O objetivo deste estudo é estimar o viés de substituição de produtos no cálculo da inflação ao consumidor, ou seja, estimar a inflação levando em conta a possibilidade da troca de bens dentro de uma cesta de produtos em resposta à mudança de preços relativos. Isso ocorre porque a fórmula utilizada atualmente pelo IBGE, tanto para o INPC quanto para o IPCA, para medir a inflação ao consumidor é a de Laspeyres modificado base móvel (índice do Bureau), que considera a mesma cesta de bens e serviços ao longo do tempo. Este índice tende a superestimar o aumento do custo de vida justamente por não considerar as trocas. Seguindo trabalhos anteriores, a estimação do viés se deu comparando um índice Laspeyres para um subconjunto do IPCA com a inflação mensurada pelo índice de Theil-Tornqvist para o mesmo subconjunto de produtos. Este índice se aproxima de um índice de custo de vida, logo, que considera a substituição de bens. No entanto, ele necessita atualizações freqüentes das cestas de bens e serviços ou das estruturas de ponderação. Como não existem no Brasil pesquisas de consumo das famílias que forneça estruturas de ponderações periódicas, estas tiveram que ser estimadas. Para tanto, foram utilizadas previsões de um modelo de sistema de demanda AIDS baseado nos microdados da POF 95-96. O viés de substituição estimado foi de 3,33 p.p. de agosto de 1999 a junho de 2006, o que equivale a dizer que a inflação ao consumidor foi superestimada em 0,31 p.p. ao ano. Pela impossibilidade de trabalhar com o nível mais desagregado do IPCA (o subitem), certamente, o viés calculado é subestimado. Caso o viés estimado fosse descontado dos reajustes dados às aposentadorias, pensões e demais auxílios concedidos pelo Ministério da Previdência e Assistência Social, o governo poderia ter poupado de 2000 a junho de 2006, aproximadamente, R$ 8 bilhões. / The objective of this study is to estimate the substitution of products bias in the calculation of consumer inflation, therefore, estimate the inflation taking into account the possibility of switching goods in a basket of products, in response to a change in relative prices. This happens because the estimation formula used by IBGE, both with INPC an IPCA, to measure consumer inflation is Laspeyres (Bureau\'s index), witch considers the same basket of goods over time. This index tends to overestimate the increase in the living cost, by not taking into account the substitution of products. Following previous works, the estimation of the bias was made comparing a Lapeyres index for a subgroups of IPCA with the inflation measure by the Theil-Tornqvist index for the same subgroups. This gets closer to an index of cost of living, which considers the substitution of goods. However, it needs frequent updates of the baskets of goods and services or of the weighted structures. As there are no surveys of family consumption in Brazil that provide periodic weighted structures, these had to be estimated. To do it, were used micro data of POF 95-96. The substitution bias estimated was 3,33 p.p of August 1999 to June 2006, which is equivalent of saying that the consumer inflation was overestimated in 0,31 p.p per year. With the impossibility of working with a more highly disaggregated level of IPCA (the sub items), certainly the calculated bias was underestimated. If the bias estimated was discounted from adjustment given to retirement, pensions and other benefits granted by the Ministry of Welfare and Social Assistance, the government could have saved, from 2000 to June 2006, approximately R$ 8 billions.

