P
HYSICAL
M
ARKET
D
ETERMINANTS
OF THE
P
RICE OF
C
RUDE
O
IL
AND THE
M
ARKET
P
REMIUM
GUILLAUME CHEVILLON CHRISTINE RIFFLART e de R echer che / 20 11 0 71 41 0 GROUPE ESSEC
centre de recherche / RESEARCH CENTER
AVENUE BERNARD HIRSCH BP 50105 CERGY
95021 CERGY PONTOISE CEDEX
essec business school.
établissements privés d’enseignement supérieur, association loi 1901,
accréditéS aacsb international - the association
Pour tous renseignements :
• Centre de Recherche/Research Center
Tél. 33 (0)1 34 43 30 91 research.center@essec.fr
• Visitez notre site www.essec.fr
DR 07020
Physical Market Determinants of the Price of Crude Oil and
the Market Premium
Guillaume Chevillon Christine Rifflart
ESSEC Business School, Sciences-Po Center for
Economic Research (OFCE)
and and
University of Oxford National Political Science
Foundation
June 2007
Abstract
We analyze the physical, i.e. non financial, determinants of the real price of crude oil by means of an equilibrium correction model over the last two decades. We find that two cointegrating relations affect the change in prices: one refers to OPEC’s cartel behavior attempting to control prices using its market power and quotas; the other to the coverage rate of expected future demand by OECD using inventory behaviours. We derive an equation for the change in oil prices which we use to assess the speculative elements of the early millennium price hike. We show that worries alien to the physical markets are the causes of the increase in oil prices and are able to quantify their impact.
Keywords: Cointegration - Forecast - Market Premium - Oil Price
JEL Codes: Q40, C53
Résumé
Cet article analyse les déterminants physiques (i.e. non financiers) du cours réel du baril de pétrole brut à l’aide d’un modèle à correction d’équilibre estimé sur les deux dernières décennies. Nous montrons que deux relations de cointégration affectent les variations du cours : l’une est reliée aux tentatives de l’OPEP de peser sur les cours à l’aide de son pouvoir de marché et de ses quotas de production ; l’autre lie le taux de couverture de la demande future des pays de l’OCDE aux comportements des stocks de brut. Nous obtenons une équation régissant la variation des cours et nous l’utilisons pour analyser les éléments spéculatifs de l’explosion des cours au début de ce millénaire. Nous montrons que des inquiétudes étrangères aux marchés physiques sont à l’origine de l’augmentation des cours et nous sommes à même d’en quantifier l’impact.
Mots-clefs: Cointégration - Cours du pétrole - Prévision - Prime de marché
Physical Market Determinants of the Price of Crude Oil and
the Market Premium
Guillaume Chevillon ∗ Christine Rifflart
ESSEC, Paris, Sciences-Po Center for Economic Research (OFCE)
and University of Oxford. National Political Science Foundation
June 2007
Abstract
We analyze the physical, i.e. non financial, determinants of the real price of crude oil by means of an equilibrium correction model over the last two decades. We find that two cointe-grating relations affect the change in prices: one refers to OPEC’s cartel behavior attempting to control prices using its market power and quotas; the other to the coverage rate of expected future demand by OECD using inventory behaviours. We derive an equation for the change in oil prices which we use to assess the speculative elements of the early millenium price hike. We show that worries alien to the physical markets are the causes of the increase in oil prices and are able to quantify their impact.
∗Corresponding author. E-mail: chevillon@essec.fr; address: Decision Sciences Department, ESSEC Business
1
Introduction
The aim of this paper is to analyze the determinants of real oil prices inphysical—i.e. non financial—markets and provide an equilibrium correction model (eqcm) thereof. We then derive a corresponding market price which, when compared to the nominal price, allows to derive the speculative—or non physical—premium.
Our choice of model is deliberate: we develop a model that relates the price of a barrel of crude oil to the situation of physical markets and discard any specula-tive elements. This allows us to quantify the latter as a difference between mar-ket clearing—based on the historical data generating process—and nominal prices. Other routes have been pursued by various authors who mostly analyzed risk premia through option markets or who focused on modelling price volatility rather than its level. Inter alia, Barros Lu´ıs (2001) developed a risk aversion measure which he used to assess the risk premium implicit in future contracts. Sadorsky (2002) considered the opportunity to hold oil options when maximizing a portfolio’s return and used an ARMAX-ARCH model to forecast that return. As for volatility estimation, Kuper (2002) used a GARCH framework for estimating markets’ historical uncertainty via price ranges.
Our decision to focus in this paper on linear models turns out quite rare in this line of research. Most of the existent literature that uses linear models tends to model the interaction between oil price and economic activity in oil importing or exporting countries. For instance, Hamilton (2000) uses the output gap as a basis for assessing the disruptive effects of oil price hikes on agents’ behaviors. Amano & van Norden (1998) exhibited a strong relation between real oil prices, exchange rates
and terms of trade in major oil importing economies. Also important is our wish to focus on determinants of market prices, not on economic elements presiding over the physical supply side, namely the level of proven reserves, itself function of wells where extraction is economically sound, which in turns depends on the price of oil. The reader is referred e.g. to Farzin (2000) for such an approach.
Surprisingly, and testimony to the difficulty of the task at hand, few authors have focused on crude oil price determination and forecasting. Philip K. Verleger (1982) was probably the first, but in his time, he could but simply focus on the link between official and spot (Rotterdam) prices. Among the recent papers, Tang & Hammoudeh (2002) use a model whereby OPEC attempts to control prices with a target range: they establish the presence of a non-linear relation between production quotas, inventory levels (or stocks), supply and demand. Unfortunately, their model brings no forecasting power. To our knowledge, Krichene (2002) and Yang et al.
(2002) are among the very few who applied an equilibrium correction model on oil demand and supply. The analysis by Krichene focuses on the whole of the 20th Century: the author derives price elasticities which can unfortunately be of no use for our analysis restricted to the last two decades. As for Yang et al., they first estimated historic volatilities in the U.S. which, in turn, they add to eqcm models for energy commodities such as coal, natural gas and crude oil. In the forecasting literature, Ye et al (2005) notably derive an easy to implement short-term forecasting model using seasonal effects and inventories.
