A "different" crossover system - 600 pips on Friday alone in EUR and GBP

How does 200 pips on the EUR and 400pips on GBP sound, all in one day? The system described below would've given that on Friday, as you can see from the charts.

Well, read on...

I've been playing with the following system, and need some help and/or ideas to further develop it.

Put a 5 period Exponential MA of the Open, and a 5EMA of the Close on a chart, on any timeframe. The above results were given from the 1hr chart.

Buy when the MA of the Close crosses above the MA of the Open, and vice versa for reversing the position to Short. This is theoretically an "always in the market" system.

The system takes advantage of the fact that when the market is in an uptrend, the closes will be above the Opens in a given timeframe. And vice versa.

Now, the system catches trends VERY WELL, both beginning and ending, as you can see below. The problem, and the reason I'm writing this post, is that it gets whipsawed (as any trendfollowing system) on sideways markets.

So I'd like your input on how we can modify/improve this system to lose less money in sideways days. Things I've thought about include, among others:

1) Introduce a longer MA, say 20EMA, and take only long trades when price is above the 20EMA, short when price below 20EMA.

2) Introduce a Higher High or Lower Low filter, entering long after a signal is given ONLY after the high of the previous candle is taken out, and vice versa.

3) Introduce an ATR or ADX filter, taking trades only when these are above a certain level - thus in theory the market is trending. I am not sure how efficient this is...

4) Introduce a price filter: wait for, say, 20 pips after a cross to take the trade. This is my favorite and probably quite efficient but would "leave money on the table", sort of speak.


Waiting to hear your thoughts. Hope we can develop this into a highly profitable system everyone can benefit from!

Regards,

Bishop

Hi Bishop,

I believe BG system of entering 20 pips after the cross is already some kind of filter .
You cannot make all the pips in a move anyway and escape the bad moves completely .
Do not be too greedy !

One says that if you get a third of a move its good performance..........

On the other hand I think in order to maximise a large move like Friday ride the trend all the way until the cross happens in the other direction
Enter 3 lots for example with different objectives :

1st to take 20 pips
2nd trailed 50% of the highest or low seen at anytime
3rd on a cross in opposite direction

Its only a suggestion , as there are so many possibilities
Exiting trades is the most difficult part of the game

Chris Johnson

Chris Johnson,

Thank you for your thoughts, I'll have a look at your recommendations today.

I agree that BG's idea of adding +20 pips to any cross to avoid sideways trends is quite efficient, if you trade it on the 30m chart.

What if you also want to take a few trades during the day from the 5m or 10m chart though? Or say re-entry trades, where you were stopped out at b/e and then the system gives you another signal? What filter would you use then?

Still looking for some more thoughts on filters to use for trendfollowing systems to avoid whipsaws in sideways markets...

Thank you,

Bishop

Introducing filters in systems like this one is very complicated issue. They can turn your system in a totally different one. For example if you use ADX filter you can see your system turning into a simple ADX crosses one… and I’m pretty sure that would be the case in majority(if not all) of the trades produced by such system.
Did you backtest the system? What are the results?

Cheers…

Seanic

Try a slightly longer time frame on the OPEN EMA - use 8 period - stops most of the whipsaws. What do you think?

TRENDY

TRENDY

You have a point. Maybe it's best to keep this a "pure" system, rather than introduce further variables as the ADX.

Haven't backtested yet, will see if I can find someone to help me program this into Metatrader, so that we can get some reliable results.

Seanic

How'd you think of using two separate values for the MA?? VERY INTERESTING OBSERVATION!! Indeed, putting a longer MA for the open smooths things out.

(also interesting is the fact that if we add a shorter MA for the open, it messes up the whole system.... will need to ponder on that some more)

Chris Johnson, any more thoughts?

Keep them coming guys, already we're making progress here! Check out how the system handled this morning's move, with Toccata's MA8 for the Open and MA5 for the Close. Around 200 pips on EUR and GBP so far, and still open position. (makes me wish I had taken the trade ... still on testing phase though! )

Have you tried combining Schaff Trend Cycle (STC) indicator with this crossover system? STC seems to filter out a lot of whipsaws in choppy times.

I am wondering what charts or trading platform you are using for this system. I have been experimenting with a similar setup where I use a 5 and 10 ema crossover as the main indicator. However, I am using netdania charts which do not let you set open or close on the ema. I think it averages the closes but I am actually not sure, since it gives you no option. It seems to work though. I am considering changing trading platforms due to the large spread of fxcm anyway, and if I chose something like cms it would include charts with more options for the indicators. Any advice on which to choose would be appreciated.

