Señal de agrupamiento de SD

La lectura principal del Análisis de Patrones — con qué fuerza se inclinan hoy las predicciones sobre acciones de IA hacia una dirección, clasificadas en desviaciones estándar (1, 2 o 3 DE) respecto de su propio historial.

The Standard Deviation (SD) Clustering Signal is a technical breadth indicator designed to measure how strongly a suite of artificial intelligence stock predictions leans in a single direction on any given trading day. By analyzing the collective output of convolutional neural network (CNN) price predictions across a fixed universe of stocks, the system calculates daily breadth statistics such as the ratio of bullish predictions (Up ratio), the ratio of bearish predictions (Down ratio), average prediction confidence, and the count of high-confidence predictions. Each of these daily statistics is then compared against its long-run historical baseline to determine how many standard deviations it sits from its historical average. A daily signal is generated when a primary direction statistic (either the Up ratio or the Down ratio) crosses a threshold of 1, 2, or 3 standard deviations above its historical mean, provided there are no conflicting signals from opposing statistics. For example, an "UP" signal is triggered if the Up ratio clears its standard deviation threshold and no bearish statistic does. If both bullish and bearish statistics cross their respective thresholds simultaneously, or if only secondary confidence metrics cross, the day is classified as neutral. The final headline signal represents the highest clean standard deviation level achieved, such as "2 SD UP" or "1 SD DOWN". For retail investors, this signal serves as a gauge of market-wide momentum and systemic sentiment shifts. Because the thresholds are built on standard deviations, most trading days naturally result in no signal, representing normal market noise. When a 2 SD or 3 SD signal occurs, it indicates an extreme statistical outlier where the AI models are in rare, highly concentrated agreement. Investors use these rare readings to identify potential market turning points, breakouts, or overextended trends, though they must remain cautious as 3 SD events are exceptionally rare and based on small historical sample sizes.

Supongamos que el Up ratio histórico de un universo de acciones seleccionado promedia 42% con una desviación estándar del 10%. Para activar una señal 2 SD UP, el Up ratio diario debe superar la media histórica en al menos dos desviaciones estándar: 42% + (2 10%) = 62%. Si en un día determinado, los modelos de IA predicen que el 63% de las acciones subirá, el Up ratio del 63% supera con éxito el umbral del 62%. Suponiendo que ninguna estadística bajista (como el Down ratio) supere sus respectivos umbrales, el día se califica como una señal '2 SD UP'.