Could Google Leverage Its Massive Database of Information For Trading Profits?

About a week ago, I wrote an article talking about how Google was going to be getting into the bond trading business.

The company has an impressive stockpile of cash (about $25 billion or so, give or take a few million) and zero debt. With interest rates being as low as they are, the company is looking for new ways to invest their ever-growing pile of cash, so they have decided to turn to the world of bond trading.

Apparently, back in the day, Sergey Brin wanted to start a "Google hedge fund" but was rejected by Eric Schmidt. Apparently Schmidt wasn't too excited about the legal ramifications of Google owning and operating its own hedge fund.

However, with Google about to take the plunge into the world of bond trading, the question needs to be asked again - should Google run its own hedge fund?

-- Illustration of google coloured puzzle coming together --Google obviously wouldn't be running other people's money - instead, it would just be using a portion of its cash reserves (say, $5 billion).

My thought process is that Google could tap it MASSIVE database of information in order to make predictions about the short-term movements of stocks.

Think about it - Google has:

1. Google Finance.
2. Google Analytics, which is placed on millions upon millions of sites.
3. The Google search index.
4. A deal with Twitter that allows them Google unlimited and real-time indexing access.
5. Google News.
6. An incredibly smart team of programmers.

Given all of this information that Google has access to, and given their unparalleled brain power, you would have to think that Google could come up with a number of edges that would help them tremendously in the running of a hedge fund.

Let's say that a stock is up 5% on the day, and suddenly Google's hedge fund algorithm picks up a surge in Twitter, Google Finance, Google News and regular Google search engine index activity pertaining to the stock. I've read about certain hedge funds that have paid programmers to develop applications that will track Twitter activity with the sole purpose of figuring out what people are buying/selling. Google could potentially do the same thing, to the "nth" degree. Not only could they track Twitter activity, but they could also incorporate a host of other things into their algorithm.

You'd have to think that there would be some potentially big edges to be found here.

Now I doubt that Google would ever do anything like this - it seems like a pretty gray area in terms of ethics (maybe I'm wrong).

However, it is certainly an interesting thought given Google's second-to-none access to information.