It's an interesting idea! And while the market does tend to discover and exploit most opportunities like this very efficiently, don't forget about Michael Burry
http://en.wikipedia.org/wiki/Michael_Burry, who was apparently one of the only people outside the major banks that realized you could read CDO documentation and pick the worst ones to bet against. So who knows, maybe this is original.
Questions: Do you have a way to automate this, or would you be doing it manually? If manually, how often are you willing to fiddle with it? Take that period (daily, weekly, monthly, etc.) and get past performance of those two funds on that interval (every day for the last six months, etc.). Then test it out: how would you do with manipulation (and maybe paying for trades) vs. buying and holding. Is it worth it?
Also I think that FORUMM has a good point - unless you set some minimums to ensure you retain some amount of each, you'll just wind up with 100% in one investment pretty soon.