A strategy becomes more useful when its assumptions can be tested. MyForexGlobal Capital provides quantitative research and strategy validation for traders, firms, and researchers who want to know how an idea behaves under explicit data, cost, and risk assumptions.
The service is designed for questions such as: Does the edge survive outside the development sample? Which parameters matter most? How sensitive is the result to transaction costs? Does the behavior change across market regimes? The objective is not to produce the most attractive backtest. It is to reduce the number of unanswered questions around the strategy.
Turn a market idea into a testable research question
Quantitative research is useful when a market idea is still a hypothesis, when a discretionary process needs to be translated into testable rules, or when an existing model has impressive historical results but weak validation. We can help structure the research question, define the test, and separate evidence from optimization noise.
What the research can examine
- Rule-based trading strategies and market hypotheses.
- Entry, exit, holding-period, and position-sizing logic.
- Parameter sensitivity and stability.
- Performance across instruments, sessions, periods, and market regimes.
- Transaction-cost and slippage assumptions.
- In-sample, out-of-sample, and walk-forward behavior.
- Drawdown, tail behavior, trade distribution, and path dependence.
- Benchmark comparisons and alternative specifications.
A typical research assignment
The engagement usually starts with strategy rules, existing research, sample data, or a clear hypothesis. We first define what can be tested without changing the question halfway through the analysis. The output may be a validation report, reproducible research notebook, test dataset, charts, code, or a combination agreed during scoping.
How the validation work is structured
- Formalize the hypothesis. Convert the idea into variables, rules, assumptions, and measurable outcomes.
- Prepare the data. Check time alignment, missing values, survivorship issues, look-ahead risk, and other data-quality concerns relevant to the test.
- Build the baseline. Establish a transparent reference model before introducing additional complexity.
- Validate outside the fit. Use time-aware holdouts, walk-forward analysis, or another structure appropriate to the data.
- Stress the result. Test costs, parameters, samples, regimes, and alternative assumptions.
- Report what survives. Separate robust findings, weak evidence, and unresolved questions.
What counts as credible evidence
Our public research pages explain several methods that may be used in an engagement, including backtesting, forward testing, strategy validation, and overfitting control. Walk-forward analysis and time-series cross-validation may also be used when they fit the data and research design. No single test is treated as universal proof.
Research outputs
Deliverables are agreed before work begins. They may include a research report, assumptions register, performance tables, robustness tests, code, notebook, charts, model comparison, or recommendations for the next research step. Results are documented so the client can see both the evidence and the limitations.
Where the research mandate stops
This service is research and analytical work. A positive historical result is not a guarantee of future performance, and a weak result is not automatically proof that an idea can never work. We do not take custody of client funds or present strategy validation as regulated investment management.
Bring us the strategy or hypothesis
If you have a strategy, hypothesis, or existing backtest that needs independent examination, describe the project and include the rules, available data, current evidence, and the decision you need the research to support.