Build your own market lens, then carry it into a governed strategy

Why does a broker need its own indicators?

Because a platform where every client computes the same twelve standard indicators offers nothing a competitor cannot offer. Proprietary research is the differentiator, and it is worth nothing if it cannot reach the place where strategies are actually built.

The usual outcome is that research lives in a spreadsheet or a notebook, gets discussed in a meeting, and never becomes an input to anything a client can act on. The gap is not analytical capability; it is the absence of a governed path from a computed series to a strategy rule.

What makes an indicator pipeline trustworthy?

Validation before activation, and versioning after it. An indicator that silently changes definition invalidates every backtest that used it, and an indicator whose data source degrades without anyone noticing produces confident nonsense.

So the workflow checks shape, history, field type, missing data and quality before a definition can be activated, and it treats each activation as a version rather than an edit. A strategy references a version, which means the strategy's evidence remains interpretable after the indicator moves on.

The four stages of a definition

How does an activated indicator get used?

Through reusable indicator codes available in charts, scanner conditions, strategy rules and backtest context, without bypassing strategy validation or execution policy.

That last clause is the point. An indicator being available does not create a route around the gates: a strategy using a custom indicator passes the same deterministic validation and the same policy checks as any other. Research gains reach without gaining exemption.

What do dynamic scanners do, and what do they not do?

Scanners compose liquidity, momentum, data-quality and ranking conditions over an eligible universe to find and rank assets. They do not decide entries, exits or position sizes.

Stating the boundary explicitly matters because scanners are the feature most likely to be mistaken for advice. A ranked list is a research output; the decision about what to do with it stays with the person building the strategy, and the platform does not blur the two.

How does research stay connected to execution?

Four surfaces share one context: discover, construct, observe and investigate. Custom research feeds the builder; the builder feeds monitoring; monitoring feeds an investigation of why a specific order was allowed, blocked, filled, rejected or paused.

Keeping those connected is what makes a research asset accountable. An indicator that contributed to a strategy that produced an order is traceable in both directions, which is the property a review needs and the one a disconnected toolchain cannot provide.

What does a worked example look like?

A desk wants a liquidity-adjusted momentum measure of its own. It declares the source and expected schema, the platform validates shape, history, field types and missing data, the desk defines the raw fields and the computed series, and an activated version exposes a reusable code usable in charts, scanners and strategy rules.

The step that earns its keep is validation. A definition that assumes twenty periods of history and receives twelve is caught before a strategy is built on it, rather than producing a series with a silent gap that turns into a backtest nobody can explain.

The step that earns its keep second is versioning. When the desk revises the measure, the previous version continues to exist, so the strategies referencing it keep producing interpretable evidence until they are deliberately moved forward.

Where an activated indicator can be used

Availability on any surface depends on organisation access and data availability for the deployment.

What does this deliberately not do?

It does not let a research output bypass a gate. An indicator being activated grants reach, not exemption, a strategy using it passes the same deterministic validation, the same risk checks and the same policy gates as any other.

It also does not turn a scanner into a recommendation engine. A ranked list of instruments is a research artefact, and the platform keeps the decision about what to do with it firmly on the person building the strategy.