Last week saw some great upsets in the NCAA Division I Men’s basketball tournament (Go Flyers and Huskies!). However, crowd-sourced earnings estimate provider, Estimize, experienced its own March Madness last week, with three notable wins, capped off by a $1.2 mln funding round reflecting the numerous advances the firm has made over the past eighteen months.
Founded in June 2011, Estimize is an open financial estimates platform based on the belief that the “wisdom of the crowds” is more accurate then estimates provided purely by sell-side analysts. The firm was established by Leigh Drogen, a former quantitative hedge fund analyst, who felt that Wall Street analysts are not correctly incented to provide accurate and unbiased earnings estimates due to inherent conflicts within the investment banking business model.
Currently Estimize collects revenue and earnings estimates from 4,232 contributors, including professional analysts from the buy-side and sell-side, as well as private investors and even students. The consensus estimates collected by Estimize cover more than 900 U.S. publicly traded stocks on a quarterly basis.
The core consensus data set generated by Estimize is open and free to everyone, regardless of an individual’s contribution level. The firm currently generates revenue by selling access to their dataset via an API for professional investors. At present, a handful of clients pay for access to its API, with almost two dozen more who are testing the data now. The company is also in talks to feed its data to a few financial media platforms. Management explains that Estimize’s revenue has grown 100% each quarter.
Deutsche Bank Report Validates Estimize Data
One of the first wins experienced by Estimize last week was an extremely encouraging report published by Deutsche Bank’s Quantitative Research Group validating the use of Estimize data in quantitative research models. Following are a few of the findings from the Deutsche Bank report:
- DB found that the Estimize consensus forecasts more accurately identified earnings surprises than the less timely IBES estimates – a factor that led to capturing more of the stock price move post earnings announcement.
- DB also found that as the number of contributors to the Estimize consensus increased, the forecast accuracy relative to IBES also increased. For example, EPS estimates for stocks with more than 20 Estimize contributors are more accurate than the IBES consensus 2/3rds of the time.
- DB also found, much to their surprise, that the EPS prediction accuracy of Estimize contributors who were finance professionals underperformed the accuracy of non-professionals.
- In addition, DB found that combining the Estimize estimates of professionals and non-professionals increased the accuracy of these estimates above the accuracy of any one of the groups alone.
In wrapping up, the quantitative group at Deutsche Bank agreed that, “In conclusion we found multiple benefits to using the Estimize dataset; especially in the case of short-term applications in which accuracy is essential. Another interesting byproduct of the analysis was the power of crowdsourcing. We found that some of the value-added in the Estimize dataset was due to the ‘wisdom of crowds’ effect as more predictions give way to greater accuracy. Moreover, the diversity of the contributors provides a greater spectrum of information which can potentially improve investment strategies based on estimates.”
Launch of Mergerize.com
Another major development for Estimize last week was the launch of a new financial information product called Mergerize which provides crowd-sourced expectations for mergers and acquisitions.
At its most basic, Mergerize (www.mergerize.com) is a platform where people can guess whether a company will get bought by another company, or whether they will buy another company, and at what price. Drogen explained, “It’s all the same set up as Estimize, but it’s more about if Company X will buy Company Y for X dollars.” The platform will eventually provide M&A expectations on more than 4,800 public companies, private companies, and even start-ups.
Drogen, a former buy-side quant analyst explained the potential value of this M&A expectations data, “How might the resulting data be used by traders? Well for the quants, I expect that they will use Mergerize data to exclude certain stocks from their trading at certain times given that there is outsized risk associate with M&A transactions in their models. If Mergerize data can mitigate this risk by putting a “no trade” tag on certain stocks, it should be very valuable. For discretionary traders, specifically of the event driven persuasion, having a set of M&A expectations should allow them to take advantage of their own beliefs if they see that they are significantly different from the crowd. If the market feels a deal will take place, and you believe there’s a premium built into a stock because of that expectation, but you don’t think that deal is going to happen, then you probably believe at some point that premium is going to come out, and you should be short. My bet is also that there will be a high correlation between the velocity of people predicting a certain deal and the implied volatility of the target in the deal. I’m sure traders will figure out how to utilize that correlation.”
Of course, Mergerize.com is a brand new expectations service and doesn’t offer many of the features found on Estimize.com, nor does it have the volume of estimates of that site. So far, a total of 38 predictions have been made, focusing on 26 predicted targets and 28 potential acquirers. However, Drogen is hopeful about the new service, explaining “Over time as we see the platform grow more resources will be put towards it.”
$1.2 mln Funding Round Completed
Capping off a great week, Estimize closed a $1.2 mln Series A-1 funding round, at approximately 3 times the valuation it received in its previous funding round 18 months earlier, reflecting the progress the company has made on building out its the product, data set, and team.
The financing round included the firm’s current venture investors from Contour Venture Partners and Longworth Venture Partners. It also included a syndicate of angel investors put together by ValueStream Labs, along with individual angels Brian Finn (former CEO of Credit Suisse), Mike Towey (Director of Research at Susquehanna), and Jason Finger (Founder SeamlessWeb).
Drogen explained his excitement about the angel syndicate put together by ValueStream Labs, “It has always been my goal to do a “crowdfunding” round at Estimize. Our platform is reliant on a network effect, and we want as many people as possible who use Estimize to be incented to grow our community with us. And while our government has made strides in opening up the legal framework for crowdfunding, it’s still touch and go in many respects when it comes to accredited vs non accredited investors. The AngelList syndicate structure gets us half way, and has allowed us to bring in 20+ amazing investors from across the financial ecosystem in a clean way for our cap table.”
The past few weeks have been particularly noteworthy for “crowd-sourced” investment research sites, including SeekingAlpha, and now Estimize. While few firms have fully embraced the view that the “wisdom of the crowds” can be even more valuable to investors than traditional investment research and analyst estimates, a small but growing number of studies have shown that this clearly could be the case.
In our mind, what is particularly exciting is the extension of the “crowd-sourcing” model to quantify new areas of “investor expectations” like M&A activity. Only time will tell if this data will be accurate and useful for investors. However, whether the Mergerize product is successful or not, we think professional investors will become increasingly more willing to consider using non-traditional research data from “crowd-sourced” investment research businesses like Estimize, Seeking Alpha, StockTwits, or other new data and research vendors that are likely to spring up in the next few years. Clearly, “crowd-sourcing” is here to stay.