O objetivo deste estudo é estimar o viés de substituição de produtos no cálculo da inflação ao consumidor, ou seja, estimar a inflação levando em conta a possibilidade da troca de bens dentro de uma cesta de produtos em resposta à mudança de preços relativos. Isso ocorre porque a fórmula utilizada atualmente pelo IBGE, tanto para o INPC quanto para o IPCA, para medir a inflação ao consumidor é a de Laspeyres modificado base móvel (índice do Bureau), que considera a mesma cesta de bens e serviços ao longo do tempo. Este índice tende a superestimar o aumento do custo de vida justamente por não considerar as trocas. Seguindo trabalhos anteriores, a estimação do viés se deu comparando um índice Laspeyres para um subconjunto do IPCA com a inflação mensurada pelo índice de Theil-Tornqvist para o mesmo subconjunto de produtos. Este índice se aproxima de um índice de custo de vida, logo, que considera a substituição de bens. No entanto, ele necessita atualizações freqüentes das cestas de bens e serviços ou das estruturas de ponderação. Como não existem no Brasil pesquisas de consumo das famílias que forneça estruturas de ponderações periódicas, estas tiveram que ser estimadas. Para tanto, foram utilizadas previsões de um modelo de sistema de demanda AIDS baseado nos microdados da POF 95-96. O viés de substituição estimado foi de 3,33 p.p. de agosto de 1999 a junho de 2006, o que equivale a dizer que a inflação ao consumidor foi superestimada em 0,31 p.p. ao ano. Pela impossibilidade de trabalhar com o nível mais desagregado do IPCA (o subitem), certamente, o viés calculado é subestimado. Caso o viés estimado fosse descontado dos reajustes dados às aposentadorias, pensões e demais auxílios concedidos pelo Ministério da Previdência e Assistência Social, o governo poderia ter poupado de 2000 a junho de 2006, aproximadamente, R$ 8 bilhões. / The objective of this study is to estimate the substitution of products bias in the calculation of consumer inflation, therefore, estimate the inflation taking into account the possibility of switching goods in a basket of products, in response to a change in relative prices. This happens because the estimation formula used by IBGE, both with INPC an IPCA, to measure consumer inflation is Laspeyres (Bureau\'s index), witch considers the same basket of goods over time. This index tends to overestimate the increase in the living cost, by not taking into account the substitution of products. Following previous works, the estimation of the bias was made comparing a Lapeyres index for a subgroups of IPCA with the inflation measure by the Theil-Tornqvist index for the same subgroups. This gets closer to an index of cost of living, which considers the substitution of goods. However, it needs frequent updates of the baskets of goods and services or of the weighted structures. As there are no surveys of family consumption in Brazil that provide periodic weighted structures, these had to be estimated. To do it, were used micro data of POF 95-96. The substitution bias estimated was 3,33 p.p of August 1999 to June 2006, which is equivalent of saying that the consumer inflation was overestimated in 0,31 p.p per year. With the impossibility of working with a more highly disaggregated level of IPCA (the sub items), certainly the calculated bias was underestimated. If the bias estimated was discounted from adjustment given to retirement, pensions and other benefits granted by the Ministry of Welfare and Social Assistance, the government could have saved, from 2000 to June 2006, approximately R$ 8 billions.

Submitted by Daniella Santos (daniella.santos@fgv.br) on 2009-10-19T12:18:21Z No. of bitstreams: 1 Dissertacao_Joao_Luiz_Ayres.pdf: 463363 bytes, checksum: 3d72501f2cf6e2f0d459c3e04d7a6cfc (MD5) / Approved for entry into archive by Antoanne Pontes(antoanne.pontes@fgv.br) on 2009-10-21T11:32:25Z (GMT) No. of bitstreams: 1 Dissertacao_Joao_Luiz_Ayres.pdf: 463363 bytes, checksum: 3d72501f2cf6e2f0d459c3e04d7a6cfc (MD5) / Made available in DSpace on 2009-10-21T11:32:25Z (GMT). No. of bitstreams: 1 Dissertacao_Joao_Luiz_Ayres.pdf: 463363 bytes, checksum: 3d72501f2cf6e2f0d459c3e04d7a6cfc (MD5) / We investigate temporary sales using a very detailed data set for 13 years of retail prices in the Brazilian economy, with price quotes collected every 10-days. We find strong evidence of co-movement between frequency and size of sales and the macroeconomic varibles. The common belief in the literature that sales do not react to changes in macroeconomic variables may be due to little macro volatility in the analysed countries so far. / Nós investigamos promoções temporárias usando uma base de dados detalhada de 13 anos sobre preços ao consumidor no Brasil, com cotações de preços coletadas decendialmente. Nós encontramos forte evidências da existência de relação entre a frequência e tamanho de promoções e as variáveis macroeconômicas. A crença comum na literatura de que promoções não reagem a mudanças nas variáveis macroeconômicas pode ser devido a baixa volatilidade do cenário macro- econômico nos países analisados até o presente momento.

