Advanced Computing in the Age of AI | Friday, June 9, 2023

Crowd-sourcing Algorithms to Beat Wall Street 

Source: Shutterstock

Looking to cash in on your algorithm expertise? Join the crowd. A three-year-old hedge fund investment company, Quantopian, which crowd-sources algorithms to guide its investing, has just received an infusion of $250 million to play with as it continues to scan the crowd for mathematical keys to success. When an algorithm is used by the company and succeeds, the algorithm developer gets a share of the profits.

Computer-driven investing, of course, is hardly new. Nor is smart mathematicians and scientists trying their hand at quantitative investing. Perhaps the most prominent example is David Shaw, whose hedge fund investing success through The D. E. Shaw Group also funded his D.E. Shaw Research, which has been extraordinarily productive in computational biology. Shaw Research designed and built Anton, a special purpose supercomputer for biosimulation (see HPCwire article, Biomedical Simulation at PSC Gets Major Performance Boost with Anton 2).

“Quantopian says it has 85,000 users signed up from 180 countries who have created more than 400,000 algorithms on the company’s free web-based platform. So far, the firm has only selected 10 of those to trade a few hundred thousand dollars on behalf of Quantopian. The platform is only for U.S. equities trading so far, but Quantopian plans to expand to other asset classes,” according to a July 27 report in The Wall Street Journal (Steven A. Cohen’s Newest Bet: Do-It-Yourself Computer Traders).

That’s a lot of algorithms to sort through (of course, there’s also an algorithm for that) and no doubt IP issues can be prickly.

The Journal article focuses on a new infusion of cash, $250M from investor Steve Cohen, who at one time was one of the riches persons in the world. His firm, S.A.C. Capital Advisors, reached a settlement of $1.8 billion in 2014 after an SEC investigation of insider trading . In January 2016, Cohen reached a settlement with the SEC on the civil charge, in which he is barred from managing money for outside investors until 2018. Now he’s making a bet on Quantopian’s ability to manage money. The Journal article, written by Bradley Hope, notes Quantopian isn’t alone:

“Several other companies, including a quantitative group within Millenium Management LLC and another startup, QuantConnect, have systems that allow amateur quants to submit their algorithms for potential use in trading. At Quantopian, the creators of winning algorithms include a mechanical engineer with a Ph.D. in computational fluid dynamics in Sydney, a data scientist at an internet mapping company in Denver and a consultant in Malta with a master’s degree in mineral and energy economics, according to the company.

“Once a would-be creator logs into the Quantopian system, they can begin playing with data and code to create an algorithm. They can test it against old data or against live data flowing into the system. Quantopian employees look for successful programs using an algorithm of their own that selects candidates based on some 50 attributes, including profitability and how an algorithm might fit into the hedge fund’s overall trading.”

Link to Quantopia: