The proportion and quality of financial non-performing assets is a direct re-flec-tion of the risk management ability of banks and financial institutions. Re-search on the auction value of non-performing assets and the construction of a reliable non-performing asset value assessment model can help improve the effi-ciency of resource allocation. Taking the first-hand data from Alibaba auction website as the data source, the main influencing factors affecting the auction value of non-performing assets are explored. Again on the basis of BP neural network to construct a model of non-performing assets auction value assessment, and fur-ther make empirical analyses for testing, the model has accuracy, enriches the research of financial non-performing assets assessment.