yes that sound like it would work- i would be keen to see the results on that as some day we have these big moves
I think you should backtest every 2 weeks with ideal stops and targets as APril, May, and June have been very good for this yet Jan, Feb, March were close to break even. I canāt really see whatās changed.
The current conditions are more important obviously so Iām not looking into it that much but it would be nice to know. ATR doesnāt seem wildly different but market was recovering from a massive downtrend at the time.
Are you factoring the spread into your calculations?
That may be the problem?
Hi,
Yes I already factor in the spread. I donāt think it has anything to do with spread. It is simple mathematics. Lets say I have ļæ½10000 starting capital and I set my intial pip value to be 1/1000 of that so thatās ļæ½1 a pip. If I win 30 pips on day 1 I will have ļæ½10030. My pip value then increases accordingly to ļæ½1.03. Now if I lose 30 pips on day 2 I am down to ļæ½9999.10.
Mickey.
I know you said any time frameā¦
Which time frame is everyone else looking at?
How long are you in the trade on average?
Is there anyway to target the price on the rebound of the initial movement?
The rebound appears to be a better target as far as maximum pips for the day.
Just some questions Iām demoing this starting today even though it seems everyone else already has!:rolleyes:
What time is 5am for me if Iām in Colorado (Mountain Time)
personal choice i look at the 30 min time frame but you can look at any upto the the 1 hour time frame to give you the overnoght highs and lows
trade lenght again depends on your target but anyway from 1 to 5 mins 10 max
you need to take away 8 hours ithink please check that for your time zone
and as for the rebound, when i think of a stratergy other than candle stick trading to confirm the turn you will be the first to know.lol
Ok so letās say that we use the HH and the LL as SR Lines.
Please look at the previously posted picture. Iām at work so I cannot edit this picture but itās pretty easy to see without and editing.
I hope that helps you.
Clint
Sorry I did not clarify on that. You can get in on the initial movement Clint and take that trade as specified by the rules.
However if you draw a line through the HH and the LL and get a positive retest of that line as a Support or Resistance ā¦ FROM WHAT I CAN TELL, this seems like a pretty good signal for a reversal, atleast pick up another 30 pipsā¦ I do think it would have to a be a manual trade
there is not always a retracment infact if you look over the last few weeks once the high or low has been broken its genrally stayed above the BO line
i was using a stratergy during jan-feb to cacth the reversals as it was entering back into the daily range but i think that system is now dead untill another day (Jan-Feb price was settling more inbetwwen the lines than above and i think thats why sanmigal has problems getting this stratergy to work back then because of the price movement were very diffrent)
Ok best of luck with that, you will probably find that changing
pip value every week, month, or at a given monatory value (every
ļæ½1000 profit?) would be more beneficial.
Any ideas as to what might have been different? I know it was recoving from a fall then but the ATR was fairly similar wasnāt it?
Not to put a negative on it but we also need to have a point in this strategy where if it isnāt working we come back and look at it. Youāll notice on the results the consecutive wins and consecutive losses for your relevant amount (average), if you get more consecutive losses than stated there then it might be an early sign.
Results
yes im always looking for any slight changes that may indicate it going back to how Janās and Febs price action was- what i would say is you will get the indication by the breakout on this strategy not being as strong also the will spend more time in the overnight range boucing from highs to lows, break out will still acor but not like they are at the moment the will move 10 pips above then come back into the range- that then is your opotunity to reverse what we are doing now and take advantage of the draw backs- the point is these 2 line are my freind and which ever way you want to do it i will always use these lines as a base.
to be honest my trailing stops actrully cam from the reverse stratergy for JAN FEB and have have kept them because they worked well back then, however in hensight i could have changed the trailing stop to my advantage or even get rid of them and apply a more hands on exit stratergy.
