Emotion recognition based on EEG signals for stock market purchase decisions analysis

Abstract

Emotion plays a significant role during a decision making process and greatly influences the investor’s behaviour. In this study, the participants’ emotions were derived from electroencephalogram (EEG) while they were trading in the stock market stimulation. This stimulation is performed using actual stock data from Bursa Malaysia via the JStock software. Stock market is a reliable barometer to measure the economic condition of a country. In the past, researchers have found that there is a significant relationship between stock traders` emotions and financial decision making, particularly if their emotions change before stock trading. The capability for emotion regulation, relates to the stock traders to invest wisely as the emotional states can have unpredictable effects on decision-making at different times. The lack of control on one’s emotion may cause losses in the stock market eventually resort to wealth erosion and widespread economic disruption. A negative emotion can reduce the rationality of the investor in decision making and may lead them to lose money. However, positive emotion and mood among investors lead towards higher stock trading returns. Therefore, this research sought to study the investors’ emotion and mood based on the stock trading stimulation and to develop a computational model for affect recognition by processing investor`s emotion in order to improve their investment returns. Another focus of this research was to analyse the model based on the stock trading stimulation. A total of 30 investors consisting of 15 active 15 active investors and 15 passive investors were chosen to participate in this experimental research. Active investors are referring to someone who have at least 2 years of experience and actively trading in the stock market investment. Whilst for the passive investors, they are someone who have a knowledge on investment or passive in the investment activities. Based on the EEG emotion recognition, the findings indicated that the emotions of active investors were dominantly shifted towards positive emotion, whereas the emotions of passive investors were most likely to be shifted towards negative emotion. Passive investors also tend to have a negative mood, and more likely to trade and experience losses. On the other hand, some of the active investors may have a negative mood but still hold onto their stocks. This research has demonstrated that the mood can be derived from these relationship and introduced the EEG estimation model in the stock market purchase decision. Last but not least, for the future works, the research in neurofinance might also focus on the time factors of stock investment. The researchers can set up a conditional situation in the investment activities to understand more about the right timing for the investors to get a profit. It is suggested that due to the limited duration of stimulations, some investors may not be able to provide better informed decisions. Future research should also look into the development of methods or tools to improve the stock market investment returns based on EEG emotions and mood.