#Exam 3 textlab code

The code for the Newey-West OLS robustness check in Appendix C, and produces the results in Table 9. It also produces the QQ plots (Figs 4-6) in the appendix and the contents of Tables 3, 4 and 5. This is the source code for the ARDL analysis and the corresponding test statistics. This script recreates the Tables 1-2, Figures 1-3 7-9.

1 for months when Gyorgy Matolcsy was the governor Matolcsy: Central bank governor dummy.1 for months when Andras Simor was the governor effective_fed_i: Effective US FED interest rates.net_hawk_agh: the Hawkish sentiment score computed with the AGH dictionary.lm_baseline: sentiment scores with dictionary baseline using the Loughran-McDonald dictionary (net positive) on the the press release corpus.nethawkish: the Hawkish sentiment score computed with our dictionary on the press release corpus.date_text: The release date of the press release.mt_cohesion: The cohesion index of the monetary council.r15y: yields of bonds with maturity rates of 15 years.r10y: yields of bonds with maturity rates of 10 years.r5y: yields of bonds with maturity rates of 5 years.r3y: yields of bonds with maturity rates of 3 years.

r1y: yields of bonds with maturity rates of 1 years.

r6m: yields of bonds with maturity rates of 6 months.

r3m: yields of bonds with maturity rates of 3 months.

long_term_resc: rescaled r5y r10y r15y factor scores.

medium_term_resc: rescaled r1y r3y yield factor scores.

out_year_resc: rescaled r1y r3y r5y r10y r15y yield factor scores.in_year_resc: rescaled r3m r6m yield factor scores.total_resc: rescaled combined yield factor scores (all time horizon).log_long_term: logarithm of the rescaled r5y r10y r15y factor scores.log_1to3year: logarithm of the rescaled r1y r3y yield factor scores.log_out_year: logarithm of the rescaled r1y r3y r5y r10y r15y yield factor scores.log_in_year: logarithm of the rescaled r3m r6m yield factor scores.log_total: logarithm of the rescaled combined yield factor scores (all time horizon).core_i_yoy: Core inflation, year on year change.The macro controls are in the data/input/data_macro.csv file. This script assembles the parts created above and merges them into the macro controls. NB: Missing data is coded as '.' (for Stata compatibility). The output of this script is the final dataset which is used for creating the figures and tables in the article. The R script transforms the PCA scores from the Stata output by rescaling and log transforming the variables. The do file creates the scores.xlsx from the data/input/yields.xlsx file, which contains the PCA scores for the short, medium and long term bond yields. The input file used for this script is data/input/mnb-mt-voting.xlsx. The scripts contains the step by step instructions for the index reproductions. It uses as input the excel sheet available at the webpage of the Central Bank of Hungary (MNB): (last accessed: 6). This script replicates the monetary council cohesion index calculation. The sentiment scores for the Newey-West OLS robustness check using the L-M and AGH dictionaries are also computed in this script. The text analysis is done in R with the quanteda package. This script replicates the sentiment scores using the press release text data and the dictionary we created. For a detailed description of each script see below. The R source code is in the scripts folder, the needed data is in the data/input and data/output folders. This repository contains the neccesary source codes to reproduce the analysis in the article. Replication materials for the 'The effect of central bank communication on sovereign bond yields: The case of Hungary' article

Exam 3 textlab