i26g Library

Discipline, Not Discretion

Some independent traders want to operate systematically — to treat trading as a disciplined, repeatable process rather than a series of discretionary decisions. They want rules they can test, execution they can trust, and an operation they can keep improving.

That ambition is not a tooling problem first. It is a question of what the operation is supposed to look like: what gets decided in advance versus in the moment, what evidence a rule has to survive before it trades real money, how risk is bounded, and what gets recorded so the operation can be improved rather than merely repeated.

The i26g Library collects that knowledge: how independent systematic trading should be done, independent of what runs it.

What It Covers

The Library is organised around the parts of a systematic trading operation:

• Data: sourcing, cleaning, storing, and checking the market data everything else depends on.
• Research and testing: turning an idea into a rule, testing it honestly, and avoiding the overfitting that makes a backtest look better than reality.
• Position sizing and risk: how much to trade, how losses are bounded, and how risk is measured across the whole operation.
• Execution and operations: running the system day to day — orders, monitoring, failures, and recovery.
• Infrastructure: what the operation runs on, and how to keep it dependable without it taking over.

The Model at Its Core

A discretionary trader can hold their process in their head. A systematic trader cannot: the whole point is that the process exists outside the trader, so it can be tested, executed consistently, and improved on evidence rather than recollection.

Trading independently makes this harder than it is inside an institution. There is no risk desk, no separate execution team, no research function — one person holds every role at once, under time pressure, with tooling built for convenience rather than control. Without an explicit model, the discipline quietly collapses back into discretion, and the trader ends up with automation wrapped around habits they never examined.

At the centre of the Library is that model: what the operation decides in advance, what evidence it requires, how it bounds risk, and what it records. Every other topic connects back to it.

Methods, Not Recommendations

The Library is about methods, engineering, and research — how to build and run a systematic trading operation. It does not recommend strategies, instruments, or trades, and nothing in it is investment advice.

Made Runnable

Knowledge on its own does not trade. The i26g Platform is what makes it operable: data, strategy, execution, and risk as one coherent system the trader owns outright — so the process defined here is the process that actually runs. The i26g Control Plane keeps that platform aligned with it through its whole life cycle.

Where an operation already running the tooling needs pieces it does not cover yet, that is i26g Custom Development.

Still Developing

The Library is not complete, and the model at its core is not meant to keep changing indefinitely — the intent is for it to stabilize as it matures. It is not there yet. It develops against a real trading operation — practice is what surfaces what it needs next, and what it gets wrong.