What do Sentiments of Budget Speech Mean for Stock Returns?
Description
In this dataset, we have compiled the unstructured data of Annual Indian Budget Speeches (AIBS) from 1991 to 2024, which has been further used to quantify the budget speech sentiments in terms of polarity and subjectivity. The AIBS data extracts the word cloud and word count, governments' term-wise policy analysis, their focus, and thematic shifts. This dataset also contains the daily SENSEX data from 01-07-1997 to 26-07-2024, which is extracted from yahoofinance using Python programming. Furthermore, both the sentiments and the SENSEX data have been merged to study the impact of sentiments on stock returns.
Files
Steps to reproduce
1. Put the following folder/ files in the Google Drive: (a) AIBS1991-2024-2025 (b) Stopwords (c) Stopwords_txt Run Python File: NLP_in_budget_speeches1991_2024EM, by opening as a Colab file and using (a), (b), and (c). Extract the zip file (a) and keep the extracted folder on the drive. Files (a) and (b) can be kept in the drive as they are. By running the codes, the following results will be extracted: (i) Figure 1 (a) and (b): Word Cloud and the word Count from the AIBS: 1991-2024 (ii) Extraction of Daily Sensex data from 01-07-1997 to 26-07-2024 using yahoofinance in the Financial Data section of the Python file. (iii) Extraction of Polarity and Subjectivity Score from filtered and raw data in the Sentiment Analysis section of the Python file (iv) Figure 5: Sentiment Scores of AIBS in terms of Polarity and Subjectivity (1991-2024) (v) After extracting the Sentiment Score of Polarity, Subjectivity, and Sensex Daily data, they are merged to create the file – Sensex_Daily.xlsx. (vi) The file Sensex_Daily.CSV, or Sensex_Daily.xlsx data is used to estimate the equations from 1-16 and produce the results as provided in Tables 2 and 3, as well as Tables 4 to 8 in Appendix B, and Figures 15-22. 2. You can create a path on your machine to run the following ‘Stata’ and Excel files: (a) Sensex_Daily.do (b) Sensex_Daily.CSV (i) By running the do file, you extract the results as shown in Tables 2 and 3, Tables 4 to 8 in Appendix B, and Figures 15-21 in Appendix B. (ii) Observing the least AIC (Akaike Information Criterion), the Excel file – AIC_Figure 6.xlsx has been created, which further produced Figure 6 3. You can create a path on your machine to run the following ‘R’ and Excel files (a) Estudy_Combineplot.R (b) Sensex_Daily.xlsx You can extract Figure 22 in Appendix B by running the ‘Estudy_Combineplot.R' file for a robustness check. 4. Using the folder 1 (a) AIBS1991-2024-2025, execute the following Python file: Keyword_Analysis_Fig2_3_4EM.ipynb. By executing this file, you can extract Figures 2 to 4 of the paper. 5. Using the folder 1 (a) AIBS1991-2024-2025, execute the following Python file: Regime_Analysis_Fig7_14EM.ipynb. By executing this file, you can extract Figures 7 to 14 of the paper. 6. The main source file to be executed is: (a) What Do Sentiments of Budget Speech Mean-RRR (b) ‘What Do Sentiments of Budget Speech Mean’ is the LaTeX file within the zip file – ‘What Do Sentiments of Budget Speech Mean-RRR’. This file has been created in Overleaf.
Institutions
- Indian Institute of Technology Jodhpur