Deep Learning-Based Stock Movement Prediction: A Case Study of Hindenburg Research's Impact on Adani Group and Nifty-50

The Nifty 50 index is a benchmark for the Indian stock market that calculates the average worth of the top 50 global companies listed on the National Stock Exchange (NSE). These 50-companies stock price momentum directly affects the Nifty 50 Index. In this work, Adani Enterprises (ADANIENT) and Adani Ports and Special Economic Zone Ld. (ADANIPORTS) stock datasets are used to train the proposed model. In addition, this letter gives a review of the Hindenburg research report (i.e., sentiment based) and its short selling in the Adani group. Due to short selling in ADANIENT and ADANIPORTS, t h e Nifty 50 was affected and came under a downtrend. The dataset used from December 2022 (1<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">st</sup> week) to March 2023 (2<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">nd</sup> week) on a 1-hour time stamp in image form, and for the image scrap, tradingview.com have been considered. Based on three experts' opinions, the dataset is classified into three categories: holding position, bearish, and bullish. MACD, BB, and candlestick pattern indicators use to make the dataset. The dataset is made for ADANIENT, ADANIPORTS, and the Nifty 50 Index. It is trained on the SCNN model and Inception V3 pre-trained model with a dropout layer. SCNN model used a categorical cross-entropy loss function and Adam optimizer. In result analysis, the model attained validation accuracy of 97% for ADANIENT, 95% for ADANIPORTS, and 94% for the Nifty 50 Index.