Stability of Price Prediction Rule for High-Frequency Trading

When using the data mining method for stock trading, it is essential to confirm the future stability of the rule found by the data mining method. Even if a rule can explain past data, in situations where the model behind the data changes, there is a risk of loss if an investment is made based on the rule learned using past data. In this study, we analyzed the mechanism behind the high-frequency trading prediction rule proposed in past research and demonstrated its stability. By demonstrating stability, this paper shows that investment using the rule can continue to be the basis for low-risk investment methods in the future