Experiment - page 272
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
What difference does it make which TF? Let's say on TF D1 five days = five sample points, and on M1 it would be 7200 points. 7200 points is better, right? There is more information, and probably less jitter because of more input data.
This was not confirmed at the beginning of the experiment when working on TF M1 because of the fact that, at that time my indicator does not work correctly on a large amount of information (more than 1000) bars, I think it is because of an error in the program. I need to correct this error.
And, here, the 3rd variant of the entry, an interesting: you can enter the market only when the trend is established, when all three lines of the indicator are combined. However, in this case the total time of staying in the market decreases. This variant should be compared with the previous one in the tester by main parameters, for example, by profitability and stability.
Yusuf, I have tested this variant long ago (I cannot even imagine how many variants and for what series I have tried to apply PNB).
The problem with variant 3 is that when it has formed, it's too late to enter.
If you could identify the moment when the transition takes place from variant 1,2 to 3 earlier, maybe you would get something useful.
Yusuf, I tested this variant a long time ago (you have no idea how many variants and for which rows I have not tried to use PNB).
The problem with variant 3 is that when it has formed, it is too late to enter.
If you could identify the moment when the switch from variant 1,2 to variant 3 would happen sooner, maybe something useful would come out.
I will think about it.
This was not confirmed at the beginning of the experiment when working on TF M1 due to the fact that, with me, the indicator does not work correctly on a large amount of information (over 1000) bars, I think due to a bug in the program. I need to correct this error.
The more points in one interval the better in my opinion.
And the sample should be selected based on at least some logic. Cycles, for example - trading session, day, etc.
A sample of 1440 minutes (24 hours) would be enough.
I had this idea.
1.Take a basket of currency pairs: EURUSD, GBPUSD, AUDUSD, NZDUSD, USDJPY, USDCHF, USDCAD.
2. Separating pairs into a currency basket: EUR,GBP,AUD,NZD,JPY,CHF,CAD, making USD a constant in the basket. The sum of the basket is always constant.
So instead of EURUSD, only EUR series will be given to the NBP input, and so on in a similar way.
Then the EA trades the basket on the currency pairs from point 1.
p/s. I will do it, if there are volunteers willing to monitor experiment no 2. Or if Yusuf will monitor it himself.
I want to hear ideas beforehand on the sampling, time of day when to open/close orders or at any time depending on the signal.
The more points in the same time frame, the better in my opinion.
And the sampling should follow at least some logic. Cycles for example - trading session, day, etc.
A sample of 1440 minutes (24 hours) would be enough.
.
Then, test this option, please. I don't have an opportunity - I am in Moscow on a visit.
Then, please test this option. I don't have the opportunity - I'm in Moscow on a visit.
The same result. Everything has been done for a long time now.
The same result. It's all been done for a long time.
Go ahead, then, to experiment number two - be the presenter.
What difference does it make which TF? Let's say on TF D1 five days = five sample points, and on M1 it would be 7200 points. 7200 points is better, right? There is more information, probably less chatter due to more input data.
We need to solve the chatter problem somehow. Maybe by pre-smoothing the price, other options... Suggest any thoughts.
Any conversion of price series is self-defeating. You are on an aeroplane and it flies you up and down. Do you want to smooth out the altitude data? If you smooth it out too much, on the next downward dash you will hit the ground and kill yourself. And on the upward roll, you'll stall and kill yourself too. Would you like to try it?
The idea of thinning the price range was also discussed. Get back on the plane with your family and on landing we will look at the radio altimeter no more than once every 10 minutes. Do you want to?
The solution is not to manipulate the price series (smoothing, thinning, etc.). By manipulating the data, we pretend that we are observing an entirely different process, and that our life, or prosperity, is directly dependent on the real process.
But, all is not hopeless, a solution to the problem of chattering was found a long time ago:https://ru.wikipedia.org/wiki/Гистерезис
Go ahead, then, to experiment number two - be the presenter.
Have you finished yet?