Skip to main content

Ledger in Python that follows corporate accounting rules.

Project description

abacus-minimal

PyPI - Version

abacus-minimal is an accounting logic library that aims to be concise, correct and expressive in implementation of double entry book-keeping rules.

Install

pip install abacus-minimal

Latest:

pip install git+https://github.com/epogrebnyak/abacus-minimal.git

Minimal example

A company receives initial shareholder investment of 500 USD in cash, acquires a stock of merchandise for 200 USD, gets a lucky chance to sell all merchandise for 400 USD, and gladly pays 150 USD as salaries to the staff.

We programmatically run the accounting workflow within one reporting period for this company and show end of period reports.

from abacus import Book, Chart, Entry

chart = Chart(
    retained_earnings="retained_earnings",
    current_earnings="current_earnings",
    assets=["cash", "inventory"],
    capital=["equity"],
    income=["sales"],
    expenses=["salaries", "cogs"],
)
book = Book(chart)
entries = [
    Entry("Initial shareholder funds").debit("cash", 500).credit("equity", 500),
    Entry("Acquired goods").debit("inventory", 200).credit("cash", 200),
    Entry("Accepted payment for goods").debit("cash", 400).credit("sales", 400),
    Entry("Shipped goods").debit("cogs", 200).credit("inventory", 200),
    Entry("Paid salaries").debit("salaries", 150).credit("cash", 150),
]
book.post_many(entries)
book.close()
print(book.income_statement)
print(book.balance_sheet)
# Some checks
assert book.balance_sheet.assets.total == 550
assert book.income_statement.net_earnings == 50
assert book.balances == {"cash": 550, "inventory": 0, "equity": 500, "retained_earnings": 50}

Accounting concepts

abacus-minimal adheres to the following interpretation of double-entry book keeping rules.

  1. The value of company property, or assets, equals to shareholder and creditor claims on the company, known as capital (or equity) and liabilities respectively:
Assets = Capital + Liabilities
  1. Company current earnings, or profit, equal to income less expenses associated with generating the income:
Current earnings = Income - Expenses
  1. Capital consists of paid-in capital (or shareholder equity), other capital accounts and retained earnings:
Capital = Shareholder Equity + Other Capital + Retained earnings
  1. Current earnings accumulate to retained earnings:
Retained earnings = Retained earnings from previous period + Current earnings
  1. Substituting we get a form of extended accounting equation:
Assets + Expenses = Shareholder Equity + Other Capital + Retained earnings + Income + Liabilities
  1. Our book-keeping goal is to reflect business events as changes to the variables in this equation while maintaining the identity between the left and the right sides of the equation, as well as following the accounting principles for recognition, measurement and reporting.

  2. Changing just one variable will break the accountign equation, so we allow postings, or entries that change two or more vairables, depending on the nature of the business transaction.

  3. Debits and credits are a common framework to keep track of the changes in accounts. It is less fundimental than accounting identity, just a clever system to avoid at least some of accounting mistakes when expressing what variables in accounting identity you are changing.

  4. Let as allow for every account to accumulate the changes as a list of debit amounts and a list of credit amounts. There will be accounts where the account balance is sum of debits less sum of credits ('debit normal'), and accounts where the balance is sum of credits less sum of debits ('credit normal').

  5. Let us designate the assets and expenses accounts as 'debit normal' and capital, liabilities and income accounts as 'credit normal'. Now instead of saying "Please do not break the accounting identity when changing the variables" we can now say "When creating and posting an entry the sum of debit changes must be equal to the sum of credit changes", which is easier to track.

  6. In a double entry you add the same amount to the debit of one account and to the credit of another account and the accounting identity never breaks. Profit!

  7. We would not want a company that just accumulates profits and never pays divident to shareholders. Sometimes you would see an equation similar to:

Retained earnings (at period end) =  Retained earnings (at period start) + (Income - Expenses) - Dividend

To pay out the dividends the company will transfer part of retained earnings to dividend due libility after announcing the dividend, and will pay out the dividend due with cash at dividend disbursement. (Think of two double entries that reflect these operations).

Accounting workflow

The steps for using abacus-minimal follow the steps of a typical accounting cycle:

  • create a chart of accounts,
  • open ledger for the current reporting period,
  • post account balances for the previous period,
  • post entries that reflect business transactions within the period,
  • post reconciliation and adjustment entries,
  • close accounts at reporting period end,
  • make post-close entries,
  • show financial reports,
  • save account balances data for the next reporting period.

