线性回归 TS (Linear regression TS)

Very simple, we need to draw a linear regression channel (LRC) from a significant high or low, we enter with fractional entries when price approach the extremes of the channel (I use an EA which buy or sell when price cross a trendline with a specific description), targeting previous high or low. Refer to the image I upload, we buy at point 1-2-3-4...This is averaging down your trades if your first entry doesn't reach target. Averaging down could be very dangerous if you don't know where to average down or if you enter with a too large size, but done properly is one of the best way to be profitable in this game.

Point of entry (POE).
We draw them by projecting the measured move of the leg which break the linear regression channel (from point 0 to 1 in the image I posted).
If you feel that the channel is weak and is not going to last you can wait its BO, projecting the POEs and enter at points where the market is more likely to turn (for example after a break of a previous low, refer to the second linear regression channel I marked in the image posted) and start building our positions from there. You can choose if be aggressive, incrementing the lot size at every POE or be conservative and mantain the same lotsize of the first POE.

Exit.
When our positions start to be profitable it's highly recommended to start trailing securing profits, draw the new LRC from the previous high or low which generated the BO of the old LRC and start the process again.

Stop loss.
We decide the SL based on the historical behaviour of price, for example in GY I would close all my positions at point 5.

Update: I do not always entry blindly scaling in, I prefer to do that only when a major low or high has been taken, otherwise your drawdown could be important. So when no major high low is taken but I notice a nice break of a LRC I just trade price action on the POEs with strict stop loss and a good risk to reward ratio (starting from 3 to 10-12, but then I help myself trailing profit drawing new LRC as price shows its way; refer to post #43).