NEW YORK (MainStreet)One analyst screams buy. Another shouts sell. So where do you place that bet? What if you could hang a giant, gold-chained Truth-O-Meter around their necks?
A free web application called TipRanks claims to do just that: tell you which analyst is right or at least which one has been correct in their stock recommendations most often. TipRanks lets you see the track record and performance of any analyst on the Web, see what top-performing analysts are saying about a stock and get real-time recommendations.
Uri Gruenbaun, the Israel-based CEO and co-founder of TipRanks is a software engineer, most recently a Senior Project manager for security company ARX, where he managed national IT projects. Two years ago current TipRanks CTO Gilad Gat and he developed a real estate service that scanned the Web detecting foreclosed assets. Users would input what type of property they were looking for and receive a text message whenever a relevant asset was found. Gruenbaun says the idea just "wasn't scalable enough for us to invest in."
But with a little tweaking of the concept, TipRanks was born.
"It started by me noticing I was making investment decisions based on opinions I was getting online, from 'financial experts,'" Gruenbaun says. "This advice is published everywhere around the Web, starting with the big financial websites quoting analyst recommendations to financial reporters telling me what are the '5 Stocks That Will Beat the Market.'"
"After relying on a few opinions that turned into bad decisions it occurred to me that there was no mechanism that allowed me to evaluate these recommendations and see how accountable are the people behind them," Gruenbaun continues. "That's how we came up with our 'financial accountability engine' a service that instantly pops-up a report with the measured performance of anyone giving sell-side advice, based on the performance of historical recommendations they've made."
Gruenbaun says the service scans the Web and analyzes financial content using "Natural Language Programs" -- algorithms his team has developed. The algorithms determine if a recommendation is embedded in Internet text and if so, who is making the call, and if it is a 'buy'or a 'sell.' "So we start to 'virtually' follow the recommendation and benchmarking it to the performance of the S&P 500 during the same time period," he said. "After scanning almost a million Web pages, we have the recommendations portfolio of over 5,000 analysts since 2009."
All of this is launched in a sidebar on your Web browser. If you begin reading an article containing a stock tip, TipRanks pops up a report with statistics on the analyst. The metrics include:
- 1. What is the analyst's success ratio -- how many of his recommendations actually outperformed the market. The calculation is based on a Z-test distribution so someone that has 80 correct recommendations out of a 100 is much better than someone who has 8 out 10.
- 2. The average return on buy/sell recommendations.
- 3. The analyst's total ranking compared to other analysts.
- 4. An option to see all of the stocks the analyst has recommended, with a link to the corresponding article.
- 5. And an option to see what more-accurate analysts say about the stock.
The service has added additional user-requested features since launching three weeks ago, including a live feed of recommendations being made by top-performing analysts and a search option where you can enter a stock symbol and see what analysts have recently recommended.
"At this stage we support only sell-side analysts," says Gruenbaun. "We are also developing unique platforms that measure the performance of bloggers around the world. By doing that, we can show users what top-performing bloggers say about any stock. Also, whenever a user reads a financial blog...he can see how well the contributor's advice has performed."
Gruenbaun says there are "a few thousand users" of TipRanks so far. While the service is free, users must request a code for access. Readers of TheStreet can use code TS123 for immediate access. Gruenbaun is considering adding premium features as a revenue generator going forward.
--Written by Hal M. Bundrick for TheStreet