Mechanical System Design Project - Join hands?

[skl99]

Hi, guys.

I’d like to share some of my thoughts. I am a professional software developer not a trader. Last few months I tried to learn how share market works and create a fully automated computerized day trading system. As I have lack of deep knowledge in financial markets probably I’m not too good at generating new ideas but I am good at picking up the existing strategies, encoding and backtesting them. I was able to test some popular strategies (gaps, Bollinger Bands, MACD crossover etc) and even some internet guru blackboxes (http://www.intradaytrades.com). But the outcome is quite disappointing so far. When you test different strategies on number of securities over period of time, the result is . In theory day trading can be automated as it’s not reallyabout zero affected by current market trend. But reality is that different strategies give you approximately same results in bearish 2001, 2002, 2007 and bullish 2005, 2006. So at the moment I decided to stop with testing and get more knowledge about the subject. But soon I understand that it can take few fulltime years to become an educated trader. Not sure I can afford it (family, mortgage etc ) Next idea was to join hands with a professional trader. I believe that two (or more) professionals can work out something valuable.

My contribution can be:
• 10 years commercial software development experience
• General understanding of how shares work and the ability to pick up and encode ideas for careful testing, examination and optimization.
• Fully automated contemporary framework with the ability to collect real-time and historical data, make decisions based on your instructions (high level programming language) and execute orders with popular brokers (US based indeed: IB, MBT etc). The framework already has most of popular indicators and functionality to backtest and optimized your strategies.
• Historical data (trades, 1 and 5 min bars) for number of instruments.
• Help with installation and set up of the framework at your side.

My target is to create one or few strategies which can be run fully on their own and give your stable low income (5-10% a year from your cash, or more using 2x or 4x margin)

I am not interested in extra money for the ATS development (I am pretty happy with my current position and salary ) but I’m definitely interested in market education and finding a professional trader to learn from.
Please don’t blame me as a novice who wants your system for nothing . I don’t ask for your “Holy Grail” trading system. As far as I know many traders have less profitable system they don’t use in everyday practice. My idea is to get some extra steady income using one of these.

Please keep in mind that some strategies (like pattern trading) heavily depend on person’s skills and experience and are extremely hard to be automated.


Regards

 

[gazelle]

Re: Join hands?

Good post SK . You are looking to apply and integrate your electronic knowledge towards constructing and designing an automated trading system
based upon certain technical criteria that you are looking at or have selected to plug into your model . I am not really experienced in this area however my trading system is partly automated in application and execution although there is a large discretionary componet as well and these two variables can and do change over time depending on the condition or phase of the market under analysis . Pure automation is a fragmented approach as you are distorting and altering the true path of the market to fit within your artificial model . Markets are rotational systems with inner and outer boundaries that expand and contract dependent upon the frequency of vibration or cycle that is running out at any given point in time . Automated systems capture a snapshot of the terrain , in essence they only decode a small section of the market . Every instrument is unique , it holds its own tone and has a unique repetitive frequency , to a degree this is similar to financial mkts so getting back to the topic how can we we design an automated trading system that has similar design charactetistics in line with our mkt of choice , well it is hard , I know very little about purely automated systems as I trade Gann so its a bit different . Certain markets move geometrically with price and time componets moving into alignment , these dimensions are always changing and seeking their centre of gravity . these ideas may seem abstract and they probably are but thats the way I look at markets , Its almost impossible to pull the veil and to tap into the cystal structure of the market but I think purely automated systems only see one side of operation .


Automation takes

 

[RichKid]

Re: Mechanical System Design Project- Join hands?

Hi, guys..........
.........
........

My target is to create one or few strategies which can be run fully on their own and give your stable low income (5-10% a year from your cash, or more using 2x or 4x margin)

I am not interested in extra money for the ATS development (I am pretty happy with my current position and salary ) but I’m definitely interested in market education and finding a professional trader to learn from.
Please don’t blame me as a novice who wants your system for nothing . I don’t ask for your “Holy Grail” trading system. As far as I know many traders have less profitable system they don’t use in everyday practice. My idea is to get some extra steady income using one of these.

Please keep in mind that some strategies (like pattern trading) heavily depend on person’s skills and experience and are extremely hard to be automated.


Regards

Hi skl99

Welcome to Aussie Stock Forums!

That's a commendable idea and one that has generated a number of threads on ASF (in some form or another). You are bound to find invaluable material in those existing threads if you use the search tool and type in key words and phrases related to your interest (eg mechanical system, drawdown, consecutive losers, scalping etc). This will provide some grounding in the basics of this field. Feel welcome to post any questions in those threads and feel free to develop this thread as your idea for the system matures. You are unlikely to give much away by discussing the specifics of basic concepts and strategies on the forum threads. It is probably advisable to spend a few weeks or months on ASF to study and reflect on the various concepts presented here- gazelle, for example, makes some observations above which you may not have considered but which are often debated between mechanical and discretionary traders.

We encourage open discussion and exchange of information on ASF and even have threads on local trading groups for those interested in face to face meetings. So read, post away and feel free to continue this as an open discussion.

Good Luck!

 

[julius]

Re: Mechanical System Design Project- Join hands?

skl99,

Questions:

- Which broker's API do you have experience with ?

- Which programming language are most familiar with ?

- Which charting platforms are you most familiar with ?

Cheers.

 

[skl99]

To RichKid

Thanks a lot for encouraging me Indeed that's what I do now. Forums give you more knowledge than books or seminars.

