Matthew Rossi is the founder, managing partner, and chief of research for MarketDNA Investment Research. In this episode of Strategic Investor Radio, Mr. Rossi explains his firm’s eponymous MarketDNA product, which takes intraday data and runs it through an algorithm to “spit out” equity trades with 80-90% accuracy. Host Charley Wright calls MarketDNA a “fascinating approach of how to invest.”
MarketDNA uses around 100 filters to ex-out “noise” and find the 1-2 big derivatives trades happening because of inefficiencies in information. Sometimes this means finding “insider trading,” and Rossi shares a story of meeting Martha Stewart’s brother on a golf course before the infamous ImClone trade that landed her in prison. According to Rossi’s story, Ms. Stewart’s brother asked him to lie on her behalf, claiming she traded on the basis of his research instead of illegal inside knowledge. Needless to say, Mr. Rossi refused!
Rossi also discussed an incident with OPRA (the “Options Price Reporting Authority” – no relation to Ms. Winfrey) when MarketDNA’s algorithms noticed its servers were down and received a phone call from a suspicious OPRA executive asking how they knew – no one else in the country had.
MarketDNA typically holds 7-10 positions, long or short, at any given time, and each of these positions is normally held for 3-5 weeks. The firm’s clients are hedge funds and institutional investors, but it has begun making a la carte research available to individual investors. These one-off recommendations come instructions to buy or sell-short with entry and exit prices, but no other research.
When asked what causes some MarketDNA trades to fail, Rossi blames short squeezes and raised guidance around quarterly earnings. Fortunately, most of MarketDNA’s trades aren’t earnings-related.
When asked what keeps him up at night, Rossi cites the excitement of his work – trying to identify a previously unobservable trade, and making sure no one out there is cheating.
Listen to the Episode
Subscribe and listen to this show: iTunes
Past performance does not necessarily predict future results.