When Leigh Drogen was working as a quantitative trader, he realized that there had to be a better way to predict which stocks would be profitable. The estimates from bank analysts, he knew, weren’t always accurate.
So Drogen decided that he wanted to provide traders and investors, whether they did it as a hobby or a full-time job, with more accurate data so they could make better decisions.
In 2012 he launched Estimize, a community that today includes 100,000 users per quarter. According to its website, the company’s data from 6,885 financial analysts “has proven more accurate than comparable sell side data sets over 69 percent of the time.” Its data is frequently referenced in Forbes, The Wall Street Journal, and CNN Money.
“The whole point of Estimize is that if you collect all of this estimate data from a much wider population of individuals who are not just at banks, a couple of things should happen,” says Drogen. “First of all, you should get a wider dispersion of estimates and a more accurate consensus based on the wisdom of crowds. When you collect the data in this manner, you get a data set that better represents the truer expectations of the market.”
At the moment the website is free, but Estimize will soon be rolling out a premium platform.
“We’re building all sorts of premium analytics, and screening and filtering derivative data,” says Drogen. “This will provide a lot of insight into the data that’s contributed to the platform.”
For now, anyone can continue to utilize the free platform and contribute their knowledge about the financial world. It doesn’t matter whether a person is a first-year student studying finance or a seasoned trader with decades of experience. What really counts is the accuracy of their predictions.
“You can validate the quality of the data not through identity of the person but through statistics,” said Drogen. “We run algorithms to see who knows what they’re talking about. We see user behavior, including how long they spent on the pages and how many times they changed their estimates, and through all of this data we build behavioral models and decide, ‘Do we trust this analyst?’ It gets rid of the need for identity.”
Photo credit: Lauren Kallen