Regards

 

[Trembling Hand]

My target is to create one or few strategies which can be run fully on their own and give your stable low income (5-10% a year from your cash, or more using 2x or 4x margin)

You lost me here. Just put your money in a high interest rate account and don't worry about the market.

 

[prawn_86]

You lost me here. Just put your money in a high interest rate account and don't worry about the market.

Have to agree here.

If your not earning over 10%pa in the market then there is no real point being in it.

Anyone can get &% in a bank account, so to me that extra 3% makes up for the risk. Obviously the higher the return the better though. And some would argue that you should be beating the index, otherwise it is better of in an index fund, however im focusing on slow steps...

 

[albi000]

Fully mechanical system described today on CompareShares.

"From 1929 to 2003, the Dogs of the Dow strategy outperformed the stockmarket (the S&P 500 index) handily, returning about 14 per cent a year compared to the market return of 11.7 per cent a year during that time. "

Maybe "Dogs (Dingoes) of the ASX " is what you are after?

 

[Trembling Hand]

In fact the system will only have to break even to get a 6.5% return. As the project would be a day trading system it will earn 6.5% interest on overnight account balances.

A often overlook advantage of day trading.

 

[skl99]

Hi

Thanks for all answers.
I am surprised a bit with the questions.

5-10% a year from your cash, or more using 2x or 4x margin

Key word here is "your cash". Using standard 4x day trading margin ROI will be 20-40%. Next it's easy to invest or borrow some extra money to increase the cash flow.
Can you do the same with bank account?

Regards

 

[skl99]

To albi000

Thanks for the link, I will study it.

Regards
Kirill

 

[nizar]

My target is to create one or few strategies which can be run fully on their own and give your stable low income (5-10% a year from your cash, or more using 2x or 4x margin)

I wouldnt bother with systems testing and design If i was aiming for 5-10% unleveraged.

That would be an underperformance of the market, many fund managers can already give me that!!

If you want to trade the stockmarket with the aim of making 5-10% a tear unleveraged, in my opinion, You are taking on too much risk for the return you are seeking.

 

[Trembling Hand]

Fully mechanical system described today on CompareShares.

"From 1929 to 2003, the Dogs of the Dow strategy outperformed the stockmarket (the S&P 500 index) handily, returning about 14 per cent a year compared to the market return of 11.7 per cent a year during that time. "

Maybe "Dogs (Dingoes) of the ASX " is what you are after?

That's not a day trading system!

 

[tech/a]

Been watching this thread for the last couple of weeks.

Seems to be dying a natutral death like all other systems design threads.

Hopefully some constructive suggestions below.

(1) You dont need to invent systems testing software or Programmes.
The best you'll get is already there. Amibroker and Tradesim Combination will do eveything you want.Your programming skills may well place you well ahead of average users like myself.BUT the full package will set you back a couple of K.

(2) http://www.quantitativetradingsystems.com/
Howard Bandy has a complete guide to systems construction in the form of his excellent book. Both Nizar and I have one and I'm sure a few others have as well.

(3) The Triggers are of course infinite for
Entry
Exit
Initial stops
Trailing stops
Position sizing
Targets etc.

The above will get you way further down the track than relying on Forum posts.

 

[skl99]

To albi000

"ASX dogs" is a longterm investment strategy and works good only on bullish market.

by the end of 2007, the strategy was really in the doghouse: our portfolio lost 6.7 per cent, compared to the S&P/ASX 50’s rise of 11.7 per cent.

I am after intraday or swing strategy.

Regards
Kirill

 

[skl99]

I wouldnt bother with systems testing and design If i was aiming for 5-10% unleveraged.

That would be an underperformance of the market, many fund managers can already give me that!!

If you want to trade the stockmarket with the aim of making 5-10% a tear unleveraged, in my opinion, You are taking on too much risk for the return you are seeking.

Well, managed funds work well only with bullish trend. I don't know AU funds which really work (trade) at the market. They just update their portfolio once a quarter and get 2% from you accurately.
At the moment I am minus 8k with my top range managed fund.

Regards

 

[skl99]

Been watching this thread for the last couple of weeks.

Seems to be dying a natutral death like all other systems design threads.

Hopefully some constructive suggestions below....

Thanks for watching and for the answer.

I posted same tread at couple of others trading forums and already have few people I cooperate with.

Main idea: they want to save time and get some extra steady income.

If you feel confident enough to automate your strategy yourself, that's great!

Regards

 

[Temjin]

skl99, check out some of the managed futures funds out there to get a feel the kind of performance the professionals are getting in the real world.

www.iasg.com for a starter, there are quite a few of these managed future brokers around on the internet.

Tech/A's advises should also get you started pretty quickly.

I wish I had your level of technical expertises in programming. Me and Nizar are involved in designing intraday automated trading systems as well and understand the fundamentals pretty well. But both of us lack the programming skills and still have a long learning curve to go.

Maybe we CAN trade some knowledge around. Good luck.

 

[skl99]

Temjin, thanks for the answer and link.

Probably we can do something together.

Regards

 

[julius]

I see a few things I don't agree with:

The objective is an INTRAday system - hence you need liquidity and volatility...realistically stocks are out of the question, so you're looking at some form of leveraged instrument (forex,futures,etc)

anyone who is familiar with day trading will tell you %return is irrelevent: the real game is not wiping yourself out, ie. low variance (risk)

leverage on futures is ~20:1, ~50:1 on forex . . . 20 x 5% = ?

intraday systems vs. medium/long term = chalk & cheese imo

5-10% is very reasonable