This paper analyzes sustainable investments of Swedish mutual funds. Morningstar’s CarbonRisk Score (CRS) - funds exposure to a future of low-carbon economy - is analysed in termsof returns, management fees and flows. The CRS measure was introduced March 2018 with ahistorical series from March 2017, without the market being aware. Analysing CRS before theintroduction is therefore greenwashing-bias free. An empirical approach with regressions findthat there is a payoff between return and alignment with a low-carbon economy future, CRS.A 1% increased abnormal return causes a 0.13 standard deviations higher CRS. Regressionsalso find no relationship between management fee and CRS. A correlation between flow andCRS is found but no causality. The shown payoff between return and CRS implies that fundswhich are well-performing are less sustainable. Fund managers maximising their returnthereby lead to unsustainable investments. To handle this, a policy of tax relief or subsidyshould be implemented for investing sustainable. The tax relief or subsidy should beproportional to the increased return renounced when investing sustainable.

Yes / The Paris Agreement, signed in 2015, sets ambitious goals for diminishing greenhouse gas emissions and restricting the rise in global temperature to achieve a less carbon-intensive and climate-resilient global economy. The Paris Agreement marked a defining moment in the worldwide response to global warming and has significantly affected the financial sector. Given this background, this research explores the effects of carbon risk on the cost of debt (CoD) in 1428 listed firms across seven economies from 2011 to 2020. The paper also reflects the post-Paris Agreement’s involvement and the ESG factors’ moderating effect in the empirical models. The study finds a significant impact of carbon risk on CoD following the implementation of the Paris Agreement. Notably, companies with higher carbon risk face higher borrowing rates. However, the effect of ESG on moderating the relationship between carbon risk and CoD is found to be insignificant. Further analyses confirm this finding, as individual pillars of ESG (governance and social aspects) also show insignificant moderating effects.

Responsible investing has become a trend throughout financial markets. As World’s economies pledge to decrease the amount of greenhouse gas (GHG) emissions, environmental policies like carbon pricing (CP) are expected to be strengthened; the above is attributed to the effort of internalizing the environmental costs of the current economic system. In the same context where asset owners have been demanding to the private sector for greater coverage of Environmental Social Governance (ESG) issues, understanding exposure and risk to carbon taxation and emission trading schemes (ETS) could be a major driver for responsible investing. Nonetheless, it has been found that this environmental policy to price emissions, falls behind from a harmonized cost per emission across sectors and geographies. Defined and assessed through a quantitative scenario analysis on scope 1 emissions, all information on carbon pricing set the basis for the model. From an investing perspective, the results showed higher exposure for the electricity sector by 2030 and 2050; nonetheless, the riskiest sector to invest at, was shown as industry. The above is based on the current and expected carbon dependency, and the expected increase in coverage from carbon pricing mechanisms respectively. In addition, aviation, which is a sub-sector from offroad transportation, showed to be the main source for this sector´s exposure and risk. It is concluded that the research carried out is a first step from a complete analysis on CP, as scope 2 emissions need to be assessed. / Att investera ansvarsfullt är en trend som ökar stadigt genom finansmarknader i världen idag. Då flera ekonomier i världen utlovar att minska mängden utsläpp av växthusgaser i linje med vissa klimatscenarion, så förstärks miljöpolicys som koldioxidutsläppspriser i ett försök att internalisera externaliteter i dagens ekonomiska system. I kontexten av när ägare av tillgångar börjar kräva större täckning i miljö, socialt ansvar och ägarstyrning, kan exponering och risk i koldioxidutsläppsbeskattning och handel av utsläppsoptioner vara en drivande faktor i ansvarsfullt investerande. En utmaning i att prissätta utsläpp genom miljöpolicys ligger i hur separerade båda mekanismerna är från en harmoniserad kostnad per utsläpp genom olika sektorer och geografier. Definierad och utvärderad genom en kvantitativ scenarioanalys av Scope 1-utsläpp, verkade all information om koldioxidutläppspriser som grund till modellen. Från ett investeringsperspektiv visade resultaten en högre exponering för elkraftssektorn till 2030 och 2050. Emellertid påvisades även att industrisektorn har störst risk för investeringar. Detta är baserat på elkraftens nuvarande och förväntade koldioxidutsläppsberoende och den förväntade ökningen i täckning från koldioxidutsläppsprismekanismer i industrisektorn. Vidare påvisades flygindustrin, som är en sidosektor av offroad-transport, vara den huvudsakliga källan för denna sektors exponering och risk. Avslutningsvis fastställs det att denna undersökning endast är ett första steg i en komplett analys av koldioxidpriser, då Scope 2 utsläpp även bör undersökas.

