ABSL has launched a highly differentiated quant fund that aims to blend the best of human insights and machine discipline to deliver more consistent investment outcomes.
The fund harnesses time-tested equity research insights (the human element) to generate its investment universe and then drives filters, weights, portfolio composition and sell decisions through well defined algos(the machine component).
The universe is obtained by including the 75 most owned stocks of each of the top 15 fund houses – which gives a number between 200-300stocks. After eliminating small cap stocks, quality and momentum filters are then applied to narrow down to a portfolio of 40-50 stocks which are then weighted to create the portfolio.
Harish believes this new fund will compete very effectively with large and midcap funds on both counts – alpha as well as risk adjusted returns. More importantly, it should be able to deliver a lot more consistency – which its back-testing has established clearly.
Pure quant always runs the risk of garbage in – garbage out and pure human effort is prone to personal biases. This jugalbandi attempts to harness the best of both worlds for the benefit of its investors.