Quote from Girlpower:
I seem to be up late again...
Ok Guys. Something I've picked up on is the question of retracements being better entry points, so I've pulled up some charts and been looking back.
It seems to me that there could be something in that, but rather as a sepparate strategy. the problems as I see it is that there is too much of a trade off between
1) A break following through and missing it completely because your waiting or having to join it at a higher level than you would otherwise have done.
2) It being a reversal and finding yourself taking the opposite position in the market (thus creating a loss)
3) It sticking and being more profitable than the original trade would have been.
So what I'm asking is basically for some ideas that will help me to put a strategy together that will have a good probability of success, that can run independantly of any other strategy. (obviously this requires commiting more of the fund than I currently do, but there is nothing wrong with having several strategies on the go at the same time...)
Natalie
I'm impressed, Natalie. You've picked up on some problems with the retracement strategy that others have taken far longer to figure out.
I wouldn't say, though, that buying retracements is necessarily "better". It's just different, and while it solves some problems, it creates others, as you pointed out. Two of the biggest problems are those that you mentioned, that there may not be a retracement unless you're using a 1m or 2m chart. There are plenty of ramps that present no entry opportunity at all, even on the 5m chart. And, if you are presented with an opportunity, you're going to have to be pretty damned good at reading the market in order to know whether you're buying a new upleg or a reversal (which, of course, you'll be late for because you're trying to go long).
There are at least two issues here, one having to do with momentum and one having to do with trend. Remember the old days when we had momentum? The retracement strategy worked great then because its psychological basis is that traders are eager to jump onto a train that they think they've missed. This is what propels price past the top of that retracement cup (I'm using the long side here for convenience), and into the next leg.
As far as trend is concerned, you have to be very sure of where you are in whatever trend you're playing in whatever timeframe you're playing it in. One way to do this is very simple: trendlines. Those trends are created by price advancing, pulling back, advancing again, pulling back no more than the highest point of the previous advance, advancing again in a "stairstep" fashion. When this begins rounding off and breaking the trendline, then you're potentially in trouble. Sometimes price simply rests and consolidates, gathering strength for another advance. But on other occasions, it's exhausted and ready to sink like a rock. Here it helps to know something of the average range for the index or contract you're trading, support and resistance levels, the demand/supply dynamic as revealed in the price/volume relationships. Or, if you're not trying to catch the very top, and there is a reasonably nearby reaction low, you can wait for price to violate the level of that low, which is Sperandeo's definition of a trend reversal. This will get you as close to the top as possible without risking a sudden re-reversal of price into another advance. Why? Because so many traders, when they see that price dropping below that last reaction low, will at the very least be unlikely to make any further buys even if they don't bail out entirely. Yes, there are occasions when somebody with the muscle can shake people out at these levels, then push the price higher, but this sort of activity generally takes place near the top anyway, so perhaps better safe than sorry.
You will also find that most people who are into the retracement strategy lean heavily on ADX/DMI. They know that the strategy won't work unless there is a strong trend, and they use ADX to define the existence of a trend and its strength. Problem here, at least these days, is that by the time ADX confirms the trend, it's mostly over. If you want to use an indicator, you may find that a MACD is nearly as good a trend indicator as ADX, and much faster. If you don't want to use any indicators at all, you can try looking for extended compressed ranges, as these are more likely to result in explosive breakouts one way or the other.
I've probably gone on too long, but perhaps what I've said will give you something to work with.
--Db
