Abstract: Currently the low-carbon economy has become one of the strategic transformations of economic development measures. Carbon financial business has become an important innovation field in the financial sector that financial institutions boost the development of low-carbon economy. While the risk control is always the key factor affecting the success of financial innovation. As a developing country, China takes part in the international carbon financial transactions relying on Clean Development Mechanism (CDM). Under the CDM, indirect finance dominates carbon financial market where commercial banks play the important intermediary role. With the emerging of low-carbon economy in China, many financial institutions, especially the commercial banks, explore a lot of carbon financial business such as credit support business as CERs usufruct pledge loan, factoring financing, CDM equipment leasing, and other services as financial advisers, financial products of carbon credits, carbon accounts hosting business. Participating in the carbon finance business, banks are facing many risks as international carbon price fluctuations, carbon trading settlement currency exchange rate fluctuations and so on. And between the multi-source risk factors, a symbiotic and complexity of correlation is exist. In this paper, the carbon financial market risk of commercial banks is investigated in this paper, and ICE CERs future price and the EURO against CNY exchange rate as two risk factors financial time series sample data, which are chosen in the websites of Intercontinental Exchange (ICE) and State Administration of Foreign Exchange (SAFE) from 2009 to 2012 are selected. First, according to financial time series' feature, ARMA-GARCH models are used to portray the characterization of carbon price risk and exchange rate risk, the treatment methods for nonlinear relationship between the risk factors on Copula function are studied, and then the integrated VaR of carbon market risk through building Copula-ARMA-GARCH model and Monte Carlo simulation is calculated. Our empirical study shows that:(1) Financial time series of carbon price and exchange rate are both volatility clustering and heteroscedasticity. (2) Compared with risk factors' VaR, carbon price risk is higher than exchange rate risk in the carbon transactions. (3) If the correlation of carbon market risk factors is ignored, risk will be overestimated. (4) Supervision on exchange rate can reduce carbon market risk in a certain extent.This article contributes to find out amulti-source risks integration measurement technique to accord with the characteristic of carbon financial assets price volatility. The carbon financial risk factors have various sources and symbiotic, therefore, portraying the characteristics of risk factors respectively then the integration provides a theoretical framework for the multi-source risk measurement research; it provides theoretical basis for commercial banks to control the carbon market risks effectively, and promote carbon finance innovation.

Key words: carbon finance, market risk, carbon price, exchange rate, Copula-ARMA-GARCH