20% ROI per month on captial above 1 Cr possible?

[marimuthu13]

Yes this system after tweeking using some filters doesnt go beyond 50%...I have tested on single stocks and also on band of 10 big stocks for 10years data
but 50% is also not bad provided you take all the signal ...for people who loves Andhadhun trading....They can use this system.

But choosing from band of 136 stock would require a very good filter and then to take out top5 is the real thing which not everyone will reveal .....Thats called turup ka ekka (A)

without technicals also we can filter the top5 from 140 F&O stocks..... guess how?

based on alphabetic order... first 5 or last 5...

based on stock prices...high to low, or low to high......

 

[sridhga]

But again I am not sure I can start looking at 136 scrips

Very simple.
Create a watchlist/portfolio of these 136 scrips on Yahoo finance. You can download the OHLC-V data every 5 minutes. You can create an excel template to check the price movement during the day. You can create it in such a way so as to track prices in comparison to 9-30 am, 9-45 am etc.(You must download those prices at 9-30, or 9-45 to compare later). You can create all the filters on the excel template. With macros you can automate several of the excel processes.
Only if some share interests you, then add it to the watch list on your stock-broker terminal.



PS: @Riskyman showed me the advantages of using Yahoo Watchlist.

 

[mohan.sic]

That’s why I said, selection criteria is only which matters here... how to select five stocks in morning which could give you 70% right trade

you mean 5 different stcoks will be selected each day out of those 136 ?

 

[marimuthu13]

you mean 5 different stcoks will be selected each day out of those 136 ?

yes ji..
that's key here..if you trade same stocks, you need 5 different systems......

 

[mohan.sic]

yes ji..
that's key here..if you trade same stocks, you need 5 different systems......

down the line it means picking the 5 potential movers of the next day.
How can anyone do that, how can we know if a stock is going to move or just play in a range through the next day ?

orb or any strategy - I think success rate should not depend on the stocks we choose. On large number of trades strategy output should be almost identical on any stock. On large data testing - strategy performance ( % in success rate ) should be independent.

Just think in this direction - If we can pick those 5 stocks- what is the edge of orb strategy ? Even a simple MA cross over can give same returns.

or am I missing something in this ?

 

[lemondew]

Can you post some more charts?

yes ji..
that's key here..if you trade same stocks, you need 5 different systems......

 

[Happy_Singh]

Good Morning

Just one more day left of the long weekend

Due to the way it was presented on this thread,
many of us would have got interested in it

Initially even the idea of targeting 20% on a biggish capital sounded quite outlandish
but the arithmetic presented convinced few of us to continue perusing the idea further

As suggested by MM, I have used most of the time during this weekend to manually scan the charts.
Not possible to scan through charts of 100's of scrip, so selected 9 arranging them on one screen
(will share the simple AFL code and screen on next post)

Will also write about my views/findings in next post
but the conclusion so far is the same as discussed by many above,
the stock selection / scanner is the master key for 20% PM
btw, I am no way even near the benchmark of 70% strike rate


For me the question is if not 20%, will 10% be good enough or how about 5%

Been very interesting 2 days, enjoyed the work put in
learned a few thing and re-learned a lot more

Thanks Marimuthu brother

.

 

[Happy_Singh]

GraphXSpace = 2.5;
RequestTimedRefresh(15);
_SECTION_BEGIN("PRICE_PLUS");
SetChartOptions(0,chartShowArrows|chartShowDates);
_N(Title = StrFormat("{{INTERVAL}} {{DATE}} O: %g, H: %g, L: %g, C: %g (%.1f%%) {{VALUES}}", O, H, L, C, SelectedValue( ROC( C, 1 ) ) ));
SetChartBkColor(colorBlack);
BarCol    =    IIf(C < O, colorRed, colorBlueGrey);
SetBarFillColor(IIf(C > O, ColorRGB(0,0,64), ColorRGB(128,0,0)));
Plot(C, "Close", BarCol, styleCandle | styleNoTitle);
_SECTION_END();


_SECTION_BEGIN("FIRST 15 MIN ORB");
// 15 MINUTES OPENING RANGE
TimeFrameSet(in15Minute);
    FB  = TimeNum() > 091500 AND TimeNum() < 093000;
    H15 = ValueWhen(FB,H);          
    L15 = ValueWhen(FB,L);
TimeFrameRestore();
BO = TimeFrameExpand(H15,in15Minute);      
BD = TimeFrameExpand(L15,in15Minute);

TL1 = 2 * BO - BD * 1;
TL2 = 3 * BO - BD * 2;
TS1 = 2 * BD - BO * 1;
TS2 = 3 * BD - BO * 2;

//    DISPLAY LEVELS
FB = Day()!=Ref(Day(),-1);        // LAST BAR
STYLE_X = styleNoRescale|styleStaircase;
STYLE_Y = styleNoRescale|styleStaircase|styleDashed;

Plot(IIf(FB,Null,BO),"",colorBlueGrey,STYLE_X);  
Plot(IIf(FB,Null,BD),"",colorLime,    STYLE_X);  
Plot(IIf(FB,Null,TL1),"",colorBlueGrey,STYLE_Y);
Plot(IIf(FB,Null,TL2),"",colorBlueGrey,STYLE_Y);
Plot(IIf(FB,Null,TS1),"",colorLime,    STYLE_Y);      
Plot(IIf(FB,Null,TS2),"",colorLime,    STYLE_Y);

_SECTION_END();

 

[marimuthu13]

Thank you Happy singh bhai..

I have shared few basic set up with few members who have actively participated in this thread..

Whatever i shared wont meet desired target , but thats the starting point..

If Anyone going through the same setup back testing manually, then it will lead to redefinement of criteria ..with that you will reach the targeted system within 6-12 months.....lot of work to do .....

 

[marimuthu13]

down the line it means picking the 5 potential movers of the next day.
How can anyone do that, how can we know if a stock is going to move or just play in a range through the next day ?

orb or any strategy - I think success rate should not depend on the stocks we choose. On large number of trades strategy output should be almost identical on any stock. On large data testing - strategy performance ( % in success rate ) should be independent.

Just think in this direction - If we can pick those 5 stocks- what is the edge of orb strategy ? Even a simple MA cross over can give same returns.

or am I missing something in this ?

I didn't understand your question brother..please elaborate with example of EMA...