All orders
Rules
Job finished
Specification
This is something I have traded for a while, but only with human intervention, so I am keen to make it into an automated EA. Although I am making some money on it I think it can be done much better, and I need your help to do it!
The trading pair is USDJPY, I currently use 1 min and 5 min bars for doing trades. The enclosed chart shows my indicator (which is generally going from -800 to 800). When the indicator changes direction – that is when I am looking to do a trade – ie a signal. The actual direction (up or down) of the indicator is not what I am looking at, just that it is changing direction.
Currently what I am looking for is to go against the trend
if I have a signal. I do not take all the signals, if there is a trend (does
not have to be big) I will counter the trend and go in the opposite direction.
I generally either take a 3-8 pip profit depending on volatility or trail it.
If there is no short term trend I do not take the trade, or if 1Min and 5min
are very conflicting to each other. This may not be the best way to trade the
signals so if one is developing a system for this it may be better to take
signals when I have not (in a non trending environment for example), or take
the trade with the trend in certain scenarios, so please test different ones.
In my experience in the signals I do trade are about 70-80%
correct that there is a correction the other way. However sometimes this
correction is very small, especially in a very strong trend. So it is sometimes
challenging to get out with a profit even if one is (very short term) correction.
Sometimes as well the indicator signals large corrections, and I am only
getting 6-7 pips.
I am sure that this can mechanically be done better than what I do today. I
usually trade the business GMT hours, but I see that I do loose some profit while
not trading 24/7.
Martingale/Grid/Averaging down: This would be great if that is an option in the EA, but the strategy has to make money on a single lot basis primarily. If you implement Martingale/Grid/Average down please send both a chart/performance report for the single lot and for the other.
Signal generation: Mixing this with other indicator signals made by you is ok, as is of course filters – but one of the main trading triggers must be my indicator, I do not want to buy a system based on something completely different.
Optimization: Some optimization is of course ok but there can not be a large number of variables that can be optimized ie purely data optimization. The indicator logic is solid on its own and therefore limited optimization should be needed with any additional indicators or filters.
Hidden TP/SL: Preferable.
Inputs that is needed in the EA: Dependent on the filters and other indicators you put in but one needs to be there – time offset. The indicator is generated in UTC, so there needs to be a function to offset the time with the “local” MT4 time.
Other tips / ideas: Ideas for filters/indicators could be (but not limited to): Trend strength, time of day, offset (some times the signals works better on a lag/lead, maybe some way to adjust for that), volatility, correlation with other pairs (although the indicator is using other currency data in its calculation), limit entry, stop entry, trailing stops, filtering the indicator itself etc etc
I am looking to pay USD 700 for a system (with source) that shows a good return profile, a good corr coefficient on the equity curve and has acceptable risk. If you are interested in doing this please apply for the job. I will then in the discussion chat send you a link to download the historical indicator data and a tool to show it in MT4 (The data is generated in a different software real-time). Then when you have coded an EA please send the equity curve and results page back in the discussion chat. I will then choose the developer that I will purchase the EA from. I may want to purchase EAs from several developers, in that case I will post a new project specific for that developer to accept (with the same price of 700 USD). I will not appoint any developer without seeing the equity chart and performance report (with a high data quality) and where I like the chart/report. Use 0.01 JPY for costs (ie 10 in spread) and high data resolution while backtesting.
Good luck I look forward to get the applications.
