Germany as oil importers with their currencies dollar and mark and the OPEC representing oil producers without an own currency. Consider an exogenous increase in the oil price. Higher oil prices shift wealth from the oil importers to the oil exporters as the elasticity of demand of the importers is assumed to be smaller than one. What the OPEC is doing with this additional wealth, is one factor influencing the exchange rate. The more American relative to German assets and goods they purchase the more the dollar is appreciating. On the other hand, the relative burden of the higher oil imports is essential. The higher this relative burden the more the currency of the respective country weakens. Krugman assumes that the spending behaviour of the OPEC depends on the time horizon. Shortly after the oil price increase, they invest mostly in assets as they are prudent and wait for prices to be stabilised at the higher level. Once they stabilise, they switch from asset purchases to goods purchases. Hence, in the short-run, the asset preferences are important and in the long-run the goods preferences matter. Let us assume for simplicity that the relative additional oil import burden is more or less the same. So, all that counts for the development of the exchange rate of the dollar against the mark are the asset and goods preferences of the OPEC. One plausible case, also suggested by Krugman, is that the OPEC has asset preferences biased towards the US and goods preferences which are stronger for German products. In this case, an increase in oil prices would increase the dollar in the short run and decrease in the long-run. But also other causal links can be constructed.
In addition, a theoretical case for a causality running from real oil prices to real economic activity can be made. Higher oil prices might slow down economic expansions due to higher input costs for various parts of the economy such as higher energy cost or transportation cost. Conversely, lower oil prices might stimulate economic expansions.
In the methodology section, we examine how to test for the suggested long-run relationship in (6) and how to estimate the long-run parameter. Moreover, we outline a causality test to assess the causal links within the long-run relationship. But, firstly, all aspects associated with the data used are presented.
There are four variables involved in our putative long-run relationship, and in this section we outline how to measure those variables.
Real world oil prices As crude oil is not a homogenous good, we have a set of varieties of crude oil. This is because there are various types of crude oil with different qualities and characteristics (Fattouh 2011, p. 20 ff.). Fortunately, there are not independent oil prices. There are reference prices for a few benchmark crude oil types, and the rest of the types are priced at a differential to these benchmark prices according to quality and their relative supply and demand (Fattouh 2011, p. 21). Moreover, if a benchmark is strongly related to the other benchmarks, it can be considered as a representative oil price for global oil prices. Thus, in order to generalise our results to global oil prices, it is necessary to work with an appropriate global oil price benchmark in our empirical analysis. For the whole sample period under consideration, there were two candidates as useful global benchmark prices. The first benchmark is the price of the crude oil type Western Texas Intermediate (WTI) produced in the US. The second is the price of Brent Blend which is a mixture of several types produced in the North Sea including the type Brent. In recent years, the WTI showed disconnections from other global benchmarks as can be seen in Figure 3 below. In the graph, we also include the price of the Dubai Crude which is extracted in Dubai and is the third most important global benchmark today.
Abbildung in dieser Leseprobe nicht enthalten
Figure 3: Global benchmarks since 2006.
Since 2007, the movement of the price of WTI is slightly different than the one of Brent Blend and Dubai Crude. The divergence becomes particularly apparent from spring 2011 onwards. These disconnections are due to the fact that the WTI price can be strongly influenced by local conditions (Fattouh 2011, p. 59). The logistical infrastructure and US specific demand developments can lead to WTI price movements unrelated from those of other benchmarks. For instance, build-up inventories of WTI cannot be shipped away. Brent Blend, however, is exportable which makes it less exposed to European specific demand conditions and oversupply. Moreover, 70% of international trade in oil is directly or indirectly priced using the price of Brent Blend (Fattouh 2011, p. 36). Thus, first, a lot of oil prices are a closely connected to the Brent Blend price. Second, also other global benchmarks such as the Dubai Crude are closely and in a constant manner connected to it. Therefore, we use the price of Brent Blend as a measure for world oil prices .
There are basically two different types of prices of the Brent Blend that we could use; spot and future prices. Unlike the futures market where prices are observable in real time, the prices on the spot market “at the terminals” must be identified. These assessments are carried out by oil pricing reporting agencies (PRA) (Fattouh 2011, p. 30 ff.). Their methodology publications reveal that they analyse closed deals as well as bids and offers reported by the market participants to determine the spot price. As changes in risk preferences over time might influence our results when using future oil prices, we use spot prices. The data for the Brent Blend spot prices is obtained from the IMF on a monthly basis. The units are in US dollars per barrel (159 litres). The series was deflated by the US CPI from the US Bureau of Labor Statistics with the base period set to 1982-1984.
