TraderJoeForex » Wed Oct 12, 2016 11:31 pm wrote:
Nope.. the only thing I can think of is with your much bigger account size you will have some proportionately bigger lot sizes than mine since I think the lot size calculation will round down rather up when looking at balance/equity and my account size is limiting some of these calculations as Zennor explained well earlier in the thread. The disadvantage for me with this is slightly higher risk on the more volatile pairs but the advantage seemingly is lower drawdown (so far).

With the exception of grid EA's I always do my testing with a #500 starting balance because for me that is a maximum amount I am comfortable losing when trading live and is the amount I usually fund my live accounts with.

If I had a 10k account to trade live I would use a much smaller lot factor.. probably around 0.5.. and then smaller basket TP and SL to match (0.25% and -5%) for example.

My demo account has been stuck on the same basket since I posted last night.. but the max drawdown shows as just over 5% and I have refreshed the trade chart indicator to double check for you. Just a waiting game now for this basket to close hopefully...

MathiasJ » Wed Oct 12, 2016 8:43 am wrote: Hi Aram!
You are right in your explanation of how the basket calculation works today.
And it is possible to implement it.
But do you really Think that this would make a difference?
Because normally your setfile should produce a higher balance anyway...
I have used a combination of trade sl and basket tp sl and if basket TP is 1% then normally the trade losses are below 0.2%, so positive end result anyway.
If you can convince me that this is really needed then I can add it, but preferably in next round as I currently try to stabilize and document what we already have so far.
BR
Mathias

• We start fresh with no trades on and boot up the EA with for instance a BasketTP = 1% and BasketsSL = 7%. Let's say we start off with a balance of 100 EUR.
• Signals are given and trades are entered.
• Quite a few trades pull the entire active basket into DD. At some point they reached the opposite CSS{H4}_Diff and are exited as the trend may continue that way. Good thing is we're ready to take the countertrend should that happen.
• Because these bad trades we're exited, we have to incur e.g. -3% in balance.
• Now indeed a few countertrend trades are taken when the signal happens and our other existing trades start to move too in the anticipated direction.
• Here comes the main point that should make the difference clear. Because our equity shrunk with -3%, our absolute basketTP profit value is going to shrink too, more precisely we'll cap a basket of 97*0.01 = 0.97 EUR. But what if volatility sets up again in the markets and we must again cut a few losers short, thus again decreasing the balance, and yet again aiming for a smaller basketTP? Now imagine a few streaks of loser-trade cuts this way. The more our balance gets cut, the smaller the basketTPs will become and the more of them we'll need to collect before we can actually grow our balance. The latter fact of collecting more basketTPs is in its turn tied to the amount of entry signals (or the duration of valid entry conditions) for trades to be entered. Therefore it could also be the case that the entry conditions don't line up and we thus can't enter trades to capture a substantial amount of basketTPs to make up for these losses. It shouldn't be too hard imagining these kinds of vicious circles.

Boulder » 13 Oct 2016 04:17 wrote:
Hey Mathias,

Thank you for getting back to me here. Yes I think very much so that having a (similar) method in place as I described a few posts earlier will be to our benefit. It will especially be rather crucial for basket-trading in combination with looking to close trades before they become even bigger losers and therefore put more strain on the basket to reach TP. Consider this scenario (which will inevitably happen as the market is a vicious beast...):

• We start fresh with no trades on and boot up the EA with for instance a BasketTP = 1% and BasketsSL = 7%. Let's say we start off with a balance of 100 EUR.
• Signals are given and trades are entered.
• Quite a few trades pull the entire active basket into DD. At some point they reached the opposite CSS{H4}_Diff and are exited as the trend may continue that way. Good thing is we're ready to take the countertrend should that happen.
• Because these bad trades we're exited, we have to incur e.g. -3% in balance.
• Now indeed a few countertrend trades are taken when the signal happens and our other existing trades start to move too in the anticipated direction.
• Here comes the main point that should make the difference clear. Because our equity shrunk with -3%, our absolute basketTP profit value is going to shrink too, more precisely we'll cap a basket of 97*0.01 = 0.97 EUR. But what if volatility sets up again in the markets and we must again cut a few losers short, thus again decreasing the balance, and yet again aiming for a smaller basketTP? Now imagine a few streaks of loser-trade cuts this way. The more our balance gets cut, the smaller the basketTPs will become and the more of them we'll need to collect before we can actually grow our balance. The latter fact of collecting more basketTPs is in its turn tied to the amount of entry signals (or the duration of valid entry conditions) for trades to be entered. Therefore it could also be the case that the entry conditions don't line up and we thus can't enter trades to capture a substantial amount of basketTPs to make up for these losses. It shouldn't be too hard imagining these kinds of vicious circles.

That's why it should be much better to let the open trades (existing ones and countertrend trades after trades in the wrong direction are cut) run instead until they reach the basketTP over the begin balance we started with (or will register everyday at some point): i.e. effectively recovering the losses plus achieving the intended basketTP. This'll probably work best if we use small enough basketTPs . Because our reference balance is static (until it gets updated with every achieved basket of profits), our lot sizes will also stay the same and not become smaller, hence helping to achieve a basketTP sooner.

