Oil and Product Price Dynamics in International Petroleum Markets Alessandro Lanza, Matteo Manera and Massimo Giovannini NOTA DI LAVORO 81.2003 SEPTEMBER 2003 IEM – International Energy Markets Alessandro Lanza, Eni S.p.A., Roma, Fondazione Eni Enrico Mattei, Milano and CRENoS, Cagliari, Italy Matteo Manera, Department of Statistics, University of Milano-Bicocca, Italy and Fondazione Eni Enrico Mattei, Milano, Italy Massimo Giovannini, Fondazione Eni Enrico Mattei, Milano, Italy This paper can be downloaded without charge at: The Fondazione Eni Enrico Mattei Note di Lavoro Series Index: http://www.feem.it/web/activ/_wp.html Social Science Research Network Electronic Paper Collection: http://papers.ssrn.com/abstract_id=XXXXXX The opinions expressed in this paper do not necessarily reflect the position of Fondazione Eni Enrico Mattei
Transcript
Oil and Product Price Dynamics in International Petroleum Markets
Alessandro Lanza, Matteo Manera and Massimo Giovannini
NOTA DI LAVORO 81.2003
SEPTEMBER 2003 IEM – International Energy Markets
Alessandro Lanza, Eni S.p.A., Roma, Fondazione Eni Enrico Mattei, Milano and CRENoS, Cagliari, Italy
Matteo Manera, Department of Statistics, University of Milano-Bicocca, Italy and Fondazione Eni Enrico Mattei, Milano, Italy
Massimo Giovannini, Fondazione Eni Enrico Mattei, Milano, Italy
This paper can be downloaded without charge at:
The Fondazione Eni Enrico Mattei Note di Lavoro Series Index: http://www.feem.it/web/activ/_wp.html
Social Science Research Network Electronic Paper Collection:
http://papers.ssrn.com/abstract_id=XXXXXX
The opinions expressed in this paper do not necessarily reflect the position of Fondazione Eni Enrico Mattei
Oil and Product Price Dynamics in International Petroleum Markets Summary In this paper we investigate crude oil and products price dynamics. We present a comparison among ten price series of crude oils and fourteen price series of petroleum products, considering four distinct market areas (Mediterranean, North Western Europe, Latin America and North America) over the period 1994-2002. We provide first a complete analysis of crude oil and product price dynamics using cointegration and error correction models. Subsequently we use the error correction specification to predict crude oil prices over the horizon January 2002-June 2002.The main findings of the paper can be summarized as follows: a) differences in quality are crucial to understand the behaviour of crudes; b) prices of crude oils whose physical characteristics are more similar to the marker show the following regularities: b1) they converge more rapidly to the long-run equilibrium; b2) there is an almost monotonic relation between Mean Absolute Percentage Error values and crude quality, measured by API° gravity and sulphur concentration; c) the price of the marker is the driving variable of the crude price also in the short-run, irrespective of the specific geographical area and the quality of the crude under analysis. Keywords: Oil prices, Product prices, Error correction models, Forecasting JEL: C22, D40, E32 The authors would like to thank Luigi Buzzacchi, Marzio Galeotti, Margherita Grasso and Micheal McAleer for helpful comments. Roberto Asti and Cristiana Boschi provided excellent editorial assistance. This study does not necessarily reflect the views of Eni S.p.A.. Address for correspondence: Alessandro Lanza Fondazione Eni Enrico Mattei Corso Magenta, 63 20123 Milano Italy Phone: +39-02-52036930 Fax: +39-02-52036946 E-mail: [email protected]
1
Oil and Product Price Dynamics in International
Petroleum Markets
(Revised: 31 July 2003)
1. Introduction
Over the last 30 years, oil prices have been closely scrutinized by applied economic
literature. Literally hundreds of applied research and policy studies have examined the
role played by oil prices in determining economic growth or inflation rates, both in
developed and developing countries.
Recently, several studies have contributed to this literature by examining the relation
between the price of crude oil and refinery products. If we exclude the specialized
literature, however, much less attention has been given to understanding the price
dynamics for different crudes, even if the quality of crude oils available to refiners (and
consequently their prices) is a critical factor in the strategies employed by refiners
around the world.
Oil is not a homogenous commodity: as a number of experts have pointed out (see,
The International Crude Oil Market Handbook, 2001) there are over 160 different
internationally traded crude oils, all of which vary in terms of characteristics, quality,
and market penetration.
Crude oils are classified by density and sulphur content. Lighter crudes generally
have a higher share of light hydrocarbons – i.e. higher value products - that can be
produced by simple distillation. Heavier crude oils give a greater share of lower-valued
products through simple distillation and require additional processing to produce the
2
desired range of products. Some crude oils also have a higher sulphur content, an
undesirable characteristic in terms of both processing and product quality.
The quality of the crude oil determines the level of processing and re-processing
necessary to achieve the optimal mix of product output. As a result, price and price
differentials between crude oils also reflect the relative ease of refining. For example, a
premium crude oil like West Texas Intermediate (WTI), the U.S. benchmark, or Brent,
the European benchmark, have a relatively high natural yield of desirable Gasoline. In
contrast, almost half of the simple distillation yield from Urals is a heavy residue that
must be reprocessed or sold at a discount as crude oil.
Refiners are in competition for an optimal mix of crudes for their refineries, in line
with the technology of the particular refinery, the desired output mix and, more
important, the relative price of available crudes. In recent years, refiners have been
faced with two opposing forces: a combination of consumers' desires for lower prices
and government regulations specifying increasingly lighter products of higher quality
(the most difficult to produce) and supplies of crude oil that are increasingly heavier, i.e.
with higher sulphur content (the most difficult to refine).
The importance of identifying the way in which a given crude is linked to a specific
crude benchmark comes directly from market considerations: the pressure of falling
margins in the oil products market, combined with some degree of flexibility in supply
decisions, obliges refiners to seek opportunities in the free market to improve their
profits. Crudes are expected to continue to become heavier with higher sulphur content,
while environmental restrictions are expected to significantly reduce the demand for
high-sulphur content fuels. As a consequence, light sweet crudes will continue to be
available and in even greater demand than today. This is why an understanding of the
3
price dynamics, and the role played by different crudes, is crucial for the modern oil
industry.
Because there are so many different varieties and grades of crude oil, buyers and
sellers have found it easier to refer to a limited number of reference, or benchmark,
crude oils. Other varieties are then priced at a discount or premium, according to their
quality. For any given crude oil, the price is considered to be linked to another crude oil
price (usually referred to as the marker). In this very simple scheme, to understand the
behaviour of a given crude oil would be sufficient to explain the behaviour of its
marker. However, the price difference between these two crudes is non-constant over
time. To enrich the relations it is necessary to include variables other than the price
marker to explain the oil price dynamics of the given crude.
In principle, several variables could affect this relation and could be used as
explanatory variables. Considering data availability, the common assumption is that
imbalances in the petroleum product price could reflect most of these missed variables.
For example: if, due to extraordinary seasonal factors, Gasoline demand were higher
than expected, this would be reflected into the relations between crudes according to
various specific characteristics.
This approach has been examined in several different papers. However the specific
economic literature on this issue is not very large. Adrangi, Chatrath, Raffiee and
Ripple (2001) analyze the price dynamics of a specific crude (the Alaska North Slope)
and its relation with US West coast diesel fuel price using a VAR methodology and a
bivariate GARCH model to show the casual relationship between the two prices.
Asche, Gjolberg and Volker (2003) make use of multivariate framework to test whether
4
there is a long-term relationship between crude oil and refined product prices in the
North Western Europe market.
Gjølberg and Johnsen (1999) analyze co-movements between the prices of crude oil
and major refined products during the period 1992-98. Specifically, they explore the
existence of long-run equilibrium price relationships, and whether deviations from the
estimated equilibrium can be utilized for predictions of short-term price changes and for
risk management.
In this paper we present a comparison among crudes considering four distinct
market areas (Mediterranean, North Western Europe, Latin America and North
America) on ten prices series of crude oils and on fourteen price series of petroleum
products.
We provide first a complete analysis of crude oil and product price dynamics using
co-integration and error correction models over the period 1994-2002. Subsequently we
use the error correction specification to predict crude oil prices over the horizon January
2002-June 2002.
The main findings of the paper can be summarized as follows.
Differences in quality are crucial to understand the behaviour of crudes.
Prices of crude oils whose physical characteristics are more similar to the marker
show the following regularities:
a) they converge more rapidly to the long-run equilibrium.
b) there is an almost monotonic relation between Mean Absolute Percentage Error
values and crude quality, measured by API° gravity and sulphur concentration.
