CoynaInstitutional Enquiries

INSIGHTS

Research with context.
Perspective with discipline.

Notes on quantitative research, market interpretation and the engineering of systematic trading.

RESEARCH NOTE

A model needs more than a backtest.

Why assumptions, evaluation and controlled deployment belong in the same research process.

Read the note ↗

MARKET PERSPECTIVE

A signal is only part of the picture.

Understanding the context around price, liquidity and changing market conditions.

Read the note ↗

ENGINEERING NOTE

Risk control is a system property.

Why independent boundaries and operational visibility matter alongside model design.

Read the note ↗

RESEARCH NOTE

A model needs more than a backtest.

A historical result is evidence to examine, rather than a conclusion on its own. The assumptions behind the result matter: the information available at each point in time, the cost of implementation and the conditions represented in the sample.

Research benefits from a separation between developing an idea and assessing it. Clear versioning and documented assumptions make it easier to ask what changed, what was tested and what remains uncertain.

Forward testing and bounded deployment add different kinds of evidence. Monitoring remains necessary after deployment because an evaluation describes observed conditions; it cannot guarantee future behaviour.

Back to insights ↑

MARKET PERSPECTIVE

A signal is only part of the picture.

The same price movement can occur in different trading environments. Liquidity, participation and variability influence the context in which a signal is interpreted and a decision is executed.

A systematic approach needs to consider that context without confusing a useful description with certainty about what happens next. Market regimes are analytical interpretations, and their boundaries can be imperfect.

Keeping analysis, execution and monitoring connected supports a consistent view of the trading process as conditions change.

Back to insights ↑

ENGINEERING NOTE

Risk control is a system property.

A model has a defined task. The wider trading system has responsibilities that extend beyond it, including constraining activity, monitoring operations and preserving the ability to intervene.

Separating those responsibilities helps keep limits independent of the logic generating a trading decision. Operational visibility complements those boundaries by making exceptions available for review.

Traceability is part of the same design. A record of the relevant model version, parameters and decisions provides a basis for investigation and controlled change.

Back to insights ↑

Original explanatory notes prepared for this website. Not live market commentary, investment research recommendations or a record of Coyna’s performance. No proprietary strategy methods are disclosed.