Talia

i just programmed and backtested the basic system without any filters and it looks to be breaking even over 1.5 years of data. if you add slippage it will be unprofitable.

id try adding an ATR filter, as you suggested. only buy when the ATR is small, to make sure you are doing the transition from a narrow channel to a wide channel.

here is a screen cap of the performance report.

http://www.forexfactory.com/perf_report1.gif

i added an ADX filter to the basic system, along with a trailing stop loss of 500*Closing Price. the system now looks much more tradable...at least on the long side..
http://www.forexfactory.com/perf_report2.gif

Where s the best site to read learn about these ,would love to read ,learn,and expand my horizons.

Louis

Hey Merlin,

wow you ran all that on tradestation. And in a very short time ,am very impressed with your knowledge of programming. What kinda of computer langauge must you learn to program something like you did about moving averages? And does it have to be tradestation. The cost is high for me. Could i do something like you did less expensively. Thanks.

once you know how to use tradestation you can program something like this in under 10 minutes.

i might actually trade this on the long side, it looks really good. and there is no form fitting, i never even optimized it.

as for a cheaper version of tradestation, the only thing i think is comprable is MetaTrader. but IMO it's limited and under supported.

Wow, eurika! 73.77% winners! $22,000 in profit, I like it. Thanks Merlin for adding the Adx filter and the 500*close price and for back testing it. Thanks Mr. Bishop (reminds me of the charles bronson movie "The Mechanic") for posting your trading system.

Thanks

The Bluemondy

Merlin you are amazing. Could you please clarify, with the ADX filter, at what point is the entry triggered? ADX >= 20? I am starting to like this strategy more and more.

haha

no, not really amazing at all you dont have to be the smartest guy on the block to do this stuff....its really more a matter of > Can you dedicate enough time and energy into gaining the tools you need to get where you want to be?

trading can make your life all you want it to be...no bastard boss telling you what to do...lots of $$$...time to spend with your family...time to hang out on the forum...time to spend doing the things you love, really love, to do.

id say thats worth dedicating a few thousand hours to, wouldnt you???

i assure you, this is all very possible (with all kinds of trading, especially forex). its all out there...there are no secrets...everything i learned i have from reading books. i havent come up with anything really innovative on my own. all i have done is follow the conventional wisdom illustrated in countless books on trading, and i have found some success just doing that.

as far as the ADX filter i used....

i am hesitant to say because i feel this simple filter is one of my most powerful. HOWEVER, today i am feeling a bit philanthropic (as i was a few weeks ago when i posted a similar system ) so here goes...

basicly i look for a low ADX, then i look to see that the ADX starts rising. the details are up to you to figure out. if you put some time into this little piece of advice you will not believe that you found it on a forum

Thank you, I think I understand now. I recently bought J. Welles Wilder's book, New Concepts in Technical Trading Systems, wherein he goes into the details of all the indicators and oscillators that he invented, including ADX, RSI, DMI, etc. Not for everyone, but if you love mathematics like I do, its fascinating reading.

Cheers,
Vince

i think i will take a look at that book, ive been meaning to read it...

its a shame i use ole Welles' indicators as much as i do and i havent read his book

Dear Merlin,

I am seeking your help for backtesting my simple strategy over TS. Moreover this may even interest you in giving a try on real account, I am sure.

I do not know whether it will be possible to backtest or not because the entry and exit rules are not 100% mechanical and requires discretion.

Here it goes :

Timeframe : 15 min
Indicators : 3 EMA and 13 EMA

Buy Condition :
Place EMA 3 (Low) and 13 EMA (Open) on the chart. Stay away when prices are inside the channel of these ema's. Once the first candle opens as well as closes outside the channel, BUY on the open of next candle.

Stop Loss : Slightly below the 3 EMA (Low) level.

Take Profit : Using trailing stops.


Sell Condition :
Place EMA 3 (High) and 13 EMA (Open) on the chart. Stay away when prices are inside the channel of these ema's. Once the first candle opens as well as closes outside the channel, SELL on the open of next candle.

Stop Loss : Slightly above the 3 EMA (High) level.

Take Profit : Using trailing stops.


Anybody else interested may just put the indicators on your chart as stated above and see the possibilities.

Merlin, let me have your comments please.

Thanks

give me some exact rules and i will do a backtest for ya...computers dont understand discretion