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

Rinse 'n Repeat
https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?t=4693
Page 30 of 56
Author:  SteveHopwood [ Wed May 11, 2016 6:17 pm ]
Post subject:  Rinse 'n Repeat

New RnR toy folks. We all know what a clever coder Rene is, so it should be a blast. http://www.stevehopwoodforex.com/phpBB3 ... 15#p140215

:xm:
Author:  Bruster400 [ Wed May 11, 2016 7:36 pm ]
Post subject:  Rinse 'n Repeat

krzysiekqq2 » Wed May 11, 2016 2:48 pm wrote: I switched to Recovery Mode yesterday, but I see that bot is still opening position as before. Is this normal behavior?
Have you checked vs the settings you've used? What did you tell it to do in recovery mode?
Author:  Bruster400 [ Wed May 11, 2016 8:29 pm ]
Post subject:  Rinse 'n Repeat

Alicat » Wed May 11, 2016 4:31 pm wrote:Just one thought, I have set MinPipProfitOnClosure = 3pips, but think it would be beneficial to be give this 2 different values, a higher one when in normal mode, but a smaller one, even 0, when in recovery mode.
Version 1b in post 1 with this idea added. MinPipProfitOnClosure is now replaced with OffsetPipProfitStandard and OffsetPipProfitRecovery. I've checked it works but that doesn't mean it helps!!
Author:  Bruster400 [ Wed May 11, 2016 9:08 pm ]
Post subject:  Rinse 'n Repeat

Just over a week on this test and some interesting results for those using recovery mode - it looks like it works only when combined with SingleSidedComplexOffset set to true.

I've tested 4 settings:

1: OOTB - (recovery ON, singlesidedcomplexoffset ON)
2: OOTB but with singlesidedcomplexoffset OFF
3: OOTB but with recovery OFF
4: OOTB with recovery AND singlesidedcomplexoffset OFF

the last setting being basically the standard original bot.

The results are below. Although 4 is the winner at this point, 1 has maintained slightly better control on DD. The relative positions vary slightly with differing market conditions but these two are clearly pulling away from the others. Test 3, with recovery off, is basically the original bot with singlesidedcomplexoffset ON and this is clearly a lot worse that with it OFF (4).

My other observation from these results is that the bot has made around 1% return on the balance that would be needed to trade all these 4 scenarios (0.01/2000) and yet the risk was >20%. If the blowout scenario doesn't happen then that's about a 50% return in year (70% if we assume just setting 4) but if it does happen..... you need to hope that you've banked enough profit by then to pay for it!
Author:  TraderJoeForex [ Wed May 11, 2016 10:00 pm ]
Post subject:  Rinse 'n Repeat

Bruster400 » Wed May 11, 2016 9:08 pm wrote:Just over a week on this test and some interesting results for those using recovery mode - it looks like it works only when combined with SingleSidedComplexOffset set to true.

I've tested 4 settings:

1: OOTB - (recovery ON, singlesidedcomplexoffset ON)
2: OOTB but with singlesidedcomplexoffset OFF
3: OOTB but with recovery OFF
4: OOTB with recovery AND singlesidedcomplexoffset OFF

the last setting being basically the standard original bot.

The results are below. Although 4 is the winner at this point, 1 has maintained slightly better control on DD. The relative positions vary slightly with differing market conditions but these two are clearly pulling away from the others. Test 3, with recovery off, is basically the original bot with singlesidedcomplexoffset ON and this is clearly a lot worse that with it OFF (4).

My other observation from these results is that the bot has made around 1% return on the balance that would be needed to trade all these 4 scenarios (0.01/2000) and yet the risk was >20%. If the blowout scenario doesn't happen then that's about a 50% return in year (70% if we assume just setting 4) but if it does happen..... you need to hope that you've banked enough profit by then to pay for it!
... nice summary of results Bruster. :party:

Now a question is can we find a way of identifying the characteristics of a run up to a potential blow out scenario and insulating ourselves from it.. through hedging during major news and weekends and/or if ADX is greater than ... and/or perhaps we could try and identify a trading range within which to let RnR do it's thing and outside of which to keep it on a leash.

