Commercial Financial Statement Analysis Project

This repository is the borrower financial analysis layer at the front of the public commercial credit-risk stack. It uses synthetic borrower statements and benchmark-style reference data to produce standardised financials, ratio summaries, working-capital metrics, and qualitative risk flags that can support both structured credit assessment and practical lending decisioning workflows.

What this repo is

This project demonstrates how borrower financial information can be turned into reusable commercial credit-analysis outputs. It is designed for recruiter and employer review, so the repo emphasises clear structure, explainable metrics, and portfolio-style documentation across both bank-style credit review and non-bank underwriting use cases.

Where it sits in the stack

Upstream inputs:

• synthetic borrower statements and benchmark reference data staged under data/

Downstream consumers:

• PD-and-scorecard-commercial
• LGD-commercial

How this is used in practice

This project can be applied in:

Bank / Institutional context

• Borrower financial spreading and ratio review for structured credit assessment
• Serviceability, leverage, and working-capital analysis for portfolio risk review
• Consistent financial inputs for downstream PD, LGD, and monitoring workflows

Non-bank / Fintech context

• Financial analysis inputs for originations decisioning and approval strategy
• Cash-flow and liquidity review for lender-side underwriting
• Early borrower quality diagnostics for portfolio performance tracking

Key inputs

• synthetic borrower financial statements
• public listed-company style benchmark extracts
• financial spreading, ratio, and qualitative assessment assumptions staged under data/

Key outputs

• outputs/tables/standardised_borrower_financials.csv
• outputs/tables/ratio_summary_tables.csv
• outputs/tables/working_capital_metrics.csv
• outputs/tables/trend_diagnostics.csv
• outputs/tables/qualitative_risk_flags.csv
• outputs/tables/pipeline_validation_report.csv

Repo structure

• data/: raw, interim, processed, and external borrower-analysis inputs
• reports/: retained PDF artifacts (reports/public_company_analysis/ and reports/portfolio_overview.pdf) for reviewer-facing examples
• src/: reusable financial spreading, ratio, and pipeline modules
• scripts/: wrapper scripts for pipeline execution
• docs/: methodology, assumptions, data dictionary, and validation notes
• notebooks/: walkthrough notebooks for reviewer context
• outputs/: exported tables, reports, and sample artifacts
• tests/: validation and regression checks

Compatibility note:

• Older setups that still contain a top-level Reports/ folder are automatically migrated to the canonical reports/ layout by src/public_company_analysis.py.

How to run

python -m src.run_pipeline

Or:

python scripts/run_demo_pipeline.py

Limitations / Demo-Only Note

• All borrower records are synthetic and are included for demonstration only.
• Benchmarking and qualitative flags use simplified portfolio assumptions rather than internal bank financial spreading rules.
• The repo is intended to show reusable analytical workflow and reporting quality, not to represent a production credit spreading platform.