Cette thèse analyse comment les fluctuations du prix du pétrole affectent les économies des pays exportateurs de produits de base. Plus précisément, l'objectif de cette thèse est d'étudier les impacts macroéconomiques, financiers et environnementaux des fluctuations des prix du pétrole, en accordant une attention particulière à l'hypothèse de la monnaie du pétrole. À cette fin, cette thèse se compose de trois chapitres. Les premier et deuxième chapitres portent sur le taux de change réel des devises de plusieurs pays exportateurs de pétrole. Le troisième chapitre explore les liens entre le prix des biocarburants et le compte courant des pays émergents et en développement exportant ou important des matières premières agricoles contrôlant l'effet non linéaire potentiel exercé par le prix du pétrole sur cette relation. Ces chapitres montrent que si le prix du pétrole a un effet macroéconomique sur les économies exportatrices de pétrole et les pays exportateurs de produits agricoles, son impact varie d'un pays à l'autre et il n'y a pas de règle unique pour décrire le fonctionnement de ces économies. / This thesis analyzes how fluctuations in the price of oil affect the economies of commodity exporting countries. More specifically, the aim of this thesis is to investigate the macroeconomic, financial and environmental impacts of oil price fluctuations, by paying particular attention to the oil currency hypothesis. To this end, this thesis is composed of three chapters. The first and second chapters deal with the real exchange rate of the currencies of several oil exporting countries. The third chapter explores the links between the price of biofuels and the current account of emerging and developing countries exporting or importing agricultural raw materials controlling for the potential nonlinear effect exerted by the price of oil on this relationship. Altogether these chapters show that while the price of oil has a macroeconomic effect on oil exporting and agricultural commodities exporting countries, its impact varies across countries and there is no one fits all rule.

In a developing oil-based economy like Libya the availability of finance is largely affected by the availability of oil revenues which are subjected to disturbances and shocks. Therefore, the decision to save and invest a certain ratio of the country's aggregate output is, to large extent, determined (and affected) by the shocks in the oil markets rather than the requirements of economic development. In this study an attempt is made to determine the optimal rate of saving and investment, both defined as a ratio of the aggregate output, according to the requirements of economic development. For this purpose, a neo-classical Ramsey-type model for Libya is constructed and applied to obtain theoretically and empirically the optimal saving and investment rate during the period (1965-1991). The results reveal that Libya was investing over the optimal level during the oil boom of 1970s and less than the optimal level during the oil crisis of 1980s. In addition, an econometric investigation of the determinants of actual investment by sector (agriculture, non-oil industry, and services) is carried out in order to shed lights on how possible it is for Libya to adjust actual investment towards its optimal level. It is found that, as expected, the most important factor which can be used in this respect is the oil revenues or, generally, the availability of finance. In addition, the study reveals that investment in agriculture is associated, during the period of study, with a very low marginal productivity of capital whereas marginal productivity was higher in both non-oil industry and services. Finally, the study investigates also the future potential saving and investment rates and concludes that the economy, which has already reached its steady state, can be pushed out towards further growth if the economy can be able to increase the level of per worker human capital, proxied by the secondary school enrolment as a percentage of population. / Secretariat of Higher Education in Libya and Libyan Interests Section in London

In a developing oil-based economy like Libya the availability of finance is largely affected by the availability of oil revenues which are subjected to disturbances and shocks. Therefore, the decision to save and invest a certain ratio of the country's aggregate output is, to large extent, determined (and affected) by the shocks in the oil markets rather than the requirements of economic development. In this study an attempt is made to determine the optimal rate of saving and investment, both defined as a ratio of the aggregate output, according to the requirements of economic development. For this purpose, a neo-classical Ramsey-type model for Libya is constructed and applied to obtain theoretically and empirically the optimal saving and investment rate during the period (1965-1991). The results reveal that Libya was investing over the optimal level during the oil boom of 1970s and less than the optimal level during the oil crisis of 1980s. In addition, an econometric investigation of the determinants of actual investment by sector (agriculture, non-oil industry, and services) is carried out in order to shed lights on how possible it is for Libya to adjust actual investment towards its optimal level. It is found that, as expected, the most important factor which can be used in this respect is the oil revenues or, generally, the availability of finance. In addition, the study reveals that investment in agriculture is associated, during the period of study, with a very low marginal productivity of capital whereas marginal productivity was higher in both non-oil industry and services. Finally, the study investigates also the future potential saving and investment rates and concludes that the economy, which has already reached its steady state, can be pushed out towards further growth if the economy can be able to increase the level of per worker human capital, proxied by the secondary school enrolment as a percentage of population.

