Many momentum systems work (moving average, relative strength, etc. ), at least over the medium or longer term. You don't necessarily want the "best" one, but rather the one that offers a reasonable reward for the attendant risk. If you are trading multiple uncorrelated markets, you can select instruments across these. If you are trading one market, you also need a filter to keep you out of the ditch. There are a lot of ways to do this.
I'm a subscriber to CXO Advisory, but otherwise have no financial interest in it. CXO has backtested several momentum systems over the years. One interesting example is a "top 7" system that each months buys and holds (for 1 yr) the 7 best performers on the NYSE over the last 6 months.
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2420743. The NYSE is a little more stable than the Nasdaq. If that's too scary for you, you can short the S&P 500 against the 7 stocks and still make decent money. Or, you can use some other filter (such as the 200-day MA of the S&P) to determine whether to trade the system or sit in cash. This is only one of many, many examples.
Momentum works over medium-term time frames (2 months to 1 year) because it capitalizes on human nature (everyone wants to pile in and out of things together, but some are slower than others). Momentum is something like owning a clothing store that sells goods to teenagers. You don't have any idea what they will wear next month or next year; you only what they are wearing now. Your best bet is to stock your store with what works at this moment. When it doesn't work any more, discount your merchandise, take the loss, and immediately buy different clothes that you know they are currently wearing. There are hills and valleys at the turnovers where you lose money, but overall with good money management, this type of stuff works. You have to live with not getting in at the bottom and not getting out at the top. No big deal; there is plenty of money to be made in the middle.