global-disclosure-data

Private firm disclosure around the world - data and descriptive analysis

This repository contains the replication package for the paper
Private firm disclosure around the world - data and descriptive analysis
by Christian Bernard, Ulf Brüggemann, and Jonas Materna.

The project provides evidence on the availability of financial statements across the world and documents stark differences in the share of disclosing firms (the disclosure rate) among private firms.

---

Requirements

This project relies on Orbis data (versions: December 2024 and August 2019), accessible via the API provided by the TRR 266 Accounting for Transparency research center.

• TRR 266 members:
  Create a file named .global-disclosure-data.env with your API credentials (a template is available as _global-disclosure-data.env).

• Non-members:
  Access to Orbis data is possible via Moody's SFTP delivery system.

A detailed list of required Orbis files can be found in:
code/01_pull_bvd_api_data.py

---

Setup

Software

• R (latest version recommended)

Required R Packages

The analysis relies on the following R packages:

arrow, countrycode, data.table, DBI, dotenv, dplyr, duckdb, extrafont,
fixest, ggplot2, glue, httr, jsonlite, logger, modelsummary,
parallel, patchwork, purrr, readr, readxl, rlog, stringr,
tibble, tidyr, utils, wbstats

---

Output and Workflow

To replicate the results, follow the scripts in the code/ folder in chronological order:

1. 01_pull_bvd_api_data.R
   Downloads the required Orbis data from the TRR 266 API.

2. 01b_pull_worldbank_data.R
   Pulls Governance and Development Indicators from the World Bank's API.

3. 01c_prep_la_porta_data.R
   Prepares data on countries' legal origin (La Porta, López-de-Silanes, Shleifer, and Vishny (1999)).

4. 02a_calc_transparency_scores.R
   Calculates firm-level transparency scores following Kim & Olbert (2022).

5. 02b_clean_transparency_scores.R
   Retains only one financial statement per firm-year based on a series of prioritization rules
   (e.g., consolidation level, reporting period length).

6. 03_get_entity_level_data.R
   Creates activity spells for each entity and merges firm-level data.

7. 04_aggregate_data.R
   Aggregates the data to country-year-legal form level.

8. 05_sample_selection.R
   Tracks the number of firms and countries dropping out in each sample selection step.

9. 06_descriptives.R
   Generates descriptive statistics for the final sample.

10. 07a_construct_validity_number_firms.R
    Construct validity test for the denominator of the disclosure rate (number of firms).

11. 07b_construct_validity_disclosure_rate.R
    Construct validity test for the numerator (number of financial statements).

12. 08_main_analysis.R
    Runs cross-country regressions and compares disclosure rates around the world.