This evidence can be motivated by considering the presence of the marker as an
5
explanatory variable: the closer the crude to the marker, the higher the
contribution of the latter in explaining and predicting the former.
The price of the marker is the driving variable of the crude price also in the short-
run, irrespective of the specific geographical area and the quality of the crude under
analysis.
This paper is organized as follows. Section 2 provides a description of the analyzed
data. Section 3 discusses the econometric methods and models. In Section 4 the
empirical results are reported and commented. The forecasting performance of the
estimated models is illustrated in Section 5. Concluding remarks close the paper.
2. Data description
Our analysis is based on ten prices series of crude oils and on fourteen price series of
petroleum products. These data cover four distinct market areas: Mediterranean (MED),
North Western Europe (NWE), Latin America (LA) and North America (NA). In the
first two areas the reference price for crude oil (marker) is represented by Brent, while
for the remaining two areas the benchmark crude is WTI. The petroleum products we
are considering belong to three different quality categories: unleaded Gasoline, Gasoil
and Fuel oil. Within the last class we distinguish between high sulphur Fuel oil (HSFO)
and low sulphur Fuel Oil (LSFO). The data frequency is weekly with the exception of
the LA market, where only monthly data are available, while the sample covers the
period 1994-2002. All crude oil prices are expressed in US$ per barrel, whilst product
prices are in US$ per metric ton. More details on the dataset are provided in Table 1.
Table 2 and Table 3 report, for both crude oils and petroleum products, the
coefficients of variation of price levels and the annualized standard deviation of price
6
changes. On average, the coefficients of variation for crude prices are the double of the
coefficients of variation of product prices, suggesting that the behaviour of crude prices
is very close to that of financial assets. Moreover, if we look at the two groups
separately, we find an inverse relation between quality (measured by API° gravity) and
the coefficient of variation. A possible interpretation is the subsidiary role played by
heavy crudes when light crudes become too expensive, while the lower-quality products
are more volatile since their price is intimately linked to the price of some specific
substitutes (e.g. natural gas).
Table 4 shows the percentage price correlations within crudes and between crudes
and products. Higher correlations occur when crudes and products similar in terms of
API° gravity are analyzed. The evidence from Tables 3 and 4 should suggest that prices
characterized by more similar coefficients of variation (i.e. light crudes and heavy
products) are more correlated. However, the coefficient of variation is a measure of
long-run volatility, whereas price change correlation captures short-run movements in
price variations. Moreover, an increase in the demand of light products has the effect of
increasing the supply of both high-quality and low-quality products (see Gjolberg and
Johnsen, 1999). Such considerations justify the presence of higher correlation between
light (heavy) crudes and the top (bottom) of the barrel.
3. Model specification
Crude oil and product prices dynamics can be modelled with an Autoregressive-
Distributed Lag (ADL) specification:
( ) ( ) ( ) ( )1 2y yc mt t t t tL p L p L p L p uα µ γ ϑ ξ= + + + + (1)
7
where L is the lag operator, ( ) 11 ... , PPL L Lα α α= − − − ( ) 0 1 ... , Q
QL L Lγ γ γ γ= + + +
( ) 0 1 ... RRL L Lϑ ϑ ϑ ϑ= + + + and ( ) 0 1 ... S
SL L Lξ ξ ξ ξ= + + + . Capital letters P, Q, R and S
represent the optimal number of lags of the polynomials α(L), γ(L), θ(L) and ξ(L),
respectively. With ctp we indicate the price of the selected crude, whereas m
tp is the
price of the marker associated with ctp , and iy
tp , i=1,2, are the prices of two products;
tu is a white noise process. All variables are log-transformed.
Recent developments in time series econometrics suggest that the first step towards
the estimation of model (1) is to check whether or not the different price series are
stationary. Augmented Dickey-Fuller (ADF) tests for unit roots have been used and all
variables have been found to be integrated of order one, or I(1), with intercept but no
trend.1
Though non-stationary, the oil and product price series may form a linear
combination which is stationary, or I(0). If this is the case, the relevant price series are
said to be cointegrated. The basic model used to test for the presence of cointegration is
given by the static regression
1 20 1 2 3
y yc mt t t t tp p p pβ β β β ε= + + + + (2)
If the residuals tε are I(0), then equation (2) provides the long-run or equilibrium
relationship between the relevant price series. When two or more variables are
1 The complete set of results is reported in Tables A1-A3 of the Appendix.
8
cointegrated, we know from the Engle-Granger representation theorem that they admit
an error correction (ECM) formulation of the type:
1 2
11 1 1
0 1 2 3 11 0 0 0
ˆQP R S
y yc c mt p t p q t q r t r s t s t t
p q r sp p p p pδ δ δ δ λε η
−− − −
− − − − −= = = =
∆ = ∆ + ∆ + ∆ + ∆ + +∑ ∑ ∑ ∑ (3)
where ( )1 20 1 2 3
ˆ ˆ ˆ ˆˆ y yc mt t t t tp p p pε β β β β= − + + + , ( )0 1
ˆ ˆˆ 1 Ppp
β µ α=
= −∑ ,
( )1 0 1ˆ ˆˆ 1Q P
q pq pβ γ α
= == −∑ ∑ , ( )2 0 1
ˆ ˆ ˆ1R Pr pr p
β ϑ α= =
= −∑ ∑ , and
( )3 0 1ˆ ˆ ˆ1S P
s ps pβ ξ α
= == −∑ ∑ .
The coefficients βi in equation (2) can be interpreted as long-run elasticities of the
crude price to the marker price and petroleum products prices. In other terms, each βi
measures the percentage variation of crude oil price due to a unit percentage variation of
each explanatory variable.
The choice of explaining oil prices in terms of petroleum product prices relies on the
theory of derived demand, which states that the price of an input should be determined
by its contribution to the market value of the output reflected in its market price (see
Adrangi, Chatrath, Raffiee and Ripple, 2001, for a test of the causal relationship flowing
from product prices to crude oil price).
Equation (3) incorporates short-run and long-run effects, captured by coefficients
ijδ and λ , respectively. In particular, λ is the so-called long-run adjustment coefficient
which measures how fast ctp converges towards the long-run equilibrium represented
by equation (2).
9
4. Empirical results
For each of the eight selected crudes we should estimate, at least in principle, as
many specifications for equation (3) as the number of combinations of products (i.e. six
models for MED and NWE, three models for LA and NA).
Given the large number of resulting models, we use a simple criterion to select the
best specification for each crude. Following Stock and Watson (1993), we estimate an
augmented version of equation (2), formed by adding one lead and one lag to all the
independent variables (DOLS estimation). In this way we obtain corrected t-statistics
for each estimated coefficient, which allow us to select the specifications of the long-run
equation with the largest number of statistically significant parameters. If two or more
long-run specifications have the same number of significant coefficients, we select the
one whose associated ECM yields the largest number of statistically significant
parameters. The final product selection for each crude is reported in the third column of
Table 5.
As it is shown in Table 5, the sum of the estimated coefficients β in equation (2)
(ignoring the intercept term) is approximately equal to one. Moreover, the null
hypothesis that this sum is equal to one is not rejected by the data in 5 cases out of 8.2
These coefficients can be interpreted as the contribution (weight) given by each
independent variable to the determination of crude oil price. The price of the marker
dominates relation (2), while product prices play a sort of compensation role, in order to
preserve the one-to-one relation between the crude and the marker. If we exclude Maya
2A corrected Wald test, based on the DOLS coefficient estimates, rejects the null hypothesis at 1%
significance level for Kern River and Thums, and at 5% for Iranian.
10
in the LA area, the β coefficients of the corresponding selected pair of product prices
have opposite signs. The contribution of each product to the market value of a particular
crude oil is such that a constant balance between price of the crude and price of the
marker is maintained in the long-run.
Specifically, 1β is always larger than one, and its magnitude increases as heavier
crudes are considered. These features show that when the price of the marker increases
the demand of heavy crude oils increases, which, in turn, forces their price to rise more
than proportionally.
Furthermore, when the MED and NWE areas are considered, the long-run
coefficients 2β and 3β have positive and negative signs, respectively. The converse is
true when we concentrate on NA. A possible interpretation of this empirical evidence is
that, while Europe is characterized by two highly demanded light products (i.e. Gasoline
and Gasoil), only Gasoline has a primary role in North America. As a consequence, an
increase in the demand for Gasoline in Europe is met using very light crudes in the
production process of Gasoline, while medium-quality crudes are employed to produce
Gasoil. On the contrary, the North American refinery system is mainly oriented towards
the production of Gasoline, which explains the positive long-run correlation between
crude and Gasoline prices.
In all areas each crude price is cointegrated with the price of the marker and the
prices of the selected pair of products, according to the ADF tests on the residuals of the
long-run equation (2) reported in Table 6.
