Signal Win Rate

The percentage of directional clustering signals whose 5-day forward SPY return went the predicted way: UP wins on a positive return, DOWN wins on a negative return.

Signal Win Rate is a metric used to evaluate the historical accuracy of directional trading signals generated by a quantitative model or technical indicator. In the context of market clustering models, it measures the percentage of active signals that correctly predicted the direction of the S&P 500 ETF (SPY) over a subsequent five-trading-day window. An UP signal is classified as a win if the five-day forward return is strictly positive, while a DOWN signal is a win if the return is strictly negative. Any flat or unchanged outcome is treated as a loss, and days with no clear signal are excluded from the calculation entirely. Investors analyze this metric to assess the reliability of a technical model before committing capital. A high win rate suggests that a specific signal level has historically been a dependable leading indicator of short-term market direction. However, traders must evaluate the win rate alongside the total sample size. A very high win rate derived from only a few occurrences is statistically less reliable than a slightly lower win rate backed by hundreds of historical data points. When interpreting Signal Win Rate, it is also crucial to consider the average forward return associated with those wins. A signal can have a high win rate but a low average return, meaning it wins often but yields small gains, or conversely, it could have a lower win rate but larger average gains. Understanding this balance helps investors manage risk and set realistic profit targets.

Over a backtested period, a quantitative model generates 120 directional signals. Out of these, 20 are classified as 'No clear signal' and are excluded. Of the remaining 100 active signals, 65 correctly predict the 5-day direction of SPY, while 35 result in flat or opposite price movements. The Signal Win Rate is calculated as 65 / 100 = 65%.