The plan of this paper is as follows. We first present the main historic features of crude oil markets in section 2. They, together with economic model for
commodi-ties, preside over our choice of variables for our model of price determination; this constitute the purpose of section 3. In section 4 we conduct a VAR and cointegra-tion analysis. We assess, there, the validity of the weak exogeneity assumpcointegra-tions that allows us to focus, in section 5, on a univariate model. We also derive a speculative premium. Finally, section 6 concludes our analysis.
2
Global oil market features
Over the last two decades, crude oil prices have mainly been determined by market mechanisms, yet this is in an imperfectly competitive market. As with any tradable good, the price of oil follows the relative evolutions of supply and demand, decreasing when the former abounds, increasing when the market situation tightens. However, oil is no standard commodity: it is marred by political concerns in an oligopolistic market dominated by a strong cartel, the Organization of the Petroleum Exporting Countries (OPEC).
Worldwide demand still originates chiefly from OECD economies: according to the data that the International Energy Agency (IEA) provides, they represent 60% of world consumption. But since the mid-1990s, emerging economies, especially China and India, have seen their consumptions surge. In 2004, China became the second biggest consumer worldwide: behind the United States but before Japan. This rapid growth seems the driving force behind the recent rise in the world’s energy demand, with an increase in demand from developing Asian economies amounting to two third of the total increase in energy consumption.
the intensity of petroleum derivatives in the GDP and on the rate of growth of the latter. Price evolutions matter very little here: short-run price elasticity of demand for crude oil is near zero as there no substitutes to its use that are readily available. This is not true for long-term fluctuations though. Following the two oil shocks in the 1970s, industrialized economies have developed policies aimed at reducing dependence on energy and diversification away from oil. Yet, the decrease in prices which started in 1985 softened the pressure and reduced the effort. Nowadays, according to most estimates—and the IEA in particular—the price elasticity of long term world demand is around 0.5, provided that long term has a meaning for a non-renewable source of energy which could deplete rapidly.
On the supply side, the oil sector is still oligopolistic: it is organized around a few global private companies and the OPEC cartel. Although it controls only 40% of pro-duction worldwide, OPEC has until recently played a crucial role on price formation owing to its policy for regulating supply. Available reserves and spare productive ca-pacities are chiefly located in the Middle Eastern OPEC countries. Originally created, in 1960, to defend the interests of its members, the cartel has progressively adapted its objectives: moving away from a policy of sustained high prices in the 1970s and 80s to a more consensual policy taking consumer countries into account. Hence, any OPEC decision to increase or reduce production quotas may lower or raise the price of crude oil.
Following a sharp reduction in market prices for crude oil in 1986, a new wave of producers has progressively emerged, in particular Former Soviet Union countries (FSU). This has let the price of a barrel of North Sea Brent stable at around US$
18-19 until 1997, a price then regarded as mostly fair by both OPEC and consumers. Yet, following the Asian crisis and the fall in demand, oil prices collapsed to about US$ 10 per barrel by the end of 1998. Hurt by this experience, OPEC decided to increase transparency. It set up a target band of US$ 22 to 28 a barrel, based on the OPEC basket of produced crude oil. In parallel, quota policy became very reactive so as to follow the variations in global demand and prevent any excess supply or lack thereof. A group of producer countries, amongst which Norway, Mexico and Russia, started cooperating with OPEC in its attempt to control production. Since 2004, however, OPEC seems to have lost its capacity to control prices: their evolution has been disconnected from relative supply and demand. This has taken place in a period of geopolitical tensions in the Middle East and global concerns for the long term production capacities. Yet, oil remains a highly strategic commodity as it is the main source of energy that sustains economic growth. OPEC owns today over 77% of the world’s proven oil reserves.
What are the main features affecting prices? OPEC’s behavior is crucial since it represents about 55% of global oil exports. Since production in this industry is very capital intensive, supply is rigid in the short-term: non OPEC producers tend to work on full capacity and cannot adjust to take advantage of higher prices. Consumption is affected by seasonal factors, especially in the Northern Hemisphere where winter are cold and summers hot. Hence, consumption of oil derivatives increases in the first and fourth quarters of each year. This is anticipated by building refinery inventories of crude oil and petroleum products in the third and fourth quarters. Speculations and unusual weather conditions hence generate ample short term volatility.
3
Determinants of the price of crude oil
To derive an econometric model for the evolution of oil prices, we need to understand what factors may intervene. Oil is a commodity which is traded in an imperfectly competitive market, with few suppliers and many consumers. This oligopolistic struc-ture implies that the price does not simply react to demand and supply as with many exhaustible commodities. Indeed, according to Hotelling (1931), the price—net of the constant marginal extraction cost—in a perfectly competitive market should be in line with the interest rate, i.e. the discount rate that stabilizes the value of future receipts so that the sacrifice of future generations who will not consume this resource is taken into account. This simple law does not apply for oil since research originated in the 1970s showed that the market is far from perfect (see inter alia Ulph and Folie (1980)).
There exist many theoretical models for the determination of world supply, de-mand and prices of oil, many of which originated from Nordhaus (1973) (see Choukri (1979) for an early review of the pre-1979 shock theories). Following Griffin (1985), we believe in a model consisting of a mix of cartel and competitive market practices for the level of production. The latter should have a longer term impact than cartel implications. Given that models of target revenue or property rights could be rejected on empirical grounds by Griffin and that non-OPEC producers have increased their significance since the mid-1980s, we do not consider these types of models.
The econometric model we derive below takes the form of an equilibrium (or error) correction model for the physical market price of crude oil. This model hence combines both shorter and longer-run effects. The rationale for their interaction stems from
Hubbard (1986) who show how intertwined they must be.
3.1
What price?
In our attempt to relate the quarterly price of crude oil to the situation of the physical market, we need to choose the relevant variables. First, what price to analyze? There exist several references in world exchanges: North Sea Brent, West Texas Intermediate (WTI) and Dubai. Each refers to an oil of high quality with a specific production or trading location. The WTI and Dubai prices are chiefly traded in the United States and Asia,whereas the North Sea Brent is often used as the world reference. In the London-based ICE Futures exchange (formerly known as the International Petroleum Exchange, or IPE), the Brent is used to specify the price of two third of crude oil exchanged worldwide. In this paper, we therefore focus on the spot price of a barrel of Brent, i.e. the price negotiated for immediate delivery on the physical market. Although the different prices do not fluctuate perfectly in line over time, they more or less follow one another.
