Abstract: In recent years, all countries in the world have focused on tackling climate change and reducing carbon emissions. Currently, it is internationally recognized that the carbon emission trading mechanism is the most effective market mechanism to deal with climate change and control carbon emissions. It can guide funds to the low-carbon industrial chain with the help of price signals. Low-carbon technology will be the strategic commanding point in the future global competition. The development of this industry requires an effective carbon financial capital market to drive a large amount of social capital to low-carbon technology industry. Carbon finance came into being in this context. It is a modern financial innovation that relies on carbon trading, promotes energy conservation and emission reduction by means of financial technology, and serves sustainable economic development. Commercial banks and other financial institutions in China are facing complex and unstable market environments when they participate into international carbon financial market. It is necessary to develop scientific models for measuring the integrated risk of carbon financial market by considering the multiple sources and interdependent relationship of different risk factors. In this paper, the nonparametric kernel estimation method is used to determine the marginal distributions of price risk and exchange risk in carbon financial market. Then, the goodness-of-fit test is used to choose the optimal Copula function that can depict the nonlinear and dynamic dependent structure of risk factors, by which the integrated Conditional Value at Risk (CVaR) is measured effectively. Some findings are drawn by Kupiec back testing and comparison with the conventional risk measurement techniques. The nonparametric Copula-CVaR model can overcome the limitations of the conventional methods in measuring the dependency of multi-source risk factors. When determining the marginal distributions, this approach can avoid model setting risk and parameter estimating error caused by parametric methods. It gives full consideration to tail risk and provides a new way for the measurement of the integrated risk in carbon financial market. The main contributions of the research work are as follows. (1) At the theoretical level, the multi-source risks of the carbon market are clarified and the deficiency of the single risk measurement theory is made up in the existing literature. (2) At the practical level, the accuracy of multi-source integration risk measure is improved in the carbon financial market, which provides decision-making reference for the financial institutions to participate in the international carbon finance business.

Key words: carbon finance, risk management, dependency, Copula function, Value at Risk