Cette thèse, divisée en trois chapitres, contribue à la vaste et récente littérature sur l'évaluation des actifs et la finance climatique. Le premier chapitre contribue à la littérature sur la finance climatique tandis que les deux derniers contribuent à la littérature sur l'évalutaion des actifs. Le premier chapitre analyse comment les politiques environnementales visant à réduire les émissions de carbone affectent les prix des actifs et la consommation des ménages. En utilisant de nouvelles données, je propose une mesure des émissions de carbone du point de vue du consommateur et une mesure du risque de croissance de la consommation de carbone. Les mesures sont basées sur des informations sur la consommation totale et l'empreinte carbone de chaque bien et service. Pour analyser les effets des politiques environnementales, un modèle de risques de long terme est développé dans lequel la croissance de la consommation comprend deux composantes: le taux de croissance de la consommation de carbone et le taux de croissance de la part de la consommation de carbone dans la consommation totale. Ce chapitre soutient que le risque de long terme de la croissance de la consommation provient principalement de la croissance de la consommation de carbone découlant des politiques et des actions visant à réduire les émissions, telles que l'Accord de Paris et la Conférence des Nations Unies sur le changement climatique (COP26). Mon modèle aide à détecter le risque de long terme dans la consommation des politiques climatiques tout en résolvant simultanément les énigmes de la prime de risque et de la volatilité, et en expliquant la coupe transversale des actifs. La décomposition de la consommation pourrait conduire à identifier les postes de consommation les plus polluants et à construire une stratégie d'investissement minimisant ou maximisant un critère environnemental de long terme. Le deuxième chapitre (co-écrit avec René Garcia et Caio Almeida) étudie le rôle des facteurs non linéaires indépendants dans la valorisation des actifs. Alors que la majorité des facteurs d'actualisation stochastique (SDF) les plus utilisés qui expliquent la coupe transversale des rendements boursiers sont obtenus à partir des composantes principales linéaires, nous montrons dans ce deuxième chapitre que le fait de permettre la substitution de certaines composantes principales linéaires par des facteurs non linéaires indépendants améliore systématiquement la capacité des facteurs d'actualisation stochastique de valoriser la coupe transversale des actifs. Nous utilisons les 25 portefeuilles de Fama-French, cinquante portefeuilles d'anomalies et cinquante anomalies plus les termes d'interaction basés sur les caractéristiques pour tester l'efficacité des facteurs dynamiques non linéaires. Le SDF estimé à l'aide d'un mélange de facteurs non linéaires et linéaires surpasse ceux qui utilisent uniquement des facteurs linéaires ou des rendements caractéristiques bruts en termes de performance mesurée par le R-carré hors échantillon. De plus, le modèle hybride - utilisant à la fois des composantes principales non linéaires et linéaires - nécessite moins de facteurs de risque pour atteindre les performances hors échantillon les plus élevées par rapport à un modèle utilisant uniquement des facteurs linéaires. Le dernier chapitre étudie la prévisibilité du rendement des anomalies à travers les déciles à l'aide d'un ensemble de quarante-huit variables d'anomalie construites à partir des caractéristiques de titres individuels. Après avoir construit les portefeuilles déciles, cet article étudie leur prévisibilité en utilisant leurs propres informations passées et d'autres prédicteurs bien connus. Les analyses révèlent que les rendements des portefeuilles déciles sont persistants et prévisibles par le ratio de la valeur comptable sur la valeur de marché de l'entreprise, la variance des actions, le rendement des dividendes, le ratio des prix sur les dividendes, le taux de rendement à long terme, le rendement des obligations d'entreprise, le TED Spread et l'indice VIX. De plus, une stratégie consistant à prendre une position longue sur le décile avec le rendement attendu le plus élevé et à prendre une position courte sur le décile avec le rendement attendu le plus bas chaque mois donne des rendements moyens et un rendement par risque bien meilleurs que la stratégie traditionnelle fondée sur les déciles extrêmes pour quarante-cinq des quarante-huit anomalies. / This thesis, divided into three chapters, contributes to the vast and recent literature on asset pricing, and climate finance. The first chapter contributes to the climate finance literature while the last two contribute to the asset pricing literature. The first chapter analyzes how environmental policies that aim to reduce carbon emissions affect asset prices and household consumption. Using novel data, I propose a measure of carbon emissions from a consumer point of view and a carbon consumption growth risk measure. The measures are based on information on aggregate consumption and the carbon footprint for each good and service. To analyze the effects of environmental policies, a long-run risks model is developed where consumption growth is decomposed into two components: the growth rate of carbon consumption and the growth rate of the share of carbon consumption out of total consumption. This paper argues that the long-run risk in consumption growth comes mainly from the carbon consumption growth arising from policies and actions to curb emissions, such as the Paris Agreement and the U.N. Climate Change Conference (COP26). My model helps to detect long-run risk in consumption from climate policies while simultaneously solving the equity premium and volatility puzzles, and explaining the cross-section of assets. The decomposition of consumption could lead to identifying the most polluting consumption items and to constructing an investment strategy that minimizes or maximizes a long-term environmental criterion. The second chapter (co-authored with René Garcia, and Caio Almeida) studies the role of truly independent nonlinear factors in asset pricing. While the most successful stochastic discount factor (SDF) models that price well the cross-section of stock returns are obtained from regularized linear principal components of characteristic-based returns we show that allowing for substitution of some