Explainable-machine-learning-based online transaction analysis of China property rights exchange capital market

The emerging China Property Rights Exchange Capital Market shows distinctive features in its organizational structure compared to the existing over-the-counter (OTC) market. In order to regulate trading behavior, regulators have imposed strict rules on the trading process in this capital market. Trading institutions have standardized the process online to enhance trading efficiency. While the size and value of bids in the market are growing, there is a problem of decreasing deal closure rates. The recommendations proposed in the current study are more indicative than specific. To investigate the impact of online standardized transaction process elements on the transaction rate within China Property Rights Exchange Capital Market, this paper undertakes research using an explainable machine learning approach to online transaction analysis. Drawing on real data from the platform, this study constructs a framework for examining prediction and influencing factors of transaction results based on explainable machine learning-based standard transactions. The best prediction model is determined through metrics such as prediction accuracy, and the SHAP value is then applied to assess the model features for interpretability. To validate the modeling method's effectiveness, the analysis results are compared with traditional empirical methods utilizing the same dataset. The results indicate that the XGBoost model has higher forecast precision and is more interpretable than traditional empirical methods with the addition of SHAP values. Using the interpretable XGBoost method based on SHAP values, it is found that the four characteristics of duration, number of working days, relisting and location have the greatest impact on the transaction turnover rate. The results of the study can provide a decision-making basis for trading institutions to improve the transaction process and increase the transaction rate, which can help to enhance the resource allocation and optimization efficiency of China Property Rights Exchange Capital Market.