The transmission of significant information through commodity futures, stock and ESG bond markets in terms of returns and actual volatility is critical for investors to effectively reduce portfolio risk by understanding the dependencies and risk spillovers among these markets. This paper uses weekly data from February 28, 2014 to March 7, 2024 and employs the GARCHSK-Vine Copula-CoVaR methodology to explore the dependencies and risk spillovers among commodity futures, stock markets and ESG bond markets from the perspective of multidimensional higher moments. Compared to other models, our approach describes the higher moment characteristics of data and extends CoVaR from a bivariate to a multivariate framework, thus mitigating the underestimation of risk spillover effects that may occur due to the overlooked indirect impacts on financial assets. In particular, the results show that CoVaR fluctuated more dramatically after the outbreak of COVID-19 and the variables generally show an increase in lower tail correlation and a decrease in positive correlation. This suggests that in extreme market conditions, asset price declines will be more consistent, increasing the risk of collective declines. But in daily market fluctuations, price movements among assets are more independent and show lower correlation.