Carbon finance - evidence of successful carbon credit projects and carbon risk in credit risk
Climate change is a global phenomenon imposing environmental, social and economic challenges to the international community. With the help of environmental regulation global warming is intended to be limited to 2 _ C. There are diverse climate policies, such as carbon taxation and carbon emission trading worldwide with different scopes with respect to GHG emissions covered and obliged entities next to voluntary carbon trading regimes. Accordingly, GHG emissions are priced in order to internalize their external effects. Thus, the cost structure of regulated companies is influenced by carbon pricing. In addition, potential climate policies and increasing carbon prices exposes companies to carbon risk. Carbon risk is related to an asset's or company's GHG emissions and the implied value destruction due to climate policies. The scope of this thesis is to investigate two carbon finance related topics. First, we identify and compare the determinants of success of carbon credit projects within a mandatory and a voluntary carbon trading regime. The carbon certificates generated under the Clean Development Mechanism (CDM) are valid certificates under the European Union Emission Trading System (EU ETS). Therefore, companies obliged to set o_ their carbon emissions under the EU ETS can invest in a CDM project in a non-Annex I country according to the Kyoto Protocol and reduce emissions in a least-developed, developing or emerging country. Thus, the CDM represents a mandatory standard. A voluntary standard is the Verified Carbon Standard (VCS). Here, companies can invest in carbon credit projects or purchase Verified Carbon Units (VCUs) in order to compensate their carbon emissions. This voluntary participation can have reputational reasons or be in preparation of a mandatory carbon credit regime. To understand the factors influencing success of carbon credit projects is important for companies investing in such projects in order to avoid losses. In addition, it is relevant for policy makers in order to adapt the market design such that the cost-efficiency is maintained. This relates to the rationale of carbon credit projects conducted in non-Annex I countries, namely, that emission reductions can be achieved at lower costs compared to those in industrialized countries. By that policy makers can secure use of carbon credit mechanisms and the benefit of technology as well as knowledge transfer. In particular, we focus on wind and hydro power, energy efficiency and biomass projects regarding the CDM and VCS. The success indicator and dependent variable in our analysis is the log quotient of actual and estimated emissions reductions. As independent variables function the projects' scale, log of actual credit issuing time, the number of involved countries, the projects' technology and their host region. We apply OLS and stepwise regression allowing the independent variables to interact and identify partly multi-collinearity in the stepwise regression results. Furthermore, we find off-setting effects in the stepwise regression results. Our results suggest that carbon credit projects underperform in general, yet, VCS projects underperform slightly less compared to CDM projects. With respect to the projects' methodology large scale CDM energy efficiency and VCS hydro power projects perform better than small scale ones within our sample. In contrast, VCS biomass projects show an increased rate of carbon credit issuance when applying the small scale methodologies. Additionally, we identify economies of scale and time regarding the overall VCS sample, VCS and CDM wind as well as CDM hydro power projects. The number of participating countries influences the success of carbon credit projects, too. However, with respect to the stepwise regressions we identify off-setting effects as disadvantages linked to the number of involved countries vanish with increasing crediting time. According to our results this is valid for CDM hydro power projects and CDM and VCS biomass projects. Concerning the host regions, Africa seems not to be advantageous for CDM energy efficiency projects, whereas CDM biomass projects located in Latin America have an increased rate of credit issuance. Our findings motivate further interesting and valuable research questions, in particular, with respect to the considered standards and geographical focus. Second, we find evidence of carbon risk in credit risk as a consequence of environmental policies. Credit Default Swaps (CDSs) are credit derivatives and their spread reflects the creditworthiness and probability of default of the reference entity. As a proxy for carbon risk we relate EUA futures prices to CDS spreads of 33 European companies while taking into account further CDS spread determinants in order to isolate the effect of carbon risk. With reference to the Merton model (Merton, 1974), which explains a firm's credit spread with the help of its debt and equity representing the firm value, we include the changes in stock prices and in stock return volatility into our regression models. The change in a market index is another explanatory variable determining the change in CDS spreads. The explanatory power of these three CDS spread determinants is advocated by Galil et al. (2014). In particular, we apply OLS company-wise regressions as well as panel regressions using the First-Difference estimator due to unit roots in the panels. With regard to the panel regressions we distinguish between the sectors Power & Heat, Oil & Gas, Chemicals and Motor industry, because we assume that carbon intensive sector face a greater carbon risk exposure compared to companies belonging to low carbon sectors. Furthermore, we take into account the companies' yearly Scope 1 carbon emissions in order to categorize the companies as "green", "gray" and "brown". Since we assume that carbon risk has a stronger effect on the creditworthiness of high emitting companies. In order to provide a longer-term perspective we consider the CDS spread maturities of one, three and five years and match the EUA futures price maturities. Our observation period starts with the third trading phase of the EU ETS in 2013, which has stricter rules with regard to the allocation on carbon allowances and the overall amount of emission allowances decreases yearly and ends in August 2018. Moreover, we consider two sub-periods, namely 2013- 2016 and 2017-2018, because we assume that carbon risk awareness increases over time and because of the steady increase in EUA futures prices beginning in 2017. Our results confirm our expectations, such that carbon risk is evident rather in the longerterm maturity. In addition, the five years maturity captures most market information as it is most liquidly traded and the CDS spread determinants of Galil et al. (2014) were found explicitly for the five years maturity. Furthermore, the awareness of carbon risk seems to increase over time as carbon risk is identified rather in the later observation period 2017-2018. This confirms that our sub-period selection is reasonable. With regard to the sector panels the panels associated to high carbon emissions are, in general, identified to be exposed to carbon risk. Whereas, "green" companies seem to benefit from increasing EUA futures prices with respect to their creditworthiness. Our empirical approach can be extended with respect to the number of companies under consideration as well as with respect to the carbon risk proxy. When changing the carbon risk proxy a measure of social cost of carbon or a carbon risk factor could be discussed. This could also solve the issue of interaction between the EUA futures prices and stock prices as well as stock return volatilities in some cases of our regression results. Furthermore, our empirical approach could be enlarged geographically taking into account further regulatory regimes next to the EU ETS. Concluding, our findings are of considerable value for companies regarding their access to finance, for policy makers concerning the regulatory influence of climate policies and actors on the credit and financial markets with respect to the incorporated carbon risk in company valuation. In this thesis we provide evidence of the success determinants of carbon credit projects, which function as a measure to hedge a company's carbon risk exposure as well as evidence of carbon risk exposure of European companies regarding their creditworthiness. We, therefore, cover two important topics in carbon finance, which are relevant to policy makers as well as companies and other actors on the carbon and credit markets.