Model Drift

Model drift occurs when the statistical properties of the target variable change over time, rendering a previously trained model less accurate. In cryptocurrency markets, this is often driven by sudden shifts in liquidity, regulatory changes, or fundamental changes in network usage.

Because models are trained on historical data, they assume the future will resemble the past. When market regimes shift, the relationship between input features and the target outcome breaks down.

Monitoring systems detect this by comparing real-time prediction error rates against historical benchmarks. If the error exceeds a defined threshold, the model is flagged for retraining or structural adjustment.

Failure to address drift can lead to significant losses in automated trading systems. It is a critical component of maintaining model relevance in highly volatile asset classes.