Cette thèse met en évidence l’aspect macroéconomique du degré de report du taux de change sur les prix dans la zone euro. Nous utilisons un large éventail de méthodes économétriques récentes afin de fournir des mesures robustes sur la transmission du taux change ainsi que sur ses déterminants macroéconomiques. Notre recherche révèle le rôle prépondérant des facteurs macroéconomiques dans le déclin récent du degré de report. Une conséquence directe de ce résultat est que la baisse du taux de transmission du change n’est pas nécessairement un phénomène structurel, et il peut être ainsi résolu par des politiques macroéconomiques conjoncturelles. Par exemple, l’adoption de régimes de politique monétaire plus crédibles avec l’engagement de maintenir une inflation faible joue un rôle important dans la réduction de la sensibilité des prix aux variations du change. Ceci est particulièrement valable pour les pays dont les politiques macroéconomiques sont historiquement laxistes. Ainsi, la poursuite de politiques économiques solide et lisible au sein de l’UEM peut être un outil efficace pour réduire le degré de report du taux de change / This thesis highlights the macroeconomic aspect of the exchange rate pass-through to domestic prices in the euro area countries. We use a wide range of up-to-date econometric methods in order to provide robust measures of the rate of pass-through as well as to shed further light on its macro determinants. The main finding of our research is the prominent role of macroeconomic forces in driving the recent declin of the transmission of currency movements. A direct consequence of this result is that the lowering in the rate of pass-through is not necessarily a structural phenomenon and it may be solved via macroeconomic policies. For instance, the shift to a more stable monetary policy conditions with credible and anti-inflationary regime would reduce the sensibility of prices to exchange rate changes. This is especially true for countries with historically poor macroeconomic policies. Thus, a better macroeconomic management with a sounder set of policies within the EMU may be an effective tool for reducing the degree of pass-through

Dans le chapitre 1, j’examine les résultats des fusions entre producteurs européens d’énergie en termes d’efficacité. Je calcule l’éco-efficacité en utilisant l’analyse de l’enveloppement des données et l’indice de productivité Malmquist-Luenberger. Je trouve que les fusions horizontales nationales, qui sont soigneusement réglementées, n’ont pas d’impact. Les fusions horizontales transfrontalières nuisent à l’éco-efficacité à court terme mais la stimulent deux ans après l’achèvement. Les fusions verticales nuisent à l’éco-efficacité. Je présente des suggestions de politiques concernant la réglementation des fusions. Le chapitre 2 est un travail conjoint avec Julian Hinz. Nous enquêtons sur les effets de l’embargo russe auto-imposé sur les importations de produits alimentaires en provenance des pays occidentaux. Nous construisons un modèle ricardien avec des liens sectoriels, des échanges de biens intermédiaires et une hétérogénéité sectorielle dans la production. L’étalonnage du modèle avec des données réelles permet de simuler les résultats de l’embargo en termes de changements de bien-être et de prix. Nous quantifions en outre l’impact sur les prix à la consommation en Russie à l’aide de la méthode des doubles différences. Le chapitre 3 est basé sur un article co-écrit avec Cristina Herghelegiu. Nous enquêtons sur l’utilisation des conditions commerciales internationales (Incoterms). Ce sont les schémas prédéfinis de la répartition des coûts et des risques entre les acheteurs et les vendeurs. Nous nous appuyons sur un ensemble de données très détaillées sur les exportations russes durant la période 2012-2015. Nous constatons que les grandes entreprises sont plus susceptibles d’assumer des responsabilités. Les gros acheteurs assument plus de responsabilités, quelle que soit la taille du vendeur, alors que les gros vendeurs le font uniquement lorsque leur partenaire est petit. C’est plus probable que les risques et les coûts sont sur les acheteurs dans les transactions de biens intermédiaires et de biens d’équipement. / In Chapter 1 I investigate firm-level efficiency outcomes of mergers between the European energy producers. I compute eco-efficiency using data envelopment analysis and the Malmquist-Luenberger productivity index. I find that carefully regulated domestic horizontal mergers do not have a statistically significant impact. Cross-border horizontal mergers hamper eco-efficiency in the short run but stimulate it two years after completion. Vertical mergers are detrimental to eco-efficiency. I put forward policy suggestions regarding the regulation of mergers. Chapter 2 is joint work with Julian Hinz. We investigate the effects of self-imposed Russian embargo on food import from Western countries. We build a Ricardian model with sectoral linkages, trade in intermediate goods and sectoral heterogeneity in production. The calibration of the model with real data allows to simulate the outcomes of embargo in terms of changes in welfare and prices. We further quantify the impact on consumer prices in Russia with the difference-in-differences estimator. Chapter 3 is based on a paper co-written with Cristina Herghelegiu. We investigate the use of International Commercial Terms. They are pre-defined schemes of repartition of costs and risks between buyers and sellers, which serve to mitigate the uncertainty. We rely on a highly detailed dataset on Russian exports over the 2012-2015 period. We find that big firms are more likely to take on responsibilities. Big buyers bear more responsibilities regardless of the seller size, whereas big sellers do so only when their partner is small. Risks and costs are more likely on buyers in transactions of intermediate and capital goods.