Can anyone backtest this system ?

[nkruger]

Good morning - I have developed a strategy to trade the Dax & Dow intra-day and I wondered if there was anyone on these boards that could back-test the system as I have no facility to be able to carry this out. It involves Parabolic SAR/Directional Movement with Candlesticks/Pivot Points/RSi..........

Anyone out there able to help ?

Nick

 

[lurkerlurker]

Good morning - I have developed a strategy to trade the Dax & Dow intra-day and I wondered if there was anyone on these boards that could back-test the system as I have no facility to be able to carry this out. It involves Parabolic SAR/Directional Movement with Candlesticks/Pivot Points/RSi..........

Anyone out there able to help ?

Nick

I'm in a similar position (having difficulty backtesting my new dow method). The problem lies in defining such things as "indicator divergence" in a mathematical model. I'm also wrestling with using CMC MarketMaker to backtest (admittedly it is a good platrform, just a little tricky). In particular, I am having problems implementing such conditions as

If A and B not C else if not C and D.​

or similar.

I am finding the inclusion of chart patterns (H/S, double top, etc) are difficult to program also.

Finally, the biggest problem I am having with this is; if I can backtest a system with a positive expectation, what is my role as a trader? Why not just let the computer execute? I do not believe that a system alone can be consistently profitable. I think the human input is important. I think I am a discretionary trader at heart, and I have simply lost my way taking advice to backtest. Maybe the rules of my new system are too rigid. I am trying to papertrade for a few days.

My main concern now is my diminished account size. For the Dow, stops < 15 or even 20 on some setups seem too close, but that is exposing my account to a fair bit of risk. Maybe I should limit myself to one losing trade per day with a 20 pip stop instead...


Oh, and would you mind sharing your system with me, since you need to make it public in order to have someone else backtest? My current system is posted in my journal if you are interested.

I currently trade the Dow intra day and am considering the Dax for the morning session (London time). Perhaps we could collaborate to an extent on a system?

--yes, I'm working late again. Ill at the moment, but that isn't stopping me from running backtest software, looking at charts, and reading as much as I can about trading. However, I will stop short of real trades until I can shift this head cold.

 

[firewalker99]

Good morning - I have developed a strategy to trade the Dax & Dow intra-day and I wondered if there was anyone on these boards that could back-test the system as I have no facility to be able to carry this out. It involves Parabolic SAR/Directional Movement with Candlesticks/Pivot Points/RSi..........

Anyone out there able to help ?

Nick

What "facility" are you refering to?

I manually backtested my system on hundreds (if not thousands) of charts from different instruments and markets to see how it held up against changing market conditions. I can tell you it cost me a whole lot of time, a lot of red eyes from watching the screen but ultimately it's the only way I got confidence in my system. By going over each chart from the past (backtesting) reading it from left to right - or using a replay function if you have software that supports such an option - I have gained the necessary confidence in my system.

 

[firewalker99]

Finally, the biggest problem I am having with this is; if I can backtest a system with a positive expectation, what is my role as a trader? Why not just let the computer execute? I do not believe that a system alone can be consistently profitable. I think the human input is important. I think I am a discretionary trader at heart, and I have simply lost my way taking advice to backtest. Maybe the rules of my new system are too rigid. I am trying to papertrade for a few days.

If you have backtested your system by computer or have a way of programming/automating your signals, then why not do it? There's no reason you have to watch the computer execute on your signals. If they are clearly defined then you don't necessarily need to leave room for a discretionary factor if you don't want to. There are people who let the computer execute trades for them... would that be so bad? :|

 

[lurkerlurker]

If you have backtested your system by computer or have a way of programming/automating your signals, then why not do it? There's no reason you have to watch the computer execute on your signals. If they are clearly defined then you don't necessarily need to leave room for a discretionary factor if you don't want to. There are people who let the computer execute trades for them... would that be so bad? :|

Remember the psychological barriers to trading success rooted in our upbringing which Mark Douglas talks about in his books? I can just about get my head around manual trading being worthy of financial reward. However, the idea of having a computer track a market (and if we are going to talk forex eventually, a computer tracking a market almost 24 hours a day) growing my net worth without significant human intervention just seems a little unbelievable. It is a little difficult to get my head around. Even if I design and implement the system myself, backtest it, find it has a positive expectation, give it a brokerage account with a few thousand in it, and let it execute on the signals without emotion, and the thing actually makes consistent profit...it would feel a little odd to me. I'm not saying I couldn't get over it eventually, just that it would feel a little odd at first.

