Research & Methodology

Research Philosophy

Mana Investment develops and uses a proprietary multi-market research engine. Its infrastructure is already operational in the internal research environment; the work extends from identifying candidate strategies to evaluating and monitoring selected configurations.

Mana is the research environment, A103 is a documented application on ES and GC, and R1 is the portfolio capitalisation and monthly sizing presentation for that application.

The objective of the research process is not to optimize for isolated historical outcomes, but to evaluate whether a strategy can preserve coherent behavior across changing intraday market environments.

Robustness, traceability and risk awareness are treated as core research constraints rather than secondary considerations.

From Research to Candidate Strategies

Mana Investment draws on more than fifteen years of research and development across quantitative finance, mathematics, statistics, software engineering and systematic market analysis. This research predates the company, incorporated on 18 September 2026. Research background and corporate milestones →

A103 is the outcome of several years of quantitative research, software development and evaluation of intraday strategies. Approaches that did not meet the research requirements for performance, consistency and risk control were revised or discarded. The underlying research covers approximately 1,500 intraday sessions across varied market conditions and includes analysis of daily results, strategy behaviour and sensitivity to indicative transaction costs. Positions are closed within the same trading session, with no overnight holdings.

The difficult work begins before a performance table is produced: identifying variables, their interactions, candidate trading rules and parameter configurations that can be evaluated under explicit risk constraints. Research and calibration are market-specific; an oil-market configuration is not mechanically copied from a Nasdaq model.

Knowledge Base & Daily Research

Mana draws on an extensive knowledge base, enriched and updated daily with market observations and model-calibration work.

The knowledge base organises market observations and calibration work to develop and examine candidate configurations.

Daily data updates and ongoing calibration of candidate models are separate from changes to the selected strategies. Proprietary variables, parameter values and internal knowledge structures are not disclosed publicly.

Validation Framework

Research validation is designed to evaluate stability and robustness across multiple market conditions, rather than rely on a single historical performance profile.

• Regime-based evaluation across multiple market environments
• Out-of-sample and walk-forward validation methodologies
• Stress testing under adverse intraday and liquidity conditions
• Parameter sensitivity and robustness analysis
• Focus on process stability rather than curve-fitting

Research methodologies are reviewed as part of an ongoing research and development process. Material strategy changes, when considered, are subject to controlled validation rather than short-term performance reactions.

Continuous Research and Strategy Continuity

The underlying A103 trading strategies were selected in late 2025. The January-August 2026 results presented use the same underlying configuration, without changes to its trading rules. Ongoing research review should not be confused with changes to the strategies used for those results.

The knowledge base continues to grow and other candidate models continue to be calibrated within the same environment. These activities do not mean that the selected A103 trading configuration is replaced or retuned daily. Candidate research and the monitoring of reference strategies remain distinct.

Selected configurations are followed as new observations become available. Material changes, when considered, are evaluated separately rather than introduced in response to the latest monthly result.

Selection focuses on consistency across observations, sensitivity to parameters, transaction costs and the risk framework, rather than the highest isolated backtest return.

Proprietary research processes, selection criteria, model parameters and internal knowledge structures are not publicly disclosed.

Behavioral Discipline

The systematic framework is designed to minimize discretionary drift and support execution consistency through rule-based processes, embedded risk constraints and predefined operational controls.

• Rule-based execution logic
• Constraint-first framework design
• Reduction of discretionary behavioral bias
• Systematic execution without discretionary signal intervention
• Predefined operational safeguards

Market Structure Research

Research activities incorporate analysis of market microstructure, liquidity conditions, volatility behaviour and intraday execution characteristics. The shared infrastructure already integrates time zones, trading sessions and instrument-specific operating conventions.

Common engineering and validation methods do not imply identical trading models. Variables, parameter configurations and risk limits are studied for each market rather than transferred mechanically from ES or GC.

Research Coverage

Documented application: ES and GC. The public A103 reports cover the selected strategies on these instruments and their R1 portfolio results.

Broader research universe: NQ, CL, EUR/USD, GBP/USD, K200, HSI and ZN, among other markets. Market-specific operating conventions are assessed within the shared research infrastructure. Strategy identification, calibration and evaluation are conducted separately for each market.

Each programme has its own calibration and evaluation. Published A103 results cover ES and GC only.

Technical Architecture

The proprietary research and execution infrastructure is already implemented and used in the internal research environment. The market-specific operating conventions have been validated internally. Software engineering, monitoring and data-integrity processes support the common research and evaluation workflow.

• Structured research and execution pipeline
• Internal monitoring of strategy behavior
• Market data integrity controls
• Post-session review and diagnostic procedures

Session Replay & Detailed Research Review

The Mana research engine supports replay of historical trading sessions, individually or for a selected set of dates, with detailed visibility into the simulated transactions generated within each session. This capability complements daily and monthly results by allowing closer examination of intraday strategy behaviour.

Focused demonstrations can be arranged for professional research discussions under confidentiality.

This section describes the internally validated research infrastructure. Its role and the basis of performance reporting are distinguished in the disclosures.

Historical Research Reports

A103 is the documented ES/GC application of the broader Mana engine. The Investor Summary presents historical research results (simulated) under R1: separate component capitalisation, monthly maximum contract limits and geometrically linked returns.

The earlier fixed-capital study remains a separate archive. Its sizing policy differs from R1; use each report's definitions when reading its figures.

The current report and Performance page set out the calculation basis, cost assumptions and observation frequency.

Review A103 performance → · Fixed-capital archive

Research Information

Mana Investment SAS is incorporated in France and does not manage third-party capital. This material is for research review, not investment advice or an offer. Full disclosures →