It is the dataset used in the paper MulFSA: Multi-level Financial Sentiment Analysis Framework for Bond Market containing 1,353,585 texts from the Chinese securities market during 2013--2023.