Value of the US dollar The value of the US dollar, , is captured by the nominal effective dollar exchange rate on a monthly basis provided by the US FED. It is constructed as a weighted average of bilateral nominal exchange rates between the dollar and seven major currencies. The weights are trade based. Thus the more intense the trade links with the US the more the respective bilateral exchange rate is accounted for. In our measure, there are only two oil producer’s currencies involved (the UK and Canada). According to data from the EIA, their average share in world oil production over the sample period under consideration is only 0.07%. Thus, the accompanying supply side effects of changes in the US dollar outlined before should not be captured to a large extent in our measure of the value of the US dollar. On the contrary, the demand effect of changes in the value of the dollar should be well captured. The countries whose currencies are used account for 43.3% of world crude oil demand for the sample period based on data from the EIA.
Global real economic activity The global real economic activity is measured by the Kilian index due to Kilian (2009) which can be obtained from Kilian’s homepage. It is based on monthly growth rates of freight rates for bulk cargoes which are deflated by the US CPI. The idea behind this measure is that ocean freight rates should be closely correlated with worldwide real economic activity. This is because global expansions drive the demand for inputs factors which are transported over sea. Thus, the stronger the economic expansion the higher the growth rates of freight rates. The measure has several advantages over worldwide GDP. First, it is available on a monthly basis. Second, the index is a direct measure of the real economic activity as no exchange rate or country weighting is necessary (Kilian 2009, p. 1056). The index is detrended to get rid of technological advances in shipbuilding. This makes the measure compatible to serve as a measure of business cycles around a long-term trend.
World real interest rates As a measure for global real interest rates, , we use the effective federal funds rate provided by the US FED on a monthly basis. The federal funds rate is an average of the interest rates at which US banks lend money to each other in order to fulfil the reserve requirements of the FED. Hence, banks with surplus balances with the FED lend to banks that need larger balances. The prefix “effective” means that this measure is based on the actual observed rates not the targets set by the FED. The effective federal funds rate is a crucial determinant of short and long-term interest rates in the US and US interest rates can be considered as the main determinants of interest rates in world capital markets (Krichene 2006, p. 7). The nominal values are deflated by the same CPI measure as above.
In addition, except for the federal funds rate and the Kilian index the variables are in logs to eliminate possible non-linearities.
Sample period The chosen sample period for the analysis ranges from January 1988 to December 2011. The end is dictated by the availability of the Kilian index. In order to see why the beginning was chosen to be 1988 some historical facts are needed. In the 1950s and 1960s, prices were set by multinational oil companies. In the period of 1973-1988, the OPEC took over the dominant position of the multinationals with their own national oil companies. The spot market of the time before the mid-1980s consisted of a small number of transactions usually done only under distressed conditions (Fattouh 2011, p. 18). In the 1980s, more and more oil fields were discovered outside the OPEC countries which meant a significant amount of crude oil reaching international markets. The new supply was offered at prices below OPEC prices in the spot market. This led to a steady decline in the OPEC market share. Eventually, the OPEC fully gave up their administered price setting in 1988. In a transition from 1986-1988, a meaningful spot market evolved and oil prices emerged in a much more competitive environment. Even though the theoretical foundation of a long-run relationship between the oil price and fundamentals does not hinge on the assumption of a perfect competitive environment, the two types of cartels (multinationals and OPEC) were so pronounced that any close connection between their price setting decisions and demand and supply factors is hard to establish on theoretical grounds. From a practical point of view, we do not possess data about the relevant prices to test for a long-run relationship between real oil prices and demand and supply factors. This is because the relevant prices are the quoted prices in the long-term contracts. If any, those prices might have been a function of world demand and supply factors. But, these prices are not available. The spot market prices of the time before 1988 cannot be used as they were not connected to fundamental developments. Therefore, the sample period excludes the time period before 1988. As we use monthly data, we still have a considerable large sample size of 288 observations.
 The rationale to treat the US Dollar as such a demand and supply factor is outlined in detail in section 3.
 Krichene writes about using a VAR approach to estimate the cointegrating relationship. Probably, the maximum likelihood estimator of Johansen (1988) is meant with this but no source is indicated.