L’objectif de la thèse est d’étudier le lien entre le prix du pétrole et les performances macroéconomiques dans les pays exportateurs de pétrole. En adoptant une perspective d’économétrie appliquée, nous recourons à plusieurs techniques récentes de l’économétrie des données de panel. Pour cela trois thèmes sont envisagés. Nous identifions d’abord les fondamentaux de la croissance économique dans ces pays exportateurs de pétrole. En tenant compte du caractère dynamique de la croissance, nos résultats mettent en avant le rôle clé de la rente pétrolière dans l’explication de la croissance économique de ces pays. L’étude des interactions met en évidence qu’à court terme, un choc pétrolier positif améliore le compte courant et bénéficie à la croissance économique mais entraîne aussi une appréciation du taux de change dans les pays exportateurs de pétrole. Pour ce qui concerne les interactions hors prix du pétrole, celles-ci sont limitées. Nous montrons ensuite que les fondamentaux du taux de change effectif réel des pays exportateurs de pétrole reposent principalement sur les termes de l’échange, la productivité relative et les dépenses publiques. L’examen des mésalignements de change montre l’existence d’une forte hétérogénéité entre les pays étudiés. Nous montrons que le régime de change fixe est plus approprié pour diminuer l’ampleur des mésalignements dans les pays exportateurs de pétrole. / The aim of the thesis is to study the link between oil prices and macroeconomic performances in oil exporting countries. Adopting an applied econometrics approach, we use recent techniques of panel data econometrics. For this, three themes are envisaged. We identify first the main determinants of economic growth of these countries. Considering the dynamic nature of growth, our results highlight the key role of the oil rent in explaining economic growth in these countries. The study of macroeconomic interactions highlights that, in the short term, a positive oil price shock improves the current account and boosts economic growth but also leads to an appreciation of the exchange rate in oil exporting countries. Regarding the interaction between macroeconomic variables excluding oil prices, we find that they are relatively weak. We then show that the fundamentals of the real effective exchange rates in oil exporting countries are mainly the terms of trade, the relative productivity and government expenditures. The study of exchange rate misalignments shows the existence of a strong heterogeneity among the countries studied. However, our investigations highlight the key role of exchange rate regimes in explaining the magnitude of these exchange rates misalignments. In particular, we show that the fixed exchange rate regime is more appropriate to reduce the magnitude of the exchange rate misalignments in oil-exporting countries.

Cette thèse de doctorat aborde le rôle de la politique dans une économie exportatrice de pétrole sous forme de trois essais. Chaque essai tente d'apporter des réponses à une problématique liée à la réponse de la politique monétaire face aux chocs externes, en particulier le choc pétrolier. A ce titre, nous construisons trois modèles dynamique et stochastique d'équilibre général (DSGE) multisectoriels que nous calibrons et estimons sur des pays producteurs de pétrole. Dans le premier essai, nous montrons que le syndrome hollandais sous ses deux effets, dépense et ressource, semble avoir lieu dans une économie exportatrice de pétrole seulement lorsque les salaires sont flexibles et les prix rigides dans le cas d'un régime de change fixe. En d'autre terme, les simulations montrent que le syndrome hollandais est évité si les prix sont rigides et les salaires sont flexibles lorsque les autorités monétaires adoptent une règle de ciblage d'inflation ; les prix et les salaires sont rigides, quelque soit l'objectif de la Banque centrale dans les deux cas : aubaine et boom. Nous montrons également, en comparant les sources de fluctuation qui conduisent au syndrome hollandais que la hausse de gisement pétrolier (boom) conduit à une plus forte désindustrialisation de l'économie comparé à l'aubaine. Enfin, le régime de change flexible semble améliorer le bien être des ménages. Dans le deuxième essai , nous comparons trois règles de politique monétaire (ciblage d'inflation, ciblage de taux de change et ciblage de l'inflation sous jacente) face à quatre chocs externes (prix du pétrole, taux de change, terme de l'échange et taux d'intérêt international) subis par un pays exportateur de pétrole. Pour ce faire, nous construisons un modèle DSGE à deux secteurs (pétrolier et non-pétrolier) estimé sur des données trimestrielles de l'économie algérienne en utilisant l'approche bayésienne. Les résultats montrent que, globalement, la réponse des variables macroéconomiques du modèle, est similaire sous les trois règles de politique monétaire. Notre principal résultat est la dépréciation du Dollar américain et la hausse du prix du pétrole