Frassie analyzes market data continuously and replicates the trading logic from quantitative AI strategies. It gives you a basis for decisions based on predictive analysis and risk minimization, regardless of which time zone you are in.
Those who manage capital remotely face a structural disadvantage: markets move continuously, but time and attention do not. Collecting, interpreting and acting on data from several exchanges and time zones yourself requires a work effort that rarely goes hand in hand with a flexible working life.
The result is often delayed decisions, or decisions that are made on incomplete basis. Few individual investors have the tools to process the same volume of data that institutional players do every day.
Frassie identifies trading strategies whose historical decision patterns meet pre-defined criteria for risk-adjusted returns, and then reflects their positioning logic in your own portfolio according to the limits you set.
Market data, order flow and volatility metrics are collected continuously from relevant trading venues.
AI models rank underlying strategies based on consistency, drawdown and risk profile over time.
Predictive analysis adjusts how much weight each signal is given, based on the prevailing market regime.
Positions are reflected in your portfolio proportionally to the capital and risk limit you have set.
The method differs from discretionary management in that every decision can be derived to a measurable parameter. It makes it possible to review, adjust and challenge the model's assumptions on an ongoing basis — rather than relying on the judgment of a single manager.
The features below are built to provide control over exposure, not just visibility into results.
Enter maximum exposure per asset class; the system automatically adjusts position size when the limit is reached.
Notice when a mirrored strategy deviates from its historical pattern of behavior, so that assumptions can be reconsidered.
Simulate how the portfolio would have reacted under previous market conditions before activating a strategy.
Each replicated position is logged with the underlying signal, timing and risk assessment that motivated it.
We report how the strategies are tested and reviewed, rather than asking you to take the results on trust.
Each strategy is tested against historical market data from multiple business cycles before being made available for replication, to avoid overfitting to a single market period.
Strategies' return profile and risk measures are continuously reviewed against predefined threshold values. Strategies that deviate are flagged for manual review before continuing to be mirrored.
Success is not defined solely as return, but as return in relation to the risk framework the user himself has set for his portfolio.
The model ranks and weights incoming signals from underlying strategies based on historical consistency and current market regime. The decision to replicate a position is made when the signal strength exceeds a threshold linked to your chosen risk profile.
Strategies are continuously monitored against their historical patterns. In case of deviation, the exposure is automatically reduced, and the strategy is flagged for manual review before possibly being reactivated.
Yes. Risk limits are set per asset class and total portfolio before replication starts, and these limits can be adjusted continuously without losing the history in the decision log.
Yes, analysis and replication happens continuously in the background regardless of your local time, which is a key reason why the platform targets remote investors.
Instead of a person's subjective judgment, the replication is based on quantifiable signals that can be reviewed in retrospect. It makes it possible to understand exactly why a position was taken.
More questions about methodology or risk management? Contact us.
Go through the methodology, set your risk limits and see how the replication works in practice before activating a strategy in your portfolio.