Appendix 8A: Chapter 8 milestone and the big picture (Ch01–Ch08)
Chapter 8 is the point where Track D moves from accounting mechanics into statistics applied to accounting data.
At this milestone you now have an end-to-end workflow that mirrors real work in small business bookkeeping / controllership:
Capture transactions in a double-entry system.
Produce financial statements.
Validate the numbers (reconciliations / controls).
Prepare analysis-ready data (tidy + rollups).
Compute descriptive statistics that reveal variability and risk.
What we have built so far
Ch01–Ch02: Accounting fundamentals + the ledger-as-data
Ch01 frames accounting as a measurement system (what we measure, why it matters, and how the accounting equation constrains the story).
Ch02 treats the general ledger as a structured database: transactions, accounts, and rules that make aggregation meaningful.
Ch03–Ch06: Statements + validation as quality control
Ch03 uses financial statements as summary statistics (income statement, balance sheet, and how line items roll up).
Ch04–Ch05 expand the accounting topics needed to support realistic analytics (assets/inventory/fixed assets/depreciation; liabilities/payroll/ taxes/debt/equity).
Ch06 formalizes reconciliations as quality control so that downstream analytics don’t silently operate on bad inputs.
Ch07: Preparing accounting data for analysis
Chapter 7 creates two “analysis-ready” datasets:
gl_tidy.csv— a line-level, tidy general ledger with consistent types and signed amounts.gl_monthly_summary.csv— a monthly rollup by account/category.
This is the bridge between bookkeeping exports and statistical workflows.
Ch08: Descriptive statistics for financial performance
Chapter 8 is the first chapter that uses statistics directly.
It produces:
gl_kpi_monthly.csv— monthly KPIs + ratios + rolling mean/std + simple z-score signals.ar_monthly_metrics.csv— A/R roll-forward style metrics (credit sales, collections, DSO approximation).ar_payment_slices.csv+ar_days_stats.csv— a small payment-lag distribution (FIFO allocation) and descriptive summaries (mean/median/ quantiles, including amount-weighted versions).
Why this milestone matters
Two “stats-first” ideas show up immediately in Chapter 8:
Skew and tails (A/R): average DSO can be misleading when a few late-paying customers create a long right tail. Median and upper quantiles (e.g., p90/p95) often communicate risk better.
Volatility signals (performance): rolling mean/std on margins and revenue growth provide a lightweight way to detect unusual months and prompt follow-up questions.
These are small examples, but they demonstrate the core theme of Track D: accounting data becomes powerful when it is treated as analyzable data.
How to reproduce the work locally
From the repo root (with the virtualenv active):
# generate the NSO synthetic dataset
make business-nso-sim
# run the bookkeeping + QC chapters
make business-ch04
make business-ch05
make business-ch06
# build analysis-ready data and then compute descriptive stats
make business-ch07
make business-ch08
# build docs locally
make docs
What remains (roadmap after Chapter 8)
Chapter 8 sets up a clean runway for forecasting and statistical modeling.
Chapter 9 will focus on visualization and reporting that doesn’t mislead (how to present KPIs and A/R risk clearly, and what common charting mistakes look like).
The following chapters can then build forecasting and regression tools using the same monthly KPI tables produced in Chapter 8.
Contributing ideas
If you want an easy contribution after Chapter 8:
Add one new KPI ratio (e.g., operating margin, inventory turnover) with a test.
Add one new diagnostic check in
ch08_summary.json(e.g., outlier month flag counts).Add a short “interpretation” paragraph in the Chapter 8 docs tied to one specific column.
The goal is to keep each chapter small, deterministic, and easy to run.