With the enhancement of global economic instability and the deepening of China's interest rate liberalization, the volatility of financial assets has been increasing. The optimal allocation of asset portfolio is still the core issue in the field of financial investment theory research and practice. Black-Litterman model has been widely used in asset allocation practice because it solves the problem that traditional mean-variance model is sensitive to parameters and allows investors' views to be incorporated into the model. However, the determination of investor perspective matrix has always been a difficulty in the theoretical and applied research of Black-Litterman model. Based on this, macro factors are integrated into the construction of investor perspective matrix. Support vector regression (SVR) model based on principal component analysis is used to estimate the viewpoint matrix, and the SVR-Black-Litterman asset allocation model with macro factors is constructed. To test the validity of this model, the model is compared with other common asset allocation models. The empirical results show that the model constructed has a good market performance.