Experiments in Technical Analysis

[aca_trader]

dear siva and ashish,

you may take up this system and join us in the system testing exercise. if you want, one of us can make the system, provided you provide the buy and sell logic in with all terms, clear enough for the system coding.

hope ashish also joins us with the testing.

murthymsr

Dear Murthy!

I am thankful for your invitation but I believe that already a few able persons are handling the system testing very efficiently and hence hence more testers are not needed.

I wrote about the Stop-loss point because of the growing unease with %age based stop-loss system which never appealed to me 'coz of its subjectivity. I hope some objective method will work wonder with the already fasinating systems.

Said so, I am always open to contribute in any productive manner needed.

Best Regards,
--Ashish

 

[oldmirage]

hi guys,

I have been following this beautiful thread occationally and it's really good to see people wokring together to make nice trading system.

One humble suggestion:

Can any of you jott down the log of this entire process and if we can keep it updated then it would really help any new member who wants to understand the process of this thread.

we can start with a the latest position of system and do some back tracking.

Just a suggestion. O/w You guys are doing awsome job.

 

[asnavale]

Hi oldmirage,

Do you mean log of the daily signals and trades to be recorded? Or are you suggesting the record of the posts? The daily signals and trades are recorded as soon as they are posted. If you are suggesting the record of all the posts then we have to to do it.

Regards

-Anant

 

[kkseal]

Hi oldmirage,

Do you mean log of the daily signals and trades to be recorded? Or are you suggesting the record of the posts? The daily signals and trades are recorded as soon as they are posted. If you are suggesting the record of all the posts then we have to to do it.

Regards

-Anant

I'd think a summary of the following would do :-

1) The salient features of the various systems conceived (in plain language & not AB etc code)
2) The parameters & objectives for the tests &
3) The status till date

Even i get a bit confused at times (being a late joiner) and have to wade through hundreds of post to find out what the original idea was.
I therefore second Oldmirage's suggestion.

Regards,
Kalyan.

 

[oldmirage]

this is what I was asking for. Thanks Kalyan for putting in more precise words. I am much more interested in the log of all devised systems and if possible their results in different systems.

I dont' have any of the systems you guys have been using, so I would not be able to contribute in it at all.

I'd think a summary of the following would do :-

1) The salient features of the various systems conceived (in plain language & not AB etc code)
2) The parameters & objectives for the tests &
3) The status till date

Even i get a bit confused at times (being a late joiner) and have to wade through hundreds of post to find out what the original idea was.
I therefore second Oldmirage's suggestion.

Regards,
Kalyan.

 

[asnavale]

Hi Kalyan and Oldmirage,

I am trying to compile all the information you suggested. Will post sometime later. Thanks for the good suggestion.

Regards

-Anant

 

[karthikmarar]

Dear Karthik and other friends who are associated in this SIE,

It seems my argument over the technical stop is not tasting good or my presentation might be wrong. I am sorry if anyone gets hurt over my argument on this issue. From my side I am not directly contributing anything to this exercise. So naturally I don't own any rights to come in between. But as I also have a little experience in these markets I thought I should do something even indirectly. But never intended to spoil the game. My only objective here is to try keeping the game on right side. But, looking into a few posts here, it seems I better go off this topic.

Dear Ajay

Sorry, I could not reply earlier as I was traveling.
Let me reiterate that there is no bad taste/Spoiling the game in your posts regarding the subject of % stoploss. We are keen learners and nothing helps like a good/heated/hard discussion. Everybody is just putting forward his/her point of view. In fact this has become a good topic for discussion and I am pleased at the response this has received. And also this is an important topic too.

More on the % stop used in the SIE later

Warm regards

Karthik

 

[karthikmarar]

A stop is a point where the trend is said to have failed. There is nothing like protection of profit or protection of capital - for a stop. The stop might hit at any time, whether you are running your trade in break-even or in profit.

Just Sharing a few thoughts

We should have different stops a Start Trade Stop/ Stoploss for capital protection and a Trailing Stop for locking in the profits.

Also there cannot be a general rule for stops for all Trading systems. For example for a Breakout trading system the Start Trade stop loss could be easily defined. But in case of a MA cross over trading system it is very difficult to pre estimate the Start Trade stop.
In such cases we can go for the following stops
1. A technical stop if we are sure that the initial drawdown will be within the risk imposed by your MM rules. A good trading system would ensure that you are technically stopped out well in time. But unfortunately most trend following system are too slow to react to sudden huge adverse moves as in case of markets behave nowadays.
2. A % stop loss based purely on MM rules.
3. A fixed amount stoploss again based on MM rules
4. The last low pivot or resistance zone. (Again one has to see if these are well within the MM rules. Otherwise your MM rules should allow such huge initial draw downs)

In case of the SIE, we could have adopted two approaches. One, have a common trading system and all volunteers would adopt the same one and apply it with their own discretion. The second approach would be each volunteer to adopt a different system and apply a set of common MM rules. Our interest was to test the systems and get the feel of a real time test instead of a back test we adopted the second method. This too has thrown up some interesting questions. So to have a common base for all system we put a % stoploss for all system based on MM rules. Also the timing of the SIE was rather perfect and with the huge drop in the market many weakness of the systems got exposed.

The trailing stop to lock in profit is another matter and it would be interesting to include it in the next round of SIE if there is one.

Warm regards

Karthik

 

[asnavale]

Dear friends,

I have posted a modified MABIUTS-HB system in the related thread 'System Implementation Excercise". I have explained the reasons for the changes and also attached the modified code there. You may access the post and the modified code here:

http://www.traderji.com/71573-post197.html

Regards

-Anant

 

[asnavale]

Just Sharing a few thoughts

We should have different stops a Start Trade Stop/ Stoploss for capital protection and a Trailing Stop for locking in the profits.
.............................
.............................
.............................

Warm regards

Karthik

Hi Karthik,

Just sharing my views on stoploss concept. I agree with your concept of different stoplosses, an intial "Start Trade Stoploss" followed by a technical stoploss.

When we enter a trade we can adopt a percentage based stoploss as per MM rules we want to follow. If the current price is against us and differs from buy price by an amount equal to or more than the SL% we should exit. If, after entry, the prices move in our favor such that the current price differs from the entry price by more than the SL%, then switch over to a trailing SL. Then our exit will be only if the trailing stoploss is hit. Even if the system gives a SELL signal, we should ignore it till the trailing SL is hit. A recent case is that of EDUCOMP. On 27-12-06, based on Closing price a SELL signal was generated. But next day, although the price initially fell further, it recovered and went higher than the previous Close and thereafter it has resumed its uptrend. If we had a trailing stoploss, the SELL signal would have been ignored and we would continue with the uptrend. The question would be how to fix the trailing stoploss. We may discuss this further. May be you or other members have different views. I am just expressing whatever I could think of.

Regards

-Anant