Hi!
I just posted some results and setfiles in my thread that could actually have been posted here...
but now you can find it here:
http://www.stevehopwoodforex.com/phpBB3 ... 55#p150055
BR
Mathias
Basket Trading BALUDA'S CSS Crossovers
-
- Trader
- Posts: 146
- Joined: Thu Mar 17, 2016 9:39 pm
- Location: Sweden
- DavidB
- Trader
- Posts: 38
- Joined: Sat Dec 20, 2014 3:13 pm
- Location: The Netherlands
Basket Trading BALUDA'S CSS Crossovers
Same thing happened to me couple of years ago and cost me a small fortune. When you select a small group of top performers from a large sample of tests, chances are you selected the statistical outliers. In other words: these were outperforming by mere chance. The more rigorous your filtering is, the higher the risk of survivorship bias. Remedy: reserve 3 months to make your first selection, then allow another 2 months to test the top performers. Then start trading small lots, grdually increase when performance is as expected. The markets are patient, you?fxstrategie » Tue Nov 29, 2016 11:24 am wrote:just want to share my results... have to switch, try another strategy... this ones are not working on live account for me. (...) After some really successfull testings during the last 2 (testing over 30 different strategies over the time on 3 Servers!) month i've decided to go live (...) The last 2 Days i've got massive SL hits, (...) Till now i've not found any Scenario wich resist for more than 2 Month and Winnings > Losses an DD<20%...
https://youarenotsosmart.com/2013/05/23 ... ship-bias/
-
- Trader
- Posts: 146
- Joined: Thu Mar 17, 2016 9:39 pm
- Location: Sweden
Basket Trading BALUDA'S CSS Crossovers
Well stated!DavidB » 29 Nov 2016 22:16 wrote:Same thing happened to me couple of years ago and cost me a small fortune. When you select a small group of top performers from a large sample of tests, chances are you selected the statistical outliers. In other words: these were outperforming by mere chance. The more rigorous your filtering is, the higher the risk of survivorship bias. Remedy: reserve 3 months to make your first selection, then allow another 2 months to test the top performers. Then start trading small lots, grdually increase when performance is as expected. The markets are patient, you?fxstrategie » Tue Nov 29, 2016 11:24 am wrote:just want to share my results... have to switch, try another strategy... this ones are not working on live account for me. (...) After some really successfull testings during the last 2 (testing over 30 different strategies over the time on 3 Servers!) month i've decided to go live (...) The last 2 Days i've got massive SL hits, (...) Till now i've not found any Scenario wich resist for more than 2 Month and Winnings > Losses an DD<20%...
https://youarenotsosmart.com/2013/05/23 ... ship-bias/
- tomele
- Administrator
- Posts: 1166
- Joined: Tue May 17, 2016 3:40 pm
- Location: Germany, Forest of Odes, Defending the Limes
Basket Trading BALUDA'S CSS Crossovers
Hi.DavidB » 29 Nov 2016, 21:16 wrote:
Same thing happened to me couple of years ago and cost me a small fortune. When you select a small group of top performers from a large sample of tests, chances are you selected the statistical outliers. In other words: these were outperforming by mere chance. The more rigorous your filtering is, the higher the risk of survivorship bias. Remedy: reserve 3 months to make your first selection, then allow another 2 months to test the top performers. Then start trading small lots, grdually increase when performance is as expected. The markets are patient, you?
https://youarenotsosmart.com/2013/05/23 ... ship-bias/
Brilliant analysis and article. It sums up why we should forward test our strategies for a very long time and not abandon the "losers" to fast. Probabilities are a long term game. And as markets change, statistics do.
Happy pippin, Thomas
It ain't what you don't know that gets you into trouble.
It's what you know for sure that just ain't so. (Mark Twain)
Keep the coder going: Donate
It ain't what you don't know that gets you into trouble.
It's what you know for sure that just ain't so. (Mark Twain)
Keep the coder going: Donate
- TraderJoeForex
- Trader
- Posts: 1157
- Joined: Fri Mar 08, 2013 10:29 pm
- Location: South London
Basket Trading BALUDA'S CSS Crossovers
Time for a quick update:TraderJoeForex » Fri Nov 11, 2016 11:12 pm wrote:TraderJoeForex » Fri Nov 11, 2016 1:32 am wrote:
I will be putting this M5H1D1 (15pair) set on a small live account next week and this will be the reference point for me to try and beat with continued testing of various further options. I plan to turn the EA off every Friday during the day and manually bring any remaining positions to a close or hedge so I don't carry any risk over weekends.
Here are my live results since 14th November. I have aimed to close flat on Fridays but had to keep a couple of positions open this past weekend (EURAUD and GBPAUD) which were in heavy drawdown and preventing Friday's basket from closing. I kept these open as my analysis suggested that they would retrace this week and they did enough to close with a small loss.
