Using Bull and Bear Index of Deep Learning to Improve the Indicator Model on Extremely Short-term Futures Trading

Abstract

Day traders, who trade over time horizons of one day or less, account for around 30% of the total transaction volume at present. Most day traders’ trading strategies are based on their own experiences or news headlines. They may also rely on technical indicators, such as the Relative Strength Index (RSI), to predict short-term trading opportunities and stock index turning points for making selling and buying decisions with respect to stock index futures. This study determined exact RSI indicators, which enhanced the accuracy of short-term stock index prediction. We then tested the proposed model’s performance during an unprecedented crisis such as COVID-19. We used artificial intelligence techniques, such as the SMO algorithm, to evaluate the performance of the proposed model and apply empirical methods on short-term stock index futures datasets to explore the impact of different RSI indicators on the turning points of the stock index futures. The results show that RSI 20 based on regular and COVID-19 periods can enable day traders to achieve higher profits compared to the RSI 30 index. Keywords: Day traders, Relative strength index, Short-term investment, Artificial intelligence, COVID-19.