On-line trading advisor. Exchange of views - page 10
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
In my opinion, such time dependencies do not carry any statistical advantage. In short, you want to regulate the non-stationary market with lazy time dependencies.
In my opinion, this is like trying to cast a fishing rod once an hour at exactly 15 min. 34 seconds exactly once an hour in the hope of increasing the catch. Don't you think it's better to do without any limits at all and fish for the full hour?
After all, as I understand it, the maximum number of orders, with or without dependencies, will sooner or later be reached anyway.
... Life will show, this test may last for several months. Which of the accounts will hold (without refilling - the most extreme method of protection against losses), and which one will show the greatest stability.
Still, it will not be an indicator, as this behaviour will be consistent with the specific time period at which the testing will take place. And waiting a few months for this to happen would be irrational in my opinion.
If it's up to you, you may perform such a test in a tester, where you can run it multiple times with optimization, and for a larger time interval as well.
Still, it will not be an indicator, as this behaviour will be consistent with the specific time period at which the testing will take place. And waiting several months for this to happen is in my opinion irrational.
As a last resort, we can test it in a tester, where it is possible to run it multiple times with optimization and over a longer time interval.
As promised, opened some accounts to test the most stable protection options.
183871651
6chfoln
183871655
8nczzlz
183871658
hbu4nie
183871661
ie7hyix
They are all set to the same currency pair, but the EA settings are different:
Lakshmi_4_293.mq4 - continuous trading, limited only by the order limit.
Lakshmi_4_293_Blocks.mq4 - cyclic trading, first, the maximum number of blocks is set, after that, until all orders are closed, no new blocks are opened.
Lakshmi_4_293_Day.mq4 - cyclic trade, new blocks are opened only during the first half of the month.
Lakshmi_4_293_DayOfWeek.mq4 - trade cyclic, new blocks open only on certain days of the week, every other day.
I am not showing screenshots yet. In a week, not to confuse with previous one, I will post all results at once, dividing them into 2 parts.
The essence of this check is to find the most stable forms of trading and to add (to the existing protections) additional protection against losses.
Stop thinking of yourself. Take it up with Vladislav. And if you decide to do a forward test on demo, do it on ECN broker in 5 figures... I have many advisors that are chopping at 4 digits and don't use martingale.
Stop being so stupid... Take it up with Vladislav. If you decide to do a forward test on the demo, do it on an ECN broker in the 5s. I have many EAs that do not use martingale.
I'm not sure what the self-delusion is. The distrust of strategy tester is confirmed by the multiple differences between on line results and tester results. And if it was just one time. I ran many currency pairs in the tester for testing. My laptop was running in constant testing mode for a couple of months, practically unplugged, but the results only caused me a deadlock and I finally gave up using it at the beginning of 2010. The Expert Advisor itself is tested daily on historical data, but with its own algorithms, not with the help of a tester.
About the ECN broker: It would be interesting to put them as well, why not. Give me the name of the broker, please. But ECN has certain disadvantages: they won't adjust their existing commission on the demo. The advantage of ECN is the speed of triggering, but for my EA it almost doesn't matter. So an order will be placed 10 seconds later, what difference does it make. Another disadvantage is the floating spread. If you understand what a lock is, you know that two opposite orders are set to breakeven. If the spread changes, the Breakeven level will change. There are two solutions to this issue: to modify constantly from 50 to 200 orders, thereby overloading the processor up to "no way". You can also fix the maximal spread (of all the historically existing ones since the EA has been installed) and modify orders by this spread only once. And what if the spread varies by a factor of 10? I've seen one broker with a floating spread change 8-9 times. It means that stops/stops will be moved away very seriously. The efficiency of such trading will decrease.
Can you show me a video of your Expert Advisors? Interesting to see, maybe I'll add something of vital importance to mine.
...The advisor itself is tested daily on historical data, but with its own algorithms, not with a tester.
...
Could you go into a little more detail here...? At least in general terms.
Topikaster already wrote about it, even posted some code. I think you should take a look at it first.
Second page of this thread, look for topikaster's post with an attachment.
Topekaster already wrote about it, he even posted some code. I think you should have a look at it first.
Second page of this thread, look for the post of the topikaster with an attachment.
Thank you.
Thank you.