AddOn Banana + BlueMonkeyEA讨论

RedRhinoLab posted:
Correlation or no correlation, it doesn't matter when you add larger lots to your losing positions waiting for a basket profit. These systems will always be a pipe dream and sadly it always results in the same outcome. It will only take 1 pair to destroy your account and by adding more to your account you are only increasing the odds of failure and not the other way around.

This is a very simplified analysis. The no correlation is important because if you set your margin stop-out to 30 and you have 4 or 5 pairs crossed with EUR (example) you will probably loose 65% your account or more if EUR drops or falls too much too fast.
Additionally, no correlation is VERY important(means essential) if you are using Amortization.
Your account will never be destroyed as you say because you can set the maximum loss.

hokie posted:

maiamrn posted:
@hokie the last week was really bad for the system. BMEA continuous to open trend even when the trend changes. This characteristic we can't change so we use protections coded in BANANA. The Amortization closes some opened trades and reduces exposure. I hardly recommend to let it ON. Why this demo account was not affected? We are trading 4 pairs in this account this month: EURUSD, EURGBP, GBPCHF, USDCHF. Pairs with low correlation between them. And yes, this account can have manual interventions but last week it was not needed.

@maiamrn

Thanks for the response. Just to make sure when you said 'I *hardly* recommend to let it ON' above, did you really mean that, or you meant to say 'I highly recommend to let it ON' ?

Regarding the choice of 4 pairs this week, is this only for this week, or you're going to use the same 4 pairs only going forward? Any particular reason why you used only 4 pairs this week, and not a dozen or so as you recommended (more diversified, to reduce risk) a few weeks ago?

This demo account will trade only 4 pairs this month because I've turned OFF Amortization. You are right - if you use amortization you need more pairs (no correlated). Remember it is a testing account...

gorydan posted:
@maiamrn could you share the settings and the time frame for the demo account 'AddOn Banana + BlueMonkey' ? because it survived without problem to this last week. The non correlation is important as you said.

This demo account:
-daily graph
-Ignore trends
-Close basket on trend change OFF
-high exposure protection ON
-exposure cut - 20
-Amortization OFF

I DO NOT recommend this settings on real accounts. Always set Amortization TRUE on real accounts.

Davey posted:
On the plus side Lots Amorization sort of worked on this account, took quite a loss but still in the black, shame it's a demo and the real account has taken a pounding.

https://www.myfxbook.com/members/Davey/addon-banana-v12-demo-22/1166390

I recommend you reduce the CAD crosses from 4 to 2 pairs. You have 5 AUD crosses, it´s too much and if AUD moves to fast you will have a big DD and you may crash. You have 7 USD crosses and it´s dangerous.
If you need to keep more than 7 total pairs you might reduce the Inilot to 0.05.
It´s possible to set the margin stop to 20 to avoid crash.
The next version improvements will help us to reduce risk.

maiamrn posted:

Davey posted:
On the plus side Lots Amorization sort of worked on this account, took quite a loss but still in the black, shame it's a demo and the real account has taken a pounding.

https://www.myfxbook.com/members/Davey/addon-banana-v12-demo-22/1166390

I recommend you reduce the CAD crosses from 4 to 2 pairs. You have 5 AUD crosses, it´s too much and if AUD moves to fast you will have a big DD and you may crash. You have 7 USD crosses and it´s dangerous.
If you need to keep more than 7 total pairs you might reduce the Inilot to 0.05.
It´s possible to set the margin stop to 20 to avoid crash.
The next version improvements will help us to reduce risk.

Yes I agree but it's only a demo, was a test to see if any pairs other pairs didn't work besides JPY, found two more. My real account trades on six pairs, with only two of each currency, and only two from each continent. Keep up the work would be great if we could reduce crashes to once a month or less and make more than we loose - one day I hope

@Davey What are the two more pairs that didn't work beside JPY?
Also, would you share which six pairs you're trading in your real account?

@hokie The two pairs that failed to close with profit that were way past the usual 10 pips, probably 100+ pips, therefore I deemed them not to work on BMEA were AUD/CHF and NZD/CHF.

I won't share the account yet because at the start I was using scripts until I got 30% drawdown, it won't help it will just muddy the waters and I don't want others making assumptions based on it.

The pairs I use are;
EUR/GBP - European pair
USD/CAD - N. American pair
AUD/NZD - Australasian pair
GBP/USD - Eur/American
NZD/CAD - Aus/American
EUR/AUD - Eur/Aus

So only use 2 pairs per country/currency - only 2 countries per continent - best spread combo for my broker

I'm going to change my strategy a little of how I use Banana & BM. Previously I have thought to start with 0.1 lots per 10,000 and try to get that down to 0.05 lots per 10,000. But now I'm going to try this instead;

We make so many correct trades then every 7-10 days we enter a trade that leads to 30% drawdown, and all the success is undone, and more. So I am going to start heavier and have a tighter drawdown, I'm aware this will make drawdown more frequent but I guess I will find out, the current strategy isn't working for me. e.g.

£50 SL per 0.01 initial trade. So I've started by risking £100 and starting with 0.02 initial trade. The hope being that I can make money more aggressively to cover the weekly £100 loss. If this works I will risk £200 and start with 0.04 initial trade, obviously adjusting BMEA to suit.

Hope it works, running out of capital, too impatient for demo's. Good luck. Also off to Fiji now so can't reply too much in the next two weeks if you have any queries.

Happy to share this untouched 9 pair demo that includes the six pairs mentioned above, just a test to see how things work on 3 countries and 3 continents. Still holding a large AUD/NZD trade.

https://www.myfxbook.com/Davey/banana-demo-9-pair/1162174

https://www.myfxbook.com/members/Davey/banana-demo-9-pair/1162174

@Davey thanks so much for the great info. It's very helpful.

The reason I was asking you about the pairs that don't work with BMEA was because I wanted to compare notes as I think GBP/AUD is not managed by BMEA too, like the JPY crosses. But your 2 pairs are different.

It will be interesting to see how your new strategy will pan out in a few weeks. Good luck, man!

In the meantime, enjoy your trip to Fiji !!

My settings now are :

Pairs traded : EURUSD, EURGBP, GBPCHF, AUDCAD

Add-on on a H4 chart
Init lot : 0.05 for 10 000
Ignore trends
Trend force 40
Close basket false
Close basket at break even true
Exposure leverage 20
Lots amortization algorithm true

https://www.myfxbook.com/members/gorydan/fxcm-banana-add-bmea/1158430

I've replaced USDCHF by AUDCAD because USDCHF is the inversed correlation of EURUSD and it caused 45% of drawdown during the last FOMC (18/03/2015). During this FOMC I was at 0.1 lot for 10 000, it's why i've reduced to 0.05 for 10 000.
I'm on H4 chart to generate a little bit more trades than on Daily chart.

You can see the good progression since a few days.

@maiamrn is there a limit date to the V1.2 ? if yes, it's the time to have the V1.3 or a version without limitation.

thank you

@maiamrn so quite lately, have you any news about development?
Cheers.

another dead project, it's a shame was really interesting :(

Exactly, we can't trust anybody. It's always the same result, people are excited and finally leave the project.