Inflation and Trading
Description
Replication Files for “Inflation and Trading” Codes: • 01a_rep_survey_data cleaning.do: cleaning raw survey data • 02a_rep_survey_data prep.do: preparing final survey dataset • 03a_rep_survey_data analysis.do: produces Figures 1-6 and Tables 1-5 and 8 • 02b_rep_bank_data prep.do: preparing final bank dataset • 03b_rep_bank_data analysis.do: produces Tables 6-7 Datasets: The folder 02_data contains survey and bank data. From the survey, we include pseudo data with the same structure as the original data needed to run the do-files 01a, 02a, and 03a, but the dataset contains only a random subsample of 300 respondents with random noise added to each continuous response. The original dataset is not available because it includes confidential information on customers of our partnering bank. • rep_survey_data raw.dta: raw survey data for a random subsample of 300 respondents and with added noise to each continuous variable. We also exclude open-ended responses at the beginning and end of survey for confidentiality reasons. These responses do not feature in the main analysis of the paper • rep_survey_data clean.dta: survey data after transformation of the raw variables • rep_survey_data final.dta: preparation of final survey dataset From the bank, we include a dataset with the same structure as the original data that allows the do-files 01b and 02b to run. The dataset includes only the necessary variables needed for the analysis, and we select a subsample of customers to match the 300 respondents randomly drawn from the set of survey respondents. The original datasets are not available since they use proprietary information from the partnering bank. • rep_bank_data sum stat pf.dta: portfolio summary statistics, coming from confidential portfolio data from the bank, and used for Table 1 • rep_bank_data sum stat trading: trading summary statistics, coming from tab6a • rep_bank_data tab1.dta: demographics data from bank • rep_bank_data tab6a: trading data from bank • rep_bank_data final.dta: final dataset from bank, which combines tab1, tab6a, and select variables from the survey for the subsample of survey respondents Runtime: We run the codes on a MacBook Pro laptop with Stata 19 MP. Runtime is below 10 minutes on real data and below one minute on pseudo data.