Cresta Patrimenza observes the markets, learns your risk appetite and automatically adjusts your portfolio strategy, without requiring data analysis skills or manual configuration.
The problem
Reading graphs, interpreting indicators and following market trends in real time requires a daily commitment that few people can afford. The result is often passive management by inertia, not by conscious choice.
The solution
Cresta Patrimenza introduces a layer of Predictive Analytics — the ability of artificial intelligence to anticipate trends by observing large volumes of historical and current data — that works in the background, updating recommendations without the user having to interpret the data themselves.
How it works
Each phase is managed by the system continuously. The user intervenes only to approve the indications received, not to produce them.
The platform collects market data, macroeconomic indicators and historical series from public and regulated sources, updating them at regular intervals during the trading day.
The model compares observed user behavior—dwell times, previous choices, reactions to volatility—with a risk profile that refines over time, without requiring repeated questionnaires.
Based on the updated profile, the system proposes an allocation consistent with the declared risk mitigation objective, leaving the final confirmation decision to the user.
The risk adaptation engine
Most platforms ask the user to declare their risk tolerance only once. Cresta Patrimenza instead observes real behavior over time, because people's preferences change with market conditions and accumulated experience.
“Risk” indicates the possibility that an investment will permanently lose value. "Volatility" indicates how much the value fluctuates in the short term, even when the underlying trend remains positive. Our model distinguishes between the two: it doesn't automatically reduce exposure just because the market fluctuates, but intervenes when it detects a structural change in risk.
Within the platform, the risk profile is represented as a continuous band, not as a fixed label such as "prudent" or "aggressive". The band gradually moves based on the observed data, and is always visible before any allocation proposal.
Personal and financial data is processed with encryption in transit and at rest. Access to the transaction confirmation functions is always subject to explicit authorization from the user: the system proposes, it does not dispose of the funds independently.
Methodology and transparency
We do not publish individual success stories, because the results depend on uncontrollable market conditions. We prefer to explain the internal workings of the models.
The system combines supervised learning models, trained on multi-year time series, with risk management rules that limit the maximum exposure allowed for each profile.
The information comes from regulated market feeds and public macroeconomic indicators. We do not use data that cannot be verified or come from undeclared sources.
Each change to the model is tested on historical periods distinct from those used for training, to verify that the observed behavior does not depend on a single market scenario.
Frequently asked questions
We do not disclose minimum thresholds on this page, as they may vary based on the type of account and applicable law. Updated entry conditions are available by contacting our team.
Initial setup requires a profile setup session. Subsequently, the system works autonomously and only sends you a notification when explicit confirmation is required, not for routine updates.
Data is encrypted both in transit and at rest. Artificial intelligence generates allocation proposals, but each capital movement requires confirmation authorized by the user: the system does not dispose of the funds independently.
We do not ask you to study graphs or learn advanced technical terminology. The initial phase consists of a short setup session, after which the system continues to work autonomously.