Unsupervised Method for Discovering Expert Traders on Social Trading Services

摘要

Social trading services are now becoming increasingly popular among people who are interested in investments. A notable feature of social trading services is that we can simply “copy” the trades of experts to achieve profits. However, identifying expert traders who achieve exceptional and consistent returns is a major challenge. In this study, we propose an unsupervised method for identifying expert traders by applying portfolio theory developed in financial engineering. Experimental results obtained with a real data set demonstrate the superior performance of our method.