The problem is this: Keynes once observed that long-term investing is all about the fundamentals, whereas short-term trading is about investor psychology.
This was almost true: long-term trading is about investor psychology too, to some extent.
What happens is this: if something's been going up for a long time and people have gained confidence in investing in that something, trend following methods will work because people will lengthen their time horizons.
If, on the other hand, something is chopping or going down, people will radically shorten their time horizons and trade in and out, if they're in it at all. Hence sideways and down markets don't respond to trend following.
So, what you need is:
1. A trend following something.
2. An extremely accurate measure of short-term momentum (hint: think about Wilder's RSI. Think some more. Remembering that he traded commodity markets with limit up/downs, what exactly was he trying to get at that he needed this for a limit up/down market, which stocks and fx aren't? Answer that, and you will be on your way to coming up with an extremely accurate momentum indicator that responds as quickly as an indicator can given its only input is past data).
3. A way of figuring out investor confidence: are investors confident enough that they're willing to lengthen their time horizons or not? This, and I state this categorically because it took me years of investigation and a lot of perspiration, is The Holy Grail. It really is. You will need reams of data, a logical mind, some math, and a second pair of eyes to check that you're not sending yourself off on some fantasy. It's probably the toughest thing on the planet to figure out.
One and two are tactics, and therefore will get you somewhere but not very far.
Three is strategy. Master the strategy, and the tactics will fall out by themselves.