The best ECM specification is attained with the product pair LSFO-Gasoline for
seven crudes out of eight (the only exception is HSFO-Gasoline for Urals NWE). The
short-run coefficient of Gasoline in the ECM equation (3) is significant, in all markets
11
and for all crudes, with the exception of Forcados. The more volatile product in the
short-run (Gasoline) is responsible of the short-run dynamics of the crude oil price. It is
well known that the refined barrel can be ideally divided in two classes of products:
high-quality (light) and low-quality (heavy) products. Hence, the best explanation of
both short-run and long-run behaviour of a crude oil price is obtained when we include
in the ECM specification the pair formed by the most representative products in each
class, that is LSFO-Gasoline (Table 7).
If we combine the information included in Table 1 with Table 7, it is easy to see that
the magnitude of the estimated long-run adjustment coefficients is sensitive to the
gravity of the specific crude, that is, with the exception of Forcados, a sort of monotonic
relation between speed of adjustment and API° emerges. Prices of crude oils whose
physical characteristics are more similar to the marker are likely to converge more
rapidly to the long-run equilibrium.
Furthermore, the price of the marker is the driving variable of the crude price also in
the short-run, irrespective of the specific geographical area and the quality of the crude
under analysis (see Table 5)3.
5. Forecasting crude oil prices
We assess the ability of the ECM specification to predict crude oil prices over the
horizon January 2002-June 2002 by computing three different sets of forecasts: static,
dynamic and simulated. With the exception of LA area, where only monthly data are
available, we split the forecasting horizon (24 weeks) into six windows of four weeks,
with the purpose of partially neutralizing potential contingent factors that could affect
3 The estimated short-run coefficients of the ECM are reported in Table A4 of the Appendix.
12
the forecasting evaluation (e.g. changes in OPEC policy). Moreover, in order to make
the calculated forecasts comparable, instead of estimating the ECM just once and using
the same estimated parameters to calculate forecast values of the dependent variable for
each of the six windows, we re-estimate the ECM six times with a rolling-sample
technique: in this way, the forecast values in each window depend on updated
coefficients estimates from samples of the same size.
While static and dynamic forecasts are self-explanatory, the procedure we use to
generate the simulated forecasts needs some explanation. The aim of this exercise is to
produce “true” out-of-sample, multistep-ahead forecasts for the crude oil price, given
the presence of marker and product prices as exogenous variables in model (3). Let’s
indicate with T the last in-sample observation for each window. Then:
i) For each variable 1 2 ˆ, , and y ymt t t tp p p ε∆ ∆ ∆ , we estimated an ARMA(1,1) model of the
type 1 1 1 1t t t tx x u uφ ϑ− −= + + , t=2,..,T. Since all estimated ARMA(1,1) models are found
to be statistically adequate to capture the behaviour of these series, for each model we
calculated the residuals ˆtu .
ii) Each ARMA residual vector ˆtu , t=2,..,T, is bootstrapped R=1000 times, to obtain
bootstrapped residuals ( )ˆb rtu , where r=1,..,R=1000 indicates the r-th replication and
superscript b denotes a bootstrapped series.
iii) Each series 1 2 ˆ, , and y ymt t t tp p p ε∆ ∆ ∆ is simulated R times out-of-sample
(t=T+1,…,T+h) using the estimated ARMA models of stage (i) and the bootstrapped
13
residuals of stage (2). That is: ( ) ( ) ( ) ( )11 1 1
ˆ ˆˆ ˆ ˆ ˆt
r r b r b rt t tx x u uφ ϑ
−
∗ ∗−= + + , t=T+1,..,T+h, where the
superscript * denotes a simulated series, and h=4 (h=6 for the crudes of the LA area,
since only monthly data are available).
iv) for each series 1 2 ˆ, , and y ymt t t tp p p ε∆ ∆ ∆ , we select, among the R simulated series, that
series whose standard deviation is closest to the standard deviation of the actual series
(this last calculated using in-sample observations).
Formally: ( )( ) ( )( )ˆmin . . . .rt t trx Std Dev x Std Dev x∗= −� , where tx� denotes the selected
simulated series.
v) we re-estimate the ECM specification (3) over the sample t=k,..,T, where
( )max , , ,k P Q R S= , and we calculate the residuals ˆtη .
vi) Residuals ˆtη are bootstrapped R times, thus obtaining ( )ˆb rtη .
vii) The dependent variable ctp is simulated R times, using the bootstrapped residuals of
the ECM model (stage vi) and the simulated exogenous series (stage iv):
( ) ( ) ( )1 2
11 1 1
1 0 1 2 3 1 ,1 0 0 0
ˆ ˆ ˆ ˆ ˆ ˆˆ ˆr
QP R Sc r c r y yc m bt t p t p q t q r t r s t s t i t j
p q r sp p p p p pδ δ δ δ λε η
−− − −∗ ∗
− − − − − − += = = =
= + ∆ + ∆ + ∆ + ∆ + +∑ ∑ ∑ ∑ �� � � �
t=T+1,..,T+h.
For crudes belonging to the MED, NWE and NA markets, we repeat this procedure for
all the 6 windows using the rolling-sample technique illustrated above.
14
After completion of the three forecasting exercises, we obtain, for the MED, NWE,
and NA areas, 24 one-step-ahead (static) forecasts, 24 (dynamic) h-steps-ahead
forecasts (h=1,..,4) and 24 (simulated) forecast distributions, each formed by R=1000
simulated forecasts. All forecasts are collected in six windows of size 4. For the LA area
we produce 6 (static) one-step-ahead forecasts, 6 (dynamic) h-steps-ahead forecasts
(h=1,..,4) and 6 (simulated) forecast distributions.
In order to evaluate the predictive ability of each ECM specifications, we calculate
the mean absolute percentage error (MAPE), the Theil’s inequality coefficient
(decomposed in bias, variance and covariance proportions) and the SR (success ratio),
which indicates the percentage number of times the forecasted series has the same sign
of the corresponding actual series. Moreover, for the simulated forecasts only, we
calculate a range of dispersion measures associated to each forecast distribution, as
follows. First, we compute the standard deviations of the distribution of forecasts in
each window and in each forecasting period (24 standard deviations). Second, we
calculate the mean of the 24 standard deviations. Third, for each window, we calculate
the mean of the standard deviations relative to the h-th forecasting point, h=1,…,4
(mean of 6 standard deviations).
Results from static and dynamic forecast are reported in Table 9. The following
comments apply.
First, due to the different data frequencies, a direct comparison between the LA
market and the remaining areas is not possible, although comments that hold for the
weekly series can be directly extended to the monthly data.
15
Second, if we rank the different crudes according to the forecasting performance of
the corresponding ECM specifications using the MAPE, the same ranking holds
irrespective of whether the forecasts are static or dynamic. The only exception is Iranian
heavy, whose dynamic forecasts seem to be relatively better than the static predictions.
Third, there is an almost monotonic relation between MAPE values and crude
quality, measured by API° gravity and sulphur concentration. Actually, among the
crudes with similar gravity, crudes with less sulphur are characterized by lower MAPE.
This evidence can be motivated by considering the presence of the marker as an
explanatory variable: the closer the crude to the marker, the higher the contribution of
the latter in explaining and predicting the former.
Fourth, from inspection of the Theil’s statistic, we experience an increase of the bias
proportion and a correspondent reduction of variance and covariance proportions when
moving from static to dynamic forecasts. Nonetheless, the values of the Theil’s
coefficient are generally quite small, indicating a good predictive fit.
Fifth, the low value of the variance proportion in the dynamic forecasts is perfectly
consistent with the values of SR.
Results from the simulated forecasts are reported in Table 10. MAPE, Theil’s
coefficient and SR are calculated on the mean of each forecasted distribution. As
expected, the forecasting performance for each model is slightly worse than in the static
and dynamic cases. Nevertheless, taking into account the crudes from the LA area, we
find that this kind of forecasts performs relatively better for heavier crudes. Actually,
MAPE values are almost five times larger than those obtained from the dynamic
forecasts in NWE, and almost twice than in NA. Conversely, the heaviest crude in LA
16
(i.e. Boscan) has MAPE values which are less than twice those of the dynamic forecast,
while Maya, the lightest crude in that area, has a MAPE value which is four times
larger.
The SR, though lower than in both static and dynamic cases, has values which are
higher than 0.50, meaning that the simulated series produce reasonable predictions of
the turning points of crude prices.
The second section of Table 10 reports several dispersion measures of the forecasted
distributions. The mean of all the standard deviations (SD) indicates that lower
predicting variability is associated with higher quality crudes. The overall coherence of
the simulation exercise is guaranteed by the values of each standard deviation, which
increase as the forecasting horizon increases.
