Saturday, 17 October 2009

The use of graphics programs

. Price - Price data used in calculating the (opening, closing ...).
. Type MA - type smoothing (moving average), which will be applied to regression (simple, exponential ...).
. Period MA - period smoothing (moving average), which is used to regression.
. The period of regression - number of bars that are used in calculating the regression line.
. The forecast period - the mean regression line may optionally be predictive of the future number of bars specified by user. If the prediction is 0, no prediction will not happen. If the forecast period greater than 0, the linear regression line is projected forward to determine the value of the regression for the bar.
. The band regression - Average strip is determined calculating the regression line, using the current bar and previous X-1 bars (X is a period of regression), and the design of the line next to the bar Y (Y is the forecast period), and then by the end point line as the values.
. The top band - band of the standard deviation, held above the regression line. Uses the number of standard deviations below. The standard deviation is calculated over the period of regression.
. Bottom line - the standard deviation band, held below the regression line. Uses the number of standard deviations below. The standard deviation is calculated over the period of regression.
. Standard Deviation (s) - How many standard deviations from the mean regression lines used for the upper and lower bands.
. Oscillator Linear Regression - Oscillator linear regression represents the number of standard deviations the current price of the line of linear regression.

Practical applications

To better understand the practical application of linear regression ?????, below the existing comments trederov who uses this tool in their daily practice.

Trader Stephen Kessler

I really like the bands of linear regression, because they are superimposed on the graph. I use the value of projection 0 (no prediction), the regression between 100 and MA between 1 (no pre-smoothing) at 1 -, 2 -, 3 - and 5-minute charts S & P, to display the likely levels of support and resistance, along with reference points, the usual recent maximum and minimum levels of recovery and Fibonacci. On the lateral market, as shown above, at 1-minute schedule can be very useful for the band, conducted by 2 standard deviations.

Trader Chad Payne

You can create a convenient indicator, which actually shows you how many standard deviations (from the linear regression line) are the current prices. This indicator will be positive when the price is above the regression line and negative when prices are lower. In the following chart shows the parameters set by the indicator and an example of this indicator, which was built in the form of the histogram below the price chart.
The indicator is installed without smoothing (the period of AI = 1), no forecast (the forecast period = 0), and the period of regression = 13.

Significance of -2.3 tells you that prices are currently at 2.3 standard deviations below the regression line (which begins 13 bars ago, and ends at this bar).
Here is an another example of use of the linear regression.

This is a daily schedule of MSFT. For the indicator is used between 13 and 1.5 standard deviations. I also added to the schedule indicator "Paint Bar". Parameters used for the indicator "Paint Bar" shown there. LRFU and LRFL - is respectively the upper and lower bands of linear regression.
As you can see the bars are displayed in black when they closed above the upper band, and blue, when they closed below the lower band.

No comments:

Post a Comment