PipStalker EA

Hi all.

I've been playing with an idea that is very simple and looks pretty good.

It's a double bounce off the EMA(14). Stoploss is at 50.

The EA does very well or breaks even with most backtests.

Enjoy.

I forgot to mention it trades on the hour bar.

I've attached an image of the GBP/USD backtest. 180% profit in 5 months.

I hope I my backtests aren't giving false positives... coding shifted bars can be tricky. PipStalker is the first EA I've ever coded.

congratulations on the 1st EA born. me junior programmer as well, hope to learn from you soon. meanwhile, are you going to code money management? so we can play around with the profit?

Added some money management......

Lot sizes will be (AccountBalance()/10,000).....

GBP/JPY looks good..... not sure when the wind will stop pushing the sail on that pair though.

Gbp/Jpy Sl=80, H1, 9/29/06-2/11/07

Can you post the entire backtest report?

Here's a report on the GBP/USD.

The first thing I'll tell you is the modeling quality bothers me. I can only get 48%.... maybe some of you can do better.

As I feared, some of my entries are off the bars.... meaning that the order can't be filled at that price. You get these fantasy fills that could never happen. I wonder how many EAs look profitable, but aren't REAL.... you know?

Anyway, this is just a hobby and I hope it can turn into something.

there you go. hope this help

edited: I've ran more backtest since 2006. not too bad at all. the only problem is, most of the time it merely reach the TP and reverse, become a stop loss. some input for you, trailing stop would help

Thanks guys. My personal preference for back testing is to use fixed lot sizing. Then it is easier to have a balanced view of how it did over time. Nonetheless, the EA results are not terrible so I will look at the EA some. Thanks for sharing!

I'm not sure if I used the right settings but they seem to match the other backtests. It looks like it depends a lot what month you start the EA in. January of 2006 was a very bad month. The EA may do better or worse in live trading. I thought you might want to look at this test and manually compare a few of the trades during the bad months to your charts to see if the EA worked as designed.

Thanks everyone.

After seeing nvwine's report, I'm not so excited anymore. I will see if a trailing stop turns some losses into gains.

A really successful EA, Pipboxer, uses trailing stops exclusively and the results are very linear.

Don't put too much stock in my report until others get the same results. I've been wrong before - many times. I did not want to discourage you. I just wanted to point out that some months the EA seems to struggle in my opinion. I thought you might be able to find out why.

nvwine:

For fun, will you backtest this revision? GBP/USD 1H please....

How'd you get 90% modeling? How'd you get data back to Jan 01? My broker (FXDD) only give me 48% quality and 5.5 months of data!!

Tom

I had to go to 25,000.00 in order to get through 2006 due to the 1 lot size and draw down. The 3 year test ended due to no money. I used the default settings. I'll give you some back testing links in a minute.

Without a trailing stop, this thing is garbage. Thanks nv.

There is a sticky thread in the MetaTrader forum on backtesting. If you want more information PM me.

In a nutshell you need to to select 'tools' / 'options' / 'charts' from the top menu bar. While in the charts tab increase the maximum bars in history to 4 million or however many you need for 1 min data. Increase the bars in charts if you want to see all the bars on charts. then

Press F2 , select the pair you want and click on 1 minute data. Then press the download button in the lower left to update history for the selected pair.

Some times it works better to exit MT and re-start go to the history center again and see if the 1 min data you wanted is there.

At the moment there are questions whether the downloaded data is hour corrected to your platform do to GMT and GMT+ server times.

Note this can increase MT4's memory usage.

Hey mate,

what's your modelling quality on that last backtest?

As with all EAs, forward testing is a must - I may put this on a chart for a month to see how it plays.
Do you see yourself refining this in the near future?

tdion,

Ok, well my computers on 24/5 anyway forward testing some other strats so I'll open up a E/U, G/U and G/Y.

Take it you're looking at SL and TP at 40?

Let's see how it goes.... thanks for sharing.

Can you post a chart detailing the method a bit more?

Cheers

This is completely OT like no tomorrow. I'm just going to say, thank you for supporting Office Space since my step-dad's in that movie and the more that you guys support the movie, the more likely I'm able to have my college tuition paid for, and thusly, able to finish college. So...Go Office Space!

/rant

Otherwise, I'm glad taht you were able to codify your system into an EA. To me personally, that is the last and utlimate step to creating a successful system. When you are able to fully mechanize it because you believe that fully in it, and you want to be able to sleep at night instead of staring at charts, you deserve the success that comes to you.