6. Conclusions
This paper presents two different exercises that need to be commented in a separate
way even if there are some common interesting features.
The first conclusion is related to the different relation between a given crude, its
area-specific market and the related petroleum products. In this paper we investigate
crude oil and products price dynamics using cointegration and ECM. Empirical
evidence shows that product price are statistically relevant in explaining short- and
long-run adjustment in petroleum markets. The relevant product mix also depend on the
specific market area and on the characteristics of the selected crude. It is also worth to
underline that the long-run adjustment coefficients are sensitive to the gravity of the
specific crude. Prices of crude oils whose physical characteristics are more similar to
17
the marker are likely to converge more rapidly to the long-run equilibrium.
Furthermore, the price of the marker is the driving variable of the crude price also in the
short-run, irrespective of the specific geographical area and the quality of the crude
under analysis.
The second conclusion is related to the part of the paper aimed at assessing the
ability of the ECM specification to predict crude oil prices over the horizon January
2002-June 2002. We computed three different sets of forecasts, namely static, dynamic
and simulated, and in general the lower predicting variability is associated with higher
quality crudes. Also in this case there is almost monotonic relation between MAPE
values and crude quality, measured by API° gravity and sulphur concentration.
Actually, among the crudes with similar gravity, crudes with less sulphur are
characterized by lower MAPE. This evidence can be motivated by considering the
presence of the marker as an explanatory variable: the closer the crude to the marker,
the higher the contribution of the latter in explaining and predicting the former.
18
References
Adrangi, B., A. Chatrath, K. Raffiee and R.D.Ripple (2001), “Alaska North Slope crude
oil price and the behaviour of diesel prices in California”, Energy Economics, 23,
29-42.
Asche, F., O. Gjolberg and T. Völker (2003), “Price relationships in the petroleum
market. An analysis of crude oil and refined product prices”, Energy Economics, 25,
289-301.
Engle, R.F., and C.W.J. Granger (1987), “Co-integration and error correction:
representation, estimation and testing”, Econometrica, 55, 251-276.
Gjolberg, O., and T. Johnsen (1999), “Risk management in the oil industry: can
information on long-run equilibrium prices be utilized?”, Energy Economics, 21
517-527.
The International Crude Oil Market Handbook (2001), Energy Intelligence Group,
Fourth Edition.
MacKinnon, J.G. (1991), “Critical values for co-integration tests”, in R.F. Engle and
Note to Table 1. Sources Platt’s and Petroleum Intelligence Weekly (2000); API° gravity and sulphur content (%) are reported in parentheses; HSFO is not traded in
Table 5. Estimation of the long-run relationship Crudes Products
(y1, y2) R2
1β 2β 3β
Urals MED LSFO, Gasoline 0.99 1.04***
(11.69) 0.12* (1.72)
-0.16** (-2.58) MED
Iranian LSFO, Gasoline 0.99 1.13***
(11.00) 0.18** (2.34)
-0.24*** (-3.34)
Urals NWE HSFO, Gasoline 0.99 1.01***
(11.54) 0.11* (1.83)
-0.13** (-2.14) NWE
Forcados LSFO, Gasoline 0.99 1.06***
(16.23) 0.01 (0.43)
-0.08 (-1.51)
Maya LSFO, Gasoline 0.95 1.85***
(4.69) -0.52* (-1.63)
-0.20 (-0.58) LA
Boscan LSFO, Gasoline 0.91 2.04***
(3.53) -0.87* (-1.85)
0.03 (0.06)
Kern River LSFO, Gasoline 0.94 1.35***
(3.77) -0.07 (-0.24)
0.04 (0.13) NA
Thums LSFO, Gasoline 0.95 1.32***
(4.70) -0.10 (-0.42)
0.03 (0.11)
Notes to Table 5. iβ i=1,..,3, are the DOLS estimates of the augmented dynamic regression
1 2 1 20 1 2 3
r r ry y y yc m mt t t t i t i i t i i t i ti r i r i r
p p p p p p pβ β β β θ φ γ ε− − −=− =− =−= + + + + ∆ + ∆ + ∆ +∑ ∑ ∑ , with r=1
(see Stock and Watson, 1993), in parentheses the rescaled t-statistics; * (**)[***] indicates significance at 10% (5%) [1%]
24
Table 6. Cointegration tests
Crudes Products (y1, y2)
a b p ADF
Urals MED LSFO, Gasoline
no no 2 -5.98*** MED
Iranian LSFO, Gasoline
no no 2 -5.98***
Urals NWE HSFO, Gasoline
no no 2 -5.33*** NWE
Forcados LSFO, Gasoline
no no 1 -4.88***
Maya LSFO, Gasoline
no no 0 -3.82 (57.96***)
LA
Boscan LSFO, Gasoline
no no 0 -3.79 (53.72***)
Kern River LSFO, Gasoline
no no 2 -5.09*** NA
Thums LSFO, Gasoline
no no 0 -5.55***
Notes to Table 6. ADF is the calculated t test for the null hypothesis of no cointegration (i.e. γ=0) in the
Augmented Dickey-Fuller regression on ε^t: 1 1
ˆ ˆ ˆpt t i t i ti
a bt vε γε γ ε− −=∆ = + + + ∆ +∑ , where ε^
t are
the estimated residuals of the DOLS regression; p is the order of the augmentation needed to eliminate any autocorrelation in the residuals of the ADF regression; * (**)[***] indicates significance at 10% (5%) [1%] on the basis of the critical values by MacKinnon, (1991); for crudes in the LA area the Johansen’s (1991) trace test is reported in parentheses.
25
Table 7. Selected products and long-run adjustment coefficients
MED NWE LA NA
Selected products
LSFO-Gasoline (Urals, Iranian)
HSFO-Gasoline (Urals)
LSFO-Gasoline (Forcados)
LSFO-Gasoline (Maya, Boscan)
LSFO-Gasoline (Kern River, Thums)
Long-run products
LSFO-Gasoline (Urals, Iranian)
HSFO-Gasoline (Urals)
- (Forcados)
LSFO (Maya, Boscan) -
Short-run products
Gasoline (Urals, Iranian)
Gasoline (Forcados)
LSFO-Gasoline (Maya, Boscan)
LSFO-Gasoline (Kern River, Thums)
Long-run
adjustment coefficients
( λ )
-0.12 (Urals, Iranian)
-0.11 (Urals)
-0.06
(Forcados)
-0.15 (Maya)
-0.09
(Boscan)
-0.07 (Kern River, Thums)
Notes to Table 7. Selected products = pair of products corresponding to the best model specifications (1) and (2); long-run products = products whose coefficients are statistically significant in the long-run relation (1); short-run products = products whose short-run coefficients are statistically significant in model (2); crudes associated with selected products, long-run products, short-run products and long-run adjustment coefficients (see equation (2)) are reported in parentheses.
26
Table 8. Static and dynamic forecast evaluation of selected ECM models
SR 0.875 0.958 1.00 0.958 1.00 1.00 0.958 0.958 Notes to Table 8. Static forecasts indicate one-step-ahead forecasts, dynamic forecasts indicate 4-step-ahead forecasts (6 steps for LA area); MAPE is the mean absolute percentage error, Theil is the Theil’s Inequality Coefficient and BP, VP, CP are the bias, variance, and covariance proportions. SR is the mean of the success ratio calculated as the percentage number of times the sign of the forecasted series is the same as the sign of the actual series. All the reported values, with the exception of those referring to LA, are mean values calculated over the 6 forecast windows.
27
Table 9. Simulated forecast evaluation of selected ECM models
SD4 0.70 0.73 0.51 0.22 - - 1.26 0.88 Notes to Table 9. Simulated forecast stands for ‘true’ out of sample 4 (6) step-ahead forecast. In order to calculate the reported measures of dispersion we proceeded as follows: i) we calculated the standard deviations of the distribution of forecasts in each window and in each forecasting period (24 standard deviations); ii) in order to obtain SD we calculated the mean of all the standard deviations of point i. (mean of 24 standard deviations); iii) in order to obtain SDk k=1,..,4 we calculated the mean by window of the standard deviations referring to k-th forecasting point (mean of 6 standard deviations).