The transform from nominal price to real price—the variable of interest for an analysis of demand and supply—is made using the GDP deflator for the G7 group of countries1. The resulting real price P
t is therefore related to the need for oil in the
production process of industrialized economies.
1Ideally, we would have used the deflator for all OECD economies, i.e. the main consumers of
oil. Owing to the lack of data for recent periods, we used the deflator for the G7. The latter closely follows the OECD deflator for the available periods.
3.2
Supply and demand
Clearing of the physical market would normally lead to a price being fixed so that demand equals supply. Unfortunately, the market for crude oil is imperfect. Global supply originates from the non-OPEC countries, which produce at full capacity at all prices, and from OPEC, where production is regulated to maintain the price of a basket of various types of crude oil within a target band. We use, here, the demand for oil from the OECD (DOECDt ), which has historically been the driving force behind market evolutions, together with non OECD demand (Dtnon OECD) and inventories (where at a given time the change in inventories—or stocks—is equal to the difference between supply and demand) both in the OECD (SOECD
t ) and outside (Stnon OECD).
3.3
OPEC behavior
Until the mid-1990s, the OPEC cartel of producers had succeeded in controlling— although somewhat loosely—world prices of crude oil via managing an important fraction of world production. Although their control has weakened, we attempt to analyze whether they still constitute a market force. In order to analyze their behav-ior, we resort to two distinct variables: one is their official production quota (Qt),
the other is their nominal price target. Until 2000, this nominal price target was implicit. It was established to achieve domestic social needs, allow economic develop-ment and prevent the expansion of other oil producers. Because the marginal cost of production is lower for OPEC than for other producers, they possess a loose capacity to gear prices downwards. Following the 1998 collapse in prices, OPEC decided to bring more transparency to their decisions. At the beginning of 2000, they decided to
announce a target price at the end of their meetings and they delineated a rule which would play automatically on production quotas, although the latter has never been put in place. Reference prices are based on a basket of seven types of crude oil that member countries produce. This basket price was at the time expected to fluctuate between US$ 22 and 28 per barrel.
The world price of a barrel of Brent remained at around US$ 18 until March 2000, shifting then to US$ 25. To reflect OPEC’s acknowledgement of their loss of marketing power, we set the nominal target to the actual price from the third quarter of 2004 onwards (we use the OPEC bulletin as a reference). The real counterpart of the nominal target (using the same deflator as previously), is denoted by Pt∗.
3.4
Graphic analysis
We present, on figure 1, quarterly observations of the natural logarithm of the vari-ables described above over 1989(1)–2005(4), consisting of 68 observations2. As noted by Geroski, Ulph and Ulph (1987) we need take account of the changes in pricing conduct. We therefore decided to restrict our analysis to the post-1987 era since the market experienced major changes at that period; in particular, market forces began to gain significance.
Four variables are clearly trending: stocks and demands, with seasonal patterns exhibited by demands—although non OECD is less regular—and less obviously by OECD stocks. Non OECD stocks are much smoother. This reflects the variable’s
2All computations were carried out using PcGive and PcGets (see Hendry & Doornik, 2001, and
Hendry & Krolzig, 2001). Tails probabilities and p-values for theχ2 distributions were obtained using GiveWin (see Doornik & Hendry, 2001).
1990 1995 2000 2005 2.5 3.0 3.5 4.0 pt 1990 1995 2000 2005 0.2 0.4 usdt 1990 1995 2000 2005 8.20 8.25 8.30 8.35 sOECDt 1990 1995 2000 2005 7.50 7.75 8.00 8.25 snon OECDt 1990 1995 2000 2005 8.2 8.3 8.4 dOECDt 1990 1995 2000 2005 7.7 7.8 7.9 8.0 dtnon OECD 1990 1995 2000 2005 10.0 10.2 qt 1990 1995 2000 2005 3.0 3.5 4.0 pt *
Figure 1: Plots of the variables entering the analysis.
larger growth: exp(.75) corresponds to an increase of about 100% compared to 20 to 30% for the other three variables.
Quotas and prices also exhibit increasing patterns over the sample, especially since 2002 for the latter. We also present usdt, a variable representing the exchange rate
of the US dollar with respect to a basket of currencies (using the weights from the IMF’s Special Drawing Rights), as we will use this variable to expand our models.
usdt looks more mean reverting than the others, but it shows long departures from its overall sample mean between 1995 and 2005.
4
Cointegration analysis
4.1
A VAR representation
Since we chose to restrict our attention to a sample consisting of 68 observations, we are constrained in our analysis and cannot resort to general unrestricted VAR modeling. Yet this does not prevent analysis on sub-versions of our model. First, we analyze the variables presented previously by means of a VAR(1) where we include an unrestricted constant. A graphic analysis of the fit is presented on figures 2 and 3 where we discarded the use of non OECD stocks as this variable does not seem to improve the estimation but lowers the precision (via increased variance of the estimators). Although it enters the VAR, we do not present here the results for p∗t to make the figures more readable. Despite a few large outliers, the fit is reasonable—although the lack of seasonal patterns for the fitted OECD demand would require additional fourth-order lags as regressors—and we conclude that the variables presented here constitute a reasonablycongruent (see Hendry, 1997) representation of the data generating process. In the analyses below, we will include indicator variables to account for the beginning of the first Iraq war (I90q3 and I90q4) and the beginning of the second Iraq war (I03q2). When performing a battery of specification tests for the VAR(1) representation, we notice that the data exhibit residual autocorrelation. Adding fourth-order lags of the endogenous variables correct for this, but for reasons of parsimony, we cannot carry out our cointegration analysis on a VAR(4). We will therefore restrict to the use of these lags to the analysis of the variables mapped to a stationary representation.
