MTS "Universum" - page 7

New comment
 
Thank you, Lukyanov . Let's see...
 
rid, please explain how this (BullsBears) EA works. Or at least when profitable positions are closed. I have an idea for a refinement, but it might be worthwhile only if I understand the EA's algorithm correctly.
 

The principle of operation is better described in Reshetov's original description https://www.mql5.com/ru/articles/1447. I won't tell you any better than it is described there.

I have taken two such Expert Advisors, based on different (but related) indicators, and combined them into one EA. They work independently of each other. They work independently of each other with different magic indicators. They are optimized separately, one by one.

Positions are closed either by a Stop Loss or a Flip. I.e. if the perceptron has changed the sign, and the resulting profit is higher than the SL, the position is reversed. 

if(Bid > (OrderStopLoss() + (sl * 2  + spread) * Point))  {               
                   if(Bulls() < 0 )  { // reverse
                  ticket = OrderSend(Symbol(), OP_SELL, Lots * 2, Bid, 3, 
                             0, 0, "Bulls L553", MagicBulls, 0, Red);

- This is very interesting in the online mode. It opens a double lot in the opposite direction.... - in short, it has to be observed. Not the point...

If we have a profit >SL, but the perceptron sign does not change, then the position is not closed. But the stop loss is retraced in increments of =SL
 

You could, of course, make a takeprofit as well. Without any flips. But it will be an additional parameter for optimization. Is it really necessary?

 
rid: If we have a profit >SL, but the sign of the perceptron does not change, the position is not closed. But the stop loss is moved with a trailing step =SL
Can we set the trailing step into optimization? I was watching the whole day on demo, for example, sell goes long to the profit, then rolls back about 2/3 of it and closes by the rollover. I think in this case it would be better to close the buy. Here's even as I'm writing this post, the profit is already 7$ and 2$ respectively. If we close now we get 9$ in total, we could close buy with 9$ profit and now sell with 2$ profit. We get 11$ in total. And so every time (see equity and balance indicator)...
 

Well, at least you're not losing, which is a good start! By the way, I optimize my Expert Advisor daily for the m5 timeframe. I take a story of 2 months to 2 weeks. I optimize all parameters in increments. I optimize all parameters in increments of 5 (except for period). It takes 20 min.

I do not understand the trailing step. In fact, it is already being optimized! Because the SL parameter here is at the same time a trailing step! If there is a separate trailing step, I will have to modify the main algorithm of the Expert Advisor. It is not difficult. But then we will lose on commission - there will be more trades. In addition, for indexes DAX (in the terminal - FDAX), FTSE,... it has been my experience that a small trawl doesn't work very well. Just open a trade on Dax, for example, on m5 and put a trawl. - Moose is inevitable! For these symbols, the trawl cannot be smaller than 155/180, and this is actually a stop loss!

Then you'd better insert takeprofit. You have to think it over. Take a guess....

 

My expert did very well today on Dax. One more thing. Try to "touch" NASDAQ (NQ), Dow Jones (YM), Standard&Poor's 500 (ES)

/ Give me your email at....

 
Lukyanov:
I think in this case it would be better to close the buy. That's even as I am writing this message, the profit is already $ 7 and $ 2 respectively. If we close now we will get 9$ in total. We could close buy with 9$ profit and now sell with 2$ profit. We get 11$ in total. And so every time (see equity and balance indicator)...

Not really. This is a seeming Opportunity. After all, after forcibly closing the trade, - yes, we will get $11/.

But after that, the trade will open then again in deficit! Against us. And in the end it will still be $ 9, not 11. I also noticed it. The Expert Advisor often has very good entries. But its getting out is a bit worse!

 
rid, the continuation of the story with the trades on the screenshot above was interesting at all. Sell closed at c/l (-4), but buy at +18. In the end it was $18-4 = $14, which I never expected. By the way, without over-optimization I made 20% in 2 days. IMHO, there is potential for the Expert Advisor. Just need to redo the code, it's not quite correct...
 

It is better to count profits and calculations in pips. Not in dollars. It's much more convenient to operate and compare. I see that you can't see in the line which version is where. Right-click on the terminal (not the chart) and check COMMENTS in the window that appears.

I sent a letter to the Mail (12:10 Moscow time) . But I am not sure if it reaches me. Please confirm receipt - rid200549@ya.ru

123456789
New comment