You’ve probably heard of crowdsourcing in a variety of sectors, from journalism and public policy, to scientific research. It’s a sourcing model that individuals or organizations use to obtain goods, services, ideas, or finances from a large, open group of internet users. It then divides work between participants to achieve a aggregate result.
So what does this have to do with hedge funds?
Crowdsourcing is actually quite a new concept for the worlds of finance and hedge funds, but it is beginning to take off. Now, several companies are crowdsourcing machine learning hedge funds. Read on to learn more about three of the top crowdsourcing AI Quant hedge funds. All of them are based in the United States.
Numerai is fairly unique in that its rewards are paid in cryptocurrency. It transforms and regularizes financial data into machine learning problems for its global community of data scientists.
This hedge fund aims to change the way Wall Street operates by promoting a collaborative approach to investment and money management. Monthly tournaments and machine learning power it. Each month, data scientists submit their predictions and best trading algorithms in exchange for a form of their cryptocurrency, Numeraire.
With $7.5m invested from companies like Union Square, Playfair Capital and First Round, Numerai manages a long/short global equity hedge fund. It is also one of the only neo hedge funds that offers full anonymity.
Quantiacs aims to simplify trading system developments so people can succeed as quants. Through its site, freelance quants can access tools, data and training. Then they create trading algorithms to compete for the prizes. Afterwards, the company matches freelance quants with hedge funds and investors can customize their profile goals and risks. Quants own their IP but license it to Quantiacs for 10% of lifetime profits.
Quantiacs raised a $1 million seed investment from Baha Holdings, a venture fund dedicated to fintech startups. Quantiacs claims that after 18 months in a closed beta, they have more than 800 trading algorithms on the platform. Students using neutral networks and other machine learning approaches have developed many of those algorithms.