I believe futuresbetting.com allows automated esignal trades (I'd need to check). I'm considering a fully hedged DMA SB account soon anyways.

I do need to backtest more though. I'll keep you posted.

May I ask if you have a "day job"? What would your attitude be to having a computer do your YM trades in your absence? Is your strategy too discretionary, or do you believe that you could formalise rules sufficient for a computer to execute for you?

And if you don't have a day job, and are a full time trader, imagine being able to go back to work and have the computer trade while working a full time job. Talk about the potential for great wealth acquisition!

 

[firewalker99]

Remember the psychological barriers to trading success rooted in our upbringing which Mark Douglas talks about in his books? I can just about get my head around manual trading being worthy of financial reward. However, the idea of having a computer track a market (and if we are going to talk forex eventually, a computer tracking a market almost 24 hours a day) growing my net worth without significant human intervention just seems a little unbelievable. It is a little difficult to get my head around. Even if I design and implement the system myself, backtest it, find it has a positive expectation, give it a brokerage account with a few thousand in it, and let it execute on the signals without emotion, and the thing actually makes consistent profit...it would feel a little odd to me. I'm not saying I couldn't get over it eventually, just that it would feel a little odd at first.

I believe this is part of the reason you are losing money. You are unable to put confidence in a system and you believe you yourself can outsmart the market on a single trade. For an example: if you had bought at the open of the DOW in the last 2 months, set a stop at -20 and a target of +50 you'd made already a nice profit. Could a computer execute this strategy? Obviously, there is nothing to it. I know this "system" isn't a realistic one over time, but for the sake of the argument I'm making this as simple example as can be.

You say you would feel "odd", that is imo because you still feel you can take control of the market or your trades. You can't. You can't know in advance which ones are going to be winners and which ones not.

May I ask if you have a "day job"? What would your attitude be to having a computer do your YM trades in your absence? Is your strategy too discretionary, or do you believe that you could formalise rules sufficient for a computer to execute for you?

And if you don't have a day job, and are a full time trader, imagine being able to go back to work and have the computer trade while working a full time job. Talk about the potential for great wealth acquisition!

No, I quit my day job some time ago. I think my system would be difficult to automate, because the basics fundamentals of my system (support & resistance) are determined by my view of the market, probably discretionary. Perhaps I should develop a program that cleary identifies zones of support & resistance. Hmm... just googled a bit and found out this site http://www.paritech.com.au/education/technical/custom/indicators/98may.asp that presents a computerized approach to finding support and resistance levels on a chart. Might have a look at that! Anyway, if you trade on indicators alone, there's no reason why a purely mechanical system can't and shouldn't be automated. In fact, it's the easiest part of getting rid of the psychological issues that effect your trading performance!

I believe most of the members here trade for freedom. Not to sit behind their screens for 16 hours a day. Personally, I don't mind, I love to sit and watch the markets but after a couple of years I'll probably put on 1 trade and quit for the day (like I did yesterday on DOW).

 

[nkruger]

Oh, and would you mind sharing your system with me, since you need to make it public in order to have someone else backtest? My current system is posted in my journal if you are interested.

I currently trade the Dow intra day and am considering the Dax for the morning session (London time). Perhaps we could collaborate to an extent on a system?

QUOTE]

Sorry for delayed reply - been in hospital..... Don't mind sharing my system at all, I think I did outline it sometime previously on these boards, can't remember now.

A bit of history to show how I got to where I am with system development. I have traded the 'DowDecoded' (www.dowdecoded.com) strategy since last September with quite good results together with a couple of EOD systems (www.elmtrader.co.uk & www.dowstomper.com) All going tickety boo, but there comes a point when one needs to be 'in charge' and work on a peronal system.

What I have developed is a system to day trade the Dow (after 3pm GMT) and now the Dax as well. It's not perfect, but it's almost there. I do feel I could do with another couple of indicators - tell me what you think ?? What I use at the moment is Parabolic SAR, RSi, MACD and 12/26 MA crossover on 3 min charts with ADX/Directional Movement as a confirmation. In the perfect world, I like to have all 5 indicating a buy/sell but I will enter a trade with 4 showing. Do I need some more indicators ? Are there any that you use (different to mine) that you think may be of use ?

And yes LL, would like to work on a collaboration, 2 heads and all that..

Fire away with questions (anyone), we've a nice break now 'til the markets open again

NK

 

[nkruger]

What "facility" are you refering to?

