Practical articles on financial data, financial documents, APIs, data extraction and the engineering behind Babylon.
We built Megargy Newsletter Service out of frustration with existing platforms. Sending a digest of Babylon’s latest articles should be straightforward: simple to set up, easy to configure and refine later, and sensibly priced. That is not what we found with incumbent vendors.
We’re changing the email address used to send Babylon newsletters.
Contract notes are usually written for people, not software. This example shows how a carefully tagged PDF can do both: remain readable as a document, while allowing babylon-table to extract its tables and turn formatted strings into clean transaction data.
Babylon has open-sourced babylon-table, a dependency-free Java library for ingesting and reshaping tabular data. Published under the Apache License 2.0 and available on Maven Central, the module grew out of Babylon’s practical financial data work. Its most distinctive feature is a small domain-specific language (DSL) for describing table transformations in a business-readable form.
A small experiment with Apache PDFBox turned into a useful discovery: creating a clear, structured PDF record of a trade is not especially difficult.
A transaction ID is a dull reference number that has to appear on a contract note, in an email, or in a support call. Who would have thought there was much to say about it? Well, there is!
EasyEquities has transformed retail investing in South Africa, but access to transaction data remains unnecessarily difficult. Transaction confirmations, machine-readable trade histories, and stable security identifiers are essential for transparency, reconciliation, and investor trust.
Refactoring a codebase into well-defined modules is one of those tasks every developer knows is valuable—but rarely prioritises.
Codex isn’t just writing code — it’s changing how I think about building software.
There’s a particular kind of joy and satisfaction that comes from seeing something work for the very first time.
Brokerage accounts provide trade history data in many different tabular formats – each with its own naming quirks, column conventions, ordering, and hidden assumptions. Yet behind this diversity lies a consistent set of trade attributes with inherent structural relationships: quantities, prices, consideration, fees, proceeds, dates, instruments, and identifiers. This is exactly where a semantic inference engine can shine.
If you scan JSE instrument lists, broker CSV files, or pricing feeds, you may notice that some shares end with “-N” in their short name. This isn’t a typo or data issue — the “-N” suffix indicates a specific, legacy class of reduced-vote ordinary shares known as N Ordinary shares, or NOrd.
TL;DR: We originally built a phrase-based parser for segment ledger quick entry, but it was too strict. Refactoring led us to anchor on the verb in the sentence (“buy”, “sell”), allowing us to identify the Subject–Verb–Object (SVO) structure. This made the parser far more flexible and capable of handling natural, messy, human descriptions — including tricky cases like “GBP”, which can be both a currency and a listing symbol!
TL;DR: On the London Stock Exchange, GBP was, until very recently, the ticker symbol for Global Petroleum Limited. This rare overlap is confusing to humans and machines alike, and serves as a reminder that context matters.
Excel includes two powerful but often overlooked features: the Stocks and Currency data types. These allow you to pull delayed market and FX data directly into your spreadsheet — no plugins or VBA required.
When investors move between tax jurisdictions, capital gains must be split between the period before and after the move. Babylon’s new gainSince parameter makes this possible. By specifying a date, you can recalculate gains as if every holding were rebased at that moment — automatically replacing pre-date transactions with a single notional trade per security per account. The result is an accurate picture of gains arising since the jurisdiction change, while preserving your original trading history.
A Segment is one of Babylon’s core concepts. It defines the natural partition of a portfolio, ensuring every security — and all of its trades — belong to one, and only one, Segment. This simple rule keeps capital gains and performance calculations coherent, consistent, and correct at scale.
Every API begins simply. A few clean endpoints, a handful of query parameters — ?symbol=AAPL, ?date=2025-09-01 — and everything feels elegant. Then the real world arrives. You need to exclude certain accounts, filter by date ranges, or select everything except one value. Suddenly that neat equality model starts to creak.
In equity prices we often encounter what appear to be strange misspellings of currencies—for example, ZAC rather than ZAR, GBX rather than GBP, or EUC rather than EUR. Why does this happen? As it turns out, they are not misspellings at all! The data provider is usually trying to express the price in minor units of the currency. For instance, a quote of 2,345 ZAC is actually a price of 2,345 South African cents, which is 23.45 ZAR. The same pattern applies to other so-called “misspellings.”