EQNX Trading
ENFR
INDEPENDENT RESEARCH & TRADING SYSTEMSGENEVA, CH

From insightto action

Research grounded in datasets we build and audit in-house. Trading systems centred on human decisions.

Explore the research

Two views of a conceptual workflow. Research starts with datasets built and audited in-house, then tests context, conditions and limits. The trade lifecycle illustrates human approval, an order request, broker execution confirmation, position monitoring in EQNX Dashboard and a trade supervisor applying protective rules. An exit request becomes a closed position only after confirmation. Trajectories are synthetic; execution and protection are in development. Use the buttons to explore either view immediately.

FIG. 01 / FROM RESEARCH TO SUPERVISIONCONCEPT MODEL

Execution & protection: in development

UNDERSTAND THE MARKETS. BUILD THE TOOLS.FUTURES / FX / CRYPTO
01RESEARCH IS THE STARTING POINT

Test the hypothesis

Building and auditing our research datasets is a substantial part of the work. We prepare sourced market records for each study, keeping provenance, transformations and known gaps visible before testing a hypothesis.

An intuition is tested. A decision is documented.

01 / Data

Establish the record

Document data provenance and preserve immutable raw snapshots. Make session boundaries, contract rollovers and gaps explicit so that every study starts from an identifiable record, with its limitations visible.

Illustrative model · Research workflow
RESEARCH NOTE / OCTOBER 2026

A result can be a reason to stop

In a recent ES/NQ study, all four predeclared primary configurations were negative after costs. None was admitted for trading. Keeping that result is part of the work.

Read the study context

The study evaluated 32 variants of an EQNX mechanical interpretation of Range Sweep Delivery. The single positive sensitivity had an uncertainty interval spanning zero. These historical simulations used minute-level execution assumptions; they do not settle the validity of every interpretation of the method. Finer execution data remains a next step.

02EQNX DASHBOARD

Detect Decide Supervise

HUMAN AT THE DECISION POINT

Live market analysis is already in place. We are extending it with multi-pair scanning, real-time position supervision and predefined protective controls. Detection informs the trader’s decision; it does not authorise an order.

EXISTING FOUNDATION / LIVE MARKET ANALYSIS

Market context, order flow, setup states and alerts already share a connected workspace. The next developments follow the decision further: into the position, its changing risk and the actions required to protect it.

FROM DECISION TO POSITION CONTROLFIG. 02 / CONTROL LOOP
  1. 01Decide
  2. 02Request
  3. 03Confirm
  4. 04Supervise
  5. 05Protect
  6. 06Review

Human approval leads to an order request; the broker must confirm execution before the Dashboard tracks an open position. The planned trade supervisor monitors that position against predefined rules and can request protective action. An exit request remains distinct from a confirmed close.

Concept animation · Execution and protective controls in development

Risk management throughout the trade

The next layer connects detection, the trader’s plan and protective rules. Suggestions support judgement; predefined protections act within the authority the trader has granted.

01

Scan across pairs

FX / CRYPTO / AFD & OTHER SETUPS

Monitor multiple FX and crypto pairs for setups as they form. The scanner development includes AFD and other defined methods, with context, confirmation and invalidation attached to each candidate.

In development
02

Calibrate exposure

POSITION SIZE / AGGREGATE RISK

Compare proposed and open positions with the trader’s risk budget. Flag over- or under-exposure relative to the defined plan, account for correlated positions and make the effect of a size adjustment visible.

In development
03

Define the stop

MISSING STOP / INVALIDATION / SIZING

Flag a missing stop-loss and suggest a level from the setup’s invalidation and volatility context. Show its distance, the implied loss and the corresponding position size for the trader to review before entry.

In development
04

Follow the open trade

SETUP EVOLUTION / PARTIAL EXITS

Follow the position, its remaining risk and the original trade plan. Identify when a planned partial take-profit or a review of protection becomes relevant, while keeping suggestions distinct from authorised actions.

In development
05

Anticipate event risk

ECONOMIC CALENDAR / ADVANCE ALERTS

Bring scheduled high-impact economic releases into the decision context. Warn ahead of relevant events, identify exposed positions and make the trader’s event-risk rules visible before volatility changes.

In development
06

Enforce protective limits

TRADE SUPERVISOR / PER-TRADE & DAILY LIMITS

Monitor predefined per-trade and daily loss limits. Request the agreed stop or close action when a limit is breached, including when the trader cannot intervene. Reconcile every request with the broker’s confirmation and retain the event trail.

In development

Protective controls are under development. Validation must cover execution routing, connectivity failures and broker confirmations before operational use.

03SOFTWARE AS THE WORKING ENVIRONMENT

Inside the tools

The research takes shape in working software: tools to examine a market, replay a decision and trace a result. Each system has a defined purpose and a distinct stage of validation.

01

Quant Lab

Data audits, explicit strategy rules, cost scenarios and trade-by-trade reviews. A record of how each experiment was built and what it found.

A completed backtest is an input to validation, not approval to trade.

