Article Preview
TopIntroduction
Contemporary asset pricing models factor investor sentiment while predicting market trends. Sentiment analysis has been a hot topic for the past decade to decode stock market trends. Empirically, it was concluded that daily changes in stock prices are not corroborated with any such change in the firm’s fundamentals (e.g., Shiller, (1981); Roll, 1988; R2, 1988; Cutler, 1991, 1988; De Long, 1990) suggesting that qualitative information content may induce stock returns.
Researchers use various methods for conducting sentiment analysis, including Natural language processing (NLP), text analysis, and computational linguistics, apart from machine learning. Though there is extant literature that reveals each methodology's relative merits, as discussed in the section on the IS literature review, it is realized that the key to any analysis lies within the meaning of the terms we use relative to a context. The concept of 'sentiment' varies from the datasets we use and is not the same. 'Sentiment' as understood in the stock market or Fin-tech domain is not the same as what we mean in the marketing/H.R. domain or customer analytics. Related articles narrate approaches to determine positive, neutral, or negative polarity from the text analyzed. In the finance domain, all positive/neutral/negative polarity is not sentiment. The polarity determined is to be segregated further in terms of ‘rational’ and ‘irrational’ components and ‘overreaction’ or ‘underreaction’ to the information content.
The term ‘investor sentiment’ used in the stock market (Wolfe, Lau and Westin, 1994) is to be understood as investors ‘irrational’ beliefs about the risk-return profile of a stock or an index with the available information. To extract sentiment, researchers use a deep learning approach taking into account the integrated market scene (Sohangir et al.,2018).
Research Question
In the Fin-Tech context, merely determining polarity is not sufficient extracting 'sentiment,' we need an approach to predict future market direction. Strategic financial managers use the prediction models to manage the sentiment favorably by initiating policies and implementing programs that pilot the sentiment in the predetermined direction. To manage sentiment, there is an underlying obligation for the analysts to determine 'irrational' elements. The concept of 'irrationality' is relative to certain stock features (comprising earning potential, book value, Price-Earnings relation, Price-earnings growth, product mix, industry affiliations, management style, and policies), phase in the business cycle, and the holistic market scene. Such clinical analysis needs a ‘multi-value-factor’ approach demanding the services of a deep learning algorithmic model. The research gap, thus identified in the article, sets the paper’s research objectives as under:
Research Objectives
The research objectives of the paper are:
- •
Develop a methodology (Zeng et al.,2018; Zhang et al.,2020; Milecki et al.,2021) to extract market sentiment from a pre-trained FINBERT model (Yang, UY, and Huang,2020) to predict the market prices/index.
- •
To adopt a deep learning model for predicting the stock price movement of NSE NIFTY50 and compare with state-of-the-art baseline techniques like Naïve Bayes and Support Vector Machines to validate its incremental contribution to accuracy in the prediction of the stock price and direction.
There is a need to perform sentiment analysis differently for financial jargon and non-financial jargon in news articles. This would help give a more authentic sentiment score to financial indicators quantified in the text and improve the prediction performance of the stock market for which the paper proposes a FINBERT model pretrained on a financial phrase bank which can perform authentic sentiment extraction from financial news and accurately predict the stock market direction.
The rest of the paper is structured as follows: the literature review is expounded below. This is followed by the methodology adopted in the paper. The results and inferences drawn from the study are illustrated. The implications of the research (both theoretical and managerial) are discussed. Finally, the conclusion and directions for future research are provided. The references are then stated.