stevehopwoodforex.com
https://www.stevehopwoodforex.com/phpBB3/
Print view

Gday: a D1 Stochastic OB/OS trading robot
https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?t=168		 Page 57 of 71
Author:  fxozgirl [ Fri Sep 14, 2012 12:31 pm ]
Post subject:  Re: Gday: a D1 Stochastic OB/OS trading robot
fok wrote:
fxozgirl wrote:Attaching my set file for anyone who is interested :D
I am trading with Cowboy Go Markets Australia (+2 GMT - although from next Monday they are changing to +3 GMT)

Just took a quick look at my charts, possible trades for this morning are:

sell - chfjpy
buy - gbpchf
Did take same pairs today morning, CHFJPY doesn't make me happy right now ...

No chfjpy is not looking real flash right now :(
Author:  Mike [ Fri Sep 14, 2012 2:53 pm ]
Post subject:  Re: Gday: a D1 Stochastic OB/OS trading robot
fxozgirl wrote:Attaching my set file for anyone who is interested :D
...
Hi fxozgirl. Thanks. I did not get any trades for some weeks, so something must be wrong with my settings. Just an observation and a question regarding yours, which I will use:
You mention, you use BE 50/2 BUT you have set "BreakEven false".
How is the bot's behaviour changed by setting "VolatilityLookBackCandles 0" (in order not to have the bot calculate sl/tp)?
Author:  garyfritz [ Fri Sep 14, 2012 6:34 pm ]
Post subject:  Re: Gday: a D1 Stochastic OB/OS trading robot
Shelley, I'm trying to understand the R:R and return on this strategy.

When you say you've banked 863 pips, does that refer to the pips earned by all 3 positions? E.g. if a winning trade that TP'd at 50 and 100, and let's say a JS exit at 200 pips, would that be a total of 350 pips banked?

Do you still run with a 200 pip stop? Which, since you have 3 positions, is actually 600 pips risk?

Let's say you trust this strategy and want to risk 2% of your account on each trade. If you traded 0.1 lot, the 600 pip risk would be $600, meaning you'd need a $600/2% = $30k account. You'd have earned $863 in 3 months, risking $600 on each trade?? That's $863/$30k = 2.87% growth in 3 months. That seems like a lot of risk for a small return.

I thought I had a fairly accurate version of Gday implemented in Tradestation. But when I dug it out and tested it, it's been losing pretty steadily. Only AUDCHF and EURAUD showed a profit in the last year. I must have something wrong. Could you PM me the list of trades you've taken since June (or as many as possible) so I can see where my problem is? And summarize any changes you've made to your original basic rules? I'm using stoch length of 8, buy when it crosses over 20 / sell when it crosses under 80, go to BE at +50, TP1 at +50, TP2 at +100, 50-pip jumping stops, SL at -200.

Thanks!
Author:  dsugums [ Sat Sep 15, 2012 5:21 am ]
Post subject:  Re: Gday: a D1 Stochastic OB/OS trading robot
This strat will work great if the sl is kept at the previous low/high. I think 200pip sl is toio ambitious. If a trade goes against more than 50pips, the trade is screwed. For GCHF, the previous low sl was about 45pips, which would saved the huge losses.

CHJ similar, sl would been only about 30pips.
Author:  Jonathan2FI [ Sat Sep 15, 2012 1:25 pm ]
Post subject:  Re: Gday: a D1 Stochastic OB/OS trading robot
Fxozgirl,
quick question - could Gday be used for 240TF & Stoch with daily trend filter and OBOS trades only in direction of trend - in other words retrenchments back in direction of trend using all same logic e.g signal X- have read first 30 pg of posts and tried backtesting 240 with AU with RSi20 on D TF - but can't get a trade with Axitrader cleaned data so have some doubts or am I being a WP - please comment to stop further banging head on wall...ooch

One further avenue to explore/develop at D or 240 TF the stoch level OBOS needs ultimately to be sensitive to strength of trend. Strong trend e.g. using NBs TMA slope >0.8 would require a lower OS signal say 30 compared to a TMA slope of 0.4 - just a thought I have used this successfully in manual trading looking for "daylight" between stoch and TMA slope.

