Valmars - page 4
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
Gentlemen, what if we introduce into the EA the time of opening orders. As far as I remember the same situation (maybe even better) occurs at 20.00 or 21.00. It does not have to be at the opening hour 0.00. By the daily candlesticks is fine. But it is not an optimal variant, as it was said above. The optimum variant may be working at 9 am for example, or at opening of various sessions. Think about it! As I see it is not difficult for a programmer to do it.
Regards!
Valmars,
At trailing minimum 3-4 pips EA will give profits from 2000 till 2006. Then it starts to stagnate and lose. Nevertheless this is, in my memory, the second EA after tsd that give profit from 2000 till now, all other EAs I have tested above 2000 were breaking even, just with different speed J. TSD with min. trailing also gives profit, but it very rarely makes a trade: 2-3 per month, and the comp. has to wait and work around the clock for it.
You said that 00:00 is a dead time and that the order should be at a certain distance from the opening. The question arises, what is the optimum hour for a certain unidirectional movement in any direction and how far away an order must be from the current price? Obviously, it can be easily estimated using an EA with the following parameters:
TakeProfit
StopLoss
TrailingStop (usual, the other ones do not make much sense)
Lots
Time 00:00 (adjustable by 1 hour step, e.g. 01:00, 02:00, 03:00 etc. )
Pips (with adjustable 1-pip increments, e.g. 1,2,3...30...40... etc. )
MagicNomber (so that the EA can distinguish its own orders and manage them only)
DepoPart (can be set from 1 to 10)
At a specified time, the Expert Advisor places 2 buy and sell orders at a specified distance from the current price in the specified points. When one order triggers, the other should not close and should not close until the first order closes at trailing stop or take and it should only exit if 1 order is really unsuccessful and a stop loss occurs. The idea is that price usually moves in and out at opening and can take profits in both directions, but consecutive triggering is required in order to prevent the price catching up with both orders and taking them both to a loss. And in case of a stop loss the opposite order must be deleted because it stays behind the price which is not logical.
DepoPart (can be set from 1 to 10) If you set 1, the Expert Advisor will automatically set the number of lots equal to the deposit, for example if the deposit is 5000, it will set 3.7 lots. 2 will be ½ of the deposit (here and further without the pledge), 3 will be 1/3 and so on up to 1/10. In case of profit or loss EA adjusts the number of lots in relation to the part of the deposit, which is allowed to be involved.
Pips strategies give good results only in the tester, on micro or ruble forex, where the machine quotes, may work for a while. Why not let the man "play" if he will eventually "sell out".
As for testing over 7 years, I don't see much point in it, as the world, discount rates, the state of countries economies, and ultimately forex, have all changed in that time. Do you want to create an EA for all times, i.e. currencies? I think that is impossible in principle. And sooner or later it will be proven mathematically (and maybe even now directly follows as a conclusion from some applications of modern sections of mathematics).
As for your last strategy, I will try it as a training task, if I have time. It seems that summer came to Vladivostok, where I am now, i.e. the sun appeared and the temperature rose above 20 degrees. Agree, we must use and not waste time on, in general, dubious ideas, especially when there is no time to check yours. I am already anticipating your next idea: to set a grid of delays. All this has already happened and some of it even had positive results.
Valmars,
When we find out how many pips from the opening and what is the optimal time, provided that this alone is not enough, we will add another thing to recalculate the candles to determine the ultra-short-term trend and act in that direction - so far such a strategy and indicator has not been found, so to speak, know-how :)
Why order on pikes is losing its relevance in 2006-7, even though that's where stops seem to get concentrated? It seems to me, that many small players have appeared, and the market dynamics have increased and even big players are not putting stops at important levels, but in the number of pips of loss they can afford.
One pipsomegger, who was making a loss before 2006 and suddenly started making a profit in 2006-7, indirectly confirms it, although his strategy is different.
when we find out that all this is useless too, we will switch to the microtrend single-colour candlestick strategy I mentioned, which I am still only thinking about. and after that we run out of ideas, then I don't know what to do : ((
Valmars,
Your mod, risk and persentrimargin are my DepoPart, so it's better to leave your variant (if they don't slow down optimization time when testing, maybe 3 parameters are slower to pick up than 1).
Your trailing stop looks cooler than usual, the more so it can easily be turned into a regular one, so it's better to leave it as it is, i.e. stepped (if it doesn't slow down the optimization time when testing).
Time 00:00 can be represented as Hour from 0 to 23, where 0 means 00:00, 1 means 01:00 etc., whichever is easier to implement, although Time 00:00 looks clearer.
Still couldn't figure out what PeriodX and MaxOrders mean. Maybe they should be removed if there's not much use in them? Again, the tester sits picking them up, wasting time. And another thing, why EA does not place more than 1000 lots, maybe it is a limitation of the broker?
Valmars,
Your mod, risk and persentrimargin are my DepoPart, so it's better to leave your variant (if they don't slow down optimization time when testing, maybe 3 parameters are slower to pick up than 1).
Your trailing stop looks cooler than usual, the more so it can easily be turned into a regular one, so it's better to leave it as it is, i.e. stepped (if it doesn't slow down the optimization time when testing).
Time 00:00 can be represented as Hour from 0 to 23, where 0 means 00:00, 1 means 01:00 etc., whichever is easier to implement, although Time 00:00 looks clearer.
Still haven't been able to figure out what PeriodX and MaxOrders mean. Maybe they should be removed if there's not much use in them? Again, the tester sits picking them up, wasting time. And another thing, why EA does not place more than 1000 lots, maybe it is the broker's limitation?
In my EAs Period_X I indicate the standard period of chart on which it should work in order to take indicator values for analysis. It is set in minutes because external parameters should have a numeric value. The Expert Advisor can be run on a period of 1 minute and it will work with the values set in the parameter . In your case only days are used and it is not needed. Max_Orders - maximum number of orders that can be opened. In your case, it is always two, so you don't need it either.
During optimization, each new value of the optimized parameter increases the number of variants by half and the optimization time increases accordingly. I don't use pure optimization by many parameters, it's too long and there will be a clear fit. I take 2-3 (for example, Stop Loss and Take Profit) and look at the pattern of changes over a year or two. If I see something interesting, I fix it and try to change others. So to say, combining optimizer methods with manual "gradient descent" method.
1000 lots is a server limitation, if you use quotes from History-Center, then Meta Quotes demo now has 5 lots max for the contest at all.
I read your terms and conditions and have questions:
Well, putting two pending orders at the beginning of an hour bar at a certain distance is ok, but what about expiration? When should I remove 1) none of them triggered, 2) one triggered and remains open 3) one triggered and closed at stop loss or take profit
If one triggered and price went in the other direction, and reached the second one, it will also trigger - otherwise what for is it needed at all and it should be deleted when the first one triggered.