linear principal components by independent nonlinear factors consistently improves the SDF's ability to price this cross-section. We use the Fama-French 25 ME/BM-sorted portfolios, fifty anomaly portfolios, and fifty anomalies plus characteristic-based interaction terms to test the effectiveness of the nonlinear dynamic factors. The SDF estimated using a mixture of nonlinear and linear factors outperforms the ones using solely linear factors or raw characteristic returns in terms of out-of-sample R-squared pricing performance. Moreover, the hybrid model --using both nonlinear and linear principal components-- requires fewer risk factors to achieve the highest out-of-sample performance compared to a model using only linear factors. The last chapter studies anomaly return predictability across deciles using a set of forty-eight anomaly variables built using individual stock characteristics. After constructing the decile portfolios, this paper studies their predictability using their own past information, and other well-known predictors. The analyses reveal that decile portfolio returns are persistent and predictable by book-to-market, stock variance, dividend yield, dividend price ratio, long-term rate of return, corporate bond return, TED Spread, and VIX index. Moreover, a strategy consisting of going long on the decile with the highest expected return and going short on the decile with the lowest expected return each month gives better mean returns and Sharpe ratios than the traditional strategy based on extreme deciles for forty-five out of forty-eight anomalies.

The central thesis of this dissertation is that by combining classical scheduling methodologies with those of inventory management and queueing theory we can better model, understand and solve complex real-world scheduling problems. In part II of this dissertation, we provide models of a realistic supply chain scheduling problem that capture both its combinatorial nature and its dependence on inventory availability. We present an extensive empirical evaluation of how well implementations of these models in commercially available software solve the problem. We are therefore able to address, within a specific problem, the need for scheduling to take into account related decision-making processes. In order to simultaneously deal with combinatorial and dynamic properties of real scheduling problems, in part III we propose to integrate queueing theory and deterministic scheduling. Firstly, by reviewing the queueing theory literature that deals with dynamic resource allocation and sequencing and outlining numerous future work directions, we build a strong foundation for the investigation of the integration of queueing theory and scheduling. Subsequently, we demonstrate that integration can take place on three levels: conceptual, theoretical and algorithmic. At the conceptual level, we combine concepts, ideas and problem settings from the two areas, showing that such combinations provide insights into the trade-off between long-run and short-run objectives. Next, we show that theoretical integration of queueing and scheduling can lead to long-run performance guarantees for scheduling algorithms that have previously been proved only for queueing policies. In particular, we are the first to prove, in two flow shop environments, the stability of a scheduling method that is based on the traditional scheduling literature and utilizes processing time information to make sequencing decisions. Finally, to address the algorithmic level of integration, we present, in an extensive future work chapter, one general approach for creating hybrid queueing/scheduling algorithms. To our knowledge, this dissertation is the first work that builds a framework for integrating queueing theory and scheduling. Motivated by characteristics of real problems, this dissertation takes a step toward extending scheduling research beyond traditional assumptions and addressing more realistic scheduling problems.

The central thesis of this dissertation is that by combining classical scheduling methodologies with those of inventory management and queueing theory we can better model, understand and solve complex real-world scheduling problems. In part II of this dissertation, we provide models of a realistic supply chain scheduling problem that capture both its combinatorial nature and its dependence on inventory availability. We present an extensive empirical evaluation of how well implementations of these models in commercially available software solve the problem. We are therefore able to address, within a specific problem, the need for scheduling to take into account related decision-making processes. In order to simultaneously deal with combinatorial and dynamic properties of real scheduling problems, in part III we propose to integrate queueing theory and deterministic scheduling. Firstly, by reviewing the queueing theory literature that deals with dynamic resource allocation and sequencing and outlining numerous future work directions, we build a strong foundation for the investigation of the integration of queueing theory and scheduling. Subsequently, we demonstrate that integration can take place on three levels: conceptual, theoretical and algorithmic. At the conceptual level, we combine concepts, ideas and problem settings from the two areas, showing that such combinations provide insights into the trade-off between long-run and short-run objectives. Next, we show that theoretical integration of queueing and scheduling can lead to long-run performance guarantees for scheduling algorithms that have previously been proved only for queueing policies. In particular, we are the first to prove, in two flow shop environments, the stability of a scheduling method that is based on the traditional scheduling literature and utilizes processing time information to make sequencing decisions. Finally, to address the algorithmic level of integration, we present, in an extensive future work chapter, one general approach for creating hybrid queueing/scheduling algorithms. To our knowledge, this dissertation is the first work that builds a framework for integrating queueing theory and scheduling. Motivated by characteristics of real problems, this dissertation takes a step toward extending scheduling research beyond traditional assumptions and addressing more realistic scheduling problems.