28
Appendix
Table A1.Unit root tests: Crudes
a b P ADF Brent yes no 1 -2.06 ∆ Brent no no 0 -15.94** Urals med yes no 1 -2.31 ∆ Urals med no no 0 -15.81** Iranian yes no 1 -2.24
∆ Iranian no no 0 -15.90**
Urals NWE yes no 1 -2.24 ∆ Urals NWE no no 0 -16.07** Forcados yes no 1 -2.18 ∆ Forcados no no 0 -15.60** WTI yes no 0 -1.69 ∆ WTI no no 0 -8.70** Maya yes no 0 -1.82 ∆ Maya no no 0 -8.33** Boscan yes no 1 -2.24 ∆ Boscan no no 0 -6.95** Kern River yes no 1 -2.26 ∆ Kern River no no 0 -15.52** Thums yes no 1 -2.11 ∆ Thums no no 0 -16.00**
Notes to Table A1. ADF is the calculated t test for the null hypothesis of a unit root (i.e. γ=0) in the series
xt from the Augmented Dickey-Fuller regression: 1 11
pt t i t ti
x a bt x xγ λ η− −=∆ = + + + ∆ +∑ ; p is the
order of the augmentation needed to eliminate any autocorrelation in the residuals of the ADF regression; * (**)[***] indicates significance at 10% (5%) [1%] on the basis of the critical values by MacKinnon, J.G. (1991) “Critical Values for Co-Integration Tests”, in R.F. Engle and C.W.J. Granger (eds.), Long-run Economic Relationships, Oxford, Oxford University Press..
29
Table A2. Unit root tests: Products, Europe
MED NWE a b p ADF a b p ADF
Gasoline yes no 1 -2.18 yes no 1 -2.15 ∆ Gasoline no no 0 -14.17** no no 0 -15.02** Gasoil yes no 1 -2.03 yes no 1 -1.84 ∆ Gasoil no no 0 -14.63** no no 0 -15.12** LSFO yes no 1 -2.50 yes no 1 -2.16 ∆ LSFO no no 0 -12.51** no no 0 -13.45** HSFO yes no 2 -2.44 yes no 1 -2.19 ∆ HSFO no no 1 -11.26** no no 0 -15.59**
Notes to Table A2. see Table A1
30
Table A3. Unit root tests: Products, America
LA NA a b p ADF a b p ADF
Gasoline yes no 0 -2.27 yes no 1 -2.73 ∆ Gasoline no no 0 -9.51** no no 0 -16.34** Gasoil yes no 1 -1.88 No no 1 -1.73 ∆ Gasoil no no 0 -7.68** no no 0 -19.25** LSFO yes no 0 -1.73 yes no 1 -2.37 ∆ LSFO no no 0 -8.92** no no 0 -14.54**
ˆP Q R Sy yc c mt p t p q t q r t r s t s t tp q r s
p p p p pδ δ δ δ λε η− − − −− − − − −= = = =
∆ = ∆ + ∆ + ∆ + ∆ + +∑ ∑ ∑ ∑ , where
P=Q=R=S; BG- stat is the LM version of the Breusch-Godfrey test for absence of first order residual autocorrelation in the regression; * (**)[***] indicates significance at 10% (5%) [1%]
NOTE DI LAVORO DELLA FONDAZIONE ENI ENRICO MATTEI
Fondazione Eni Enrico Mattei Working Paper Series Our working papers are available on the Internet at the following addresses:
SUST 1.2002 K. TANO, M.D. FAMINOW, M. KAMUANGA and B. SWALLOW: Using Conjoint Analysis to Estimate Farmers’ Preferences for Cattle Traits in West Africa
ETA 2.2002 Efrem CASTELNUOVO and Paolo SURICO: What Does Monetary Policy Reveal about Central Bank’s Preferences?
WAT 3.2002 Duncan KNOWLER and Edward BARBIER: The Economics of a “Mixed Blessing” Effect: A Case Study of the Black Sea
CLIM 4.2002 Andreas LöSCHEL: Technological Change in Economic Models of Environmental Policy: A Survey VOL 5.2002 Carlo CARRARO and Carmen MARCHIORI: Stable Coalitions CLIM 6.2002 Marzio GALEOTTI, Alessandro LANZA and Matteo MANERA: Rockets and Feathers Revisited: An International
Comparison on European Gasoline Markets ETA 7.2002 Effrosyni DIAMANTOUDI and Eftichios S. SARTZETAKIS: Stable International Environmental Agreements: An
Analytical Approach KNOW 8.2002 Alain DESDOIGTS: Neoclassical Convergence Versus Technological Catch-up: A Contribution for Reaching a
Consensus NRM 9.2002 Giuseppe DI VITA: Renewable Resources and Waste Recycling KNOW 10.2002 Giorgio BRUNELLO: Is Training More Frequent when Wage Compression is Higher? Evidence from 11
European Countries ETA 11.2002 Mordecai KURZ, Hehui JIN and Maurizio MOTOLESE: Endogenous Fluctuations and the Role of Monetary
Policy KNOW 12.2002 Reyer GERLAGH and Marjan W. HOFKES: Escaping Lock-in: The Scope for a Transition towards Sustainable
Growth? NRM 13.2002 Michele MORETTO and Paolo ROSATO: The Use of Common Property Resources: A Dynamic Model CLIM 14.2002 Philippe QUIRION: Macroeconomic Effects of an Energy Saving Policy in the Public Sector CLIM 15.2002 Roberto ROSON: Dynamic and Distributional Effects of Environmental Revenue Recycling Schemes:
Simulations with a General Equilibrium Model of the Italian Economy CLIM 16.2002 Francesco RICCI (l): Environmental Policy Growth when Inputs are Differentiated in Pollution Intensity ETA 17.2002 Alberto PETRUCCI: Devaluation (Levels versus Rates) and Balance of Payments in a Cash-in-Advance
Economy Coalition Theory Network
18.2002 László Á. KÓCZY (liv): The Core in the Presence of Externalities
Coalition Theory Network
19.2002 Steven J. BRAMS, Michael A. JONES and D. Marc KILGOUR (liv): Single-Peakedness and Disconnected Coalitions
Coalition Theory Network
20.2002 Guillaume HAERINGER (liv): On the Stability of Cooperation Structures
NRM 21.2002 Fausto CAVALLARO and Luigi CIRAOLO: Economic and Environmental Sustainability: A Dynamic Approach in Insular Systems
CLIM 22.2002 Barbara BUCHNER, Carlo CARRARO, Igor CERSOSIMO and Carmen MARCHIORI: Back to Kyoto? US Participation and the Linkage between R&D and Climate Cooperation
CLIM 23.2002 Andreas LÖSCHEL and ZhongXIANG ZHANG: The Economic and Environmental Implications of the US Repudiation of the Kyoto Protocol and the Subsequent Deals in Bonn and Marrakech
ETA 24.2002 Marzio GALEOTTI, Louis J. MACCINI and Fabio SCHIANTARELLI: Inventories, Employment and Hours CLIM 25.2002 Hannes EGLI: Are Cross-Country Studies of the Environmental Kuznets Curve Misleading? New Evidence from
Time Series Data for Germany ETA 26.2002 Adam B. JAFFE, Richard G. NEWELL and Robert N. STAVINS: Environmental Policy and Technological
Change SUST 27.2002 Joseph C. COOPER and Giovanni SIGNORELLO: Farmer Premiums for the Voluntary Adoption of
Conservation Plans SUST 28.2002 The ANSEA Network: Towards An Analytical Strategic Environmental Assessment KNOW 29.2002 Paolo SURICO: Geographic Concentration and Increasing Returns: a Survey of Evidence ETA 30.2002 Robert N. STAVINS: Lessons from the American Experiment with Market-Based Environmental Policies
NRM 31.2002 Carlo GIUPPONI and Paolo ROSATO: Multi-Criteria Analysis and Decision-Support for Water Management at the Catchment Scale: An Application to Diffuse Pollution Control in the Venice Lagoon
NRM 32.2002 Robert N. STAVINS: National Environmental Policy During the Clinton Years KNOW 33.2002 A. SOUBEYRAN and H. STAHN : Do Investments in Specialized Knowledge Lead to Composite Good
Industries? KNOW 34.2002 G. BRUNELLO, M.L. PARISI and Daniela SONEDDA: Labor Taxes, Wage Setting and the Relative Wage
Effect CLIM 35.2002 C. BOEMARE and P. QUIRION (lv): Implementing Greenhouse Gas Trading in Europe: Lessons from
Economic Theory and International Experiences CLIM 36.2002 T.TIETENBERG (lv): The Tradable Permits Approach to Protecting the Commons: What Have We Learned?
