Stock Market Forecasting Using Deep Learning and Technical Analysis: A Systematic Review

Abstract

Stock market forecasting is one of the biggest challenges in the financial market since its time series has a complex, noisy, chaotic, dynamic, volatile, and non-parametric nature. However, due to computing development, an intelligent model can help investors and professional analysts reduce the risk of their investments. As Deep Learning models have been extensively studied in recent years, several studies have explored these techniques to predict stock prices using historical data and technical indicators. However, as the objective is to generate forecasts for the financial market, it is essential to validate the model through profitability metrics and model performance. Therefore, this systematic review focuses on Deep Learning models implemented for stock market forecasting using technical analysis. Discussions were made based on four main points of view: predictor techniques, trading strategies, profitability metrics, and risk management. This study showed that the LSTM technique is widely applied in this scenario (73.5%). This work significant contribution is to highlight some limitations found in the literature, such as only 35.3% of the studies analysed profitability, and only two articles implemented risk management. Therefore, despite the widely explored theme, there are still interesting open areas for research and development. INDEX TERMS Deep learning, profitability metrics, risk management, stock market forecasting, systematic review, technical analysis, technical indicators.

Literature Survey and Research Agenda of Risk Determinants in Indian Equities and Machine Learning

“…Another key consideration is to assess predictive strength of Linear, Nonlinear, and Hybrid models to forecast stocks and indexes' equity performance through this literature review. The existing work in literature on stock price prediction may broadly be grouped under OLS or simple Regressing, Panel data Methods [26], [87], Timeseries Models (such as ARIMA) [72], [74], [88], Granger Causality [31], [43], [33]) and Machine Learning and Deep learning [68], [71], [76], [84] [69], [76]. Finally, Underlying the investment decision and risk decisions or selection of risk determinants are the psychological bias experienced by investors.…”

BovDB: A data set of stock quotes for Machine Learning on all companies from B3 between 1995 and 2020

“…In the financial area it is possible to forecast lots of scenarios using these data set, for instance, calculations of investment risk [Basak et al 2019], predictions of stocks [Hu et al 2021] [Thomaz et al 2021] and technical analysis [Li and Bastos 2020] of stocks. The use of neural networks as stock classifiers makes it possible to predict the behavior of one or more stocks and suggest the purchase or sale of a given asset at a given moment, indicating a high/moderate uptrend, stability, or moderate/ sharp drop.…”

Know Sure Thing based Machine Learning Strategy for Predicting Stock Trading Signals

“…Although there were numerous intelligent trading methods devised in the literature [20], none of them was based on concepts used by technical analysts for forecasting stock trading signals.…”

“…Time series modelling approaches like autoregressive integrated moving average (ARIMA) [20][21], and machine learning strategies like logistic regression (LR) [22][23], support vector machine (SVM) [24][25], random forest (RF) [26][27][28], decision trees [27][28], etc. are widely used by researchers to predict stock market.…”