Just curious. Have you ever thought that maybe it's not the trend line itself that's providing you with the "edge," but rather something else that is coincident?
It matters if you're trying to understand why you're winning or losing. Otherwise, you may as well throw darts at the WSJ.
kj, Yes I have thought about it and looked at traditional support and resistance, fibs, pivots (floor, open, and camarilla) and mas as potential support. I have found different forms to be effective to differing extents depending on what tradable I am looking at and I have found behaviour of market participant to vary around the SR or clusters of SR. For me, all support and resistance, is about finding the simplest way of detecting where enough market participants will enter/exit to make a trade entry/exit worthwhile. I don't actually care why; I only really care that its statistically useful. And, like old db, I take the approach that when support "fails" thats why my stop is there. In some cases these are coincident with the channels support (actually if you look hard enough ... lol) but I have found the channels (as I use them) to be the simplest trading tool. So in the occam's razor approach, they work. Its important to note that I deliberately chose to trade things that trend on the timeframe that I am trading. If they dont trend then I dont trade them or I change the timeframe. If you didnt do this then other forms of support and resistance might be more important to you.
There is no better place to display the difference between retail and professionals than on this subject. Retail traders want an absolute answer ("yes it always works, or no it never works"). In contrast, professionals are looking to evaluate how often it works, and under what conditions. It is the approach to using trendlines that "makes" them useful. As Peter said, trendlines can be drawn to connect higher highs or or lower lows. After you do that, the premise is that price will "respect" the extension of the line through time and space. What you need to do is to look at the historical record for the tradeable issue you are working with. This is called "research" and is the basis for your future success. How do you do it. Well you go back through intraday charts drawing trendlines and looking at the result. Did price respect the trendline? for how long? Once price broke through, did it signal a significant breakout. Did this behavior happen every day, or just on Tuesdays and Thursdays? Do trendlines work only in August? or throughout the year? Finally, one needs to understand that whatever result you get, you have obtained an edge. If after testing you come to the conclusion that trendlines don't work for the tradeable issue you are work with, then you have learned something important. I hope some of this helps somebody out there. Lefty
Hello Kent, Very interesting. Just want to make sure I am understanding your commentary correctly: trendlines in your interpretation are of a finite length, hence ending in the time domain at the last available data point of the dataset in question. If this is what you mean, would it not be possible instead to use a line of infinite length (ray, or vector), extending in either or both directions, extrapolated into the future (and past) to determine possible area(s) of support/resistance? In this case, the trendline does not in fact have an endpoint, being that it extends infinitely. In my analysis of past data, trendlines serve to indicate those areas of support and resistance, and cannot be used in the strictest sense to predict future price direction, but rather indicate a pivot point, where a move away from the pivot point is most likely. Instead of 'predicting' future price movement, could perhaps the trendline be used as a low-risk entry point to enter a trade? For example, if your analysis defines a trendline as a possible strong point of support, and price reaches that support, you enter long. If the support fails, you take your stop. Then, if what was previously support acts as resistance, you now have a low-risk entry on the short side should the price 'pivot' away from the new resistance upon a retest. A big risk here, it would seem, is that the trendline defined by the technician is not honored strictly as a point of separation but rather a place where the price action chops through the line repeatedly, hence invalidating the usefulness of the trendline as a quality entry point for a trade. In my research, it seems that there are enough circumstances where, through proper analysis, a trendline is honored enough to be useful as a legitimate edge to extract profits from the markets, particularly on larger issues such as index futures, or commodities futures... I don't want to start posting charts - that is the fun of this game - doing the research yourself to find your own edge. But I will say this, if you take the time to research the charts, it will become clear that a tool as simple as a trendline can in fact be used to find entries, should the trendline be honored, or even broken. Hence, it is a tool used not to always predict (one does have to 'predict' to be able to speculate whether a trade is to work or not, no?), but rather to define a pivot point where price will move away from in either direction. It seems to me that it is actually rather easy to find trades in this game, but somehow there will always be enough people willing to 'screw it up' somehow - the right trade can be screaming at you "trade me! easy entry!! easy money!!", literally given to you on a silver platter, but still, many traders will find a way to screw it up - taking profits too soon, for example, as a biggie (particularly in my case). Case in point, I have found so many entries, but didn't have the kahunas to simply HOLD the winning trade until there was a compelling reason to reverse to the other side of the trade.. rather I have tried in the past to daytrade or scalp when my analysis clearly defined a great entry on a longer-term time frame. Mind you, this is for my ventures into discretionary trading, which admittedly is probably more effective in the long run invoked on a longer-term timeframe (discretionary trading, that is).. less stress, less emotional involvement, more time to prepare for the trade. Sorry to digress a bit here folks. I am interested in anyone elses constructive commentary either embellishing upon what I am saying, or refuting what I am saying with evidence to back it up. Trading isn't just about analysis, obviously - trendlines are just a tool in the kit of the technician.. and besides, to say that a given form of analysis will work ALL the time is folly, obviously - nothing in trading is 100%.. anything can fail - any chart pattern, support/resistance line, whatever, can fail - but does that invalidate the usefulness of the tool in the future? I guess only by testing one's edge (be it trendlines, or some type of statistical analysis, or..) through a large enough past dataset is the only way to have an idea of your edge's long-term efficacy in the future.. Thanks, ei
As usual Lefty, nice work. I'm a decent technician. Been doing my own work for many years. My trendline work/analysis SUCKS! Not part of my personal arsenal I subscribe to a couple of guys who really do a good job with tradeable channels. I won't go so far as to say trendlines fall into a "hard right edge" (sic) but on a whole while I'm cognizant of major trendlines I would much prefer using a myriad of MA's instead.
Just a few thoughts on trendlines. When one says that trendlines "work" or trendlines "don't work", it implies that one is attaching a set of expectations to a trendline. Interesting in light of what Edwards and Magee in "Technical Analysis of Stock Trends" has to say about trendlines in terms of strategies. Basically their idea is that the breaking of a trendline is not "actionable". In other words, just because a trendline breaks does not mean you go long, or short as the case may be. Rather, all it implies is that the trend "may" be ending. So if the trend has stopped going down for instance, it does not mean that the trend "must" go up. But it may then mean that the stock would go through some bottom building, or perhaps a consolidation as a prelude to further decline. Therefore, as Lefty and others have said, a trendline give you information. It's up to you to evaluate the information. But I believe that you will find that most trendline breaks don't immediately lead to sizeable moves in the opposite direction until other types of price action have taken place along with the trendline break. Typically, for a sizeable move to take place their must be some top-building/bottombuilding where the "smart money" can gradually accumulate their position. OldTrader