Two years ago, the government put an end to a covert—and lucrative—Wall Street practice of surveying stock analysts for their opinions on companies beyond what they’d put in their research reports—a practice regulators dubbed “Insider Trading 2.0.”
Rather than the typical buy/sell/hold recommendations and price targets, the surveys asked analysts to answer a series of questions about the companies they covered—say, to rank them in order of competitive position, or rate their likelihood of an earnings surprise on a scale of 1 to 9. Quantitative investors would then plug the data into their trading algorithms to in order to beat the market and make money. That didn’t sit well with the New York Attorney General, who thought the traders were getting an unfair advantage. Blackrock (BLK) didn’t admit to any wrongdoing, but paid $400,000 to the New York AG to “cover the cost of the investigation,” and agreed, along with 18 other financial firms, to stop soliciting analysts’ responses or answering the surveys.
There was just one problem: Hedge funds had relied on the analysts’ stock rankings; certain fund managers still swear it was the biggest source of alpha they’d ever had. And it was gone.
Now, a startup thinks it’s come up with an above-board way to obtain the data that’s just as valuable to investors—and this time, anyone can put money on it. Estimize, a company that specializes in crowdsourcing earnings estimates that it says are more accurate than Wall Street’s consensus 74% of the time, is launching Forcerank, a game some are describing as the stock market version of fantasy sports sites FanDuel and DraftKings.
The app, which opens Thursday to the several thousand people already on its wait list as well as the general public, asks players to rank sets of 10 stocks in order of which they think will perform best in the upcoming week. Forcerank players whose rankings most accurately matched the stocks’ returns relative to each other at the end of the week can win cash rewards ranging from a few bucks to $4,000 for first place in a large prize pool game the company plans to hold in the next few weeks.
Unlike Estimize, which allows anyone to post earnings forecasts for free and doesn’t offer rewards for accuracy, Forcerank contests cost $5 to $100 to enter (with free contests offered occasionally). That’s partly to encourage people to “compete better,” says CEO Leigh Drogen, as well as just to make the game more fun. And while the entry fees are helping to fund the app’s development for now—in addition to the $7 million Estimize raised last year—”that’s not our business,” says Drogen. He expects to make a lot more money selling the data gleaned from the rankings to hedge funds and other investors, who may find it as useful in predicting stocks’ movements as the analyst surveys were, as Forcerank players’ attitudes about companies could reflect the market’s as a whole.
“I think companies like Estimize have proved that crowdsourced data has value, and we’re a big believer in that,” says Matt Ober, co-head of data strategy at major hedge fund WorldQuant, which was Estimize’s first client as well as an investor in the startup through WorldQuant’s venture capital arm.
While the stock-ranking game seems similar to the way FanDuel gamers “draft” a roster of their top football players, Forcerank’s awards are simply based on stocks’ relative performance, regardless of how well or poorly the stocks actually performed. In a technology contest on Forcerank, for example, competitors can rank Amazon (AMZN), Netflix (NFLX) and Alibaba (BABA), along with seven other e-commerce stocks, according to which they believe will post the highest returns next week. If someone ranked Amazon first, the player would receive 100 points if Amazon’s shares rose, say, 5% to beat the rest of the stocks. But if Netflix bested them all with a 6% return, players would get 50 points if they’d ranked it second, another 50 points for being one spot off on Amazon, and so on (with fewer points awarded the farther away they were in their stock placements).
Fantasy sports prizes, on the other hand, are heavily tied to whether teams win or lose. Estimize also contends that there’s another important difference insulating Forcerank from the gambling accusations that have recently plagued fantasy sports sites, along with failed market betting ventures like Intrade: Forcerank is not a game of chance—the legal definition of gambling. Nor does it facilitate trades or sell derivatives, financial instruments that could constitute illegal trading in such an unregulated marketplace.
Stock trading posing as "fantasy," what could go wrong? https://t.co/mjFbq5ek9Y
— Dustin Gouker (@DustinGouker) February 19, 2016
https://twitter.com/LDrogen/status/700741810432434177
“This is clearly game of skill; we’ve proved that beyond a shadow of a doubt,” says Drogen, adding that certain competitors turned out to be repeat high-scorers during the app’s six-week beta testing period, demonstrating an aptitude for market predictions. Of course, the daily fantasy sports sites, and poker websites before them, have tried to make the same argument.
Nonetheless, Drogen is taking special care to avoid the mistakes and problems that have befallen the fantasy sports companies; Estimize employees aren’t allowed to compete in Forcerank contests, and the game limits players to one entry per contest. “We’ve done a lot of homework to make sure we didn’t trip up on this question, because we know that there’s going to be a certain amount of loud and idiotic commentary from the peanut gallery,” says Drogen, “Because it is an interesting thing that we’re rolling up, and to some people does look close to what they’re doing.”
While Estimize has so far resisted clients’ requests to offer buy-sell-hold recommendations and ratings on stocks in addition to simple earnings forecasts, Drogen thinks Forcerank’s data could meet the same need, believing that once the contests gain momentum they could “legitimately replace buy sell hold within a couple of years.” And if that works? He envisions creating a Forcerank ETF so that any investor could literally bet on the data.