1990 2000 2.5 3.0 3.5 4.0 pt Fitted 3.0 3.5 4.0 2.5 3.0 3.5 4.0 pt× Fitted 1990 2000 −2 0 2 residuals: r pt (scaled) −2.5 0.0 2.5 0.1 0.2 0.3 0.4 0.5 Density r:pt N(0,1) 1990 2000 9.9 10.0 10.1 10.2 10.3 qt Fitted 10.0 10.2 9.9 10.0 10.1 10.2 qt× Fitted 1990 2000 −2 0 2 residuals: r qt (scaled) −2.5 0.0 2.5 0.2 0.4 0.6 Density r:qt N(0,1)
Figure 2: Fit of a VAR(1) model on the five variables
1990 2000 8.20 8.25 8.30 8.35 stOECD Fitted 8.20 8.25 8.30 8.20 8.25 8.30 8.35 stOECD× Fitted 1990 2000 −2 0 2 r:stOECD (scaled) −2.5 0.0 2.5 0.2 0.4 Density r:stOECD N(0,1) 1990 2000 8.2 8.3 8.4 dt OECD Fitted 8.2 8.3 8.4 8.2 8.3 8.4 dt OECD× Fitted 1990 2000 −2 0 2 r:dtOECD (scaled) −2.5 0.0 2.5 0.2 0.4 Density r:dtOECD N(0,1) 1990 2000 7.7 7.8 7.9 8.0 8.1
dtnon OECD Fitted
7.7 7.8 7.9 8.0 8.1 7.7
7.8 7.9 8.0 dt
non OECD× Fitted
1990 2000
−2 0 2 r:dt
non OECD (scaled)
−2.5 0.0 2.5 5.0 0.2
0.4
Density
r:dtnon OECD N(0,1)
Figure 3: Fit of a VAR(1) model on the five variables
We are now ready to perform cointegration tests and we use Johansen’s trace test (see Johansen, 1991 and 1996) which leads to rejecting the presence of fewer than one stationary relation (at a p-value lower than 1%) and to accepting the existence of two cointegrating vectors.
Cointegration test over the sample 1989(2)-2005(4)
rank log-likelihood eigenvalues Trace Statistic Prob.
0 828.0655 – 180.84 [0.000] ** 1 859.9639 0.61411 117.04 [0.001] ** 2 886.8835 0.55227 63.204 [0.150] 3 900.8864 0.34164 35.198 [0.442] 4 910.1124 0.24073 16.746 [0.666] 5 916.2178 0.16661 4.5354 [0.852] 6 918.2215 0.058059 0.52792 [0.467]
We also test the same model when allowing also for a linear trend to enter the cointegrating space. This leads to the same conclusion of a cointegration rank of 2. Restrictions of the coefficients of the cointegrating vectors allow for an identification of the stationary relations:
c1,t = pt−p∗t + 0.63qt (1)
c2,t = sOECDt −d OECD
t + 0.00172t (2)
these vectors are presented—demeaned with superscript µ—on figure 4. Their inter-pretation is that (c1) OPEC attempts to control the difference between the observed
price and its target by modifying its quota; (c2) there is a link between OECD stocks
1990 1995 2000 2005 −0.25 0.00 0.25 0.50 c 1, t µ 1990 1995 2000 2005 −0.05 0.00 0.05 c2, tµ
Figure 4: Cointegrating vectors corrected for the mean. cµ1 links OPEC quotas to the difference between price and target. cµ2 relates OECD stocks and demand, and the continous decrease in their ratio.
properly later—and their difference (in logs) has been decreasing at a constant rate over the last fifteen years, hence the necessity to add a linear trend for stationarity. Relations between the variables entering the cointegrating relations are presented on figure 5.
4.2
Weak exogeneity
Given the low number of observations at our disposal, we would like to pursue a univariate analysis of the determinants of the real price of crude oil. For this to be possible, we need to ascertain that we are entitled to model the distribution of the
1990 1995 2000 2005 2.5 3.0 3.5 4.0 pt pt* 1990 1995 2000 2005 8.1 8.2 8.3 stOECD 0.0017*t dOECDt
variable of interest without having to consider the marginal distribution of the condi-tioning variables. This can be done provided that the regressors areweakly exogenous
for the parameters of the conditional regression, see Engle et al. (1983). In addition to parameters being variation free (which is clear here), we need to assess that the loading factors—or impact coefficients, see Johansen (1996)—of the cointegrating vec-torc1 be zero for all the equations of the VAR, but that governingpt (and necessarily p∗t given its definition from 2004 on). We proceed to a Likelihood ratio test for these restrictions. The log-likelihood of the unrestricted model islU R = 867.048; imposing 5
impact factors to be zero leads tolR= 865.668.The statistic is then 2(lU R−lR) = 2.76
which we need comparing to the critical values of aχ2[5] : this leads to a p-value of 74%. The same test, but not restricting the loading factor ofqt leads to a p-value of
60%. It seems difficult to consider that OPEC quotas are weakly exogenous for the setting of prices, but we yet consider that we can marginally disregard the process governing the behavior of OPEC. This therefore allows for a univariate analysis of price determination, which we will do in the next section.
5
A univariate analysis
5.1
An equilibrium correction model
Under weak exogeneity of the regressors for the parameters of the conditional regres-sion, we proceed to univariate modelling of the change in oil price. With the help of the variables presented above. We also create an “expected demand coverage rate” variable in order to assess whether the current level of stocks is sufficient to satisfy
expected demand for the next period given the highly seasonal pattern of the lat-ter. This variable is constructed as the ratio of current crude oil stocks in the OECD countries over expected OECD demand for the next quarter, computed as the current value to which we add the previous year’s corresponding quarter-to-quarter variation. This is thus defined asdT+1|T =ET [dT+1] following the linear rule:
ET [dT+1] =dT +L4∆dT+1 =dT + ∆dT−3, (3)
where dt is the log of the OECD demand Dt. The expected demand coverage rate
hence becomes:
CRt= St Dt+1|t
, (4)
where St is the current level of OECD stocks. This new variable closely mimics cµ2,t
but with a forward looking element.
