In this paper, we follow the framework of measuring systemic risk proposed by Acharya et al (2012). We use the non-parametric Fourier transform method proposed by Malliavin and Mancino (2009) to construct the dynamic continuous-time model which aims to estimate the form of capital shortfall to provide a systemic risk indicator for financial firms. At the same time,we evaluate this indicator in several aspects. We measure the systemic risk (SRISK) of financial firms in the United States, Taiwan and China to verify that the indicator has a certain ability to explain and predict financial system risk. Regulators can access this indicator to catch the whole systemic risk of domestic financial markets and the contribution of the individual enterprise to the system risk.