Price movements in the same direction?

[uasish]

Basing purely on yesterday's price this is the probability displayed by Mkt:

http://www.traderji.com/90711-post26.html

Basing purely on yesterday's price this is the (crude)probability displayed by indicator(coded on 19th May'07 in Metastock):

http://www.traderji.com/90985-post57.html

The cache here is ,to-day in mkt hrs if it trades below 'Bear's Zone' (for more than 30mins),a Short position initiated(on rise)has got above 50% chance (with 'Bear's zone' as SL) tomorrow it will be Black candle,hence (as the daily price range is above 2%) a reasonable > 2% from entry to exit.For my style (of 50% liquidation at 1st chance of profit) enters the next indicator 'Bear's Grip',the 1st tgt (a place to book 50%),if it closes below 'Bear's Grip' then keeping open the balance 50%.
Now as Savant_Garde & kkseal both pointed out (based on their rich experience) we can take recourse to other indicators to filter which scrip has more chance of falling tomorrow ,in those cases (pre filtered) the Bear's Zone
price may be penetrated in 1st 30mins.
As these are not tested in actual mkt situation (coded just yesterday)& based on purely EOD data,so i am having 'Believer Bias'.

 

[uasish]

Now for filtering ,depending purely on yesterday's O H L C ,2 more indicators:

http://www.traderji.com/91001-post59.html

Here the 'Predicting Tomorrow's Probability' & 'Markov Chain' (coded yesterday & today in Metastock) will help.

Thks to CV for conceptually hitting me 1 month back,with 'Random'; in process a bug was ignited & to Ajay in channeling the Quest.At least now i have a crude (pipe-gun) level tool to fight this out.
Lots of backtesting & forward testing is to be done,until then MACD / RSI /MA .

 

[oxusmorouz]

If you are testing only OHLC, then instead of just one day's relationship, you will also have to test channel breakouts, n-bar H-L, ORBs, Inside/Outside Bars, range breakouts etc as they are all dependent on OHLCs.

Regards

Yes you're right, so far we have been concentrating on 1 day price action alone. I hope the studies of more complextity can be dealt with as time proceeds

 

[oxusmorouz]

Basing purely on yesterday's price this is the probability displayed by Mkt:

http://www.traderji.com/90711-post26.html

Basing purely on yesterday's price this is the (crude)probability displayed by indicator(coded on 19th May'07 in Metastock):

http://www.traderji.com/90985-post57.html

The cache here is ,to-day in mkt hrs if it trades below 'Bear's Zone' (for more than 30mins),a Short position initiated(on rise)has got above 50% chance (with 'Bear's zone' as SL) tomorrow it will be Black candle,hence (as the daily price range is above 2%) a reasonable > 2% from entry to exit.For my style (of 50% liquidation at 1st chance of profit) enters the next indicator 'Bear's Grip',the 1st tgt (a place to book 50%),if it closes below 'Bear's Grip' then keeping open the balance 50%.

Wonderfully explained sir

 

[oxusmorouz]

Does stock market movements behave like an outcome of a random coin toss over a long run?

Data Used : DJIA from 1897 - daily data comprising of 28795 bars.

Results in spreadsheet is attached.
I request members to comment on this.

Obliged.

 

[oxusmorouz]

Results for the above post:

 

[uasish]

This is a 'Conceptualy Enriching' thread.
Ajay's Beta test report point out now he had to incorporate 28795 bars instead of his initial 8000 odd bars,gosh! the RARERITY.
However this is a report on NSE on 500 bars for your perusal.


http://www.4shared.com/file/16310464/134a278c/Report_of_500_Days_NSE_Scrips.html

 

[kkseal]

Results for the above post:

Not sure what the various 'expected bars' stand for.

Regards,
Kalyan.

 

[uasish]

I believe Ajay wanted to term 'Success' as say yesterday's close less than today's close.
What is the 'Certainity' of this 'Success' ? To find this out Ajay has fallen back on the
science of Probability.
To find the probability of THIS 'Certainity' of 'Success' :=
Here [ 1 ] in (colE of Ajay's report)denotes that the probability of CERTAINITY
(say yesterday's close < today's close) is observed ONCE in 28795 bars.
Mathametics ,hence we have to accept it.

The cache here is Ajay has taken into a/c of 'Random Variable' whose outcome is either
[ 0 ] or [ 1 ] as in 'Coin Toss'

http://www.traderji.com/88911-post965.html

After this post i had a deep introspection 'Conceptually' & accepted that indeed it is
a 'Random Variable' but may not be 'Coin Toss' ,becoz coin toss has no previuos
"MEMORY" but PRICE certainly has,hence it is a random variable with previous memory.

Memorylessness is often misunderstood by us taking recourses to probability:
the fact that Pr(X > 13 | X > 10) = Pr(X > 3) does not mean that the
events X > 13 and X > 10 are independent; i.e. it does not mean that
Pr(X > 13 | X > 10) = Pr(X > 13).
To summarize: "memorylessness" of the probability distribution of the number
of trials X until the first success means:=
Pr(X > 13 | X > 10) = Pr(X > 3)

not this

Pr(X > 13 | X > 10) = Pr(X > 13)
(That would be independence. These two events are not independent.)


Ajay is not trying to reinvent the wheel ,in trading "Gaussian" "Random Walk" "Z- Test"
all this concepts are known,only to tackle the probability.(many indicators are built by famous peoples on these concepts)

My dilemma ,contradiction is not with Ajay's brilliant effort & findings
but i still belive Price is definitely a Random variable but not without 'MEMORY'as in
'Coin Toss' but as in a game of Poker 'With Memory'

A 'Stochastic process' amounts to a sequence of 'Random' variables known as
a 'Time series' (CV often mentions Price to be FINANCIAL time series).

So i stand to be corrected conceptually ,which will only help me to develop self.
System Traders Success is on his own system built by his concepts,understanding,experience.

 

[uasish]

Ajay has MS 10 (where the exploration col no more than my MS 9)

This is the % of close > yesterday ,if after consequetive 2/3 days close being higher ,(in sheet 2).

http://www.4shared.com/file/16330688/fd3bc127/Certainity_.html