Hello, everybody,
I must have a comment on the bactesting effort and parameter searching. I have done a plenty of backtesting of various systems myself before - automated, that is, allowing me to easily and quickly to test a big number of parameter combinations, once I have coded the calculations needed. There is one typical trap that I have noticed people usually fall in, me included. The process goes like this:
[ol]
[li]You invent or find a promising system.[/li][li]You identify a set of parameters that impact the system, including the ideal instruments and timeframes.[/li][li]You perform backtests that lead you thinking about some interesting tunings of the system, along with a couple of new parameters.[/li][li]So you improve your system and perform more backtests, getting still better results.[/li][li]You select the best of the tests and iterate steps 3-5 still a bit more. (Note that you are all the time typically using the same set of history data.)[/li][li]Finally you have a brilliantly backtested system and are ready to take it into use on real data. The system starts failing from the beginning or at least lags a lot from what the backtests have kept on promising.[/li][li]Alternatively, you take a different period of history data, and to your great surprise the performance is far from what you just tested.[/li][/ol]At least for me it was very eye-opening to realise the explanation why the failing happens. I had at maximum some 10 parameters for one system, and invented a clever iteration mechanism where I optimised one parameter after another, until I had a perfect set of optimised parameters. The problem with this approach is that you optimise the parameters for your historical data, and nothing else. There is no guarantee that the future data will resemble the historical data even remotely. Sure, there is some probability for that, but no guarantee. The more you have optimisation, the subtler the system is to market changes. To my dismay, the same problem can manifest itself as easily when you only have a couple of parameters.
To prove my point, try this: a) squeeze the last pip out of your system for a selected period of one month, b) try the same parameters for the next month, c) optimise the parameters for that second month. Youāll realise that: i) the best optimised parameters differ a lot for the two months, and ii) when the same parameters are used, the results differ a lot.
I nowadays think that you should not spend too much time in optimising the parameters. Altough SanMiguel got considerably better results for the other variations and parameterisations of the system than the original one, I am afraid that these extremes are those that most easily fall apart when the market changes a bit. I personally like the simplicity and predictability of the original approach.
Also the position sizes can have importance. If you have a system that promises to yield 1000 pips within a period as opposed to another system promising 5000 pips, at first glance it seems that the second system is five times better. But if your risk management allows you to invest, say, 3 times more in the first system, the difference is not that big any more. And, the 5000-pip system could vary a lot in performance while the 1000-pip could yield pretty consistent results.
SanMiguel, I apologise if you read the above as critique against your backtesting. That is not my purpose at all - I rather see it as pretty valuable material here. I would just like to point out some caveats that one may run at with backtesting, and raise some thoughts. And I wouldnāt be happier here if somebody proves that I am wrong!
i dont think anybody here is going to run of and use this live based on the back testing results,
i have traded this sytem for nion six months with no ea, changes in the market is a ongoing thing so when i released this startergy it was not for the sole purpose of creating a EA, more maximising profits and tweaking, backtesting is good to give you an idea but i have to be honest i dont always trust the result 100% for the above reasons you have mentioned however if we can get a programme that has a base who know were we can take it, anyway this is something extra to what i expected when i wrote this thread, personnally i find it kind of exciting, but dont get a false sence that this thread is just about the EA, its not if i wanted a EA i would have payed somebody to do one and forgot about discussing ideas,
Itās critique but I donāt mind that at all otherwise I wouldnāt be here 
Of course you are correct but when looking at backtesting we are looking at the fundamentals of the system. Of course an EA is rigidā¦I mean if you are targeting 45 pips but price stalls at 44 the EA will not do anything yet manually you can close it out. You can also see what is happening to price on a daily basis. But I do believe backtesting is important. On every system you should backtest trades taken and see if your RR is still valid or if it should be reduced/increased or if something is not working out.
Time will tell for these systems I think
Today would have been a nice 2fer and I actually thought about taking that 2nd trade. Oh well 
Not much movement today is there? NFEC report?
Iāve moved the short a little lower beyond the recent S&R lines circled though am a bit wary of the news.
Yesterday when you didnāt take the second tradeā¦ i did and got 134 pips on the downhill slide. today I got 23 going short just before the strategy said I should.
so I didnāt follow the strategy exactly as prescribed, but tend to try and follow a trend when I see one. a lot of that is dumb luck though cause Iām still a newbie
got 20 pips on the 29 and 27 pips today. I am not following the strategy 100% but using S&R lines from 4h go down to 30m make my lines from 18 to 5, but depending of the trend I might adjust the lines up or down.
But be careful, I am a newbie 
Have a good day and thanks for all your contributions I am learning a lot
terachon
Hi SanMiguel,
on your pic, did you paint the nighttime in or is it your EA doing it?