CLIM 37.2002 K. REHDANZ and R.J.S. TOL (lv): On National and International Trade in Greenhouse Gas Emission Permits CLIM 38.2002 C. FISCHER (lv): Multinational Taxation and International Emissions Trading SUST 39.2002 G. SIGNORELLO and G. PAPPALARDO: Farm Animal Biodiversity Conservation Activities in Europe under
the Framework of Agenda 2000 NRM 40.2002 S .M. CAVANAGH, W. M. HANEMANN and R. N. STAVINS: Muffled Price Signals: Household Water Demand
under Increasing-Block Prices NRM 41.2002 A. J. PLANTINGA, R. N. LUBOWSKI and R. N. STAVINS: The Effects of Potential Land Development on
Agricultural Land Prices CLIM 42.2002 C. OHL (lvi): Inducing Environmental Co-operation by the Design of Emission Permits CLIM 43.2002 J. EYCKMANS, D. VAN REGEMORTER and V. VAN STEENBERGHE (lvi): Is Kyoto Fatally Flawed? An
Analysis with MacGEM CLIM 44.2002 A. ANTOCI and S. BORGHESI (lvi): Working Too Much in a Polluted World: A North-South Evolutionary
Model ETA 45.2002 P. G. FREDRIKSSON, Johan A. LIST and Daniel MILLIMET (lvi): Chasing the Smokestack: Strategic
Policymaking with Multiple Instruments ETA 46.2002 Z. YU (lvi): A Theory of Strategic Vertical DFI and the Missing Pollution-Haven Effect SUST 47.2002 Y. H. FARZIN: Can an Exhaustible Resource Economy Be Sustainable? SUST 48.2002 Y. H. FARZIN: Sustainability and Hamiltonian Value KNOW 49.2002 C. PIGA and M. VIVARELLI: Cooperation in R&D and Sample Selection Coalition Theory Network
50.2002 M. SERTEL and A. SLINKO (liv): Ranking Committees, Words or Multisets
Coalition Theory Network
51.2002 Sergio CURRARINI (liv): Stable Organizations with Externalities
ETA 52.2002 Robert N. STAVINS: Experience with Market-Based Policy Instruments ETA 53.2002 C.C. JAEGER, M. LEIMBACH, C. CARRARO, K. HASSELMANN, J.C. HOURCADE, A. KEELER and
R. KLEIN (liii): Integrated Assessment Modeling: Modules for Cooperation CLIM 54.2002 Scott BARRETT (liii): Towards a Better Climate Treaty ETA 55.2002 Richard G. NEWELL and Robert N. STAVINS: Cost Heterogeneity and the Potential Savings from Market-
Based Policies SUST 56.2002 Paolo ROSATO and Edi DEFRANCESCO: Individual Travel Cost Method and Flow Fixed Costs SUST 57.2002 Vladimir KOTOV and Elena NIKITINA (lvii): Reorganisation of Environmental Policy in Russia: The Decade of
Success and Failures in Implementation of Perspective Quests SUST 58.2002 Vladimir KOTOV (lvii): Policy in Transition: New Framework for Russia’s Climate Policy SUST 59.2002 Fanny MISSFELDT and Arturo VILLAVICENCO (lvii): How Can Economies in Transition Pursue Emissions
Trading or Joint Implementation? VOL 60.2002 Giovanni DI BARTOLOMEO, Jacob ENGWERDA, Joseph PLASMANS and Bas VAN AARLE: Staying Together
or Breaking Apart: Policy-Makers’ Endogenous Coalitions Formation in the European Economic and Monetary Union
ETA 61.2002 Robert N. STAVINS, Alexander F.WAGNER and Gernot WAGNER: Interpreting Sustainability in Economic Terms: Dynamic Efficiency Plus Intergenerational Equity
PRIV 62.2002 Carlo CAPUANO: Demand Growth, Entry and Collusion Sustainability PRIV 63.2002 Federico MUNARI and Raffaele ORIANI: Privatization and R&D Performance: An Empirical Analysis Based on
Tobin’s Q PRIV 64.2002 Federico MUNARI and Maurizio SOBRERO: The Effects of Privatization on R&D Investments and Patent
Productivity SUST 65.2002 Orley ASHENFELTER and Michael GREENSTONE: Using Mandated Speed Limits to Measure the Value of a
Statistical Life ETA 66.2002 Paolo SURICO: US Monetary Policy Rules: the Case for Asymmetric Preferences PRIV 67.2002 Rinaldo BRAU and Massimo FLORIO: Privatisations as Price Reforms: Evaluating Consumers’ Welfare
Changes in the U.K. CLIM 68.2002 Barbara K. BUCHNER and Roberto ROSON: Conflicting Perspectives in Trade and Environmental Negotiations CLIM 69.2002 Philippe QUIRION: Complying with the Kyoto Protocol under Uncertainty: Taxes or Tradable Permits? SUST 70.2002 Anna ALBERINI, Patrizia RIGANTI and Alberto LONGO: Can People Value the Aesthetic and Use Services of
Urban Sites? Evidence from a Survey of Belfast Residents SUST 71.2002 Marco PERCOCO: Discounting Environmental Effects in Project Appraisal
NRM 72.2002 Philippe BONTEMS and Pascal FAVARD: Input Use and Capacity Constraint under Uncertainty: The Case of Irrigation
PRIV 73.2002 Mohammed OMRAN: The Performance of State-Owned Enterprises and Newly Privatized Firms: Empirical Evidence from Egypt
PRIV 74.2002 Mike BURKART, Fausto PANUNZI and Andrei SHLEIFER: Family Firms PRIV 75.2002 Emmanuelle AURIOL, Pierre M. PICARD: Privatizations in Developing Countries and the Government Budget
Constraint PRIV 76.2002 Nichole M. CASTATER: Privatization as a Means to Societal Transformation: An Empirical Study of
Privatization in Central and Eastern Europe and the Former Soviet Union PRIV 77.2002 Christoph LÜLSFESMANN: Benevolent Government, Managerial Incentives, and the Virtues of Privatization PRIV 78.2002 Kate BISHOP, Igor FILATOTCHEV and Tomasz MICKIEWICZ: Endogenous Ownership Structure: Factors
Affecting the Post-Privatisation Equity in Largest Hungarian Firms PRIV 79.2002 Theodora WELCH and Rick MOLZ: How Does Trade Sale Privatization Work?
Evidence from the Fixed-Line Telecommunications Sector in Developing Economies PRIV 80.2002 Alberto R. PETRUCCI: Government Debt, Agent Heterogeneity and Wealth Displacement in a Small Open
Economy CLIM 81.2002 Timothy SWANSON and Robin MASON (lvi): The Impact of International Environmental Agreements: The Case
of the Montreal Protocol PRIV 82.2002 George R.G. CLARKE and Lixin Colin XU: Privatization, Competition and Corruption: How Characteristics of
Bribe Takers and Payers Affect Bribe Payments to Utilities PRIV 83.2002 Massimo FLORIO and Katiuscia MANZONI: The Abnormal Returns of UK Privatisations: From Underpricing
to Outperformance NRM 84.2002 Nelson LOURENÇO, Carlos RUSSO MACHADO, Maria do ROSÁRIO JORGE and Luís RODRIGUES: An
Integrated Approach to Understand Territory Dynamics. The Coastal Alentejo (Portugal) CLIM 85.2002 Peter ZAPFEL and Matti VAINIO (lv): Pathways to European Greenhouse Gas Emissions Trading History and
Misconceptions CLIM 86.2002 Pierre COURTOIS: Influence Processes in Climate Change Negotiations: Modelling the Rounds ETA 87.2002 Vito FRAGNELLI and Maria Erminia MARINA (lviii): Environmental Pollution Risk and Insurance ETA 88.2002 Laurent FRANCKX (lviii): Environmental Enforcement with Endogenous Ambient Monitoring ETA 89.2002 Timo GOESCHL and Timothy M. SWANSON (lviii): Lost Horizons. The noncooperative management of an
evolutionary biological system. ETA 90.2002 Hans KEIDING (lviii): Environmental Effects of Consumption: An Approach Using DEA and Cost Sharing ETA 91.2002 Wietze LISE (lviii): A Game Model of People’s Participation in Forest Management in Northern India CLIM 92.2002 Jens HORBACH: Structural Change and Environmental Kuznets Curves ETA 93.2002 Martin P. GROSSKOPF: Towards a More Appropriate Method for Determining the Optimal Scale of Production
