Overview of Regression Trend Channel (RTC)
by Marc Rinehart
Since the early days of technical analysis the use of a "trendline" has been an acceptable technique to help define price support and/or resistance. One form of a trendline developed from higher level mathematics is known as a "linear regression trendline.
A linear regression trendline is the straight line mathematical measurement of the relationship between sets of price data that plot out as a trendline.
In many ways it is similar to a moving average. The difference is there is no lagging the linear regression trendline and, secondly, it is helpful because it often identifies a change in trend quicker than a moving average.
According to theory, prices can vacillate above and below the regression trendline and still the overall trend is maintained. By using the Regression Trend Channel (RTC) you can better identify that outer range of that overall trend pattern.
A Regression Trend Channel (RTC) is simply the mathematically standard deviation of the linear regression price data. It is calculated using the actual prices of the bars in the trend. The idea is to draw an upper and lower channel from the linear regression line by using the standard deviation of the prices. The break of a Regression Trend Channel is usually used as an entry or exit signal.
The RTC is similar to Bollinger Bands which draw bands using standard deviations of a moving average. However, instead of using a moving average, by using the RTC tool, Advanced GET is simply allowing you to use the linear regression line of a swing you identify. (In case you where wondering, the Advanced GET RTC uses the data bar midpoint for its calculation.)
Find the Regression Trend Channel (RTC) in the Advanced GET studies selection or highlight the RTC icon. To activate the RTC default menu, click your right mouse button.
The Advanced GET defaults the trendline source to "High-Low flip." The H-L flip indicates that the Automatic Trend Channels should be calculated using the Low of the bars when the trend is up, and the High of the bars when the trend is down. Our studies indicate the H-L flip generates the best regression trendline. You can change to a different calculation method if you have a different preference.
The program also defaults with the Standard Deviation turned "On" for the Upper and Lower Channel. With the standard deviations checked off you will no longer have a regression trend channel, but you could use this tool as a channeling technique. We prefer the standard deviation to be turned "On."
The "End Bar" is a useful selection in the RTC menu that allows you to have a visual reference or reminder of where the ending calculation is on the RTC. For example, say you are calculating the early stages of a new sequence using the RTC tool. By turning "On" the "end bar" you will see a vertical line in that RTC. You also have the option to change this identification point to any color you select. Later, if you want to recalculate a new RTC based on current action, you now will have a reminder of where those RTC's calculations were taken from. This is also a useful feature when you are using multiple Regression Trend Channel lines when interpreting price action.
The "Pearson's R" On/Off button indicates if the Pearson's value will be shown at the end of the Regression Trend Channel. Showing Pearson's value helps identify the "slope" in the price movement.
As the Pearson's R value gets closer to the value of 1, the calculated trend line is matching the actual value of the data. In most instances a very high Pearson's value will be identifying the desirable "single slope" price movement we should be monitoring for as Advanced GET users. This means that the regression line is "fitting" the trend very well. As the Pearson's R value gets closer to the value of 0, the trend line is doing a poor job at matching the data action. This means that the regression line does not "fit" the trend very well. However, a low Pearson's may still be useful if you want to help identify the potential outer ranges of price movements in a non-trending market, or wider swing movements. Think of the "Pearson's R" value as a percentage-- a 0.900 (90 percent) match is a good fit; while a 0.060 (6 percent) match is very poor.
If you use the RTC enough you tend not to use the Pearson's feature because they can help clutter a screen and you can pretty much guess at the "fit" when you see them enough times.
The most effective use of the Regression Trend Channel (RTC) would be when a "single sloped" market is identified (in simple terms: when a market that is moving in a very linear manner) to begin to apply the RTC for a potential breakout pattern. The use of Pearson's helps identify that linear relationship. If a market is "curve fitting" well-- some would say a Pearson's value of 0.90 to 0.93, or greater, is what you want to see for a "tight fit"-- you can now begin to setup for a trade opportunity of the break of channel line in that tight fit.
As Advanced GET users we want to be monitoring for Elliott Wave Type One and Two Buy or Sell trading opportunities. The question to ask: "Is the Type One or Two pattern moving in a straight line advance or decline, or not?" If it is, use the RTC to help confirm the trade opportunity. If the price is not moving in a single slope, then use the 6/4 DMA's to confirm the trade setup.
In a Type One setup the RTC calculation begins at the end of a major Wave Three and should end at the current major Wave Four. In a Type Two setup you want to focus your calculation on the major Wave Four to the current high.
It is important to remember you will always increase the probability of success for your trading opportunity when you combine the use of the RTC with other Advanced GET tools, such as the Oscillators, MOB, Ellipse, XTL, and so forth.
Before ending, let me share with you a few experimental ideas you may want to try on new uses of the Regression Trend Channel tool in Advanced GET:
(1.) When you see a Type One or Two setup where you are using a Regression Trend Channel, try re-issuing that chart to a lower time frame to better frame up the trade opportunity. For example, take a look at again at the Disney Type One buy opportunity setup on the Daily chart. Now, either issue to a new Hourly chart, or re-issue the Daily to an Hourly and calculate a new RTC from this new time frame. By applying this cross-referencing technique you may be better able to identify a breakout entry point more quickly, one which has improved risk/reward parameters.
by Marc Rinehart
Since the early days of technical analysis the use of a "trendline" has been an acceptable technique to help define price support and/or resistance. One form of a trendline developed from higher level mathematics is known as a "linear regression trendline.
