PipStalker EA

Find back testin with tick data (99%) modeling.

That was a letdown but curve fitting is the name of the game when you have enough back data. I'm just glad we got a 99% MQ test and think it maybe be possible to have a profitable backtest at 99%.

I bet tic data for 99% MQ is expensive. Anyone got any input into that?

Congratulations on 99%
How come at the top of your statement it says the test is for 1970 with the same start and end time? Is your history data from the guy that was collecting tick data on TSD?

Can someone point me in the right direction?

What's all the hoopla about. I backtested this EA a week or 2 ago and got these exact results: 25% prifitable trades with nothing to show for after 2 years with net profit of -1400 it struggles to stay afloat.

Let's see at 154 trades at $20 spread per trade dealers will make $3080. Wow!
My advice, don't waste your time on this pipe dream of someone who calls himself EA master, but haven't learned that you can at least indent your EAs..
I wasted 6 months looking for a holy grail of mechanical trading without any success. So I changed my tactics and now I prowl this forum for tidbits from wise people who have been successful in this and generally spend my time waiting for right setups to enter into. In other words, it's hard work. I haven't give on EAs, but only as crutch to help you manage going trades while you sleep.

I deleted the EA from 2 weeks ago and started over from scratch.

Yeah I don't indent my EAs.... so what? I have a B.S. in Computer Science from The College of Charleston and am getting an M.S. in Computer Science from The Citadel. I was a computer programmer for the Department of Defense for 3 years. What credentials do you have?

Do you suggest we never build EAs? I guess a Holy Grail system just happens with no failures. I guess hard work and effort aren't what EA programming is about.

You said what you had to say and so did I. I'll go all day long if you want to.

Some people think that they can manipulate a couple of variables and get magic results.

Sorry, I don't have 2 years of data handy right now.

Somehow mine looks better at 90% MQ.

You don't have time to test different settings, but you do have time to waste by posting garbage like that on my thread.

People that choose to participate in this project will, and people who don't won't. While you are naysaying, I AM DOING.

I signed up for Intellicharts with FXTrek and am porting my code into it....

Modeling quality should be about 99....

I will post my results later.

I'd like to point out a few things about my method.

First, it does very well in trending markets, but does not get crushed during sideways markets.

I have spent many hours looking at various money management strategies for when drawdowns inevitably occur. This has been a good lesson.

Basically, this method makes a fortune during tight trends. To captilize on this, I have modified the EA to adjust the trigger to 20 when drawdown is below 20%. If drawdown is above 20%, the EA sets the trigger to 40 (and plays defense: doesn't trade as much)

Also, the lot size per trade decreases by 20% every time the account falls 1%. It adds 20% to the size every time the account gains 1%.

In closing, I realize there are a lot of naysayers out there that pass through the threads lazily and freely unload criticism. I am planning to trade this EA with real money and hope there are a few people that are interested in taking it seriously.

Here is the latest EA with money management.

I call this a "working man's" EA because I am designing it to trade in a data warehouse in California while I go to my job everyday. 5 T1 lines, gas powered generators, all that jazz...

Anyway, th graph posted in #69 is from 2001 - 2007. You can see that in 2006 the EA made no money. This is a long term method with "retirement" written on it.

Hi, tdion! Do you still test your EA on GBP/USD? Have you tried to test Pipstalker on other trendy pairs?

The EA gives similar results with GBP/USD, but commissions eat all the profits because you are paying 2 extra pips/trade.

The EA is hardcoded for a 2 pip spread, so don't be fooled by a profitable GBP/USD backtest!

This is a long term investment EA. I'm posting the year-by-year backtests of what $10,000 would have done each year. If you are looking for a quick buck, you might get lucky or you might not because there are some losers here.

Note that in 2001, when the account dropped below $7,000 the EA traded less frequently and with a smaller lot size (to protect capital)

Not too shabby

Another unproductive year.

Lightning in a bottle..

Insane

Breakeven

My conclusion is that the EA gives you an opportunity to ride the lightning. The most it lost in a year was $5,000, but there were 2 100K years in the past 6.

Past performance doesn't dictate the future, but I like what I see.

T-

Started forward demo testing Pipstalker tonight (eurusd, h1)...will give it some time before getting in the money. Just wondering if you have a second to look at a custom indicator I am working on. It is a simple concept to go along with a bollinger band strategy i have been working. i am only conceptually working on it, have not started thinking about coding it into an EA, but i want to manually test it with custom indicator. anyway, the specifics:

1) Histogram showing the current Close Price minus the Close price of X # of bars in the past (x is designated as an ext variable that can be changed)
2) line bar (in the same indicator window) showing the current Low minus the Low of X # of bars in the past (this x should be the same as the x of the histogram)

That is basically it -- pretty simple. Though i can't make it calculate historically. I think my error has something to do with an array, but I am only a beginner when it comes to coding EA's and indicators. Thank you for your help.