The resulting model for the physical determination of crude oil prices is as follows (standard errors in brackets):
d ∆pt = − 0.309 (0.125) crt− 2.11 (0.454) ∆snon OECD t + 4.53 (1.35) ∆dnon OECD t−1 − 3.29 (0.813) ∆2dnon OECD t−1 − 0.203 (0.0596) cµ1,t−1− 2.28 (0.444) ∆4cµ2,t−1+ 1.31 (0.479) cµ2,t−1 + 0.34 (0.0892) I90q3t+ 0.528 (0.991) I90q4t− 0.259 (0.0858) I01q4t− 0.36 (0.0878) I03q2t (5)
with estimation analysis;
RSS= 0.3462, σˆ= 0.0816, R2 = 0.742, R¯2 = 0.693,
lnL= 163.920, AIC= −4.855, HQ= −4.707, SC= −4.480, T = 63, p= 11, FpNull= 0.0000, FpConst = 0.0000; and misspecification tests:
FAR(1−4) 1.074 [0.38]
FARCH(1−4) 0.686 [0.61]
χ2
normality 0.125 [0.94]
χ2hetero 22.96 [0.19]
Thus this model satisfies almost all the specification tests, has a ¯R2 of about 70%,
and we can deem it congruent. Figures 6 and 7 present visual representations of the fit and of the distribution of the residuals.
Now the model (5) shows that the serial autocorrelation exhibited in the VAR(1) can be corrected by the fourth order difference of the cointegrating vector c2,t. It
could have also been a simple consequence of the many outliers. We, hence, added an indicator variable for the post-September 11, 2001 trauma. According to our equation, the price is a decreasing function of current levels of OECD stocks via crt.
Changes in inventories also play a role, either as non OECD or via annual changes in
c2,t−1. Increasing non OECD demand also drives prices up but in a non-linear fashion
(with diminishing impact as the acceleration becomes larger). OPEC also achieves its aims via an equilibrium correction feature involvingc1,t−1.
The model presented above in (5) constitutes a valid representation of the deter-mination for the price of crude oil according to the physical market. This accounts
1990 1995 2000 2005 −0.25 0.00 0.25 0.50 actual fitted −0.4 −0.2 0.0 0.2 0.4 −0.25 0.00 0.25 0.50 fitted ×∆pt 1990 1995 2000 2005 −2 −1 0 1 2 residuals (normalized) 1990 1995 2000 2005 1 2 3 4
squared residuals (normalized)
Figure 6: Univariate modeling of the changes in oil prices via an equilibrium correction mechanism. Graphs above present the actual and fitted values, and the graphs below record the normalized residuals.
1 5 9 −0.5
0.0 0.5
1.0 Correlogram ACF−∆pt PACF−∆pt
0.0 0.5 1.0 0.05 0.10 0.15 0.20 Spectral density ∆pt −40 −30 −20 −10 0 10 20 30 40 0.01 0.02 0.03 0.04 Density ∆pt N(s=11) −2 −1 0 1 2 −2 −1 0 1 2 QQ plot ∆p t× normal
Figure 7: Analysis of residual distribution from fitting an eqcm to the change in oil prices. The figure presents estimates of the ACF and PACF (panel (a)), spectral den-sity (b),density (c) and QQ-plot against the Normal. The four panels are consistent with the residual being Normal white noise.
1990 1995 2000 2005 2.5 3.0 3.5 4.0 pt p ^t 2.50 2.75 3.00 3.25 3.50 3.75 4.00 2.5 3.0 3.5 4.0 pt× ^pt 1990 1995 2000 2005 −0.1 0.0 0.1 pt− ^pt −0.15 −0.10 −0.05 0.00 0.05 0.10 0.15 2.5 3.0 3.5 4.0 pt× (pt− ^pt)
Figure 8: Implications of the eqcm for the level of oil prices, and corresponding forecast errors (panel (c)).
for about 70% of price variation, leaving the remaining 30% to financial markets and unexpected innovations. Figure 8 presents the corresponding forecasts for the price level.
5.2
Alternative specifications
To prevent overfitting our model given the low number of observations, we attempted to determine which variable account for most of the fit, via a smaller version thereof. This is presented in (6) with a fit of ¯R2 = 63% but specification tests lead to rejecting
the hypothesis of non-autocorrelated errors with a p-value of 0.7%. This not need be a problem if we follow Hendry and Krolzig (2005) who hold that when performing
multiple, sayn, specification tests at a nominal sizeη, the probability of rejecting at lest one is 1-(1−η)n≈nη so that congruency tests should be carried at the 1% level to avoid almost one-hundred percent chance of rejection and yet some rejection is to be expected. This smaller model (6) which excludes non-linear patterns therefore constitutes a satisfactory yet non-congruent representation.
g ∆pt = − 0.488 (0.115) crt− 1.91 (0.426) ∆snon OECD t − 1.44 (0.348) ∆4cµ2,t−1− 0.23 (0.0639) cµ1,t−1 + 0.411 (0.0648) (I90q3t+ I90q4t)− 0.292 (0.0657) (I01q4t+ I03q2t) (6)
We have also tried non linear models whereby the model is of the general form
∆pt=vt−1 −α(pt−1−wt−1) + X z γzLkzt+εt ! ,
for some linear combination wt of the z regressors and with vt a measure of historic
volatility (scaled by its geometric mean so that the product of past vt is unity) but
this model does not lead to a better fit.
Additionally, as our interest lies in observing how the price relates to its physical demand, we tried to compare its value to that of its currency of denomination. For this purpose, we construct an exchange rate for the United States dollar with respect to a basket of currencies, namely the Euro, the Pound Sterling and the Japanese Yen, where the weights are chosen to follow those of the respective currencies in the valuation of the IMF Special Drawing Rights (SDRs). But this variable seemed not to matter either.
5.3
Contributions
Here we need to compute the contributions of the different variables via (5). Figure 9 records contribution to changes in pt since 2000q1 owed to the various regressors.
Because we deal with an eqcm, we need to take full account of the dynamics: here we use 2000q1 as a reference observation and compute the dynamic impact of the regressors. We use the fact that the eqcm takes the form
∆pt=−α(pt−1−wt−1) +
X
z
γzLkzt (7)
for some linear combination wt of the z regressors which we disregard first. Re-write
(7) as
pt = (1−α)Lpt+αwt−1+
X
z
γzLkzt
so that the total dynamic impact of one of thez regressors onto pt is
+∞ X i=0 ((1−α)L)iγzLkzt=γzL k +∞ X i=0 (1−α)iLizt
so that we define the impact on ∆pt as
C∆p(z, t) = γzL k τ X i=0 (1−α)iLi∆zt=γz(1 +L)L k τ X i=0 (1−α)iLizt
for some τ . Cumulating the Cz(i), starting from some observation of reference T0,
provides the cumulated dynamic impact of z onp overT0, ..., tas
Cp(z, t) = t−T0 X
i=1
C∆p(z, T0+i)
which we compare topt−pT0.For the impact of the cointegrating relation, we simply
report the cumulative sum of−α(pi−wi) fori=T0, ..., t.
