It appears the OP read a recent article regarding this strategy wherein buying both sides of the market did in fact produce a decent profit during this upward move since the March '09 bottom. The question is, will it produce a profit if the market churns sideways or declines. The responses to this thread are interesting.
One very important thing to remember too, especially with three times leveraged ETFs, is that you need to check the historical prices, not the prices on the charts. Because they head for zero, they also tend to do reverse splits at some point. The ultimate destination for both is zero. To buy both, by betting on a very big short term move is like buying a strangle or straddle with options. You hope one side will go to zero, and the other side will move big. So it could work, but it's dangerous if things do rally very strongly in the right direction. Much like time decay on options, your little triples both start to work against you. For example, when the financial markets were crashing, you would have lost money buy going long FAZ and long FAS. They both just went to zero. Shorting both is the better option, but you'll find them hard to short, plus you'd better be ready to sit through huge drawdowns, and you'd better hope you don't get forced to sell at the worst possible time.
some monte carlo simulations were performed on this a bit back, showing why it's not a very good idea. http://intelligenttradingtech.blogspot.com/
Take a look at the day the ES was at 1040. Buy on close an equal dollar amount of fas and faz and tell me if you're in the black today. Quick answer is yes, by quite a bit. The 'battle' against the decay takes place when you own both in a TREND. i.e. Pick any ... iwm , qld, sds... whatever and their corresponding partner, and do the same thing. These things only work in trending markets. Chop is where the "rebalancing" really takes effect and the money that came out of the trend is lost. These are designed for long term holders and not short or intermediate term tranactions.