Current standard carbon footprint metrics attribute responsibility for a firm’s green house gas (GHG) emitting activities equally between an entity’s equity and debt. This study introduces a set of novel duration-weighted metrics which take into consideration the length of financing provided. These measure show promise for reporting footprints of debt portfolios, but further study of methodological robustness should be performed before they can be adopted widely. The measures are also attractive from a risk perspective as they are linearly dependent on duration and therefore are sensitive to yields. A factor portfolio is constructed using the new carbon intensity measure, and corporate yields are studied in a linear factor model. Other factors included derive from Nelson-Siegel parameterizations of US Treasury rates and the USD swap spread curve. Following the Fama-MacBeth procedure, the carbon factor is found not to persist over the 10-year period. / Nuvarande standardmått f ̈or koldioxidsavtryck i en portfölj tilldelar ansvaret för ett företags emitterande aktiviteter av växthusgas lika mellan aktier och skulder, där finansieringens längd inte beaktas. Ett ny durationsviktat mått introduceras i denna studie och dess lämplighet som metrik för rapportering undersöks. Studien visar att detta mått har potential för rapportering i kreditportföljer, men ytterligare studier av hur robust metoden är bör utföras innan den tillämpas brett. Måttet har attraktiva egenskaper eftersom den är linjärt beroende på durationen och därmed känslig gentemot obligationsavkastningen. En faktorportfölj konstrueras med hjälp av det nya kolintensitetsmåttet, och i en linjär faktormodell studeras företagsobligationsavkastning. Andra faktorer som inkluderas i modellen härstammar från Nelson-Siegel-faktorisering av US Treasury och USD swap- spread kurvorna. CO2-faktorn utvärderas med hjälp av Fama-MacBeths tvärsnittsmetod, och det konstateras att faktorn inte visar signifikans under den 10-åriga studieperioden.

In context of the forthcoming Corporate Sustainability Reporting Directive, this study examines how scope 3 emissions and the reporting thereof affect the cost of debt. Further, it investigates how scope 1 emissions affect the cost of debt and how the two scopes differ in materiality. As a theoretical foundation, this thesis uses previous research on environmental risk management, carbon risk premium, scope 3 emissions and cost of capital. By collecting a sample of 1710 firm-year observations for publicly listed European companies during the period 2019-2022, this quantitative study utilizes fixed effect regression models to find the relationship between scope 3 emission and cost of debt. No evidence of a relationship between scope 3 emissions and cost of debt is found. When looking at scope 1 emissions, the results show that companies with lower scope 1 emissions are rewarded by creditors with a reduced cost of debt. Regarding reporting of scope 3 emissions, we find no evidence suggesting that scope 3 disclosure lowers the cost of debt.

