SWF Financials: Virtual Company Data (1996-2025)
Description
Overview This dataset provides a synthetic "Income Statement" (Profit & Loss) for a virtual mega-corporation, covering 30 years of business operations from 1996 to 2025. The data is designed to simulate a realistic corporate financial lifecycle, useful for financial time-series analysis, forecasting, and educational purposes. Dataset Composition The dataset contains 120 quarters of financial data divided into two distinct periods: 1996–2011: Projected and modeled data (64 quarters) based on historical financial trends. 2012–2025: Aggregated financial data derived from real SEC EDGAR filings (56 quarters), representing a consolidated view of large-scale corporate performance. Key Metrics Included Each record includes standard accounting metrics: Revenue and Cost of Revenue Gross Profit and Operating Expenses Operating Income and Net Income Calculated margins (Gross, Operating, and Net Margins) Data Quality Flags indicating the source (Projected vs. Complete) and validity. Files Included swf_financials.csv: The primary dataset. SWF_FINANCIALS_SCHEMA.md: Detailed column definitions and data types. SWF_ASSUMPTIONS.md: Documentation of the methodology used for data aggregation and projection. financial_preprocessing_pipeline.py: The Python script used to process and clean the raw data. examples.ipynb: A Jupyter Notebook demonstrating basic analysis and visualization of the dataset. Potential Use Cases Training machine learning models for financial forecasting. Backtesting algorithmic trading strategies. Academic teaching of accounting principles and margin analysis. Benchmarking corporate performance against synthetic "average" mega-corporations. Important Research Note Please note that while the 2012–2025 data is based on real SEC filings, this is a synthetic aggregation and does not represent a single specific real-world company. It is intended for research and educational applications only.
Files
Steps to reproduce
The data for the "Complete" period (2012–2025) was aggregated from real-world corporate SEC EDGAR filings using the provided financial_preprocessing_pipeline.py script. The "Projected" period (1996–2011) was generated using financial modeling techniques to simulate a 30-year lifecycle based on historical trends. Detailed methodology and accounting assumptions are documented in SWF_ASSUMPTIONS.md . To reproduce the analysis, open the examples.ipynb notebook in a Python environment with Pandas and Matplotlib installed.
Institutions
- Multimedia University - Cyberjaya CampusSelangor, Cyberjaya