Experiments in Technical Analysis

Hi

I dont see any thing much promising with this simple maco system with manual backtest i see exit is a major flaw in this case also author of this system didnt explained what kind of MM need to be adpoted or he adpot's to make it consistent trading system . Good lucks and good trading to all

Ps: dont waste too much time on left hand charts and fooled by birdeye view and curve fitting and over optimisation it wont take u any were better . Also fancy indicators too just read price action :d and dont spend too much time on it

To be brief how i developed own system

1. select market you want to trade with ( stock or futures or options or bonds etc )

2. select the time frame wrt to you risk tolerance ( for eg a day trader or even a swing trader cant place stop based on weekly as correction on weekly levels is crash on 4hr or daily ) so it have no meaning if u look for weekly even if u r system is colorful and promising on weekly basics

3. select you trading tools ( just keep it fixed if u trade real money and trading professional ways ) for learning and demo trading use separate trade box dont mix and modify both

4. Develop system based on tool box or tool list you have also brief out yourself why you selected each tools and how it really add edge really to you again dont get fooled by fancy color on charts

5. Apply tool to u r chart setting and develop strategies 1st i dont think a newbie can develop a MTS straight unless he is well expertise in finance , maths and programming all together

6. Optimze indicator settings till you feel ok and add you mech rules for ENTRY , EXIT , TRIAL STP And STP a complete system needs all always and must work consistently atleast on mkt and time frame you develop

7. Now backtest you system with atleast 2 - 3 yrs of tick data ( most recommended ) if not aleast on 1 min data also backfill must be highly reliable without any gaps and noise which decide quality of back test and its reports . Dont depend on stimulated backtest which is useless 100% and only give rosy results but merely loose on real time forward testing

8. Optimize your money managment and settings on backtest to see line of best fit

9. Forward test it for atleast 2 or 3 quaters on real market conditions and calculate cost of trading and other factors etc

10. IF still sucessful trade it and smell sucess and make consistently

IMO just playing with indicators and guess work leads ppls no were i dont want to complain such thing wont work such systems like described here was tested already by many 1000s of trader around world already . This is up to members to decide how they develop themself . Good lucks to all

ps : but for demo trading and learning this thread might be a start

Dear Amarnath,

very nice post for everyone. If you share your thoughts in the form of trading system ( especially coded in Metastock) with instruction what to do & what should not do by a small investor. It help lot.

Regards,

(moneypick)

Hi Amarnath

Nice to see you posting on this thread. A few clarifications so that the intent of this thread is not misunderstood. There was not even the faintest hint that these are great trading systems making load of money. The following para from my first post leaves no doubt

Also there was never an attempt to include the MM part in this thread. You would definitely agree that MM cannot be universal and has to adapted to one trading style, flavor, risk tolerance profile and capital. Saint has been covering the MM aspects rather well in his educative thread. This thread is restricted to only the TA part.

Thanks for listing the steps to develop a trading system. This will definitely help the members.

mmm…Contradictory isn’t it, Amarnath? Backtest Testing on the Left hand charts ..meaning data….

Right Amarnath..learning …. That was the very purpose of this thread.

Hope to see more postings from you on trading systems.

Warm regards

Karthik

Hi,

Yes well the sucess of this system depends on MM employed what i said it if we use long = green and short = red as it as i dont its much profitable as it minimise profit lowest extent possible . If its really better on your forward test and if its not discertionary i am more happy than you.

Yes the concept of MM is subjective depends only employ fixed ratio or fractional ratio etc but i belive as a author of this thread you can highlight some eg like which kind of MM you prefer for this system as you are developer of this system i prefer if you do it than any other seniors here even if there is some other threads regarding this subject bcoz i again to say only developer understand their system better than newbies here

Backtesting with tick to tick is like stimulating real market but at just faster rate for scanning past conditions for our enviroument

what i tried to explained here is dont focus on developing better indicators ( like MARSI , MACDRSI , KAMA , AMA etc ) already few 1000s of indicators do exits . All such can avoid just little lag it dont take you any were better

ps : i dont use metastock and i use AB but dont code it much i prefer Tradestation ( Omega prosuite platinum 2000i ) and for coding i depends on my programmers and i just design strategy rest is done by programming ppls to make it as codes ( why i want to interefere with their jobs )

Hi,

Yes well the sucess of this system depends on MM employed what i said it if we use long = green and short = red as it as i dont its much profitable as it minimise profit lowest extent possible . If its really better on your forward test and if its not discertionary i am more happy than you.

Yes the concept of MM is subjective depends only employ fixed ratio or fractional ratio etc but i belive as a author of this thread you can highlight some eg like which kind of MM you prefer for this system as you are developer of this system i prefer if you do it than any other seniors here even if there is some other threads regarding this subject bcoz i again to say only developer understand their system better than newbies here

Backtesting with tick to tick is like stimulating real market but at just faster rate for scanning past conditions for our enviroument

what i tried to explained here is dont focus on developing better indicators ( like MARSI , MACDRSI , KAMA , AMA etc ) already few 1000s of indicators do exits . All such can avoid just little lag it dont take you any were better

ps : i dont use metastock and i use AB but dont code it much i prefer Tradestation ( Omega prosuite platinum 2000i ) and Metatrader. For coding i depends on my programmers and i just design strategy rest is done by programming ppls to make it as codes ( why i want to interefere with their jobs )

Hi all,

I offer one of my trading system screenshot hope it helps all and how better is it than 9 , 13 cross over system

Attached Images

gold.gif (98.6 KB, 258 views)

The major issue over here is how to develop a proprietary system without experimentation. Everywhere we do read, "do optimize, develop your own system" but in the same breath the warning is not to experiment leaving the traders perplexed.

Further, there are two ways of improvement in business as well in life. Either, you fix anything only when it's broke and another is to always strive for continuous improvements. All the successful businesses have realized that the second strategy works better and hence strive to achive success thorugh continuous innovation and improvemnt in processes.

Now if we have to treat trading as business how can we avoid the path of continuous improvement and further how to stop experimenting with a goal of continuous improvement?

Also, if we want to experiment how do u do it? By backtesting, is not it? Despite all the limitations is there any way out?

Hope to be educated wherever I have made a mistake.

Best Regards,
--Ashish

Hi,

I dont say not to experiments and develop but develop and experiment in right and professional ways . This thread way intented to experiment 9 13 cross over but later there way slowly diversified to custom indicators which finally mislead peoples to think that custom indicators are better tools and as usual a good thread die it natural death over time that's why i insisted to stick to one fixed tool box and develop on same line correct me if i am wrong

Hi,

Can you explain the system u used to get these signals?...so tht we can learn from it...

thanks and regards,
Prakash

Custom indicators may or may not a better tool depending upon the indicator, its use and most importantly the understanding of the market and its ways of the system developer.

But I do agree in one aspect that the newbies should start with some simple system and then they can graduate in the indicator building exercise as we do in school going from nursey to KG and so on. If the newbies read this thread to learn the gradual shift from an very easy to understand and build system to complex systems, it would certainly add more value to them.

Best Regards,
--Ashish