This paper examines the performance of "clean," "brown," and "dirty" stocks in the S&P 500 from January 2010 to September 2022 using panel random effect estimation and factor models. It also uses cointegration analysis to assess the long-term relationship between risk premiums and two carbon risk factors: "brown minus clean" and "dirty minus clean." Finally, we use random walk tests to examine whether carbon risks are priced, and therefore, the S&P 500 market is weakly efficient. Findings indicate that the brown portfolio outperforms the clean portfolio in factor models, likely due to market trends where energy, driven by rising oil and gas prices, outperforms all other sectors. The results also show that the two carbon risks and the political risk have a negative and significant impact on the risk premium and that the excess return series do not follow random walks and are weak form inefficient.