Avelixio applies predictive analysis models to historical market series to build verifiable strategies. Each recommendation is traceable back to the data that generated it.
The system does not make predictions based on intuition. It processes historical data through three sequential phases, each verifiable and documented, before proposing any operational indication.
The engine collects price, volume and volatility data from multiple market sources, normalizing it to eliminate anomalies and discontinuities that could distort subsequent analysis.
Statistical models analyze the correlations between macro variables and asset behavior, isolating recurring signals from the market noise typical of short-term fluctuations.
The final indications take into account the historical volatility of the asset and the declared risk profile, avoiding excessive concentrations on individual positions.
Technical validation: each strategy is tested on historical time windows distinct from those used for model calibration, to reduce the risk of overfitting on past data.
The following features reduce the operational complexity typical of quantitative analysis, maintaining transparency on the criteria used for each recommendation.
The system constantly updates the analysis of market conditions, reporting significant changes in volatility without requiring constant manual control by the user.
Each user profile defines maximum exposure thresholds per asset and overall portfolio, automatically applied to each indication generated by the engine.
Each recommendation is accompanied by the historical parameters that generated it, which can be consulted at any time to verify the underlying logic.
We do not show case studies of individual users: we show how the model behaves compared to market history, within the limits of available data.
The solid line represents the simulated performance of a strategy filtered by the model over a defined historical period. The dotted line represents the raw volatility of the reference asset in the same range.
The comparison is not a prediction of future performance: it illustrates how the predictive filter reduces exposure to swing peaks compared to an unmanaged exposure.
Transparency Note: Simulated historical results do not guarantee future results. Backtesting data is subject to the intrinsic limitations of the available historical series and the market conditions of the period analyzed.
The typical profile of the university user requires time efficiency and risk containment, rather than frequency of operations.
The engine automatically processes variables that would require hours of manual analysis, leaving only the final allocation decision to the user.
The risk thresholds adapt to limited investment capital, avoiding logic designed for institutional portfolios.
The proposed strategies favor consistency over medium-long periods over high-frequency trading, reducing implicit transaction costs.
Avelixio was created to offer students and young professionals structured access to quantitative analysis, historically reserved for institutional operators with greater technical resources.
Our work focuses on the quality of the input data and the testability of the models, rather than the promise of high returns in a short time.
Access the platform to consult the backtesting methodology and configure your risk profile before making any operational decisions.
The Analysis begins Do you have questions about the methodology? Consult the FAQ