K Score - US - Non Financial

Approach

Harnessing years of historical data and a variety of data sources, Kavout takes in structured and unstructured data including fundamental information, pricing and trading volumes, technical signals, and alternative data, to deliver an actionable and easy-to-use equity rating score from 1 to 9.

The development of K Scores apply the best-of-breed learning algorithms, and encompass methods such as regression, classification, model selection, deep learning and more to produce a rating to rank stocks. Our state-of-the-art machine learning (ML) models identify the intrinsic relationships, uncover the dependency structure and calculates predictive analytics for future performance.

Factors Evaluated

Pairing fundamental research with quantitative analysis, our ML model evaluates over 200 factors and signals. Some factor groups include but are not limited to

1. Value, Quality, Growth, Momentum, Profitability, Volatility, Size and more.
2. Technical indicators and scores such as MACD, RSI, Z-Score, M-Score and more.
3. Price patterns such as chart and candlestick patterns.
4. Time series of the features above as well as price-related data.
5. Sentiment information such as insider and options trading activities, and text analytics.

Methodology

On any given day, Kavout processes millions of diverse data sets, and runs models encompassing many traditional and advanced financial engineering methods such as regression, classification, deep learning, and reinforcement learning to produce a predictive rating to rank stocks.