Ideas & Concepts: Stock Pick Aggregator

If we build it we should call it "Stockr"

Ok seriously though, I've never seen anything like this. I think it's a great idea, but is there really an adopted standard for stock picks used by all the analysts? That's the real hiccup.

Yeah, it'd be great if there were something that aggregated Google Finance, the Motley Fool, Bloomberg, and all the other tickers. But would something like this not already exist? Where do traders get their tips now?

A couple issues back, Business 2.0 mentioned a company that has developed different algorithms to predict the weather. They're now licensing their technology to many media outlets.

They based their predictions upon the weather of yesteryear, traffic conditions, news, etc... And are supposedly able to pretty accurately nail their predictions several months ahead of the date.

Something like that for stocks, to not only aggregate other sources but analyze them as well and present a prediction based upon a formulaic process, would be pretty damn cool.

With enough brainstorming and a brilliant engineer, this could really come to life.

Yeah, such a site/service would demand quite an audience. Unfortunately, www.justahunch.com is taken. lol

Posted By: EricDamn, that would have been a great name

So...anyone know a good programmer interested in working for a share of future profits? :)

Heh, maybe - I know quite a few (being one myself by training), but we'd need to flesh this out a bit more. :-)

It looks like we've got some rather different ideas about this. I was talking about looking at proven stock pickers recommendations and providing that aggregated information to people who would pay us for that information in a convenient package (the reasoning being they have shown over time sustainable abnormal returns).

It sounds like you guys are more interested in coming up with automated algorithms that can capitalize on the volatilty of the markets. I've done that in the past with my company and while I am very interested in its potential I think something like that takes a fair bit of dedication to get things working really well. I toy with making something like that a Master's topic in Finance or Math so I can get paid to work on it :-)

Travis, I must have missed that article because that's very interesting - meterological prediction is view the holy grail by many mathematical and statistical types due to the inherent nonlinear noise in such a system. My company's technology did some early work on this in terms of pollution propagation after a contamination leak from a plant, given nearby water, land, weather conditions, etc. If they have been able to predict so far ahead with good accuracy, then they should definitely turn that on its head and use it in the financial realm (where the randomness has become much more accurately defined compared to weather models).

Anyway, I'd be interested in a) seeing if we can actually aggregate these sources without being sued, and then b) seeing if we can use this aggregation to our own advantage programmatically. I personally think we could use their historical performance on stock picks the day, week, month, etc. after and in a fairly manual way outperform the markets. Adding automatic analysis and decision making would only help.

Hi Audarius,

Thanks for your comments. What you describe is pretty close to Marketocracy's model. While I think there's a place for such types of services, it's not where my head's at.

In my mind, one of the biggest drawbacks to such schemes in fnance is the fact that the more people who operate on the same strategy in the same marketplace, the less effective said strategy is at delivering superior performance. This is why a lot of stocks have a quick jump when a popular stock picker recommends them. Good stock picks (and pickers) provide good returns over the long term, so there's no need to get in that first day. Part of what I'm proposing would involve providing historical time series information on previous picks for various pickers. This way you could see how much stocks drop for various pickers after the initial flurry of activity, so determine an appropriate time to get into the stock.

While what I describe is simple, that's part of the reason I like it - it is something that seemingly few (if any) systems taking advantage of the collective information already out there (at least that is accessible to retail investors). We are adding value simply by making it easier to access, analyze and synthesize this vast array of information. If it works really well, there's a good chance I'd rather just use the system as is for myself and others who are involved since we'll probably make more money using it ourselves then we could get by selling the information to others.

As for some of the people I am worried about being sued for: I was thinking like if I went and subscribed to some of TheStreet's analysts and then passed on that information to other consumers, if that would violate my subscription agreement (kinda like violating music copyright).