Well - we haven’t really managed to code an EA that is durable in the long run, so I just thought it would be good to have it in for testing purposes. To figure out what work and what doesn’t. But if you are sure that dd, based on balance will work like a charm I guess there is no point in keeping it 
no, and until now it has been calculated from equity, and not from balance (as the trading system dictates)
so would be better if we can get the EA to follow the trading system first, and [B]THEN[/B] tweak it [B]IF[/B] we encounter any problems 
I have now manually set up two (extreme) scenarios to test primers. I start at 1.5000 and move straight up to 1.5600 where a long primer is triggered. Then back to 1.5100 in order to trigger a short primer, and then straight up again.
With one short primer open (lock in at +10 pips when SECOND primer is opened), I hit the 0.06 drawdown at 1.5600 as before (where the first needed long primer is already waiting). Adding a new primer every 30 pips, always keeping the primer lot size the same (4K in this example), I can extend the channel 900 pips straight up before the growth in shorts triggers another 0.06 warning at 1.6500. At/before 1.6500 larger longs would have had to be calculated for complete safety.
Same scenario, but I allow 2 primers to remain open (lock in at +10 pips when THIRD primer is opened). At 1.5600 i hit 0.06 again. Adding a new primer every 30 pips, always keeping the primer lot size the same (4K in this example), I can extend the channel 300 pips straight up before the growth in shorts triggers another 0.06 warning at 1.5900.
900 versus 300 pips is a pretty massive difference, so why is keeping two primers open against the trend a good thing?
Re
D
PS! “The trend would never move like this” is not relevant for the math.
im looking at your excel and im confused…
are you still opening oposite trades once you started primers??
because if i remember correct, the math in the system is based on that you stop normal trading once you start primers, until the primer run stops…
I really can’t see the problem.
In You example, You will just get an upper primer zone and a lower primer zone and normal trades in between.
So even Your extreme case will just make some money to the channel.
Actual the primer size should also increase in each end, as it should be open lot in opposite direction divided by tree, rounded up to nearest lot. and that calculation should be made with every single primer. Else the maths doesn’t work.
So even a few extreme swings should be safe and when the rate then starts to move normally again. then it should earn some good money.
Remember the system is not made for fiction but for real markets. And no countries would let they currency go berserk like that. That would make them look untrustworthy.
I’m not looking for realism here, but the algorithm so we can code this correctly. Once the math is correct for the extreme cases, the math is correct. Before that, we’re close to something that will probably work, in most cases, unless… What would you prefer?
Calculating new lot size for each primer is new information, and I don’t see how that would work, at least not one by one. Using the regular 12.000 units difference again and dividing with 3, you would get P1: 4.000, P2: 2.000, P3: 2.000, P4: 1.000 units. That’s not enough to maintain balance. Not even if you don’t place opposing trades as ZeppDK suggestes (also new to me).
If you don’t place regular shorts while having long primers, then you’ll get stuck with a potentially large gap once the rate finally turns. If 10 long primers are hit before retracement, there will be 270 pips of no trades before you get back to pre-primer rates. That can’t be right? No regular longs while long primers I get, no shorts, I don’t.
Re
D
Sorry for not being specific enough.
The primers is calculated every time it has been deactivated.
That means, that if you have some buy primers and the rate drops enough to activate sell primers. Then it should be calculated again. and if the rate goes back up to the upper primer zone, then new primers should be recalculated and maybe even adjusted with additional primers to fit the new size
Normal trades is suspended in the primer zone, and only if the rate goes back behind first primer, then the normal trades is reactivated that is the reason that the primers do not need to be recalculated as long as they are active.
And this system do work in extreme cases. If we just stick to the rules of the system, instead of inventing our own.
But of cause it should be possible to adjust the primer size. But the standard value should stick to the rules.
So you’re saying that no regular shorts should be placed, and no regular longs should be placed while primers are active (in profit)?
(Sorry to be harping on this, I’m just very surprised and want to make sure I get it right. It’s not at all what I have understood from the system dox).
Re
D
yes that is exactly what I say.
We stopped updating the system docs, as there were too many people that started a channel, only knowing the basics. And when we reach the primer zones, we got bombed with request for helping people calculating primers over and over again. And a lot of people also misunderstood the plain text a lot of times.
The documents is only meant to give an idea of how the system works, and is not complete, as Michael and I, have made a lot of adjusting and calculating afterwards.
I changed the Excel sheet to no regular trades while primers are active, and that seems to work really well. I also realized that the gap between primers and regular trades won’t be as big as I initially thought, assuming you would still add both long and short regular pending orders to “newly opened up areas” as each primer reaches TP.
Ex:
Say I trigger a long primer at 1.5604. If it reaches 1.5704 and closes in TP, then a pending regular buy will be added at 1.5604 (there’s already a short at 1.5600)
As P2 (opened at 1.5634) reaches 1.5734, nothing changes.
As P3 (opened at 1.5664) reaches 1.5764, a short pending is opened at 1.5650 and a long at 1.5654.
You would then need to retrace a maximum of 134 pips or a minimum of 104 pips to get into profitable trades again (ignoring primer SL for ease), but your channel would be expanded safely.