Looking at the profit/loss for each of the pairs I am quite comfortable with the 15pairs I chose to run. There are no extreme losers here which is interesting.
This week so far just one new basket close then and have a feeling it may be a bit slower from here until January. I will keep this running though for the next couple of weeks and then probably call it a day until after Christmas.
For me this set does have it's frustrations seemingly entering just before a retrace and initially going into drawdown but that is the price of waiting for three timeframes to line up. Fortunately there seems to be enough momentum at certain points during the week to always bring a profit close and that is why I am not using SL or news filter with this strategy as that will be a sure way to turn this profitable strategy into an unprofitable one. The way I am managing risk is using a small account size.
I have continued testing many different approaches on demo including some non-basket approaches and have nothing that improves on this M5H1D1 set yet for a small account. I have had some success using H4 and H1 HGI as a trigger just in the direction on D1 diff and delta and am using these to add some extra trades alongside my manual D1 trading and I do look forward to trying some of the set files posted by others here that also look promising.
My time online is going to be limited in December as I have a lot going on including a house move so will aim now to report back in with anything new in January. Wishing everyone a joyful festive season and end to this year.
You do not have the required permissions to view the files attached to this post.
Last edited by TraderJoeForex on Wed Nov 30, 2016 4:47 pm, edited 1 time in total.
-
- Trader
- Posts: 118
- Joined: Mon Dec 05, 2011 2:04 pm
Basket Trading BALUDA'S CSS Crossovers
Excellent Read and recommendation!DavidB » Tue Nov 29, 2016 3:16 pm wrote:
Same thing happened to me couple of years ago and cost me a small fortune. When you select a small group of top performers from a large sample of tests, chances are you selected the statistical outliers. In other words: these were outperforming by mere chance. The more rigorous your filtering is, the higher the risk of survivorship bias. Remedy: reserve 3 months to make your first selection, then allow another 2 months to test the top performers. Then start trading small lots, grdually increase when performance is as expected. The markets are patient, you?
https://youarenotsosmart.com/2013/05/23 ... ship-bias/
thanks
-
- Trader
- Posts: 118
- Joined: Mon Dec 05, 2011 2:04 pm
Basket Trading BALUDA'S CSS Crossovers
All,
FYI....I just got this message from my broker concerning Margins.
Please note margin requirements will be adjusted after 16:00 GMT on December 2nd, 2016 due to the Italian Referendum. Please ensure that you have enough available margin to support your open positions. If you have any questions or concerns, please contact an FXCM representative.
Thanks
Lionsden
FYI....I just got this message from my broker concerning Margins.
Please note margin requirements will be adjusted after 16:00 GMT on December 2nd, 2016 due to the Italian Referendum. Please ensure that you have enough available margin to support your open positions. If you have any questions or concerns, please contact an FXCM representative.
Thanks
Lionsden
- Gertje
- Trader
- Posts: 1936
- Joined: Mon Dec 12, 2011 1:17 pm
- Location: Middle of the Netherlands
Basket Trading BALUDA'S CSS Crossovers
Just to add to the confusion, here are my results for the month of November, trading a live cent account.
The indi at the right top says 37-ish % profit, this is from the current balance.
It compares to almost 60% from the starting balance.
I used the H4/H1 set, and added the D1 today to see if it counters the high DD that occures now and then.
.
The indi at the right top says 37-ish % profit, this is from the current balance.
It compares to almost 60% from the starting balance.
I used the H4/H1 set, and added the D1 today to see if it counters the high DD that occures now and then.
.
You do not have the required permissions to view the files attached to this post.
- Gertje
- Trader
- Posts: 1936
- Joined: Mon Dec 12, 2011 1:17 pm
- Location: Middle of the Netherlands
Basket Trading BALUDA'S CSS Crossovers
Just to impress; I traded appr. 1.800 full lots during the month of November.....
-
- Trader
- Posts: 644
- Joined: Thu Nov 17, 2011 12:53 am
Basket Trading BALUDA'S CSS Crossovers
I can also report good results with this set, (zennor's H1/H4) seems a consistent performer, but like you said sometimes DD is high.
please keep us in the loop regarding the D1 filter
trader
modification
please keep us in the loop regarding the D1 filter
trader
modification
Gertje » Wed Nov 30, 2016 9:10 pm wrote:Just to add to the confusion, here are my results for the month of November, trading a live cent account.
The indi at the right top says 37-ish % profit, this is from the current balance.
It compares to almost 60% from the starting balance.
I used the H4/H1 set, and added the D1 today to see if it counters the high DD that occures now and then.
.