I manually backtested my system on hundreds (if not thousands) of charts from different instruments and markets to see how it held up against changing market conditions. I can tell you it cost me a whole lot of time, a lot of red eyes from watching the screen but ultimately it's the only way I got confidence in my system. By going over each chart from the past (backtesting) reading it from left to right - or using a replay function if you have software that supports such an option - I have gained the necessary confidence in my system.

As I don't use any 'Charting Package' I am not really up to speed on whether there is an in-built facility to backtest home-grown systems to save time. I have done what you describe and stayed up all night testing what I thought was a eureka moment, and as you say, it is time consuming....

I think I need to to do more backtesting as i think ultimately you are correct, if I have faith in my system, I do not need any more indicators. Thank you for your words of wisdom.

NK

 

[nkruger]

Oh, not sure how I forgot this, but I also use candlesticks as well. I'm still learning about all the different signals they can throw at me......

 

[rols]

And if you don't have a day job, and are a full time trader, imagine being able to go back to work and have the computer trade while working a full time job. Talk about the potential for great wealth acquisition!

Hi there

I love your enthusiasm.

Please read this. It may save you much time and money.

http://www.ensignsoftware.com/tips/tradingtips42.htm#In

 

[investforwealth]

For automated system testing, it may be worthwhile looking at TradeSim software, written by CompuVision in Australia. It can be used with MetaStock, TradeStation, AmiBroker, BullCharts and can even import trades detailed in Excel spreadsheets.

I've found it invaluable in testing various strategies I've played around with.

http://www.compuvision.com.au/

 

[lurkerlurker]

Hi there

I love your enthusiasm.

Please read this. It may save you much time and money.

http://www.ensignsoftware.com/tips/tradingtips42.htm#In

Thanks rols. It's okay. Really. Read my other posts, and my journal. I'm not looking for a grail. I'm fully dedicated to trying my own system manually, and my system employs discretion. I am trying to focus on consistency while paper trading right now before I risk more of my capital. I appreciate the warning however.

I note you have an animated avatar. How so? I can't upload mine.

 

[rols]

Thanks rols. It's okay. Really. Read my other posts, and my journal. I'm not looking for a grail. I'm fully dedicated to trying my own system manually, and my system employs discretion. I am trying to focus on consistency while paper trading right now before I risk more of my capital. I appreciate the warning however.

I note you have an animated avatar. How so? I can't upload mine.

Make sure your avatar is the correct max pixel size or it won't load.

Sorry to be a bit abrupt previously.

Backtesting in my experience is for the birds.

I wasted soo much time and money backtesting, automating systems in the past.

You've just gotta do it .

There's nothing like real time trading with real money and real losses. Put the costs down to education. After all how much are folks paying now to go to Uni for a few years?

Difference is these guys aren't necessarily gonna get an income but if you stick at the trading and beat your own drum then you will succeed.

Belief then action then success.

Treat failure as your friend. It's a necessary part of the journey to success. The benefits are salutary too.

And BTW don't spend so much time on these boards. It saps you in the end.

 

[Supermu]

Nick,

hope this finds you well..

I came across your post as I was looking for reviews on elmtrader - impressive if one is to believe the stats. I was looking at dow fs1 2008, but it looks too good to be true.

I'd like to check out your system, but as I'm new here, I'm not sure how to view your journal.

Rgds,
Jesper

 

[nkruger]

Nick,

hope this finds you well..

I came across your post as I was looking for reviews on elmtrader - impressive if one is to believe the stats. I was looking at dow fs1 2008, but it looks too good to be true.

I'd like to check out your system, but as I'm new here, I'm not sure how to view your journal.

Rgds,
Jesper

Hi Jesper, I've PM'd you.

 

[playhouse2005]

Good morning - I have developed a strategy to trade the Dax & Dow intra-day and I wondered if there was anyone on these boards that could back-test the system as I have no facility to be able to carry this out. It involves Parabolic SAR/Directional Movement with Candlesticks/Pivot Points/RSi..........

Anyone out there able to help ?

Nick

Hi Nick
I can assist you and anyone in backtesting your system using a fully automated trade engine but you pay for coding your logic (for which the cost will depend on how much work is involved) as well as data feed if required depending on what you intend on trading and I will backtest for a duration of time and forward the results to you.

I hope this is helpful as it will reduced your workload and knowing that you can alter your strategy without having to sit and trade it for some time.
If your interested email me at: [email protected]