Accounts survive these transformations quite differently, here is the best summary I can give:

Step Assets Expenses Shareholder Equity Retained Earnings Current Earnings Income Liabilities
Account type permanent temporary permanent permanent temporary temporary permanent
1. In empty ledger all accounts have zero balances 0 0 0 0 0 0 0
2. Opening balances restore permanent accounts + 0 + r 0 0 +
3. Business entries affect all accounts except earnings + e + r 0 i +
4. Closing income and expense to current earnings + closed + r i-e closed +
5. Closing current to retained earings + + r + i -e closed +
6. Saving permanent account balances for the next period + + + +

End-to-end example

In this example we use more abacus-minimal features including:

  • contra accounts specification,
  • opening ledger with account balances at the start of reporting period,
  • posting multiple entries and using various types of syntax for an entry,
  • showing proxy income statement and balance sheet before closing,
  • saving and loading data to JSON files.

The complete code is in readme.py.

1. Create chart of accounts

Steps involved:

  • specify names of the current earnings and retained earnings accounts,
  • add account names for assets, capital, liabilities, income and expenses,
  • add contra accounts (e.g. refunds is a contra account to sales).

Code example:

from abacus import Chart

chart = Chart(
    retained_earnings="retained_earnings",
    current_earnings="current_earnings",
    assets=["cash", "ar"],
    capital=["equity"],
    liabilities=["vat_payable"],
    income=["sales"],
    expenses=["salaries"],
)
chart.offset("sales", "refunds")
chart.name("ar", "Accounts receivable")

Chart class is a pydantic model, which means it is easily converted to a JSON file. You can save or load a chart from a file.

chart.save("chart.json")
chart = Chart.load("chart.json")

2. Start ledger

Steps involved:

  • create a data structure that represents state of accounts (a 'book'),
  • record account starting balances from the previous period,

Let's create a book with opening balances known from previous period:

from abacus import Book

book = Book(chart)
opening_balances = {
    "cash": 10_000,
    "equity": 8_000,
    "retained_earnings": 2_000
    }
book.open(opening_balances)

At this point the book is ready to record new entries.

3. Post entries to ledger

Steps involved:

  • record entries that represent business transactions,
  • show state of ledger (as trial balance or as account balances) at any time.

Each entry has a title and directions to alter the accounts that are called debits and credits. The sum of debits should match the sum of credits for a valid entry. The Entry class provides several ways to record the composition of an entry as shown below.

from abacus import Entry

entries = [
    Entry("Invoice with VAT").debit("ar", 6000).credit("sales", 5000).credit("vat_payable", 1000),
    Entry("Cash payment").debit("cash", 6000).credit("ar", 6000),
    Entry("Cashback").double(debit="refunds", credit="cash", amount=500),
    Entry("Paid salaries").set_amount(1500).debit("salaries").credit("cash"),
]

# Post entries to book
book.post_many(entries)

Note: there are no reconciliations, adjustments and post-close entries in this example.

4. Inspecting ledger

After posting entries you can inspect the trial balance or account balances:

# Show trial balance and account balances
print(book.trial_balance)
print(book.balances)

# Check account balances match expected values
assert book.balances == {
    "cash": 14000,
    "ar": 0,
    "equity": 8000,
    "vat_payable": 1000,
    "sales": 5000,
    "refunds": 500,
    "salaries": 1500,
    "current_earnings": 0,
    "retained_earnings": 2000,
}

5. Closing accounts

Closing accounts at period end involves:

  • closing contra accounts related to income and expense accounts, and
  • closing income and expense accounts to retained earnings.

See section below for code for closing accounts.

Note: account closing was a surprisingly hard part of abacus-minimal code that I had to refactor several times.

6. Reporting financial statements

Financial reports are typically displayed after account closing, but proxy reports are available before closing as well:

  • The income statement will be the same before and after closing.
  • The balance sheet before closing the will contain current earnings account and retained earnings from previous periods.
  • After closing the current earnings account will be transferred to the retained earnings account and removed from the ledger; it will not appear in the balance sheet.

Expect to see a lot of dictionary-like data structures in code output below:

print("=== Before closing ===")
print(book.income_statement)
print(book.balance_sheet)
assert book.balance_sheet.capital["current_earnings"] == 3000

# Close accounts at period end
book.close()

print("=== After closing ===")
print(book.income_statement)
print(book.balance_sheet)

# Check account balances match expected values
print(book.balances)
assert book.balances == {
    "cash": 14000,
    "ar": 0,
    "equity": 8000,
    "vat_payable": 1000,
    "retained_earnings": 5000,
}

7. Saving data for the next period

You can save the list of entries and the period end account balances to JSON files, unless these files already exist. In that case you will need extra precaution – for example, save files to a different folder or under a different name.