1. 2000-2001: after a rise in 2000, the price of oil were driven downwards, in the next year, by market forces. Following the collapse of oil price in 1997-98, a very sharp cut in OPEC production quotas from July 1998 to April 1999 was aimed to dry the market out and gear prices to the cartel’s target band. The nominal Brent price jumped back to US$ 26.8 at the beginning of 2000 (our reference point)—having fallen to US$ 11 at the end of 1998—and shot up to $ 30.4 in the third quarter of that year. In this context, OPEC began to soften its policy in 2000. Oil production was allowed to increase quickly. OECD inventories nevertheless remained at low levels. Not until 2001 did supply tensions (with respect to demand) progressively disappear. But a slowdown in demand in 2001 alleviated the pressure on prices, leading OPEC to react and cut dramatically its production quotas. The coverage rate of stocks recovered rapidly in 2001, thus explaining the negative contribution to the change in oil price attributed to OECD supply and demand in 2000 and, particularly, in 2001. Prices declined sharply to $ 19.3 at the end of 2001 in the September, 11 aftermath. This is captured in our equation by an indicator variable whose effect does not need presenting in figure 9.
2. 2002-2003: increasing non OECD demand and low non OECD stocks, combined with a lowering OECD cr led to an increase in oil prices. At this stage, OPEC stayed firm and maintained quotas at a very low level to prevent excess supply in 2002. Then, in the first half of 2003, OPEC became much more accommodative. During the period preceding intervention of the US troops in Iraq—starting on March, 20—and throughout the active war, price volatility was high. This
explains the presence of the indicator variable in the 2003q2. Indeed, with the combination of a break in Iraqi oil production and of a stronger world demand, markets feared a shortage. To reassure consumers and soften prices, OPEC largely loosened its quotas, and Saudi Arabia committed to satisfy demand. Yet, prices ended up stabilizing at a relatively high level.
3. 2004-2005: a strong disconnection emerged, prices exploding while all contri-butions would have driven them in the opposite direction. The situation on the physical market was improving. Despite a demand shock in 2004, supply stayed above demand. Producers hence managed to accumulate inventories and increase their coverage of expected future consumption. In 2005, a demand slowdown met rising production quotas, the latter even reached all time highs. Yet, despite vanishing physical tensions and an accommodative OPEC policy, markets did not calm and the price did not decrease.
5.4
Market premium
The analysis above enables us to derive the premium on oil prices with respect to physical markets determination. This is done using the difference between the current price and the cumulated residuals from estimation starting with a reference observa-tion, here 2000q1. It leads to the graph presented on figure 10 where we notice that the premium has been exploding since 2004. Notice as presented on figure 8 that the premium we derive is different from the residuals from the computed level of prices. Here, the idea is to construct a dynamic forecast analysis with respect to the physical elements of the market. We see that most of the price increase observed in 2004 and
2000 2001 2002 2003 2004 2005 2006 −0.4 −0.2 0.0 0.2 0.4 0.6
OPEC quota
S&D OECD
S&D non OECD
p
Figure 9: Cumulated dynamic contributions of theeqcm regressors to the changes in the log of real oil prices. All series are presented as the variation since 2000q1. S&D represents the combined effect of supply and demand.
2005 can be attributed to non market forces. The premium, thus computed, increased by up to US$ 20 between the ends of 2001 and 2005.
Several factors can explain this. The demand shock in 2004 highlights the inad-equacy of the supply structure both on the upstream market and the downstream market. Over several years, investment was weak. Shortages of some specific prod-ucts appeared in 2004. Spare capacities had never been so tight. Moreover, despite the slowdown in non OECD demand, combined worldwide demand growth was above its recent trend. At the same time, several disruptions were affecting supply from non OPEC countries. Although markets were largely supplied, concerns were rising about the difficulties faced by suppliers in coping with such a dynamic demand. The geopolitical context fed these fears. Besides a worsening Iraqi situation, there were spreading tensions in the Middle East, conflicts in Nigeria, political instability in Ecuador, Venezuela and Russia. Expectations concerning the risk of a global short-age rose, leading to a rise in prices, increased speculation and high volatility. In this context, the debate on the depletion of world oil reserves emerged. According to some pessimistic studies, the world approaches the all time maximum of oil production, from which prices increase very quickly.
Neither the improvements in the situation of the physical market or an easing OPEC policy could calm markets. Volatility was extreme in 2004 and 2005, fostered by a high level of financial liquidity and speculative activity. As a consequence, the cumulated sum of residuals increased away from zero: this constituted a premium on the physical market price of oil.
-10 0 10 20 30 40 50 60 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 actual dynamic estimate premium
Figure 10: Dynamic forecast of the nominal crude oil price and associated market premium
6
Conclusions
In this article, we derived an econometric model for the determination of the real price of a barrel of crude oil. This model is based entirely on physical aspects of the market situation. Two cointegrating relations matter here, relating to OPEC’s attempts at controlling prices and a tendency—also witnessed in other industries—towards a reduction in inventories. Using a dynamic forecast derived from the equilibrium correction mechanism, we were able to compute a resulting premium on the non-physical elements; we call it a market premium since it embodies elements which are not part of historic physical market behavior. This premium accounts for about half the increase in oil price between 2000 and the end of 2005.
Our model also allows for a decomposition of the various contributions of the regressors to the change in prices. We can thus observe what factors are driving the evolutions at any point in time. We observed that between 2004 and 2005, the increase in market price was indeed unrelated to the physical situation, whereas in the two years before, the increase was caused by emerging market (i.e. non OECD) demand and a tightening of OPEC quotas.
In this paper, we restricted our analysis to the post-1988 era, when oil prices started being driven by market forces and we did not include aspects which markets only started to focus on, such as spare production capacity and “the end of oil”. As with any econometric modeling, we need a sample that is large enough for any pattern to be estimated. But these issues will have to be estimated when enough data is available.
Unfortu-nately, periodicity is an issue. Here, our model is quarterly and we could not have increased the frequency. Most models based on financial aspects focus on higher fre-quencies, and it is not sure whether the influences still matter from a quarter to the other.