Have I got it right now?
Re
D
that was exactly how i remembered it… weee… i can remember things
And I spent my entire Easter vacation trying to program the wrong thing… weee…
Re
D
sorry i havent noticed it before, but it hit me when looking at your excel sheet… and then its also pretty clear why primers would not work correct, if you keep adding to the drawdown, but keep the same primer size, then the ratio between the two gets smaller all the time…
Now You a bout to get a hold of it… :45:
And another little detail is that when the normal trades, due to increase to volume is at the same size as the primers, then the primers should be reset and then new primers (calculated from the same rules) will be made when the 0,06% is reached.
If we go back to CC’s first attempts (before all the focus on equity), there was an idea of ChannelTop and ChannelBottom. I would think that adding code to change these two variables as primers open and close should bring us very close to the system as I now understand it. That might tell us about margins as well, which I haven’t been able to calculate manually. Spreading Print statements of margin and equity around would, I assume eventually give us a pretty good picture of lot sizes for primers (as in maybe divide by 2.8 works better).
Even without any regular trades in the primer zones, we would still need to cover the rising negative P&L for the trades below. I’ve been reading up on margin and different brokers have different margin settings for hedged trades, but I think the spikes in free margin that happens when a positive trade is closed is a large part of what has started the unravelling of the system so far. With a zero cost setting (no margin used for equally hedged trades), closing a primer will immideatly cause a large fall in free margin to cover the now “un-hedged” opposing trade. Assuming this spike gets you below 90% margin or whatever the broker insists on (even for an instance), other oppposing trades will automatically be closed to cover the difference. Then you end up with a “pulling a thread in a carpet” type situation and everything unravels. (Once again, I think. Not sure, not trying to invent a new system, just thinking and looking for clues).
Re
D
Primers are only primers while larger than CurrentLotSize, got it. I think that’s been included in several of CC’s versions already.
Re
D
why would that be a big problem? we have covered the margin until the first primer is set, and we only close this primer if we have futher primers opened, so i cant really follow you on this one, and when the trend turns we still have 2 primers open…
could you please explain it in more detail what you mean?
Hmm… One question:
As I understand it from reading Dennis post the channel can still expand - i.e. whenever we reach primer no. 3-4-5 etc. that opens up the primer-1-2 for ‘regular trading’.
I hope that is how it works, as that would allow for slowly expanding the channel into new areas which might be the new standards (say over a 2-3 year period the basic economic relationship between e.g. EUR/USD could change fundamentally).
However I am not entirely sure that that is what is the intention from Zepp and Webzone? I.e. will the channel always stay fixed or will it expand as per Dennis’ suggestion above?
Looking forward to see new version of EA
then you have not been listening to what webzone just said
first normal trading until the 0.06% drawdown… then normal trades stops and primers start, when the drawdown has been turned around, either by going back to the “normal trading zone” from earlier, [U][I][B]OR[/B][/U][/I] if the normal tradesizes become the same size as the primers… this means, that if we are in a place that was a primer zone, but has collected enough capital to trade normaly with that lot size, then we trade normaly until we get to a new 0.06%, and have now expanded the channel
Not really, which is the problem 
I’ve run back tests on all of CC’s versions and several I made myself, and I’ve noticed that “[no enough money]” pops up almost immideatly before every single crash. Occasionally the test will recover for awhile, but usually not. The equity line on the charts generally fall slowly (with some spikes) until both equity and balance takes a massive dive. If you open the log file and look through the details, you’ll find messages like these:
22:11:15 2011.10.07 08:48 Tester: not enough money for sell 0.10 GBPUSD at 1.54727 sl: 0.00000 tp: 1.54227 [2011.10.07 08:48]
22:11:15 2011.10.07 08:48 Tester: PrevBalance: 14212.57, PrevPL: -13374.17, PrevEquity 838.40, PrevMargin: 1058.45, NewMargin: 1207, FreeMargin: -368.58
22:11:15 2011.10.07 08:48 Tester: pending order is deleted [no enough money]
22:11:19 2011.10.07 09:18 Tester: not enough money for sell 0.10 GBPUSD at 1.55252 sl: 0.00000 tp: 1.54752 [2011.10.07 09:18]
22:11:19 2011.10.07 09:18 Tester: PrevBalance: 14260.83, PrevPL: -13779.53, PrevEquity 481.30, PrevMargin: 1214.24, NewMargin: 1366, FreeMargin: -885.04
22:11:19 2011.10.07 09:18 Tester: pending order is deleted [no enough money]
22:11:48 2011.10.07 14:29 06_split_Greenlander_Lots: stopped because of Stop Out
I don’t know what all of that means yet, and I don’t know if changing the EA to not placing any regular trades in the primer zones will remove this problem. I have however gone through lots of posts and blogs discussing hedged trading systems, and they all refer to margin as the killer. Hence I’m waiting for the next version in order to test out margin levels. Hopefully there will be no problem, but in general I don’t like risking money on things I don’t understand, so I try to figure out all aspects of this system.
Re
D