Dueling investing apps: How to best forecast earnings
A better way to get earnings estimates: crowdsource 'emMost of the time we use analyst consensus earnings -- an average of all the different Wall Street opinions. The problem with using an average is that earnings estimates on certain stocks diverge pretty significantly and taking the average isn't an entirely accurate way of trying to gauge earnings. There are a couple of ways to do this better/smarter right now.
Estimize: crowdsourcing earnings estimatesLaunched earlier this year, Estimize uses a game system to encourage its users to input their own estimates for revenue and earnings estimates.
A user who's the most accurate in forecasting earnings gets to display a badge -- a fun and rewarding non-monetary reward for users inside Estimize. Bragging rights, you know.
Users can view what the community is forecasting for some of their favorite stocks and if they have something to contribute, they can quickly input their own estimates for the upcoming quarter's earnings and revenues.
More advanced users are encouraged to contribute a small write-up -- a rationale -- behind their projections.
What's nice is then users are given access to a graph that displays actual earnings performance compared with Wall Street consensus estimates and Estimize estimates. You can see the graph below for Apple ($AAPL).
I assume there will be stocks where crowdsourced estimates do a better job than pros while other stocks, which require a different skillset, would be less prone to be crowdsourced well.
Not all stocks are tracked but as Estimize grows its userbase, more stocks are being added and according to an interview with Techcrunch, Estimize users are 67% more accurate than Wall Street estimates
Trefis: crowdsourcing financial modelsTrefis is one of those tools that's easier to comprehend when you see it than when you explain it. Built by guys from MIT, Trefis created a platform to analyze the components that make up a company's revenues and earnings. So, using Apple (AAPL) as an example, that could include iPads, iPhones, Macs, iTunes, Apps, iPod, Apple TV and a return on cash.
Playing around with any one of those products can have a big impact on Apple's overall earnings and revenues. Trefis tries to display visually how all these legs of a firm's business impact overall share price.
Users can get even more granular by forecasting not only the revenues of the entire company but any one of these drivers of the business.
So, here we're drilling down on iPhones and we can input estimates on pricing declines, market share, the global market for phones, and even margins.
Essentially, users are building a similar investing model to what professionals would use when they dissect a company. It requires a lot more work to even come up with the estimates in this fashion (because it's so much more granular) but the output and just potential to learn from other users is so powerful.
Trefis allows you to see aggregate user ratings on a company and compiles a Trefis price target on the companies it covers. You can also identify talented analysts by ranking them by how close they came to reality.
Estimize vs. Trefis
Estimize requires a lot less work for investors. If you're looking for a better earnings consensus number, this should probably be your first stop. It's light, easy, informative and fun to use actually. If you're a contributor, going for badges and bragging rights and getting visibility for your work must be rewarding. Plus, it's beating Wall Street consensus numbers - bingo!
Trefis, and its modeling software for all inputs into an investment model, is an extremely powerful tool. I feel like it's the type of thing professionals could use to benchmark their own models, see how others diverge, and qualify their work. For your average investor, it may be overkill. Which is why we're beginning to see mashups of these next generation investment tools (see my interview with James Berman, author of the Berman Value Folio, who built an investment newsletter on top of Trefis, infusing his research with interactivity and the power of crowdsourcing).
Other tools for estimates
Starmine: This is owned by Thomson Reuters and integrated into their research tools. Starmine does a variety of things well and one of them is monitoring the best, most accurate analysts in their industries. So, instead of following bland consensus numbers, investors could drill down on the truly talented. When combined with other screens, Starmine predictive models are getting really accurate (100% accurate this past quarter) in determining earnings surprises.
What do you use for estimates? Let me know in the comments.