Notes

IAN D. GOW · RESEARCH & PRACTICE

Ideas, evidence,
and a little code.

Notes on accounting, finance, data, and the tools we use to understand them. Explore a topic, follow a question, or find something useful.

Explore the collection ↘ Read the blog →

{n}

Data curation, databases, and software tools

Note	Topics
ACNC Registry data: Arrow version	Australia, Arrow
Data curation and the data science workflow	Data curation, Australia, ASX, SIRCA
Data curation: The case of Call Reports	Data curation, Polars, DuckDB
Data collection (with spreadsheets)	Data curation, Spreadsheets
Working with Form AP data	PCAOB, Form AP, R, DuckDB, Data curation
Getting SEC EDGAR XBRL data	SEC, XBRL, EDGAR
Working with DERA Notes data using Polars	SEC, DERA, Polars
Working with JSON data: The case of SEC submissions data	SEC, EDGAR, JSON
Processing SEC submissions data	SEC, EDGAR, JSON, Datetimes, Python, DuckDB
Working with date and times	Datetimes, DuckDB, SEC
Responsive open-source software: Two examples from dbplyr	Data curation, dbplyr, SQL, DuckDB
Shared code	Research, web data
GitHub for accounting researchers	GitHub, Git, Reproducibility, Research methods
Why aren’t more data people talking about ibis?	Python, Ibis, DuckDB, SQL
Writing better SQL without writing SQL	SQL, dbplyr
SIRCA ASX End of Day (EOD) collection	Australia, SIRCA, ASX, CSV, Parquet
SIRCA Mergers and Acquisitions collection	Australia, M&A, ASX, SIRCA, Parquet
Converting lazy data frames into Parquet files	R, Parquet, db2pq, WRDS, Tidy Finance
Converting lazy data frames into Parquet files (Python version)	Python, Parquet, db2pq, Polars
Benchmarking local PostgreSQL-to-Parquet export paths	PostgreSQL, Parquet, benchmarking, Polars
Getting data out of Postgres: an update (8 October 2026)	PostgreSQL, Parquet, benchmarking, Polars
Improving performance of SQLite data	Tidy Finance, SQLite

Python, finance, and data workflows

Note	Topics
Ball and Brown (1968): Replication using Python Polars	Accounting research, Replication, Python, Polars
Data management ideas for researchers	Python, Parquet, db2pq, WRDS, Polars
Data management ideas for researchers (R version)	R, Parquet, db2pq, WRDS
Some benchmarks with comp.g_secd	SAS, WRDS, CRSP, Parquet, Python
The best of both worlds: Using modern data frame libraries to create pandas data	WRDS, Polars, Ibis, pandas
Using SAS to create pandas data	SAS, pandas, wrds2pg
Explaining some .assign() idioms	Python, pandas, R, Data cleaning
Calculating betas using DuckDB	Tidy Finance, DuckDB, WRDS, Finance
Trading days per year (crsp.dsf)	CRSP, WRDS
Stock returns on Yahoo Finance	Yahoo, finance
Adding delisting returns to monthly data	SAS, Stock returns, CRSP

Research methods, replication, and commentary

Note	Topics
How common are direct-method cash flow statements?	Cash flows, Accounting, SEC, XBRL, R, dera.pq
Business valuation with ITCs	Sunnova, Accounting, Valuation, Investment tax credits, Noncontrolling interests
Growth by industry	Valuation, Finance, Compustat, R, Data visualization
Trump’s tariff formula	Economics, Trade, Tariffs
Digging into Tesla’s reporting on PP&E	Tesla, Accounting, PP&E, Cash flows
Should Bao et al. (2020) be retracted?	Research methods, Machine learning
What a Journal of Finance retraction gets right	Replication, Reproducibility, Research methods
But not all research is bad	Research methods, p-hacking
Proposal for teaching data analytics	Education, Data analytics, R
Taxonomy of researchers	Research methods, Education, Academia
Observations on accounting education in 2021	Accounting, Education
Solutions for ‘Beaver (1968)’	Accounting research, Replication, Python, Polars
Solutions for Ball and Brown (1968)	Accounting research, Replication, Python, Polars
Solutions for Panel data	Accounting research, Replication, Python, Polars
Missing Form APs?	Research methods, PCAOB, XBRL, SEC
The elephant in the room: p-hacking and accounting research	Research methods, p-hacking
The Gino-Colada Affair	Reproducibility, Research methods
Reproducible data collection	Reproducibility, Research methods
Does @Beardsley_2021 show anything?	Research methods
Analysis of IPOs on the ASX	Australia, IPOs

R, graphics, and miscellaneous data notes

Note	Topics
A quick look at City of Melbourne bike data	Melbourne, transport
Comparing skylines	Melbourne, R, Data curation
Data visualization challenge	Data visualization, ggplot2
Making plotnine more ‘Pythonic’	Python, plotnine, graphics, Polars
Retail sales	Retail, ABS, R
The Retail Sales Data Set	Retail, ggsql, SQL, Visualization
Defining winter and summer in Boston	Weather, Boston
Defining winter and summer in Melbourne	Weather, Australia, Melbourne
Defining winter and summer in Sydney	Weather, Australia, Sydney
Defining winter and summer in Oxford	Weather, Oxford, Python
Sunrise and sunset times	Datetimes, Weather, Australia, Melbourne, Boston