Basket Trading BALUDA'S CSS Crossovers

KeithM · Post by **KeithM** » Wed Oct 05, 2016 6:36 am

KeithM » Wed Oct 05, 2016 6:26 am wrote:
Great work Zennor. Did you go into drawdown yesterday? I applied your set file and hit a 1.2% basket and then went into 19% DD. It is currently sitting at 11% drawdown. Will continue to monitor to see if it recovers.

Happy to report that despite the large drawdown the Zennor 24/5 H1+H4 basket finally recovered and hit TP. At its peak it had over 300 trades running. It is starting a new basket just prior to LO. Will sit back and watch.

Post by **Zennor** » Wed Oct 05, 2016 7:18 am

KeithM » Tue Oct 04, 2016 9:26 pm wrote:
Great work Zennor. Did you go into drawdown yesterday? I applied your set file and hit a 1.2% basket and then went into 19% DD. It is currently sitting at 11% drawdown. Will continue to monitor to see if it recovers.

Yes Keith I did get some DD with the EUR and CHF pair making sharp rises and mostly with trades that I had going in the opposite direction. Did not observe such a high DD as you, only 13.5% but that is to be expected as it will depend on when you put the EA on.

I was however extremely happy to see such a negative move as it would really test the method and see if it had the capability to trade it's way out of the hole..

Which it did...

one.jpg

Post by **Gertje** » Wed Oct 05, 2016 9:36 am

548 trades closed for nearly 5%

MathiasJ · Post by **MathiasJ** » Wed Oct 05, 2016 9:40 am

Gertje » 05 Oct 2016 11:36 wrote:
548 trades closed for nearly 5%

That is really cool, but at the same time it tells me that we need to be better in filtering what trades to take.
but that s part of the journey as well I assume

Post by **Gertje** » Wed Oct 05, 2016 10:25 am

MathiasJ » Wed Oct 05, 2016 9:40 am wrote:That is really cool, but at the same time it tells me that we need to be better in filtering what trades to take.
but that s part of the journey as well I assume

I started really late yesterday, and picked a lot of GBPXXX sells a few hours before the GBP reversed.
The EA/ strategy showes it is easily capable of 1 basked with x% every day, why not be happy with whats offered and stop trading after that?
Start @LO like Zennor first suggested, and close shop after basket hit TP.

EDIT: or maybe 2 smaller baskets, you get the idea.

Post by **Zennor** » Wed Oct 05, 2016 11:39 am

Gertje » Wed Oct 05, 2016 11:25 am wrote:
MathiasJ » Wed Oct 05, 2016 9:40 am wrote:That is really cool, but at the same time it tells me that we need to be better in filtering what trades to take.
but that s part of the journey as well I assume
I started really late yesterday, and picked a lot of GBPXXX sells a few hours before the GBP reversed.
The EA/ strategy showes it is easily capable of 1 basked with x% every day, why not be happy with whats offered and stop trading after that?
Start @LO like Zennor first suggested, and close shop after basket hit TP.

EDIT: or maybe 2 smaller baskets, you get the idea.

I think that the sample we have so far is far too small for any filtering at this stage.. For now I am happy to trade 24/5 and for the EA to take as many trades as it can for later analysis.

What has surprised me and albeit from a small sample size and only time will either confirm or deny early observation but the Asian session is so far proving to be the most profitable.

My future intentions is to run as is for a few weeks and then remove the least profitable pairs should the results warrant it. Then run four different instances of all pairs trading only one of the four sessions and then do some analysis, remove the least profitable pairs from each session and then set the EA up to trade the best pairs in the best sessions..

Now that Mathias has given us a versatile tool all this testing can be accomplished with no intervention and with great ease.. :!!:

MathiasJ · Post by **MathiasJ** » Wed Oct 05, 2016 11:59 am

Zennor » 05 Oct 2016 13:39 wrote:
I think that the sample we have so far is far too small for any filtering at this stage.. For now I am happy to trade 24/5 and for the EA to take as many trades as it can for later analysis.

What has surprised me and albeit from a small sample size and only time will either confirm or deny early observation but the Asian session is so far proving to be the most profitable.

My future intentions is to run as is for a few weeks and then remove the least profitable pairs should the results warrant it. Then run four different instances of all pairs trading only one of the four sessions and then do some analysis, remove the least profitable pairs from each session and then set the EA up to trade the best pairs in the best sessions..

Now that Mathias has given us a versatile tool all this testing can be accomplished with no intervention and with great ease..

If we know what News to look out for then we can actually filter out those without removing the pair itself. Because I assume we are not News trading.
And that is another investigation, what to keep away from and how early before / late after should we not trade that symbol and should we even close trades for security reason etc.
So, more to investigate along the road.
BR
Mathias

Post by **TraderJoeForex** » Wed Oct 05, 2016 1:14 pm

TraderJoeForex » Tue Oct 04, 2016 9:47 am wrote:I am travelling at the moment and back home tonight. I have managed to log into my VPS to setup the CSSDiff set I posted with restricted trading times ... Today will be a dry run to see if I have the trading times right and if not I can adjust tonight when I get home. I will repost my set file then.

I think I've sorted the session times for my CSSdiff set and after seeing the drawdown from my test yesterday I have added CSSdelta and restricted the EA to only open trades for the first 3 hours of each session.

The session times are based on Global Prime so for anyone who also wants to try this if you set your VPS or PC clock to match broker time then you won't have to adjust the session time in the setfile to your timezone.

I have set this up on a £500 demo with £5 basket close and no SL to give it some room.

Post by **Jemook** » Wed Oct 05, 2016 1:18 pm

Just a tip - one thing I've learned when it comes to basket trading TMA/CSS strategies is that simply selecting every single pair throws the weight heavily in one direction for two reasons.

1. Correlation - trading long AUDNZD, AUDCHF, AUDCAD and then short EURAUD, GBPAUD is throwing your weight big time behind AUD strength. TMA will almost always report that your AUDxxx pairs are trending one way and the xxxAUD pairs are going the other way. Same applies to GBPxxx and xxxGBP etc.

2. Trading GBP pairs (as an example) or pairs with huge ADRs can also throw things off if you are aiming at total pips / PnL in a basket rather than % win/loss per individual trade. Watch out whenever you add GBPJPY, GBPAUD, GBPNZD etc pairs into the mix.

See if you guys can work out a way to still trade a basket but balance your portfolio of chosen pairs you've added. All you need is one day for the GBP monsters to go in the wrong direction to really throw things off.

Post by **TraderJoeForex** » Wed Oct 05, 2016 1:34 pm

TraderJoeForex » Wed Oct 05, 2016 1:14 pm wrote:
I think I've sorted the session times for my CSSdiff set and after seeing the drawdown from my test yesterday I have added CSSdelta and restricted the EA to only open trades for the first 3 hours of each session.

The session times are based on Global Prime so for anyone who also wants to try this if you set your VPS or PC clock to match broker time then you won't have to adjust the session time in the setfile to your timezone.

I have set this up on a £500 demo with £5 basket close and no SL to give it some room.

Oooops... max trades per pair on the same candle should be 1 not 10 for this. Now corrected above and here: