Hit Rate
Hit Rate measures how often a forecast correctly anticipates the direction or outcome of a market movement over a defined period.
Rather than focusing on the size of forecast errors, hit rate answers a simpler question: how frequently was the forecast right in directional or categorical terms.
What hit rate captures
Hit rate is commonly used to assess whether a model consistently signals the correct direction or regime. Depending on the definition, a “hit” may refer to:
- correctly predicting price direction
- correctly identifying an up or down move
- correctly classifying outcomes above or below a threshold
Because of this, hit rate reflects consistency rather than precision.
Strengths and limitations
A high hit rate can indicate that a model captures underlying market signals. At the same time, hit rate alone does not account for the magnitude of errors or the economic relevance of misses. Small correct calls and large incorrect ones are treated equally.
As a result, hit rate is most informative when interpreted alongside other performance measures.
Hit rate in commodity forecasting
In commodity markets, hit rate is often used to evaluate short-term directional forecasts or regime signals, particularly in volatile or event-driven environments. It helps distinguish signal quality from noise without relying solely on point accuracy.
At Datasphere Analytics, hit rates observed across our forecasts are consistently higher than those of standard benchmark models, supporting signal quality assessment alongside other metrics.