References
Amano, R. A., and van Norden, S. (1998). Exchange rates and oil prices. Review of International Economics,6, 683–694.
Choukri, N. (1979). Analytical specifications of the world oil market: A review and comparison of twelve models. Journal of Conflict Resolution, 23, 346–72. Engle, R. F., Hendry, D. F., and Richard, J. F. (1983). Exogeneity. Econometrica,
51, 277–304.
Farzin, Y. H. (2000). The impact of oil price on additions to U.S. proven reserves. Working paper, Department of Agricultural and Resource Economics, Univer-sity of California, Davis.
Geroski, P. A., Ulph, A. M., and Ulph, D. T. (1987). A model of the crude oil market in which market conduct varies. Economic Journal, 97, S77–86.
Gil-Alana, L. A. (2001). A fractionally integrated model with a mean shift for the US and the UK real oil prices. Economic Modelling, 18, 643–658.
Griffin, J. M. (1985). OPEC behavior: A test of alternative hypotheses. American Economic Review, 75, 954–63.
Hendry, D. F., and Doornik, J. A. (1996). Empirical Econometric Modelling using PcGive for Windows. London: International Thomson Business Press.
Hendry, D. F., and Krolzig, H. M. (2005). The properties of automatic gets modelling.
Economic Journal,115, C32–C61.
Hendry, D. F. (1997). On congruent econometric relations: A comment. Carnegie-Rochester Conference Series on Public Policy, 47, 163–90.
Hendry, D., and Krolzig, H.-M. (eds.)(2001).Automatic Econometric Model Selection using PcGets. London: Timberlake Consultants Press.
Hotelling, H. (1931). The economics of exhaustible resources. Journal of Political Economy, 39, 137–55.
Hubbard, R. G. (1986). Supply shocks and price adjustment in the world oil market.
Quarterly Journal of Economics, 101, 85–102.
Johansen, S. (1991). Estimation and hypothesis testing of cointegration vectors in Gaussian vector autoregressive models. Econometrica, 59, 1551–80.
Johansen, S. (1996). Likelihood-based inference in cointegrated vector auto-regressive models. Oxford: Oxford University Press.
Krichene, N. (2002). World crude oil and natural gas: a demand and supply model.
Energy Economics, 24, 557–576.
Kuper, G. H. (2002). Measuring oil price volatility. Working paper, Department of Economics & SOM, University of Groningen.
Lu´ıs, J. B. (2001). The estimation of risk-premium implicit in oil prices. OPEC review, 25, 221–260.
Nordhaus, W. B. (1973). The allocation of energy resources. Brookings Papers on Economic Activity,4, 529–76.
Philip K. Verleger, J. (1982). The determinants of official OPEC crude prices.Review of Economics and Statistics,64, 177–83.
Sadorsky, P. (2002). Time-varying risk premium in petroleum futures prices. Energy Economics, 24, 539–556.
Tang, L., and Hammoudeh, S. (2002). An empirical exploration of the world oil price under the target zone model. Energy Economics, 24, 539–556.
Ulph, A. M., and Folie, G. M. (1980). Exhaustible resources and cartels: An intertem-poral Nash-Cournot model. Canadian Journal of Economics, 13, 645–58. Wickham, P. (1996). Volatility of oil prices. Working paper, International Monetary
Fund.
Yang, C. W., Hwang, M. J., and Huang, B. N. (2002). An analysis of factors affecting price volatility of the US oil market. Energy Economics, 24, 107–119.
ESSEC
CENTRE DE RECHERCHE
LISTE DES DOCUMENTS DE RECHERCHE DU CENTRE DE RECHERCHE DE L’ESSEC
(Pour se procurer ces documents, s’adresser au CENTRE DE RECHERCHE DE L’ESSEC)
LISTE OF ESSEC RESEARCH CENTER WORKING PAPERS
(Contact the ESSEC RESEARCH CENTER for information on how to obtain copies of these papers) RESEARCH.CENTER@ESSEC.FR
2004
04001 BESANCENOT Damien, VRANCEANU Radu
Excessive Liability Dollarization in a Simple Signaling Model
04002 ALFANDARI Laurent
Choice Rules Size Constraints for Multiple Criteria Decision Making
04003 BOURGUIGNON Annick, JENKINS Alan
Management Accounting Change and the Construction of Coherence in Organisations: a Case Study
04004 CHARLETY Patricia, FAGART Marie-Cécile, SOUAM Saïd
Real Market Concentration through Partial Acquisitions
04005 CHOFFRAY Jean-Marie
La révolution Internet
04006 BARONI Michel, BARTHELEMY Fabrice, MOKRANE Mahdi
The Paris Residential Market: Driving Factors and Market Behaviour 1973-2001
04007 BARONI Michel, BARTHELEMY Fabrice, MOKRANE Mahdi
Physical Real Estate: A Paris Repeat Sales Residential Index
04008 BESANCENOT Damien, VRANCEANU Radu
The Information Limit to Honest Managerial Behavior
04009 BIZET Bernard
Public Property Privatization in France
04010 BIZET Bernard
Real Estate Taxation and Local Tax Policies in France
04011 CONTENSOU François
Legal Profit-Sharing: Shifting the Tax Burden in a Dual Economy
04012 CHAU Minh, CONTENSOU François
Profit-Sharing as Tax Saving and Incentive Device
04013 REZZOUK Med
2005
05001 VRANCEANU Radu
The Ethical Dimension of Economic Choices
05002 BARONI Michel, BARTHELEMY Fabrice, MOKRANE Mahdi
A PCA Factor Repeat Sales Index (1973-2001) to Forecast Apartment Prices in Paris (France)
05003 ALFANDARI Laurent
Improved Approximation of the General Soft-Capacitated Facility Location Problem
05004 JENKINS Alan
Performance Appraisal Research: A Critical Review of Work on “the Social Context and Politics of Appraisal”
05005 BESANCENOT Damien, VRANCEANU Radu
Socially Efficient Managerial Dishonesty
05006 BOARI Mircea
Biology & Political Science. Foundational Issues of Political Biology
05007 BIBARD Laurent
Biologie et politique
05008 BESANCENOT Damien, VRANCEANU Radu
Le financement public du secteur de la défense, une source d'inefficacité ?