Actually come to think about it: you don't need to register the balance everyday at a certain point. Instead it should be more than sufficient to simply take the begin balance at first initiation of the EA and work with that until a basketTP is reached. When a basketTP is hit, that amount should be added to the balance and the EA should work with that until the 2nd basketTP comes in etc etc. This way you're also actually compounding your profits too and have a rather clear grasp on the way your balance grows, i.e.:

by the N-th basketTP hit (which is expressed in decimal form, e.g. basketTP = 0,01 as being 1%), our balance will be: Balance(N) = StartingBalance*[(1+BasketTP)^N]

Hope it is a bit more clear now and convinces you Mathias.

Cheers mate

Aram

Boulder » Thu Oct 13, 2016 2:17 am wrote:
Hey Mathias,

Thank you for getting back to me here. Yes I think very much so that having a (similar) method in place as I described a few posts earlier will be to our benefit. It will especially be rather crucial for basket-trading in combination with looking to close trades before they become even bigger losers and therefore put more strain on the basket to reach TP. Consider this scenario (which will inevitably happen as the market is a vicious beast...):

• We start fresh with no trades on and boot up the EA with for instance a BasketTP = 1% and BasketsSL = 7%. Let's say we start off with a balance of 100 EUR.
• Signals are given and trades are entered.
• Quite a few trades pull the entire active basket into DD. At some point they reached the opposite CSS{H4}_Diff and are exited as the trend may continue that way. Good thing is we're ready to take the countertrend should that happen.
• Because these bad trades we're exited, we have to incur e.g. -3% in balance.
• Now indeed a few countertrend trades are taken when the signal happens and our other existing trades start to move too in the anticipated direction.
• Here comes the main point that should make the difference clear. Because our equity shrunk with -3%, our absolute basketTP profit value is going to shrink too, more precisely we'll cap a basket of 97*0.01 = 0.97 EUR. But what if volatility sets up again in the markets and we must again cut a few losers short, thus again decreasing the balance, and yet again aiming for a smaller basketTP? Now imagine a few streaks of loser-trade cuts this way. The more our balance gets cut, the smaller the basketTPs will become and the more of them we'll need to collect before we can actually grow our balance. The latter fact of collecting more basketTPs is in its turn tied to the amount of entry signals (or the duration of valid entry conditions) for trades to be entered. Therefore it could also be the case that the entry conditions don't line up and we thus can't enter trades to capture a substantial amount of basketTPs to make up for these losses. It shouldn't be too hard imagining these kinds of vicious circles.

That's why it should be much better to let the open trades (existing ones and countertrend trades after trades in the wrong direction are cut) run instead until they reach the basketTP over the begin balance we started with (or will register everyday at some point): i.e. effectively recovering the losses plus achieving the intended basketTP. This'll probably work best if we use small enough basketTPs . Because our reference balance is static (until it gets updated with every achieved basket of profits), our lot sizes will also stay the same and not become smaller, hence helping to achieve a basketTP sooner.

Actually come to think about it: you don't need to register the balance everyday at a certain point. Instead it should be more than sufficient to simply take the begin balance at first initiation of the EA and work with that until a basketTP is reached. When a basketTP is hit, that amount should be added to the balance and the EA should work with that until the 2nd basketTP comes in etc etc. This way you're also actually compounding your profits too and have a rather clear grasp on the way your balance grows, i.e.:

by the N-th basketTP hit (which is expressed in decimal form, e.g. basketTP = 0,01 as being 1%), our balance will be: Balance(N) = StartingBalance*[(1+BasketTP)^N]

Hope it is a bit more clear now and convinces you Mathias.

Cheers mate

Aram

TraderJoeForex » Thu Oct 13, 2016 10:36 am wrote:
haha this makes sense to me now too.. cheers Aram

TraderJoeForex wrote:No progress for me either past 48hours.... after such a good run last week and start of this one it's been a surprise to see this one basket struggle to close.. Drawdown has stayed low but it just ain't going anywhere. I have a feeling the restricted number of currency pairs may have something to do with this but drawdown may have been much bigger with all the pairs too..

Here is a new set file I have put together and have now put this on one of my accounts to see if it will get these positions closed. It may not offer the very fast quick win baskets that the previous did but trading only when H1 and D1 agree direction it stands more chance of pulling through by the end of this week. I have kept my other demo running with the same basket and settings so I can see which set digs out of the hole first.

Rules are simple with this one.. only trade when H1 is breaking out in the direction of D1 defined by diff and delta as before. This time using closed candles for the conditions to be true (diff shift 1), Delta > 0.2 and DIff > 2.0 for both. I have increased max trades to 3 per pair so that my existing trades can be hedged and/or added too whichever way price moves from here.

I have added back all the currency pairs and the basket SL as -100% so there is room to breathe until I know what a sensible SL will be.

If anyone wants to try this and add some early exit conditions that would be interesting to compare.