Units VOL 94.2002 Scott BARRETT and Robert STAVINS: Increasing Participation and Compliance in International Climate Change
Agreements CLIM 95.2002 Banu BAYRAMOGLU LISE and Wietze LISE: Climate Change, Environmental NGOs and Public Awareness in
the Netherlands: Perceptions and Reality CLIM 96.2002 Matthieu GLACHANT: The Political Economy of Emission Tax Design in Environmental Policy KNOW 97.2002 Kenn ARIGA and Giorgio BRUNELLO: Are the More Educated Receiving More Training? Evidence from
Thailand ETA 98.2002 Gianfranco FORTE and Matteo MANERA: Forecasting Volatility in European Stock Markets with Non-linear
GARCH Models ETA 99.2002 Geoffrey HEAL: Bundling Biodiversity ETA 100.2002 Geoffrey HEAL, Brian WALKER, Simon LEVIN, Kenneth ARROW, Partha DASGUPTA, Gretchen DAILY, Paul
EHRLICH, Karl-Goran MALER, Nils KAUTSKY, Jane LUBCHENCO, Steve SCHNEIDER and David STARRETT: Genetic Diversity and Interdependent Crop Choices in Agriculture
ETA 101.2002 Geoffrey HEAL: Biodiversity and Globalization VOL 102.2002 Andreas LANGE: Heterogeneous International Agreements – If per capita emission levels matter ETA 103.2002 Pierre-André JOUVET and Walid OUESLATI: Tax Reform and Public Spending Trade-offs in an Endogenous
Growth Model with Environmental Externality ETA 104.2002 Anna BOTTASSO and Alessandro SEMBENELLI: Does Ownership Affect Firms’ Efficiency? Panel Data
Evidence on Italy PRIV 105.2002 Bernardo BORTOLOTTI, Frank DE JONG, Giovanna NICODANO and Ibolya SCHINDELE: Privatization and
Stock Market Liquidity ETA 106.2002 Haruo IMAI and Mayumi HORIE (lviii): Pre-Negotiation for an International Emission Reduction Game PRIV 107.2002 Sudeshna GHOSH BANERJEE and Michael C. MUNGER: Move to Markets? An Empirical Analysis of
Privatisation in Developing Countries PRIV 108.2002 Guillaume GIRMENS and Michel GUILLARD: Privatization and Investment: Crowding-Out Effect vs Financial
Diversification PRIV 109.2002 Alberto CHONG and Florencio LÓPEZ-DE-SILANES: Privatization and Labor Force Restructuring Around the
World PRIV 110.2002 Nandini GUPTA: Partial Privatization and Firm Performance PRIV 111.2002 François DEGEORGE, Dirk JENTER, Alberto MOEL and Peter TUFANO: Selling Company Shares to
Reluctant Employees: France Telecom’s Experience
PRIV 112.2002 Isaac OTCHERE: Intra-Industry Effects of Privatization Announcements: Evidence from Developed and Developing Countries
PRIV 113.2002 Yannis KATSOULAKOS and Elissavet LIKOYANNI: Fiscal and Other Macroeconomic Effects of Privatization PRIV 114.2002 Guillaume GIRMENS: Privatization, International Asset Trade and Financial Markets PRIV 115.2002 D. Teja FLOTHO: A Note on Consumption Correlations and European Financial Integration PRIV 116.2002 Ibolya SCHINDELE and Enrico C. PEROTTI: Pricing Initial Public Offerings in Premature Capital Markets:
The Case of Hungary PRIV 1.2003 Gabriella CHIESA and Giovanna NICODANO: Privatization and Financial Market Development: Theoretical
Issues PRIV 2.2003 Ibolya SCHINDELE: Theory of Privatization in Eastern Europe: Literature Review PRIV 3.2003 Wietze LISE, Claudia KEMFERT and Richard S.J. TOL: Strategic Action in the Liberalised German Electricity
Market CLIM 4.2003 Laura MARSILIANI and Thomas I. RENSTRÖM: Environmental Policy and Capital Movements: The Role of
Government Commitment KNOW 5.2003 Reyer GERLAGH: Induced Technological Change under Technological Competition ETA 6.2003 Efrem CASTELNUOVO: Squeezing the Interest Rate Smoothing Weight with a Hybrid Expectations Model SIEV 7.2003 Anna ALBERINI, Alberto LONGO, Stefania TONIN, Francesco TROMBETTA and Margherita TURVANI: The
Role of Liability, Regulation and Economic Incentives in Brownfield Remediation and Redevelopment: Evidence from Surveys of Developers
NRM 8.2003 Elissaios PAPYRAKIS and Reyer GERLAGH: Natural Resources: A Blessing or a Curse? CLIM 9.2003 A. CAPARRÓS, J.-C. PEREAU and T. TAZDAÏT: North-South Climate Change Negotiations: a Sequential Game
with Asymmetric Information KNOW 10.2003 Giorgio BRUNELLO and Daniele CHECCHI: School Quality and Family Background in Italy CLIM 11.2003 Efrem CASTELNUOVO and Marzio GALEOTTI: Learning By Doing vs Learning By Researching in a Model of
Climate Change Policy Analysis KNOW 12.2003 Carole MAIGNAN, Gianmarco OTTAVIANO and Dino PINELLI (eds.): Economic Growth, Innovation, Cultural
Diversity: What are we all talking about? A critical survey of the state-of-the-art KNOW 13.2003 Carole MAIGNAN, Gianmarco OTTAVIANO, Dino PINELLI and Francesco RULLANI (lix): Bio-Ecological
Diversity vs. Socio-Economic Diversity. A Comparison of Existing Measures KNOW 14.2003 Maddy JANSSENS and Chris STEYAERT (lix): Theories of Diversity within Organisation Studies: Debates and
Future Trajectories KNOW 15.2003 Tuzin BAYCAN LEVENT, Enno MASUREL and Peter NIJKAMP (lix): Diversity in Entrepreneurship: Ethnic and
Female Roles in Urban Economic Life KNOW 16.2003 Alexandra BITUSIKOVA (lix): Post-Communist City on its Way from Grey to Colourful: The Case Study from
Slovakia KNOW 17.2003 Billy E. VAUGHN and Katarina MLEKOV (lix): A Stage Model of Developing an Inclusive Community KNOW 18.2003 Selma van LONDEN and Arie de RUIJTER (lix): Managing Diversity in a Glocalizing World
Coalition Theory
Network
19.2003 Sergio CURRARINI: On the Stability of Hierarchies in Games with Externalities
PRIV 20.2003 Giacomo CALZOLARI and Alessandro PAVAN (lx): Monopoly with Resale PRIV 21.2003 Claudio MEZZETTI (lx): Auction Design with Interdependent Valuations: The Generalized Revelation
Principle, Efficiency, Full Surplus Extraction and Information Acquisition PRIV 22.2003 Marco LiCalzi and Alessandro PAVAN (lx): Tilting the Supply Schedule to Enhance Competition in Uniform-
Price Auctions PRIV 23.2003 David ETTINGER (lx): Bidding among Friends and Enemies PRIV 24.2003 Hannu VARTIAINEN (lx): Auction Design without Commitment PRIV 25.2003 Matti KELOHARJU, Kjell G. NYBORG and Kristian RYDQVIST (lx): Strategic Behavior and Underpricing in
Uniform Price Auctions: Evidence from Finnish Treasury Auctions PRIV 26.2003 Christine A. PARLOUR and Uday RAJAN (lx): Rationing in IPOs PRIV 27.2003 Kjell G. NYBORG and Ilya A. STREBULAEV (lx): Multiple Unit Auctions and Short Squeezes PRIV 28.2003 Anders LUNANDER and Jan-Eric NILSSON (lx): Taking the Lab to the Field: Experimental Tests of Alternative
Mechanisms to Procure Multiple Contracts PRIV 29.2003 TangaMcDANIEL and Karsten NEUHOFF (lx): Use of Long-term Auctions for Network Investment PRIV 30.2003 Emiel MAASLAND and Sander ONDERSTAL (lx): Auctions with Financial Externalities ETA 31.2003 Michael FINUS and Bianca RUNDSHAGEN: A Non-cooperative Foundation of Core-Stability in Positive
Externality NTU-Coalition Games KNOW 32.2003 Michele MORETTO: Competition and Irreversible Investments under Uncertainty_ PRIV 33.2003 Philippe QUIRION: Relative Quotas: Correct Answer to Uncertainty or Case of Regulatory Capture?
