Using Kronos Forecast Quantiles to Set Trade Entries and Exits

Summary

This automated strategy uses the Kronos financial time-series model to forecast a future price distribution. It compares the forecast median with the current price to decide whether the projected opportunity is large enough to trade, while the lower and upper forecast quantiles define stop levels. A target is set from the median forecast; the position closes at its target or stop, or when the forecast horizon expires. The document describes one-at-a-time position handling, paper and live modes, long-only spot trading, and two-way futures trading.

The supplied material explains deployment requirements and includes partial implementation details for prediction, trade state, and monitoring. It reports that the model was trained on more than 12 billion candlesticks, but provides no strategy backtest results or evidence of predictive or trading performance. The code excerpt is incomplete, so key thresholds, risk sizing, and execution behavior cannot be fully assessed from the available text. Live use also depends on local model setup and exchange integration.

Key ideas

• The strategy uses forecast median and quantile bounds to define trade direction, targets, and stops.
• Positions close when a target, stop, or forecast expiry condition is reached.
• The described modes support paper and live trading, with different direction rules for spot and futures.
• The document gives no strategy performance results, and its code excerpt omits details needed to assess the full method.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.