 Conceptually, this last factor is confusing. It is not a factor which influences the unobservable supply but rather a (crude) proxy for supply.
 The Kilian index is presented later in the data section.
 Unfortunately, they do not report any hypothesis tests of the significance of the long-run parameter estimates.
 Despite conceptual differences, from a pure quantitative view the nominal and real exchange rate are still strongly correlated.
 To avoid any confusion, a demand and supply factor is considered as a “fundamental” in this paper.
 Throughout, we use the expressions “oil price” in singular and “oil prices” in plural interchangeable. This is admissible even though there is no single world oil price. In the data section, we investigate the pricing system of crude oil and show that there are a few benchmark prices which can be seen as representative for all oil prices. If we refer here to a single oil price we refer to such a benchmark price. Especially in this section, we frequently use the term in singular as we consider a single market setup.
 Note that the specification in (1) assumes that the demand function is linear and additive.
 This is because a market situation with excess demand cannot be sustained as it implies long-run positive profits. Conversely, over supply cannot prevail as supply which is not purchased or stored at non-prohibitive costs means losses. Note that the statement that the market clears completely is based on the assumption of a perfect competitive market and more precisely zero-profits in the long-run.
 Clearly, this is a simplification as no market in real life resembles such a market structure with perfect competition and market clearing. These assumptions are made to keep the theoretical framework simple. We could substitute the link between fundamentals and the market price by some other mechanism without losing the theoretical argument for a long-run relationship between the oil price and the fundamentals. At least as long as this mechanism relates changes in the supply and demand factors to the price.
 This effect does not take place if the currency of the importing country’s currency is pegged to the dollar. However, the currencies against which the dollar is evaluated in our measure of the value of the dollar are all not pegged to the dollar. So we can ignore this point.
 For a similar approach see Golub (1983).
 Krugman does not analyse the origins of these increases.
 According to the Australian National Bank, there are more than 160 crude oil types traded internationally.
 This also allows us to use the singular (oil price) and plural (oil prices) interchangeable.
 Today it is a mixture of four types (Brent, Forties, Oseberg und Ekofisk).
 Both are vulnerable to undersupply due to disruptions in the logistical system. However, in the recent past, this was more a problem for the WTI than for the Brent Blend system.
 All of our results are robust to changing the measure from Brent Blend to WTI.
 The spot price is the price a buyer has to pay “on the spot” when he buys crude oil at one of the terminals where Brent Blend can be loaded, for example in Sullom Voe. The future price, on the contrary, is the price charged for receiving a certain amount of crude oil at some future point in time and these contracts can be traded.
 The data was also cross-checked with the Brent Blend spot prices from the EIA. There were only negligible differences. As it is not published from which source these spot prices stem from it could be that the differences are due to the fact that they are from different PRAs.
 The currencies are: euro, Canadian dollar, yen, British pound, Swiss franc, Australian dollar and Swedish krona.
 There are broader measures of the value of the dollar including more currencies. However, these measures are not used in the literature and their use gave implausible results. An extensive investigation of how the measures were calculated would be necessary. However, this is beyond the scope of this paper.
 In the remainder of the text we sometimes simply use the short hand “dollar” to refer to the value of the US dollar for brevity.
 Already in 1986 the cartel was breaking apart, but the final end was in 1988.
 The prices were kept secret between the transaction parties.
 Spot market prices in the period 1973-1988 are found to be stationary (on the 10% level), whereas the demand and supply factors are I(1). This precludes any long-run relationship.
Master's Thesis, 70 Pages
Seminar Paper, 26 Pages
Pre-University Paper, 18 Pages
Master's Thesis, 82 Pages
Research Paper (undergraduate), 75 Pages
Doctoral Thesis / Dissertation, 237 Pages
Bachelor Thesis, 61 Pages
Term Paper (Advanced seminar), 33 Pages
Research Paper (postgraduate), 9 Pages
Research Paper (postgraduate), 25 Pages
GRIN Publishing, located in Munich, Germany, has specialized since its foundation in 1998 in the publication of academic ebooks and books. The publishing website GRIN.com offer students, graduates and university professors the ideal platform for the presentation of scientific papers, such as research projects, theses, dissertations, and academic essays to a wide audience.
Free Publication of your term paper, essay, interpretation, bachelor's thesis, master's thesis, dissertation or textbook - upload now!