TRADEMEM
English

Backtested Trading Signals — Every Signal Has a Track Record

Most signal providers show you predictions. TRADEMEM shows you proof. Every signal is validated against historical data with transparent metrics: return, win rate, drawdown, trade count and position duration.

Access Live Signals

What Backtest Metrics Are Provided

For every asset and every timeframe, TRADEMEM computes: total return (%), win rate (%), maximum drawdown (peak-to-trough), number of trades executed, and cumulated time in position. These metrics are computed on held-out data — never on the training set.

Negative Returns Override Predictions

If a backtest shows negative total returns for a given timeframe, the prediction is automatically overridden to sell with 100% confidence. This is a hard rule — the model does not fight its own track record.

Held-Out Validation Methodology

Every KANNOT-1M model is evaluated on a 20% held-out test set. We report accuracy, precision, recall, F1 score and full confusion matrix. No model ships without statistical proof of performance.

Why Backtesting Matters for Traders

Backtesting is the difference between faith and evidence. A signal with a 65% win rate and controlled drawdown tells you something actionable. A signal without a track record tells you nothing. TRADEMEM makes the choice obvious.

FAQ

Are backtests run on historical data the model has seen?

No. All backtests are computed on held-out test data that the model was never trained on.

How often are backtest metrics recalculated?

Backtest data is refreshed with configurable staleness thresholds — typically every 4 to 24 hours depending on the timeframe.

What happens when a backtest shows losses?

The prediction for that timeframe is automatically overridden to sell. The original prediction is shown struck-through for transparency.

Trading involves significant risk. TRADEMEM provides AI-generated signals for informational purposes only — not financial advice. Past performance is not indicative of future results.

TRADEMEM

© 2026 Michael Vergoz. All rights reserved. Made in 🇨🇭