Robots are not new to the world of finance, where algorithms are used to spot trends. Credit: Getty Images/iStockphoto/blackdovfx
There’s a robot in the Manhattan offices of investment firm Stripes that enjoys reading. It consumes everything it can on the 13 million private companies it tracks. When it finds one it likes, about 0.006% of the time, human co-workers start the often years-long process of relationship building.
High frequency trading this isn’t. But the unusual marriage of algorithms and shoe leather has scored Stripes some big wins. Now, it’s getting closer to perfecting the art of not losing.
Robots in finance aren’t new. Hedge funds have been using them to beat humans for decades, building algorithms able to dip in and out of an investment several times before you finish this sentence. Yet in private equity, where swagger and a monogrammed shirt are still the most reliable way to source deals, artificial intelligence hasn’t caught on.
Stripes’ recent cyborg successes could make a strong case for that to change.
The firm’s investments in Swiss sneaker maker On Running and workplace software business Monday.com have, respectively, netted Stripes 18 and 22 times their original investment. Money transfer app Remitly Global and online education platform Udemy have delivered similarly splashy returns.
But while the wins are impressive, it’s the losses — or rather the lack of them — that stand out. Since 2015, the robot — known internally as SONAR — has been learning not only the art of catching winners, but also how to avoid duds.
The scoring formula is tightly held, but the robot pulls from a diverse universe of information, including credit card scans, company filings, news reports, consumer-engagement numbers, employee-retention data and cohort performance.
The latest LI business news in your inbox Monday through Friday.
By clicking Sign up, you agree to our privacy policy.
Privacy Policy | Terms of service | Subscription terms | Your ad choices | Cookie Settings | California Privacy Rights | About Us | Contact Newsday | Reprints & permissions | Advertise with Newsday | Sitemap | Help