Skip to content

Quantitative systems · Switzerland

Quantitative architecture. Built for systematic clarity.

We develop, test and protect trading systems for MetaTrader 4/5 and FIX API: expert advisors, indicators, connectors, news readers and monitoring. For over 25 years, traceable, reproducible, risk-aware.

MODEL ── NULL ┈┈ ▲ ENTRY · SIMt 095 · -0.34
Illustrative chart. Not actual performance.
25+
years of experience with forex systems
validation stages from hypothesis to monitoring
  1. Hypothesis
  2. Data
  3. Validation
  4. Robustness
  5. Monitoring
Every stage must be passed before the next one begins. See the whole chain of checks →
FIX
API, MT4 and MT5: connectivity to brokers and platforms

Offer

What we do for you

We develop trading software to your rules, test finished and half-finished products in our test environment and protect your source code. We develop systematic trading strategies and systems for algorithmic trading, from research to broker connectivity.

  • 01

    Custom development

    Expert advisors, indicators, connectors and copy solutions for MetaTrader 4/5 and FIX API, including conversion and extension of existing programs.

    To development →
  • 02

    Testing your product

    We test your product on real tick data, on request with your parameters and optimisation. You receive a full report, on request with ready-to-use set files.

    To product testing and sample report →
  • 03

    Protecting your source code

    Encryption, expiry date, binding to the account holder's name or to the account number. Singly or combined.

    To source code protection →
  • 04

    Custom news reader

    Filter by currency and keyword, windows before and after the event, bank holidays. A protection layer, not a signal source.

    To the news reader →
  • 05

    Fast instead of tick churning

    We find processes that compute needlessly on every tick and make the EA lean, fast and resource-friendly, in backtest and in operation.

    To speed →
  • 06

    Latency and order routing

    We measure how long requests, broker response and order processing take, optimise the routing and show how good your broker really is.

    To latency measurement →
  • 07

    Monitoring in operation

    Watchdog panel with sources, backlog and service status, messages to Telegram or WhatsApp.

    To monitoring →
  • Send an enquiry

Capabilities

From research to broker connectivity. One team, one standard.

  • Research and strategy development

    Formulate hypotheses, test them against a null model and turn them into tradable rules, including arbitrage systems.

  • Expert advisors, indicators and tools

    In-house development for MetaTrader 4/5: from indicator to expert advisor to analysis tool.

  • FIX API and connectors

    Direct connection to liquidity providers and platforms via FIX API, plus data and copy solutions.

  • Data, execution and risk

    Clean data handling, verified execution, robustness tests and continuous monitoring.

See development and testing

A global team

Traders, developers and mathematicians, spread across several countries. Working languages are German and English. Everyone works under the Sentinel Systematics brand to the same testing standard.

Method

Five stages. None is skipped.

01 / 05

Hypothesis

Every strategy starts from a reasoned market assumption, not from a parameter search. Without a causal explanation there is no test.

Key questionWhy should this effect exist?

See the full chain of checks

Technology

Modular, from data to order.

  1. 01Data
  2. 02Research engine
  3. 03Signal modules
  4. 04Risk
  5. 05Execution
  6. 06Monitoring
Watchdog panel with sources, backlog, service status and Telegram connection
From operation: the watchdog panel of an integration. Anonymised.
Modular system architecture
Shared infrastructure, separate strategy logic. Every module carries its own hypothesis.
To the principle →
Data processing
Tick and bar data, versioned preparation, every join checked for match rate.
To the chain of checks →
Automated workflows
Reproducible runs with an engine fingerprint: data and code state are tied to every result.
To the test infrastructure →
Interfaces and copy solutions
Connection to MetaTrader 4/5 and further platforms through defined interfaces.
To monitoring →
Test infrastructure: job queue and scheduler distribute test runs across four isolated instances
Many tests run in parallel without disturbing the workstation. Illustrative.

Research

Research areas

  • 01

    Market regimes

    When does a market change its behaviour, and how early can that be measured?

    We measure regime changes only with data that was available at decision time, and test every classification against a null model.

  • 02

    Volatility structure

    Distribution, clustering and time-of-day profiles of price movement.

    Time-of-day and session profiles are measured per instrument and never taken over from another.

  • 03

    Model robustness

    How strongly does a result depend on parameters, years and single trades?

    A result only counts once it survives Monte Carlo runs, cost surcharges and a split by market phase.

    Monte Carlo distribution of possible outcomes
  • 04

    Execution quality

    Spread, slippage and fill behaviour as the cost axis of every strategy.

    Costs decide whether a logic holds. That is why the cost axis runs in every evaluation.

    Cost stress: result under rising costs
  • 05

    Risk analysis

    Drawdown distributions, position sizing and shutdown criteria.

    We look at the whole distribution of losing streaks, not only the worst case of the past.

What failed the checks

Not every idea passes. Two examples we rejected, and why.

  • Reaction to price levels

    In the first evaluation the idea looked strong. After testing against a null model and without phantom positions, no edge was left. No Expert Advisor was built.

  • Breakout variant with good-looking numbers

    The signal could not be reproduced in the Expert Advisor the way the research model specified it. Good numbers do not replace parity, so it was not released.

Governance

Risk notice

Trading financial instruments involves substantial risk of loss. Simulations and past results are not a reliable indicator of future results.

This website is for information only. It contains no investment advice, no offer and no promise of return or performance.