AlphaCapture is a project that I have developped over the years based on my experience in manager selection, stock picking, portfolio construction and risk management.
AlphaCapture is essentially a system that dynamically selects a portfolio of international stocks based on the regulatory filings, public announcements, and disclosed actions of some talented investors and company insiders. All the information used by Alpha Capture is publicly available.
Unlike some of its peers, our system uses (i) a smart selection of the investors and insiders that it monitors, based on a thorough analysis of their past performance over various periods; and (ii) a smart selection of the information (from filings, announcements and actions) it uses based on some dynamic rules and analytics.
The current version of Alpha Capture is 100% automated, long only, mostly U.S. focused, and went live in January 2020. It has not been "optimized". It would have delivered a 9% annualized alpha over the S&P 500 since March 2008, with similar risk characteristics.
I am currently working on creating a long-short version, adding some non-U.S. markets, and incorporating a machine learning component to analyse some aspects of the regulatory filings.