On why "market data" is not one thing — and what happens when data engineers treat it like it is.
The Two Trees
Most financial platforms organise their instruments using one of two patterns:
- By asset class — equities in one tree, fixed income in another, FX in a third.
- By instrument type — stocks, bonds, swaps, options, each a separate category regardless of underlying asset.
Both are valid. Both are used in production at major banks and data vendors. The confusion starts when a system pretends they're the same thing.
Why It Matters
When you conflate asset class and instrument type, you get schema decisions that look clean on day one and break by day ninety. A convertible bond is both fixed income and a derivative. A currency option is both FX and an option. If your column is called asset_class but actually stores the instrument type, every downstream join becomes a lie.
The fix isn't hard: use two columns. One for the asset class (what underlying market does this reference?), one for the instrument type (what kind of contract is this?). They answer different questions, and conflating them answers neither correctly.
A Simple Rule
If you're building a financial data model and you're not sure which tree you're climbing, ask yourself: "Would a quant and a portfolio manager categorise this the same way?" If the answer is no, you need two columns.