Internal research
02

Chart Lab

A visual environment connected to EQNX historical datasets, with comparative charts, indicators, drawings and replay.

Live execution and the launch of Quant Lab backtests are not connected.

Working prototype
03

Order-flow tools

C++ tools, including CVD Sentinel, for cumulative delta, pressure analysis and contextual confirmation signals.

Compiled and tested in code; behaviour in market replay remains to be assessed.

Replay validation next

Development snapshot — October 2026. Internal tools and prototypes; no product release is announced.

04THE ENGINEERING BEHIND THE RESEARCH

Trace every result

The dataset is part of the experiment. We preserve the market source records, document our preparation and internal checks, and connect each result to the data and conventions used.

  1. 01

    Source

    Market records

  2. 02

    Snapshot

    Preserved inputs

  3. 03

    Audit

    Internal checks

  4. 04

    Dataset

    Defined conventions

  5. 05

    Experiment

    Traceable results

DATA → RESEARCH → REVIEW

Built and audited in-house

Our futures workflow turns market records into EQNX research datasets. Internal audits cover the preparation conventions, sessions, contract rolls and known gaps. The source snapshots remain available alongside the resulting datasets.

01

Preserve provenance

Immutable snapshots and content hashes identify the data used in an experiment. Research outputs can be traced to their source and configuration.

02

Respect time

Historical replay exposes only completed bars up to the selected clock time. Incomplete periods and missing records remain explicit. Historical data-reception timestamps are not reconstructed.

03

Make costs explicit

Fees and slippage are declared in the simulation; funding is included where the research requires it. Model assumptions and unresolved execution constraints travel with the result.

04

Separate the validation steps

Software tests, historical evaluation, replay and live operation answer different questions. Passing one stage does not establish the next.

THE WORKING STACK

Research

Python / SQL

Data preparation, internal audits and reproducible experiments. SQL queries connect each study to its research records.

Data

Parquet / DuckDB

Parquet holds the columnar datasets. DuckDB provides the analytical engine and catalogue, querying those files without duplicating the full history.

Systems

C++ / TypeScript / React

C++ powers native market-analysis components. TypeScript and React shape the Dashboard interface and its working tools.

05DIFFERENT MARKETS, DIFFERENT REALITIES

Explore across markets

Different markets expose different mechanisms. Our research spans futures, digital assets and foreign exchange, with coverage expanding as data and methods are qualified. The examples below illustrate current work; they do not define the limits of the research universe.

01 / INDICES / ENERGY / AGRICULTURE & BEYOND

Futures

Contract expiries and the transfer of activityILLUSTRATIVE MODEL

Conceptual animation: activity shifts from a nearby futures contract to the next expiry, with both contract histories kept distinct.

Data & strategy research

Study contract structure, trading sessions and the transfer of activity between expiries. Dataset preparation preserves each contract’s prices and rollover conventions rather than treating successive expiries as one unchanged instrument.

Selected examples: ES, NQ and DAX. Further contracts enter the research universe as their data is qualified.

02 / SPOT / PERPETUALS / CROSS-VENUE

Digital assets

Spot–perpetual basis and fundingILLUSTRATIVE MODEL

Conceptual animation: spot and perpetual price paths move relative to each other; a funding interval marks a separate carrying cost. These are synthetic trajectories, not market prices.

Historical evaluation

Explore the relationship between spot and perpetual markets, fragmented liquidity and the cost of carrying a position. Historical studies declare fees, slippage and funding rather than assuming the quoted spread is an available return.

Selected examples: BTC, ETH and SOL. These are research examples, not an exhaustive list of assets or pairs.

03 / CURRENCY PAIRS / SESSIONS / LIQUIDITY

FX

Two-sided quotes across trading sessionsILLUSTRATIVE MODEL

Conceptual animation: a bid/ask corridor changes through the Asia, Europe and Americas sessions. The paths illustrate quote dynamics, not a fixed pattern or live data.

Data qualification

Investigate bid/ask behaviour and liquidity across overlapping trading sessions. Collection and normalisation are under development; continuous qualified coverage remains a prerequisite for strategy evaluation and the planned multi-pair scanner.

Selected examples: EUR/USD, GBP/USD and USD/JPY. Pair selection follows the hypothesis and the quality of available data.

Selected research examples · Non-exhaustive coverage · October 2026

06RESEARCHERS / ENGINEERS / TRADERS

Build with us

We welcome conversations with researchers, engineers and traders about market data, strategy evaluation, trade supervision and protective controls. Human judgement sets the questions, the assumptions and the rules within which a system may act.

contact@eqnx.trading
WITHIN EQNX
EQNX Trading
Research, engineering and trading systems
EQNX Quant Lab
Experiments, evaluation and validation
EQNX Dashboard
Market analysis and the developing supervision interface
EQNX Stratum
The common platform architecture under development
EQNX Capital
The longer-term institutional ambition
© 2026 EQNX TRADINGBack to top
Risk information

Trading involves a risk of loss. Historical simulations depend on their data and assumptions and do not establish future performance. The research and development examples shown here are not investment recommendations.