The present document determine the existence of multiple bubbles in China's corporate debt, through the estimation of a Bayesian Hidden Markov Model, using Hamiltonian Markov Chains and Non-U-Turn Sampler (NUTS) algorithm. This is novel, since up to now there are no empirical studies that analyze China's corporate debt, nor bubbles in the credit market using Hidden Markov Models. This methodology is highly pertinent in the context of credit booms, since it allows modeling the explosive behavior of a bubble considering the non-linearity of the dynamics of China's corporate debt. In addition, Bayesian algorithms allows to date stamp and estimate the bubbles simultaneously, and also effectively address the problems associated with the low frequency of the data.