Selecting a programming language or languages for an algorithmic trading system is not easy and takes some thinking. The key factors are performance, facility to create, resilience and tests, separation of concerns, familiarity, maintenance, availability of the source code, licencing fees, and library maturity.

The advantage of a distinct architecture is that it enables languages to be" plugged in" when requirements for various elements of a trading stack change. A trading system is a dynamic tool, and any language choices will almost certainly change in lockstep with it.