We use machine-learning tools to identify prevailing trends that have historically affected stock performances and use those patterns to estimate probabilities of upcoming outperformance on an individual stock basis.
In addition, we analyze the interaction and correlation of different individual stocks to optimize portfolio risk profiles and diversification efficacy.

Synthetically, we group the factors that affect stock’s price performance in three global groups:

(i) macroeconomic and market variables, from inflation to policy rates to long-term treasury yields;
(ii) trading technical, like stock price momentum and trailing average prices; and,
(iii) company fundamentals, like earnings and leverage.

Our system analyzes a decade of data in each of these drivers, identifying static pattern and dynamic trends that relate the data set with each stock’s performance. As a first step, it allocates each stock’s price an individual dependence on each driver both on direction and intensity (e.g., a particular stock may be more sensitive to rates, whereas another depends on price momentum at a particular point in time).
Subsequently, the platform ranks each stock by positive exposure to prevailing market conditions at a particular point in time.

We have bifurcated goals for the platform:

(i) on an individual stock basis, and to provide weekly trading ideas, we seek to highlight stocks that are likeliest to outperform the index (typically the S&P500) in the short term; and,

(ii) on a portfolio basis, identify groups of stocks that would optimize stock price outperformance while containing the volatility (or reducing maximum drawdowns) of the investment.