constituent la principale source de fluctuation cyclique de l'économie algérienne. Aussi, et comme prédit par la théorie, la dépréciation du dollar américain a significativement contribué à la détérioration des termes de l'échange et du compte courant. Dans ce cas, la Banque centrale peut adopter une politique de dévaluation pour éviter les effets de ces chocs. Dans le troisième essai, nous considérons un échantillon de 16 pays exportateurs de pétrole que nous divisons en deux sous échantillons afin de comparer l'occurrence du syndrome hollandais dans deux principales catégories : les pays fortement dépendants du pétrole et les pays faiblement dépendants du pétrole. Nos principaux résultats montrent que six parmi huit pays fortement dépendants subissent les effets du syndrome hollandais. Dans l'autre sous échantillons, seul un pays sur huit subit le syndrome hollandais sous ses effets dépenses et ressources. Toutefois, concernant la règle de politique monétaire à adopter face au syndrome hollandais, les résultats obtenus ne sont pas tous identiques dans tous les pays. Il semblerait que, l'adoption d'une politique monétaire appropriée dépend essentiellement des caractéristiques structurelles de chaque pays. / The aim of this thesis is to analyze the impact of external shocks on oil exporting economies and the role of monetary policy in this context. It consists of three essays. In the first essay, we build a Multi-sector Dynamic Stochastic General Equilibrium (DSGE) model to investigate the impact of both windfall (an increase in oil price) and boom (an increase in oil resource) on an oil exporting economy. Our model is built to see if the two oil shocks (windfall and boom) generate, in the same proportion, a Dutch Disease effect. Our main findings show that the Dutch disease effect under its two main mechanisms, namely spending effect and resource-movement effect, occurs only in the case of flexible wages and sticky prices, when exchange rate is fixed. We also compare the source of fluctuations that leads to a strong effect in term of de-industrialization. We conclude that the windfall leads to a stronger effect than a boom. Finally, the choice of flexible exchange rate regime helps to improve welfare.In the second essay, we estimate, by using the Bayesian approach, a DSGE model for Algerian economy investigating the dynamic effect of four external shocks (oil price, real exchange rate, international interest rate and foreign inflation), and examining the appropriate monetary policy rule. Our main findings show that, over the period 1990Q1-2010Q4, core inflation target is the best monetary rule to stabilize both output and inflation. In the third essay, we investigate the impact of the recent increase of oil price on a small open oil exporting economy. For this, we estimate a Dynamic, Stochastic, General equilibrium (DSGE) model for some oil producing countries using the Bayesian approach. We consider, in this essay, a sample of 16 oil exporting countries (Algeria, Argentina, Ecuador, Gabon, Indonesia, Kuwait, Libya, Malaysia, Mexico, Nigeria, Oman, Russia, Saudi Arabia, United Arab Emirates, and Venezuela) over the period from 1980 to 2010, except for Russia where our sample begins in 1992. In order to distinguish between high-dependent and low-dependent countries, we use two indicators : the ratio of fuel exports to total merchandise exports and the ratio of oil exports to GDP. We estimate the median for each ratio on our 16 studied countries. Countries above (below) the median are considered as high (low) oil dependent economies. We verify if the first group is more sensitive to the Dutch disease effect. We also assess the role of monetary policy. Our main findings show that in the first sample, namely high oil dependant economies, 6 countries are affected by the Dutch disease (decrease in the manufacturing production). Low oil dependant countries, are less affected by the fluctuation of oil price. Indeed, only one country has suffered a Dutch disease effect after the shock. Nevertheless, Regarding the appropriate monetary policy rule, we find that both inflation targeting and exchange rate rules may be effective to contain the size of the Dutch disease effect. Our results suggest that in Algeria and Saudi Arabia, inflation targeting offers better performances. We observe the opposite in Gabon, Kuwait, Oman, and Venezuela. Such results are consistent with economic theory. Indeed, we see that in more open economies and smaller countries (in terms of economic size), the exchange rate rule is preferable to inflation rule. Venezuela seems an exception. Such country does not fulfill the traditional criteria favoring the choice of the exchange rule. In fact, this exception is only apparent. First, if we consider the volatility, we see that Venezuela is among the most volatile economy. Second, Venezuela suffers from a fiscal dominance effect: both inflation rate and fiscal deficit are the highest relative to other studied countries.