# Save JSON files
book.store.save("./entries.json")
book.balances.save("./end_balances.json")

Architecture

Core library

In abacus-minimal there is a small core library that consists of:

  • ChartDict maps account names to their types,
  • Ledger class maps account names to debit normal and credit normal T-accounts, and
  • Posting type that represents a double or a multiple entry.

Ledger is created from ChartDict and incoming entries change the state of ledger. ChartDict allows to create closing entries at accounting period end. Trial balance, account balances, balance sheet and income statement reports reflect the state of ledger.

User interface

As a user you do not have to interact with the core directly. abacus-minmal exports Chart, Entry and Book classes that we used in the examples. The Book class holds together a chart, a store of entries, and a ledger and allows posting entries, closing the accounts, creating reports and saving and loading JSON files.

Limitations

Several assumptions and simplifications are used to make abacus-minimal easier to develop and reason about.

The key assumptions are:

  • one currency,
  • globally unique account names,
  • one level of accounts in chart and no account aggregation for reports,
  • no treatment of other comprehensive income,
  • no changes in equity and cash flow statements.

See core.py module docstring for more details.

Alternatives

abacus-minimal takes a lot of inspiration from the following projects:

Accounting knowledge

If you are totally new to accounting the suggested friendly course is https://www.accountingcoach.com/.

ACCA and CPA are the international and the US professional qualifications and IFRS and US GAAP are the standards for accounting recognition, measurement and disclosure.

Part B-G in the ACCA syllabus for the FFA exam talk about what abacus-minimal is designed for.

Textbooks:

Project conventions

I use just command runner to automate code maintenance tasks in this project.

just test and just fix scripts will run the following tools:

  • pytest
  • mypy
  • black and isort --float-to-top (probably should replace with ruff format)
  • ruff check
  • prettier for markdown formatting
  • codedown to extract Python code from README.md.

examples/readme.py is overwritten by the just readme command.

I use poetry as a package manager, but heard good things about uv that I want to try.

Changelog

  • 0.11.1 (2024-11-06) abacus.core now feature complete.
  • 0.10.7 (2024-11-02) Posting type is a list of single entries.
  • 0.10.5 (2024-10-27) Handles income statement and balances sheet before and after close.
  • 0.10.0 (2024-10-24) Separates core, chart, entry and book code and tests.

Roadmap

Using upstream

Implanting abacus-minimal as a dependency to:

New features

  • Business event layer Event("Invoice", net_amount=200, vat=40)
  • Book.increase() and Book.decrease() methods
  • Entry.explain() method

Application ideas

  • real company - eg Walmart accounts
  • business simulation layer - stream of entries
  • more examples from textbooks
  • chart repository and conversions between charts of accounts as requested in #4.
  • a quiz based on Book class - play entries and expect the use picks correct debit and credit

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

abacus_minimal-0.11.1.tar.gz (18.7 kB view details)

Uploaded Source

Built Distribution

abacus_minimal-0.11.1-py3-none-any.whl (14.8 kB view details)

Uploaded Python 3

File details

Details for the file abacus_minimal-0.11.1.tar.gz.

File metadata

  • Download URL: abacus_minimal-0.11.1.tar.gz
  • Upload date:
  • Size: 18.7 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: poetry/1.8.3 CPython/3.12.1 Linux/6.5.0-1025-azure

File hashes

Hashes for abacus_minimal-0.11.1.tar.gz
Algorithm Hash digest
SHA256 483826cc780cb531f23b0aac5a66d59927d02b214323bb7a2326180ed3c5e399
MD5 ad2809ce649f3dbdd9e8e356a6968af0
BLAKE2b-256 b26b829c4d1ebc370e68e9e4b88f212629937edfcf16ff4504feb454bb1fa6db

See more details on using hashes here.

File details

Details for the file abacus_minimal-0.11.1-py3-none-any.whl.

File metadata

  • Download URL: abacus_minimal-0.11.1-py3-none-any.whl
  • Upload date:
  • Size: 14.8 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: poetry/1.8.3 CPython/3.12.1 Linux/6.5.0-1025-azure

File hashes

Hashes for abacus_minimal-0.11.1-py3-none-any.whl
Algorithm Hash digest
SHA256 1792e5e68c7fa1dab34e76600ba5f445fc4f1a3fde09766d566a09d15e1df926
MD5 7cd38bb21d3f653cd8256c2f16958202
BLAKE2b-256 37ee2a9bc8cd85aee737aa8d0823636b9388c4def7159260a394da4f78fa2e89

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page