Using Standard Deviation Channels

Hi,

Adding this to my toolbox just wondering if anyone know if the instituations use it in their anaylsis personally I think its better than support and resistance lines.

Thanks.

I used to get my junior traders to learn this just so they will learn the hard way that any channel-based strategy can work with the right money management and trade entry strategy...

... and that simple manually-drawn horizontal S/Rs work even better than channels.

Manual support and resistance lines are good but its a sideways market and its hard to draw trendlines that connect especially longer term check out standard deviation setting of 88.6 on the euro/usd monthly chart on 1st May 2011 before the big drop its almost perfect projection.

I did say "HORIZONTAL" S/Rs

Anyways, it's good that you think you have found an edge using channels and made it work for you.

For me and my boys, however, we prefer not to have to optimize too many indicator settings. Agreed this is a sideways market, which is all the more reason to just draw horizontal S/Rs around a consolidation zone and just wait for the breakout, which I can almost guarantee will be worth the wait.

Good luck!

Yeah, bit of a non indicator fan myself but sometimes they help pick out stuff you have missed then delete the indicator and use your own judgement to draw manually.

Good luck.

Won't need trend lines at all in a truly horizontal market. The real challenge that puts people off balance is not the bullish, or the bearish, or the horizontal market - but the transverse market, or the market in transition from the vertical into and/or through the horizontal. Or, one that moves from vertical and then does not stop to set-up shop on the horizontal, as it makes its way into the opposing vertical side.

I call these transitional markets, or transverse markets.

They are absolutely necessary, very predictable, and more easily recognized than most traders realize. Yet, they are also the very same place where most traders got whipsawed half to death and where most profits are lost for the average trader. Without the ability to detect them, then can be dangerous places to trade.

The standard deviation channel becomes just another lagging indicator, as its adjustment is predicated essentially on rate of expansion in price over time - thus is must lag. In fact, the only way to cause it not to lag, would be to remove the input that causes the deviation itself. Of course, that would be a fixed channel and such channels are pointless.

There are only two ways to trade when you break it all down:

1) You are either going attempt to predict the next significant pivot

or

2) You are going to attempt to predict whether or not the trend continues


Either way, you are still making a prediction. Which is one of the reasons why I laugh out loud when I hear somebody saying that they are "not making predictive trades." Such statements quantify in a nutshell, the lack of understanding about what the trader is actually doing. We all predict, one way or the other.

The so-called "trend" does not really exist, because it is predicated on that which has already happened. There can be no trend without the prior existence of market data points at some time in the past. And, no one can trade a market movement in the past. One can only predict that the market will resemble the past, and yet said past does not have to project in a straight line. The "past" (upon which the prediction is made) can actually be a curved projection that equates to pivot into a significant market swing in the future. Thus, projecting and predicting a pivot in the market.

Therefore, we are always trading the Present based on a prediction about the Future, which is in turn taken from the Past. In the present we make a prediction about the future, predicated on the past.

Do Institutions use them? 100% absolute yes they do (some of them). They are human beings just like the rest and if they have not done the research, testing, development and all the technical evaluation necessary to produce their own custom indicators, then you can bet your last dollar that they are using some of the same 60 year old conventional technical analysis that others use as well.

In fact, there is a well known fund manager who routinely appears on Bloomberg, discussion how Stochastics show this, and Stochastics show that, and Stochastics shows the other. And, these guys are managing billions each day. Some of them, Hundreds of Billions each day.

IMO SD channels give a good rough indication of overbought/oversold. Their median is a regression line, which means that (if candle closes are normally distributed) approx 95% of them should occur inside 2SD channel boundaries. Much the same theory applies to Bollinger bands.

However, it's possible to use SDCs as the basis of a very simple 1-2-3 pullback entry strategy, that can potentially work on any pair/TF, thus:

#1. Channel slope = trend direction. Trade only in the direction of the trend. The steeper the slope, the stronger the trend.
#2. Pullback. Wait until price has pulled back into the opposite half of the channel, i.e. the upper half (overbought) in a downward sloping channel; the lower half (oversold) in an upward sloping channel.
#3. Resumption. Look for a strong candle that suggests the trend is likely to resume, following the pullback.

That is the basic setup. There are a number of different ways to time the entry; and of course, any combo of TPs (in a weak trend, take profits earlier) and/or trailing SLs (in a strong trend, be more willing to let profits run) may be used for the exit.

The fact that SDCs repaint is not really a problem (unless you're visually backtesting). The real problem lies in calibrating the indicator, i.e. how many candles in the channel window. Different settings can yield vastly different results; as has been pointed out already, that's much the same dilemma posed by any indicator.

I concur with skenobi that horizontal S/R frequently provides some low risk, 'deadly accurate' entries. There are orderflow-based reasons for this, e.g. self-fulfilling prophecy, trapped traders. S/R confluences with pivots, fibo retracement levels, round numbers, trendlines, i.e. studies that are also closely watched by heavyweights, and subsequently the large number of pending orders placed at these levels, almost invariably cause some stalling (more details here). However, these reasons are less likely to apply to the upper and lower SDC boundaries; channels constructed using technical criteria (e.g. pairs of trendlines, pitchforks, etc) may work better in this regard than those constructed using purely statistical methods.

Just my 2c,
David

Nice post link. Good stuff!

Do these channels have to be moved physically like dragged or are there any that move with chart days? Can they be set for periods?

http://forum.mql4.com/54193