A linear regression trendline is the straight line mathematical measurement of the relationship between sets of price data that plot out as a trendline.
In many ways it is similar to a moving average. The difference is there is no lagging the linear regression trendline and, secondly, it is helpful because it often identifies a change in trend quicker than a moving average.
According to theory, prices can vacillate above and below the regression trendline and still the overall trend is maintained. By using the Regression Trend Channel (RTC) you can better identify that outer range of that overall trend pattern.
A Regression Trend Channel (RTC) is simply the mathematically standard deviation of the linear regression price data. It is calculated using the actual prices of the bars in the trend. The idea is to draw an upper and lower channel from the linear regression line by using the standard deviation of the prices. The break of a Regression Trend Channel is usually used as an entry or exit signal.
The RTC is similar to Bollinger Bands which draw bands using standard deviations of a moving average. However, instead of using a moving average, by using the RTC tool, Advanced GET is simply allowing you to use the linear regression line of a swing you identify. (In case you where wondering, the Advanced GET RTC uses the data bar midpoint for its calculation.)
Find the Regression Trend Channel (RTC) in the Advanced GET studies selection or highlight the RTC icon. To activate the RTC default menu, click your right mouse button.
The Advanced GET defaults the trendline source to "High-Low flip." The H-L flip indicates that the Automatic Trend Channels should be calculated using the Low of the bars when the trend is up, and the High of the bars when the trend is down. Our studies indicate the H-L flip generates the best regression trendline. You can change to a different calculation method if you have a different preference.
The program also defaults with the Standard Deviation turned "On" for the Upper and Lower Channel. With the standard deviations checked off you will no longer have a regression trend channel, but you could use this tool as a channeling technique. We prefer the standard deviation to be turned "On."
The "End Bar" is a useful selection in the RTC menu that allows you to have a visual reference or reminder of where the ending calculation is on the RTC. For example, say you are calculating the early stages of a new sequence using the RTC tool. By turning "On" the "end bar" you will see a vertical line in that RTC. You also have the option to change this identification point to any color you select. Later, if you want to recalculate a new RTC based on current action, you now will have a reminder of where those RTC's calculations were taken from. This is also a useful feature when you are using multiple Regression Trend Channel lines when interpreting price action.
The "Pearson's R" On/Off button indicates if the Pearson's value will be shown at the end of the Regression Trend Channel. Showing Pearson's value helps identify the "slope" in the price movement.
As the Pearson's R value gets closer to the value of 1, the calculated trend line is matching the actual value of the data. In most instances a very high Pearson's value will be identifying the desirable "single slope" price movement we should be monitoring for as Advanced GET users. This means that the regression line is "fitting" the trend very well. As the Pearson's R value gets closer to the value of 0, the trend line is doing a poor job at matching the data action. This means that the regression line does not "fit" the trend very well. However, a low Pearson's may still be useful if you want to help identify the potential outer ranges of price movements in a non-trending market, or wider swing movements. Think of the "Pearson's R" value as a percentage-- a 0.900 (90 percent) match is a good fit; while a 0.060 (6 percent) match is very poor.
If you use the RTC enough you tend not to use the Pearson's feature because they can help clutter a screen and you can pretty much guess at the "fit" when you see them enough times.
The most effective use of the Regression Trend Channel (RTC) would be when a "single sloped" market is identified (in simple terms: when a market that is moving in a very linear manner) to begin to apply the RTC for a potential breakout pattern. The use of Pearson's helps identify that linear relationship. If a market is "curve fitting" well-- some would say a Pearson's value of 0.90 to 0.93, or greater, is what you want to see for a "tight fit"-- you can now begin to setup for a trade opportunity of the break of channel line in that tight fit.
As Advanced GET users we want to be monitoring for Elliott Wave Type One and Two Buy or Sell trading opportunities. The question to ask: "Is the Type One or Two pattern moving in a straight line advance or decline, or not?" If it is, use the RTC to help confirm the trade opportunity. If the price is not moving in a single slope, then use the 6/4 DMA's to confirm the trade setup.
In a Type One setup the RTC calculation begins at the end of a major Wave Three and should end at the current major Wave Four. In a Type Two setup you want to focus your calculation on the major Wave Four to the current high.
It is important to remember you will always increase the probability of success for your trading opportunity when you combine the use of the RTC with other Advanced GET tools, such as the Oscillators, MOB, Ellipse, XTL, and so forth.
Before ending, let me share with you a few experimental ideas you may want to try on new uses of the Regression Trend Channel tool in Advanced GET:
(1.) When you see a Type One or Two setup where you are using a Regression Trend Channel, try re-issuing that chart to a lower time frame to better frame up the trade opportunity. For example, take a look at again at the Disney Type One buy opportunity setup on the Daily chart. Now, either issue to a new Hourly chart, or re-issue the Daily to an Hourly and calculate a new RTC from this new time frame. By applying this cross-referencing technique you may be better able to identify a breakout entry point more quickly, one which has improved risk/reward parameters.
Comment