PipHunter Improvement suggestions

[AlexPali]

I'm running Pip-Hunter on two FXDD and FXCM-UK accounts, with 7 pairs, 0.2 lot size
The FXCM-UK account was started with $10K portfolio and increased to $18K with manual trading Pip-Hunter’s way. Thus the final output does not reflecting performance of Pip-Hunter EA but the robot traded more actively here than with the FXDD counterpart.

The two brokers are using 2 hours different server times. The above situation doesn’t affect systems trading systems based lower, up to H1 TF because the candles are starting and ending exactly at the same physical time despite the server’s time differences.
Yet it matters in the case of H4 and higher (D1 in case of Pip-Hunter) time frame because the starting/ending time is overlapping and each hour of difference could deliver a big gap of the upper/lower value readings of the D1 candle for each broker. As a result, Pip-Hunter is spotting entry set-ups shifted hours or even days.

For example, I started FXDD and FXCM-UK platforms on Sunday, June 7th, 2009; 21:00 GMT running on the same settings or seven pairs.
Due to the two hours differences of the starting/ending time of the daily candles, Pip-Hunter spotted entry set-ups and engaged trades on the first trading day on the FXCM-UK platform, while on the FXDD platform trades were executed the next day. As well, the executed trades on the two platforms were not matching the same currency pairs and of course delivered different results. The FXCM-UK secured almost $2’000 of profit while FXDD executed just one TP1 (50% of the 0.2 size trade) securing $177 only. As we see, the differences in candle readings happened to be favorable for FXCM this time. However, I don’t think that this is any solid clue that FXCM is any better than FXDD, simply because the better output occur due to favorable market trend for the 2 hour postponed D1 TF.

During previous and the last trading week in particular, I noticed that at least for 3 days in a row, both accounts were gaining significant floating profit ranging from $300 up to $1’200 before going into various $-300, $-600, $-1’500 and above $-2’300 DDs.
It is common with this strategy the situation when the trade is approaching the TP1 (1st take profit) level and reverses back to the starting (or even below) level and few hours (or the next day) it again goes the proper way. Due to the above happenings, the original trading strategy is using large 100 – 175 pips S/L levels to stay in trade hoping that for positive reversal.
I believe that we could lower the risk by preventing the trades from going into such a deep DD while collecting smaller gains during such situations. In most cases, the negatively pair recovers and continue following the direction previously forecasted by the strategy. As soon as that occurs and the previously determined entry set-up conditions are still valid, the robot can reopen the trade and maintain it accordingly. We would not only lower our lose risks but harvesting bonus pips when enhancing Pip-Hunter with such capability.

Here is my modification vision:

Priority No 1:
As soon as an open trade achieves adequate level of floating profit, the robot should modify S/L values to brake-even level.

Priority No 2:
Assuming, that the trade positively distanced itself from the break-even S/L point, the robot keeps moving the S/L to match/follow the gaining altitude.

Since the pair prices are fluctuating with variable ranges due to different market volume activity at particular times, determining the optimal space between S/L level and the current price would be very important. Properly managing the above task will prevent from going into deep DD as well as from premature (unnecessary) termination of the trade that happen to be on a winning path.

For the reason that the optimal S/L against current price stage stretch is uneven during different trading-times/market-conditions as well as it varies for individual currency pair, conceivably the device controlling this duty could be aided by indicators such as Bluto’s Chop_Whacker, Chopemeter or FerruFx's ProbabilityMeter etc or by other potential solutions. After such a controller is synchronized with supportive indicators, the time/pair sensitive variables will be optimized with series of back and forward testing.

I understand that the above request is not an easy task because it brings another level of complexity to the code, however I believe it is as important as the main strategy that the robot is based on.
I hope this is achievable and worth to try.

[FreddyFX]

Alex, I also noticed that the TP was not activated and it lost all its profit.
Maybe a little glitch in the programing, but at least it was taking trades.

Problem is I cannot watch it all the time, as I am trading live again.
For that reason I drop you a message on Skype when I notice something, as I know you are right on top if this baby all the time.

To be clear:
I have PH running on ODL, while I trade with IBFX, so don't watch the ODL screen all the time.

[FreddyFX]

Just downloaded and installed version 2.
Anxious to see what will get traded this week, running EU and GU.

[xux99]

I shall add a breakeven feature. The stop loss already moves following the LL3/HH3 level. I thik the order should be closed not by stop loss but preferably by a possible top/bottom. iMAX cross or ZZ could be used.

Priority No 1:
As soon as an open trade achieves adequate level of floating profit, the robot should modify S/L values to brake-even level.

Priority No 2:
Assuming, that the trade positively distanced itself from the break-even S/L point, the robot keeps moving the S/L to match/follow the gaining altitude.

For the reason that the optimal S/L against current price stage stretch is uneven during different trading-times/market-conditions as well as it varies for individual currency pair, conceivably the device controlling this duty could be aided by indicators such as Bluto’s Chop_Whacker, Chopemeter or FerruFx's ProbabilityMeter etc or by other potential solutions.