K Score - US - Non Financial

Overview

K Score is a predictive equity rating powered by machine learning (ML) technologies. Kavout takes in vast and diverse datasets including fundamentals, pricing, technical indicators, and alternative data, and then uses a combination of statistical analysis and ML techniques with ranking algorithms to deliver an actionable equity rating score with values ranging from 1-to-9.

A higher K Score (7-9) assigned to a stock indicates a higher probability of outperformance, whereas a lower K Score (1-3) indicates a lower probability of outperformance in the next month.

Data Highlights

  1. For quantitative users, overlay K Score as a signal in investment models.
  2. Generate new equity long/short ideas that weren't on your radar before, or validate research as a screening parameter.
  3. Mitigate risk in portfolio construction by avoiding stocks with low K scores (1-3).

Performance Showcase

To test the performance of the K Score, we constructed two portfolios – Top Picks and Bottom Picks. The Top Picks portfolio is comprised all the tradable stock equities on the US stock exchanges with a K Score of 9, while the Bottom Picks portfolio have stocks in K Score of 1.

Cumulative Return of Portfolios (2012-01-03 to 2019-08-01)

Statistics CAGR Alpha Volatility Beta Sharpe
Top Picks 15.87% 5.40% 13.17% 1.03 1.19
Top 500 US Stocks by Market Cap 9.74% 0.00% 11.07% 1.00 0.90
Bottom Picks 6.68% -3.78% 14.87% 1.14 0.51