Abstract
This study examined the impact of the security, trust, and reliability of internet-based financial transactions, including internet banking, online securities brokerage, foreign exchange (FX), cryptocurrency, and smartphone-specific securities, on customer satisfaction. SHAP (SHapley Additive exPlanations) of machine learning was adopted to avoid assuming a particular stochastic distribution. The main results were: (1) Security, trust, and reliability were essential to customer satisfaction in using internet-based financial transactions overall. (2) The importance of secu-rity/trust was confirmed, especially in internet banking and cryptocurrency. Three key items of security and trust, which encompass a full range of security measures, effective management, and company credibility, significantly impact customer satisfaction, particularly in internet banking. On the other hand, the effect of company credibility was most pronounced in the cryptocurrency.
(3) The effect of usefulness was larger than that of security, trust, and reliability. Cryptocurrency is a new technology that is expected to support the e-society of the future, but it also entails risks. Therefore, customers are concerned about the company’s credibility and require the disclosed information to be clear. The contribution of this study is to compare the particularities among internet-based financial transactions in Japan from the viewpoint of customer satisfaction by machine learning.