Jonathan
Author:  Jonathan2FI [ Sat Sep 15, 2012 2:11 pm ]
Post subject:  Re: Gday: a D1 Stochastic OB/OS trading robot
From testing Gday and optimising - not curve fitting!
Stoch 6,5,2, OS 22 OB 86 SL 300 JSL 50 (orig based on FXozgirl recent set but opti over 2 years)
Results Net profit 20% RDD8% 15 100% profit trades. Arrow straight upward equity curve. results files attached. Jonathan ..... could do with high freq of trades to be statistically significant!
Still a nice bot.
Jonathan
Author:  tomazzo [ Sun Sep 16, 2012 1:34 pm ]
Post subject:  Re: Gday: a D1 Stochastic OB/OS trading robot
What is the latest version of the EA? It's the one in the post 1
Author:  fxozgirl [ Sun Sep 16, 2012 8:58 pm ]
Post subject:  Re: Gday: a D1 Stochastic OB/OS trading robot
Mike wrote:
fxozgirl wrote:Attaching my set file for anyone who is interested :D
...
Hi fxozgirl. Thanks. I did not get any trades for some weeks, so something must be wrong with my settings. Just an observation and a question regarding yours, which I will use:
You mention, you use BE 50/2 BUT you have set "BreakEven false".
How is the bot's behaviour changed by setting "VolatilityLookBackCandles 0" (in order not to have the bot calculate sl/tp)?

Hi Mike, the exit strategy is hard coded into the EA at 50/2 where it will close the first part of the trade.

Changing the VolatilityLookBackCandles makes the TP dynamic based on the number of candles entered in lookback.
Author:  fxozgirl [ Sun Sep 16, 2012 9:39 pm ]
Post subject:  Re: Gday: a D1 Stochastic OB/OS trading robot
garyfritz wrote:Shelley, I'm trying to understand the R:R and return on this strategy.

When you say you've banked 863 pips, does that refer to the pips earned by all 3 positions? E.g. if a winning trade that TP'd at 50 and 100, and let's say a JS exit at 200 pips, would that be a total of 350 pips banked?

Do you still run with a 200 pip stop? Which, since you have 3 positions, is actually 600 pips risk?

Let's say you trust this strategy and want to risk 2% of your account on each trade. If you traded 0.1 lot, the 600 pip risk would be $600, meaning you'd need a $600/2% = $30k account. You'd have earned $863 in 3 months, risking $600 on each trade?? That's $863/$30k = 2.87% growth in 3 months. That seems like a lot of risk for a small return.

I thought I had a fairly accurate version of Gday implemented in Tradestation. But when I dug it out and tested it, it's been losing pretty steadily. Only AUDCHF and EURAUD showed a profit in the last year. I must have something wrong. Could you PM me the list of trades you've taken since June (or as many as possible) so I can see where my problem is? And summarize any changes you've made to your original basic rules? I'm using stoch length of 8, buy when it crosses over 20 / sell when it crosses under 80, go to BE at +50, TP1 at +50, TP2 at +100, 50-pip jumping stops, SL at -200.

Thanks!

Yes, Gary you are right in your R:R comments...if the trade is closed at -200 SL it is actually -600 because you have 3 times your position, however this is an extreme, keep your shirt SL, whereas most trades that don't reach BE will close at the cross of the stoch. I'm not sure how you would work out the R:R on that basis, but in all my testing of this strategy (nearly 1 year) I have only had 1 trade go to a full SL (disclaimer - I have not traded non stop during this time)

As for the 863 pips, I was actually being a bit lazy and just taking the results from myfxbook these were the pips if only 1 position (they don't multiply out ...it is actually much more due to the multiple positions taken as well...1111 pips.


This morning though, both my trades have closed at the cross of the stoch with drawdowns...
chfjpy -177.2 pips (-531) and gbpchf -85 pips (-255) so total pips now banked = +325

With that being said, both of these trades were very marginal and is something I am looking to tweak to improve entries.

As for your Tradestation results...I am currently using stoch of 7,2,2 all other parameters are as you have listed. Can you please run your test with the updated stoch settings?

Cheers

Shelley
Author:  fxozgirl [ Sun Sep 16, 2012 9:41 pm ]
Post subject:  Re: Gday: a D1 Stochastic OB/OS trading robot
dsugums wrote:This strat will work great if the sl is kept at the previous low/high. I think 200pip sl is toio ambitious. If a trade goes against more than 50pips, the trade is screwed. For GCHF, the previous low sl was about 45pips, which would saved the huge losses.

CHJ similar, sl would been only about 30pips.
Most trades never go anywhere near the 200 SL (this is a shirt protection stop), if trades don't make it to BE, they are closed at the cross of the stoch, which in most cases is much lower than 200 pips.
All times are UTC Page 57 of 71

Powered by phpBB® Forum Software © phpBB Limited