KNOW 34.2003 Giuseppe MEDA, Claudio PIGA and Donald SIEGEL: On the Relationship between R&D and Productivity: A Treatment Effect Analysis
ETA 35.2003 Alessandra DEL BOCA, Marzio GALEOTTI and Paola ROTA: Non-convexities in the Adjustment of Different Capital Inputs: A Firm-level Investigation
Governance: the Role of Ownership Structure and Investor Protection CLIM 38.2003 Rolf GOLOMBEK and Michael HOEL: Climate Policy under Technology Spillovers KNOW 39.2003 Slim BEN YOUSSEF: Transboundary Pollution, R&D Spillovers and International Trade CTN 40.2003 Carlo CARRARO and Carmen MARCHIORI: Endogenous Strategic Issue Linkage in International Negotiations KNOW 41.2003 Sonia OREFFICE: Abortion and Female Power in the Household: Evidence from Labor Supply KNOW 42.2003 Timo GOESCHL and Timothy SWANSON: On Biology and Technology: The Economics of Managing
Biotechnologies ETA 43.2003 Giorgio BUSETTI and Matteo MANERA: STAR-GARCH Models for Stock Market Interactions in the Pacific
Basin Region, Japan and US CLIM 44.2003 Katrin MILLOCK and Céline NAUGES: The French Tax on Air Pollution: Some Preliminary Results on its
Effectiveness PRIV 45.2003 Bernardo BORTOLOTTI and Paolo PINOTTI: The Political Economy of Privatization SIEV 46.2003 Elbert DIJKGRAAF and Herman R.J. VOLLEBERGH: Burn or Bury? A Social Cost Comparison of Final Waste
Disposal Methods ETA 47.2003 Jens HORBACH: Employment and Innovations in the Environmental Sector: Determinants and Econometrical
Results for Germany CLIM 48.2003 Lori SNYDER, Nolan MILLER and Robert STAVINS: The Effects of Environmental Regulation on Technology
Diffusion: The Case of Chlorine Manufacturing CLIM 49.2003 Lori SNYDER, Robert STAVINS and Alexander F. WAGNER: Private Options to Use Public Goods. Exploiting
Revealed Preferences to Estimate Environmental Benefits CTN 50.2003 László Á. KÓCZY and Luc LAUWERS (lxi): The Minimal Dominant Set is a Non-Empty Core-Extension
CTN 51.2003 Matthew O. JACKSON (lxi):Allocation Rules for Network Games CTN 52.2003 Ana MAULEON and Vincent VANNETELBOSCH (lxi): Farsightedness and Cautiousness in Coalition FormationCTN 53.2003 Fernando VEGA-REDONDO (lxi): Building Up Social Capital in a Changing World: a network approach CTN 54.2003 Matthew HAAG and Roger LAGUNOFF (lxi): On the Size and Structure of Group Cooperation CTN 55.2003 Taiji FURUSAWA and Hideo KONISHI (lxi): Free Trade Networks CTN 56.2003 Halis Murat YILDIZ (lxi): National Versus International Mergers and Trade Liberalization CTN 57.2003 Santiago RUBIO and Alistair ULPH (lxi): An Infinite-Horizon Model of Dynamic Membership of International
Environmental Agreements KNOW 58.2003 Carole MAIGNAN, Dino PINELLI and Gianmarco I.P. OTTAVIANO: ICT, Clusters and Regional Cohesion: A
Summary of Theoretical and Empirical Research KNOW 59.2003 Giorgio BELLETTINI and Gianmarco I.P. OTTAVIANO: Special Interests and Technological Change ETA 60.2003 Ronnie SCHÖB: The Double Dividend Hypothesis of Environmental Taxes: A Survey CLIM 61.2003 Michael FINUS, Ekko van IERLAND and Robert DELLINK: Stability of Climate Coalitions in a Cartel
Formation Game GG 62.2003 Michael FINUS and Bianca RUNDSHAGEN: How the Rules of Coalition Formation Affect Stability of
International Environmental Agreements SIEV 63.2003 Alberto PETRUCCI: Taxing Land Rent in an Open Economy CLIM 64.2003 Joseph E. ALDY, Scott BARRETT and Robert N. STAVINS: Thirteen Plus One: A Comparison of
Global Climate Policy Architectures SIEV 65.2003 Edi DEFRANCESCO: The Beginning of Organic Fish Farming in Italy SIEV 66.2003 Klaus CONRAD: Price Competition and Product Differentiation when Consumers Care for the
Environment SIEV 67.2003 Paulo A.L.D. NUNES, Luca ROSSETTO, Arianne DE BLAEIJ: Monetary Value Assessment of Clam
Fishing Management Practices in the Venice Lagoon: Results from a Stated Choice Exercise CLIM 68.2003 ZhongXiang ZHANG: Open Trade with the U.S. Without Compromising Canada’s Ability to Comply
with its Kyoto Target KNOW 69.2003 David FRANTZ (lix): Lorenzo Market between Diversity and Mutation KNOW 70.2003 Ercole SORI (lix): Mapping Diversity in Social History KNOW 71.2003 Ljiljana DERU SIMIC (lxii): What is Specific about Art/Cultural Projects? KNOW 72.2003 Natalya V. TARANOVA (lxii):The Role of the City in Fostering Intergroup Communication in a
Multicultural Environment: Saint-Petersburg’s Case KNOW 73.2003 Kristine CRANE (lxii): The City as an Arena for the Expression of Multiple Identities in the Age of
Globalisation and Migration KNOW 74.2003 Kazuma MATOBA (lxii): Glocal Dialogue- Transformation through Transcultural Communication KNOW 75.2003 Catarina REIS OLIVEIRA (lxii): Immigrants’ Entrepreneurial Opportunities: The Case of the Chinese
in Portugal KNOW 76.2003 Sandra WALLMAN (lxii): The Diversity of Diversity - towards a typology of urban systems
KNOW 77.2003 Richard PEARCE (lxii): A Biologist’s View of Individual Cultural Identity for the Study of Cities KNOW 78.2003 Vincent MERK (lxii): Communication Across Cultures: from Cultural Awareness to Reconciliation of
the Dilemmas KNOW 79.2003 Giorgio BELLETTINI, Carlotta BERTI CERONI and Gianmarco I.P.OTTAVIANO: Child Labor and
Resistance to Change ETA 80.2003 Michele MORETTO, Paolo M. PANTEGHINI and Carlo SCARPA: Investment Size and Firm’s Value
under Profit Sharing Regulation IEM 81.2003 Alessandro LANZA, Matteo MANERA and Massimo GIOVANNINI: Oil and Product Dynamics in
International Petroleum Markets
(l) This paper was presented at the Workshop “Growth, Environmental Policies and Sustainability” organised by the Fondazione Eni Enrico Mattei, Venice, June 1, 2001
(li) This paper was presented at the Fourth Toulouse Conference on Environment and Resource Economics on “Property Rights, Institutions and Management of Environmental and Natural Resources”, organised by Fondazione Eni Enrico Mattei, IDEI and INRA and sponsored by MATE, Toulouse, May 3-4, 2001
(lii) This paper was presented at the International Conference on “Economic Valuation of Environmental Goods”, organised by Fondazione Eni Enrico Mattei in cooperation with CORILA, Venice, May 11, 2001
(liii) This paper was circulated at the International Conference on “Climate Policy – Do We Need a New Approach?”, jointly organised by Fondazione Eni Enrico Mattei, Stanford University and Venice International University, Isola di San Servolo, Venice, September 6-8, 2001
(liv) This paper was presented at the Seventh Meeting of the Coalition Theory Network organised by the Fondazione Eni Enrico Mattei and the CORE, Université Catholique de Louvain, Venice, Italy, January 11-12, 2002
(lv) This paper was presented at the First Workshop of the Concerted Action on Tradable Emission Permits (CATEP) organised by the Fondazione Eni Enrico Mattei, Venice, Italy, December 3-4, 2001
(lvi) This paper was presented at the ESF EURESCO Conference on Environmental Policy in a Global Economy “The International Dimension of Environmental Policy”, organised with the collaboration of the Fondazione Eni Enrico Mattei , Acquafredda di Maratea, October 6-11, 2001
(lvii) This paper was presented at the First Workshop of “CFEWE – Carbon Flows between Eastern and Western Europe”, organised by the Fondazione Eni Enrico Mattei and Zentrum fur Europaische Integrationsforschung (ZEI), Milan, July 5-6, 2001
(lviii) This paper was presented at the Workshop on “Game Practice and the Environment”, jointly organised by Università del Piemonte Orientale and Fondazione Eni Enrico Mattei, Alessandria, April 12-13, 2002
(lix) This paper was presented at the ENGIME Workshop on “Mapping Diversity”, Leuven, May 16-17, 2002
(lx) This paper was presented at the EuroConference on “Auctions and Market Design: Theory, Evidence and Applications”, organised by the Fondazione Eni Enrico Mattei, Milan, September 26-28, 2002
(lxi) This paper was presented at the Eighth Meeting of the Coalition Theory Network organised by the GREQAM, Aix-en-Provence, France, January 24-25, 2003
(lxii) This paper was presented at the ENGIME Workshop on “Communication across Cultures in Multicultural Cities”, The Hague, November 7-8, 2002
2002 SERIES
CLIM Climate Change Modelling and Policy (Editor: Marzio Galeotti )
VOL Voluntary and International Agreements (Editor: Carlo Carraro)
SUST Sustainability Indicators and Environmental Valuation (Editor: Carlo Carraro)
NRM Natural Resources Management (Editor: Carlo Giupponi)
KNOW Knowledge, Technology, Human Capital (Editor: Dino Pinelli)
MGMT Corporate Sustainable Management (Editor: Andrea Marsanich)