For example in the screenshot below once there is a breakout either side we have far more risk with this EA than inside the range so why not just trade inside. Upon a breakout the overall position could be hedged until a new range is defined and then we close that hedge neutrally and let RnR carry on harvesting pips until the next breakout.

I am using the horizontal "breakout lines" function of the TDMTL indi to mark out a range...

...thoughts?
Author:  krzysiekqq2 [ Thu May 12, 2016 8:13 am ]
Post subject:  Rinse 'n Repeat

Have you checked vs the settings you've used? What did you tell it to do in recovery mode?
Hi Bruster,
Both were set to true. Thanks for explanation.
Krzysztof
Author:  TraderJoeForex [ Thu May 12, 2016 9:17 am ]
Post subject:  Rinse 'n Repeat

TraderJoeForex » Wed May 11, 2016 10:00 pm wrote:
Upon a breakout the overall position could be hedged until a new range is defined and then we close that hedge neutrally and let RnR carry on harvesting pips until the next breakout.
Here is a practical example of what I mean using GJ. Currently RnR has 4 buy trades and 13 sell trades and is trading within a range on M15.

If price breaks lower than the red line at the bottom of the current range we are happy since we have more sell trades than buy so this would be in our favor and we continue to let RnR trade the move down.

If price breaks above the blue line at the top and moves up quickly it will get messy. And depending on the strength of the move it could turn into a leaky toilet. So what I am suggesting is that once a break above is confirmed (using TDMTL breakout qualifier rules) a trade is placed to hedge the overall position. Currently that would be 0.09 lots buy for example since RnR has 13 sells and only 4 buys. RnR can then place no more trades until the hedge is taken off.

If the breakout is true and price continues up this hedge is held until the top of the new range is marked out by the blue line above price (showing a highest high has occurred with at least one candle either side). At this point the hedge can be closed and used to offset the biggest losers and bank our minimum pips. RnR then continues trading the new range as before.

Now the risk is that the breakout is false and after price has confirmed a breakout the next candle breaks lower back into the range. In this case the hedge position may be closed at a loss unless there are sufficient buys to offset it. That is ok - this occasional loss would be the cost of this protection.

I have found TDMTL breakout qualifier rules quite robust in my trading. They eliminate a lot of false breakouts. A good summary of the rules are explained here: http://www.marketinout.com/stock-screen ... Qualifiers

This approach can be applied manually by turning RnR off while a hedge is placed and back on after the hedge has been closed on whatever timeframe you like to trade. The higher the timeframe used the higher the potential drawdown that RnR can accrew but obviously the less often a hedge is needed.

Adding what is essentially a breakout trading strategy to what is currently the most effective automated range trader I have ever seen could just be a worthwhile paring. ....
Author:  TraderJoeForex [ Thu May 12, 2016 9:45 am ]
Post subject:  Rinse 'n Repeat

.... just to update the above example did breakout so I have temporarily turned off RnR and hedged as below:
Author:  TraderJoeForex [ Thu May 12, 2016 10:34 am ]
Post subject:  Rinse 'n Repeat

TraderJoeForex » Thu May 12, 2016 9:45 am wrote:.... just to update the above example did breakout so I have temporarily turned off RnR and hedged as below:
Once a top of new range is confirmed, the hedge is closed banking a nice profit which could be used to offset some biggest losers and RnR turned back on.
Author:  drswing [ Thu May 12, 2016 1:44 pm ]
Post subject:  Rinse 'n Repeat

Latest update.
2016-05-12_RnR.png
It is very obvious from my general testing that this little bugger loves ranges but hates breakouts! And it works better in ranges with Offset and Recovery OFF. Draw-down is better handled with Offset and Recovery ON, but a sustained move on a volatile pair like GJ will burn you account! It is also much easier to handle the DD on demo than it will be on a live account, wondering if you'll get that retrace ;) .
All times are UTC Page 30 of 56