We developed our recommendation engine algorithm in collaboration with the researchers at the University of Connecticut Finance & Analytics Department. These recommendations are generated based on two key aspects, which we believe inform the investing decisions of customers:
Customer interest: What would the customer like to invest in? This is generated by identifying other customers that are similar to the customer at hand based on demographics and behavioral characteristics. Then recommending holdings of similar customers to the customer at hand.
Portfolio Optimum: What would add optimum value to a customer's portfolio? This is generated by using a dynamic efficient frontier algorithm to iterate over the universe of securities and calculating which securities improve key ratios of the customer's portfolio (e.g., Sharpe ratio, Treynor ratio).
Ultimately, the generated recommendations are blended to create a list of recommendations for each customer that incorporates their interests as well as optimal asset/portfolio characteristics.