2006
06001 CAZAVAN-JENY Anne, JEANJEAN Thomas
Levels of Voluntary Disclosure in IPO prospectuses: An Empirical Analysis
06002 BARONI Michel, BARTHELEMY Fabrice, MOKRANE Mahdi
Monte Carlo Simulations versus DCF in Real Estate Portfolio Valuation
06003 BESANCENOT Damien, VRANCEANU Radu
Can Incentives for Research Harm Research? A Business Schools Tale
06004 FOURCANS André, VRANCEANU Radu
Is the ECB so Special? A Qualitative and Quantitative Analysis
06005 NAIDITCH Claire, VRANCEANU Radu
Transferts des migrants et offre de travail dans un modèle de signalisation
06006 MOTTIS Nicolas
Bologna: Far from a Model, Just a Process for a While…
06007 LAMBERT Brice
Ambiance Factors, Emotions and Web User Behavior: A Model Integrating and Affective and Symbolical Approach
06008 BATISTA Catia, POTIN Jacques
Stages of Diversification and Capital Accumulation in an Heckscher-Ohlin World, 1975-1995
06009 TARONDEAU Jean-Claude
Strategy and Organization Improving Organizational Learning
06010 TIXIER Daniel
Teaching Management of Market Driven Business Units Using Internet Based Business Games
06011 COEURDACIER Nicolas
06012 AVIAT Antonin, COEURDACIER Nicolas
The Geography of Trade in Goods and Asset Holdings
06013 COEURDACIER Nicolas, GUIBAUD Stéphane
International Portfolio Diversification Is Better Than You Think
06014 COEURDACIER Nicolas, GUIBAUD Stéphane
A Dynamic Equilibrium Model of Imperfectly Integrated Financial Markets
06015 DUAN Jin-Chuan, FULOP Andras
Estimating the Structural Credit Risk Model When Equity Prices Are Contaminated by Trading Noises
06016 FULOP Andras
Feedback Effects of Rating Downgrades
06017 LESCOURRET Laurence, ROBERT Christian Y.
Preferencing, Internalization and Inventory Position
06018 BOURGUIGNON Annick, SAULPIC Olivier, ZARLOWSKI Philippe
Management Accounting Change in the Public Sector: A French Case Study and a New Institutionalist Perspective
06019 de BEAUFORT Viviane
One Share – One Vote, le nouveau Saint Graal ?
06020 COEURDACIER Nicolas, MARTIN Philippe
The Geography of Asset Trade and the Euro: Insiders and Outsiders
06021 BESANCENOT Damien, HUYNH Kim, VRANCEANU Radu
The "Read or Write" Dilemma in Academic Production: A European Perspective
2007
07001 NAIDITCH Claire, VRANCEANU Radu
International Remittances and Residents' Labour Supply in a Signaling Model
07002 VIENS G., LEVESQUE K., CHAHWAKILIAN P., EL HASNAOUI A., GAUDILLAT A., NICOL G., CROUZIER C.
Évolution comparée de la consommation de médicaments dans 5 pays européens entre 2000 et 2004 : analyse de 7 classes pharmaco-thérapeutiques
07003 de BEAUFORT Viviane
La création d'entreprise au féminin dans le monde occidental
07004 BOARI Mircea
Rationalizing the Irrational. The Principle of Relative Maximization from Sociobiology to Economics and Its Implications for Ethics
07005 BIBARD Laurent
Sexualités et mondialisation
07006 VRANCEANU Radu
The Moral Layer of Contemporary Economics: A Virtue Ethics Perspective
07007 LORINO Philippe
Stylistic Creativity in the Utilization of Management Tools
07008 BARONI Michel, BARTHELEMY Fabrice, MOKRANE Mahdi
Optimal Holding Period for a Real Estate Portfolio
07009 de BEAUFORT Viviane
One Share - One Vote, the New Holy Graal?
07011 TIXIER Maud
Appreciation of the Sustainability of the Tourism Industry in Cyprus
07012 LORINO Philippe
Competence-based Competence Management: a Pragmatic and Interpretive Approach. The Case of a Telecommunications Company
07013 LORINO Philippe
Process Based Management and the Central Role of Dialogical Collective Activity in Organizational Learning. The Case of Work Safety in the Building Industry
07014 LORINO Philippe
The Instrumental Genesis of Collective Activity. The Case of an ERP Implementation in a Large Electricity Producer
07015 LORINO Philippe, GEHRKE Ingmar
Coupling Performance Measurement and Collective Activity: The Semiotic Function of Management Systems. A Case Study
07016 SALLEZ Alain
Urbaphobie et désir d'urbain, au péril de la ville
07017 de CARLO Laurence
The Classroom as a Potential Space - Teaching Negotiation through Paradox
07019 ESPOSITO VINZI Vincenzo
Capturing and Treating Unobserved Heterogeneity by Response Based Segmentation in PLS Path Modeling. A Comparison of Alternative Methods by Computational Experiments
07020 CHEVILLON Guillaume, Christine RIFFLART
Physical Market Determinants of the Price of Crude Oil and the Market Premium
07021 CHEVILLON Guillaume
Inference in the Presence of Stochastic and Deterministic Trends
07023 COLSON Aurélien
The Ambassador, between Light and Shade. The Emergence of Secrecy as the Norm of International Negotiation
07024 GOMEZ Marie-Léandre
A Bourdieusian Perspective on Strategizing
07025 BESANCENOT D., VRANCEANU R.
P
HYSICAL
M
ARKET
D
ETERMINANTS
OF THE
P
RICE OF
C
RUDE
O
IL
AND THE
M
ARKET
P
REMIUM
GUILLAUME CHEVILLON CHRISTINE RIFFLART e de R echer che / 20 11 0 71 41 0 GROUPE ESSEC
centre de recherche / RESEARCH CENTER
AVENUE BERNARD HIRSCH BP 50105 CERGY
95021 CERGY PONTOISE CEDEX
essec business school.
établissements privés d’enseignement supérieur, association loi 1901,
accréditéS aacsb international - the association
Pour tous renseignements :
• Centre de Recherche/Research Center
Tél. 33 (0)1 34 43 30 91 research.center@essec.fr
• Visitez notre site www.essec.fr
DR 07020
