If your using EFS, enter the highest return you've had...

[derekg]

Originally posted by spike500
I'm glad for you that you did that. 5 years is long enough to have a mix of all kind of situations.

Now if ESignal would just 1) provide at least 5 years of data or, 2) allow us to import historical data. Until the company stops refusing to do either, EFS will never be considered a viable backtesting platform.

[spike500]

Originally posted by derekg
Now if ESignal would just 1) provide at least 5 years of data or, 2) allow us to import historical data. Until the company stops refusing to do either, EFS will never be considered a viable backtesting platform.

I use a Tradestation. I have my own data that i can import. I use Esignal only for the datafeed.

How did Bakedwafer backtest on 5 years data in Esignal? It seems to be impossible if i read your posting?

[BakedWafer]

This system has been in development for the past 6 months, but it was based on QQQ. On recently, did we change to futures but the fundamentals are still the same.

Originally posted by spike500
I'm glad for you that you did that. 5 years is long enough to have a mix of all kind of situations. My impression that you were going too fast was apparently wrong.

[spike500]

TraderJO

TraderJO,

I looked into your performance strategy and analysed some of your trades.
My first thought was that you should add something that was still missing , a trendindicator. I think that you trade in a very short time frame ( 3 minute, 5 minute or 15 minute bars). These timeframes are to me too short to give a reliable trend.

So what I did: I took an trendindicator on hourly base that I use, and accepted only your trades that were in the same direction as my indicator.

This gave for the first 100 trades that you did the following information:
27 loosing trades were eliminated. So the losses diminished with 6460.50$
17 winning trades were eliminated. So the profits diminished with 5892 $.
The total gave us 44 trades less and 568.50 $ more profits.
The accumulated profits for the period went from 5037.5$ to 5606$ . This means an increase of 11.28%.

The fact that the improvement is not better is because that my indicator and your trading system are not in line with each other. I think that, if you develop an hourly trendindicator, the improvement can be quit spectacular.

I only took the first 100 trades because it takes too much time to me to check all the trades. But 100 trades can give you already an idea about what you can expect.
The only person that will not be happy is your brooker because you will generate 44% less commission for him.
Less trading with bigger profits and smaller losses, that’s the general idea.

[fflaque]

i've had a very high return but can't backtest the darn thing

[BakedWafer]

Can anyone add a trailing stop to this attached code?

[spike500]

The following link has a lot of good information for people that are building systems. Take a look, it's for free.

http://members.aon.at/tips/moneyMan.htm

[FibbGann]

The fgalfalse.efs has been greatly improved, but is still under testing and improvements. A long work in progress that is getting close to completion.

[derekg]

Wow, this thread really died once Brad left. Is he still out there with the hedge fund?

[Doji3333]

I'm here..

I got back late Sunday... What a long trip - never had time to adjust to the time difference. Slept only about 5 hours a night (thanks in part to my wife).

I'm here.. I've got something else I've been working with too. Don't have all the rules worked out yet, but it looks good so far.

B

[BakedWafer]

Figured we'd fire this thread back up again. This is an EFS I've been working on in conjunction with the previous EFS. The curve is not as forgiving as the previous one but the # of trades is much lower and the profit is higher.

[chrisf95003]

Thank you everyone who has posted on this thread. I have learned alot by reading it. For those people that have fully automated systems. In the coding do you call studies into the system or is everything written into the efs? For example with the stops of the entry trigger.

Does anyone have the framework for an automated system or a link to one that I could look at and learn from? I am not asking for your production trading systems, just an example of the logic for a fully automated system so I can see how the entry, exits, stops, etc interact with each other.

Thank you,
Chris

[chrisf95003]

I might have overstated my request there. The end of that first paragraph didn't make sense either. If I want to use a non built in study like the LSMA, is it better to have the whole formula in the automated script or just call it?

Thanks,
Chris

[shogun]

chris
have a look at this site there are several free efs examples there you can download, also you can use dynaorder for free for two weeks to see if it is for you,

http://www.dynaorder.com

look under Tips & Examples and click on "Programming for trade with Interactive Brokers or Paper Trader (DOTS5.DLL). "

[Doji3333]

inclusion of indicators..

This is really a judgement call on your part...

In most cases, it is simple enough to include these indicators using CALL functions. This makes your code cleaner and in some cases more effective.

Now, if you need to do something with the indicator to improve your system (for example, customize the indicator's actions based on other conditions or other indicators), then it might be smart for you to include the indicator into your code..

It just really depends on what you intend to do with the indicator

B