Cette thèse passe en revue certains facteurs de risques économiques (risque de revenu, risque de la finance parallèle, et risque carbone) en utilisant de nouvelles sources de données et méthodologies. Le premier chapitre examine comment la réponse de la consommation face au risque de capital humain affecte la finance des ménages. A partir de données conjointes sur la consommation, les revenus et les actifs des ménages américains, ce papier documente le lissage excessif de la consommation comme un facteur essentiel pour le choix de portefeuille et montrer qu’il peut expliquer les énigmes financières observées chez les ménages américains. Par ailleurs, le papier formalise l’effet du lissage excessif sur le choix de portefeuille à l’aide d’un modèle de cycle de vie où un ménage est confronté à un risque de revenu salarial idiosyncratique. Le modèle est calibré de façon à correspondre aux observations sur le cycle de vie de la détention d’actifs risqués des ménages américains. Le deuxième chapitre évalue le transfert de risques des banques dans les activités bancaires non réglémentées. En exploitant les variations dans les risques discutés par les banques dans leur rapports financiers et en utilisant les outils de l’analyse textuelle, ce document fournit une nouvelle mesure de l’activité bancaire non-réglementée. Le papier montre empiriquement que (1) les banques sont plus susceptibles de contourner les régulations lorsque leurs contraintes de fonds propres deviennent contraignantes, (2) il existe une relation positive entre le transfert de risque et le risque extrême des banques. Par la suite, le papier rationalise ce transfert de risque en utilisant un modèle macroéconomique avec un secteur financier. Dans le modèle, l’événement de défaut de paiement et la présence d’externalités dues à une application imparfaite de la réglementation encourage les banques à s’engager dans une stratégie de transfert des risques. Enfin, le papier utilise ce cadre pour étudier la régulation optimale. On montre qu’une taxe sur l’activité sectorielle réduit efficacement le transfert des risques des banques par rapport à d’autres politiques comme la réglementation des fonds propres de la banque. Enfin, le troisième chapitre aborde l’effet du risque carbone sur la stabilité économique. Nous étudions ce risque à l’aide de données de panel pour 50 États américains au cours des années 1998 à 2018. De plus, nous supposons une dépendance transversale des facteurs communs non observés (par exemple, les liens commerciaux, l’intégration financière) entre les États. En utilisant une approche d’émissions de carbone basée sur la consommation, ce chapitre montre qu’une diminution d’une unité des émissions de carbone est associée, à long terme, à une croissance de la production logarithmique par habitant de 4,5 points de pourcentage. En outre, nous trouvons des impacts différentiels dans la distribution du revenu par habitant des États. Ces résultats éclairent le débat sur la voie de transition optimale vers une économie sobre en carbone. / This thesis reviews some economic risk factors (labor income risk, shadow banking risk, and carbon risk) using new data sources and novel methodologies. The first chapter investigates how the response of consumption to human capital risk affects household finance. Using joint data on consumption, income, and assets of representative US households, I document the excess smoothness of consumption as an essential factor for portfolio choice and show that it can explain household finance puzzles. Furthermore, I formalized the effect of the excess smoothness on the portfolio choice using a structural life-cycle model where a household faces an idiosyncratic wage income risk. The model is calibrated to match relevant aspects of the dynamics and the life cycle of risky asset holding from the PSID. The second chapter assesses banks’ risk-shifting in the non regulated banking activity, also called shadow banking. Exploiting variations in risks disclosed by banks in their financial reports and using textual analysis tools, this document provides a new measure non regulated banking activity. The paper empirically documents that (1) banks are more likely to shift risk out of the regulator’s reach when their risk-based capital constraints become binding, (2) there is a positive relationship between risk-shifting and tail risk of banks. The paper then rationalizes banks’ risk-shifting behavior using a macroeconomic model with a financial sector. In the model, the event of default on debt and the presence of externality due to imperfect regulation enforcement encourage banks to engage in risk-shifting strategies. As a result, banks behave as cross-sector arbitrageurs. Finally, the paper uses this framework to study optimal regulation. We show that a tax on sectoral activity effectively reduces banks’ risk-shifting compared to other bank’s equity regulation policies. Finally, the third chapter studies the effect of carbon risk on economic stability using a consumption-based carbon emissions approach for 50 U.S. states over the years 1998 - 2018. The paper assumes a cross-sectional dependence from unobserved common factors (e.g., trade linkage, financial integration) between the states. Under this assumption, we find that one unit decreases in carbon emissions is associated with 4.5 percentage points decrease in the per capita output growth over the long run. Besides, we find differential impacts across the distribution of per capita states income. These findings inform the debate over the optimal transition path toward a low carbon economy.

The impact of climate change on the financial performance of companies is of concern to bank credit processes. The main objective of this research was to develop a South African contextualised credit process that incorporates environmental risk. The research methodology comprised of a mixed-method being content analysis – the qualitative portion and the Probability of Default prediction using a Merton Model and the Hoffmann and Busch (2008) carbon risk analysis model - the quantitative portion. A content analysis of the banks’ Annual Reports, Integrated Reports and Sustainability Reports showed that, while South African banks follow a qualitative approach to embedding environmental risk into their credit process, none of the four banks that formed part of the study divulged their quantitative approach to embedding environmental risk. The study used a proximity matrix method to examine the level of embedding. The second part of the study, which used prior studies as the benchmark, adopted the following: (1) a simulated carbon tax regime as a proxy for an environmental risk, and (2) the Hoffmann and Busch (2008) carbon risk analysis tool and the Merton Model (1974) as the bank credit process proxies. The second part of the study used a sample of 33 JSE-listed Carbon Disclosure Project reporting companies out of a population of 107. The carbon risk analysis showed that the companies in the materials and energy sector have a high carbon risk. However, the results from the Merton Model showed that the companies have enough profit to cushion the additional carbon tax liability, given the insignificant shift in probability of default between the three scenarios, where financial data had (1) no carbon tax, (2) was adjusted for a carbon tax with incentives, and (3) adjusted for carbon tax without incentives. Triangulation of the results from the content analysis, carbon risk analysis and the probability of default analysis confirms that South African banks do not fully integrate environmental risk across the credit value chain or process in the 2010 to 2017 period. However, the carbon risk analysis shows a heavy dependency on carbon sources for critical inputs into the South African companies’ production processes, which if not checked, will affect the credit portfolios of banks. / Finance, Risk Management and Banking / D. Phil (Management Studies)

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.