Regression Channel double

Trading robot for a real account. It is a continuation of the idea used as the basis of the popular Regression Channel EA. Unlike its predecessor, it uses a bidirectional trading algorithm. The robot trades inside a channel formed by regression curves. Deals are always by stop losses.

Strategy Features

The robot trades inside a trend channel bound by the curves of a polynomial regression. The trend direction and strength are determined by the gradient of the channel cures during the trading. Market entries and exits are performed near the bordering regression curves. The deals are only opened in the trend direction, so most of them are closed with profit. Losing positions are averaged using additional deals, which are calculated so that their aggregate sum could also be closed with profit. Presence of a trade cycle in one of the directions while the criteria are met allows for the beginning of the second cycle in the other direction. This reduces the drawdown during averaging and increases the price movement usage coefficient. Due to money management the trade volume increases with the growth of the deposit.

Results

The default parameters are optimized to trade with the minimal risk on EURUSD H1. The EA was tested with the default parameters using real ticks. Testing results: starting with the initial deposit of $100 on EURUSD H1, the robot earned $840 from 01/01/16 to 21/04/17, net profit for 15 months of trading comprised 840%; with the initial deposit of $1000 and initial lot increased to 0.05 it earned $8000, net profit comprised 800%.

Recommendations on usage

The EA operation does not require any indicators. Additional i-Regr indicator might be useful for visual tracking of regression curves calculated by the EA. Launch the indicator on the same chart with the EA. The indicator file is provided on the Comments tab. The degree, kstd, bars indicator inputs should be identical to the appropriate robot ones (default parameters are matched). The robot is easily optimizable for other currency pairs and timeframes. Parameters recommended for optimization are marked by (*). Optimization technique is shown in the video.

Parameters

Main Parameters

• Language messages - English, Russian;
• Magic number - magic number for orders;
• Entry criteria - the distance to the trend line (% of the channel width) to enter/exit a trade;
• Gradient criteria - minimum gradient of the regression channel, with which market entry is allowed, pts;
• Offset for determining Gradient criteria - offset of the historic bar, relative to which the gradient is determined;
• Distance to belay purposes - distance in points.

Regression channel parameters

• Оrder of regression (degree) - order of regression;
• Channel width (kstd) - channel width;
• Number of bars in the channel (bars) - amount of bars for the channel regression line calculation.

Averaging parameters

• Step of averaging positions - step in points (*);
• Coefficient of change of the lot - lot increase ratio;
• Maximum number of steps for averaging positions - the maximum number of averaging step (if exceeded, close all) (*);
• Coefficient of stop - the proportion of the averaging step, determining the stop behind the level of the last order, 0.1 ... 1) (*);
• Specified profit with averaging - preset averaging profit, $ per 0.01 lot.

Additional Parameters

• Clearance - gap above distinctive spots (points);
• Use money management - apply money management (Yes, No);
• Initial lot - initial lot (*).

Interface parameters allow to customize the display style of on-screen information and do not require comments.

The robot can be used on 5 and 4 digit quotes. Automatically detects the number of digits in quotes. At the same time, the input parameters specified in points should always be set as for 5-digit quotes (set by default).