This paper proposes state space modeling and Kalman filtering as the theoretical foundation to explore the impact of China’s monetary policy on the offshore RMB market and capital flows. In the state space model, the relationship between observable and unobservable variables is established, with the Hidden Markov Model representing the state vector as a first-order Markov process to estimate unobservable variables. Additionally, a variable parameter state space model is proposed to alter the relationship between explanatory variables and the dependent variable. Kalman filtering, as the core of the state space model solution algorithm, calculates the likelihood function through prediction error decomposition to estimate unknown parameters. The optimal lag of the model is determined to be 2 after constructing the VAR model in the empirical analysis, and the first-order autoregressive model is used to calculate variable parameters. The elasticity coefficient of the depth of development of the offshore RMB market on capital flows is the most significant variable, with a value of 0.96. Following this, the RMB exchange rate has an elasticity coefficient of 0.78.