Nine groups of mixed frequency risk variables during the period of January 2014 to March 2022 from macroeconomy, traditional finance industry, internet industry, and internet finance industry are adopted to timely and systematically measure the systemic risk of internet finance using the frequency-converting method and dynamic factor model. This research also use the Markov-switching model to identify the risk movement property and conduct out-of-sample forecasting using fixed parameter and rolling window forecasting methods. The results show that the mixed frequency risk variables have strong explanatory power in explaining the movements of internet financial risk; Internet finance risk has a significant regime-switching property with the risk level switching from high to low; Internet financial risk level exhibits a downward trend and will experience a moderate increase in the near future. However, it still resides in the low-risk regime in general. As internet finance return to its “finance” nature, the pro-cyclicality of risks and the resonance effect with traditional financial risks are increasingly emerging.