Nifty Positional Trading - Super Trend Indicator

[mastermind007]

Hello Cubt

When we take back testing results we forget to check if we would get a fill at the price the back tests are showing . . .

Besides the trading Cost, Slippage is a big concern but that is know and acknowledged by everyone.
I have seen many people back testing do not even consider Forwardization charges/costs?

2008 to 2013, 5 years => 60 months => that will cost you 1500 to 1800 points to shift your futures position from one expiry to next
2100 trades should cost you 6300 points (1.5 points per trade entry & exit) in brokerage & taxes
minimum slippage can be consider at 4200 points ( avg slippage 1 point per trade i.e. entry & exit)

Without all these how does one arrive at a realistic estimation?

Cheers


Happy

Happyji


I have worked out a full cost ratio which is approx 0.1176% of trade cost in cash market Intraday. Delivery traders are 5 times Intraday.

NFO and Options are lot more complicated but when you shrink all numbers down, it comes at 0.09% and improves in your favour with lot size growth.

Both are approximations but idea is to budget for this costs; set aside few extra points (but not too much) so that if ur AFL says X% profit, you will get at least X%.

Also, I play around with entry prices. Besides enter at candle close or enter-at-next-open or permutation off O H L C such as (2 * H + L) / 3 for long positions and (H + 2 L)/ 3 will be more likely to get realistic in trade prices.

Fortunately, given the way I have structured my ATS, my slippages are low
and less than 0.5%.

 

[rh6996]

Big blunder with my Amibroker Intraday settings. In Database - Intraday settings I have mentioned start and end time as 9:30 and 16:00 instead of 09:15 and 15:30.

I have made this settings already but instead of using the correct database today, i mistakenly selected the wrong database and traded. Resulted to unnecessary trades and big loss



so as per Super Trend- the position that was shorted yesterday around 6215 is still on hold with Stop Loss as 6185.

Do you mean to say that the first Data candle in your chart was Starting from 9:30??? Then it means levels for 9:15 to 9:30 was not being displayed in your chart, but this would be obvious to you at the start of the day!? Dont think you can miss this in RT !
Cubt, is it possible to post for next 8/10 trades the Candle on which you are taking a trade, example : Going long at 6140, on close of 10:12 candle on 30th Oct , SL at entry at 6090 being ST value . Just ignore it if it means a lot of hassle.
Also can you post the start and end dates when we had a consecutive 500+ loosing trades !

 

[Cubt]

Yes Sure, Rh... I shall post the charts from next trade onwards. And the consistent loss that I told u was during 2010 October I guess, not sure. I will check n update u.

 

[Cubt]

And I was thinking the difference was because of my rt data when u posted long @ 6184.

Edit : Matter of fact , my system is still showing SL @ 6189.66 for the Short triggered @ 6194.9. Don't know why the diff. . :annoyed:

Check what ur 10 mins time frame says, the trade I took at 6215 was by 10 mins tf not 12 mins.

I have already tested 12 mins time frame, it gave good returns of 11200 points in 950 trades for 6 years of data.

10 mins time frame gave 11150 points in 1050 trades for 6 years.

But in last few days of trades, am missing the gap up or gap down if traded with 12 mins, but 10 mins tf is able to capture those gaps in advance. I go mad when I miss those profitable gaps. So thinking of trading in 10 mins time frame, but when I get signal to sell/short I should follow what 10 mins says.

But when system give buy signal in 10 mins, I should wait for 12 mins as well to give buy signal. Bcoz many times market touches n could breakout 10 mins tf and start falling again.

But when its sell signal, market falls rapidly so instead of waiting for 12 mins confirmation, we should sell immediately when 10 mins says so. When market moves up it doesnt move that much fast, thats why I said to wait for 12 mins confirmation for buy signal.

I have test this for couple of trades, if am comfortable with it I will follow it. Or else, I will stick to 10 mins time frame.

 

[VJAY]

IMO if trading backtested method must trade @ least for 1 year in same tf and rules...keep journal&notes of every trade...then after 1 year you can tweek some points after observe your notes ......otherwise switching in timeframes are always hurt the method....it will make switching mentality every time we miss some good trades OR we see contineous loss trades ....My views

 

[Cubt]

Thanks Vijay, I respect ur views. Henceforth, I will continue to trade in same lots, same time frame, same system. No more changes.

Time frame - 10 Mins
Method - Super Trend with parameters 4 and 6.
Lots - 4 lots.

Will start posting my trades here with charts. Will also add my observations here, so that I can refer it anytime, any where.

Edit - May be I should reduce the lot size to the SIZE that I dont care whatever happens. So that I will be disciplined enough to trade even with consistent loss.

Change in lot size - 2 lots.

 

[rkkarnani]

Check what ur 10 mins time frame says, the trade I took at 6215 was by 10 mins tf not 12 mins.

I have already tested 12 mins time frame, it gave good returns of 11200 points in 950 trades for 6 years of data.

10 mins time frame gave 11150 points in 1050 trades for 6 years.

But in last few days of trades, am missing the gap up or gap down if traded with 12 mins, but 10 mins tf is able to capture those gaps in advance. I go mad when I miss those profitable gaps. So thinking of trading in 10 mins time frame, but when I get signal to sell/short I should follow what 10 mins says.

But when system give buy signal in 10 mins, I should wait for 12 mins as well to give buy signal. Bcoz many times market touches n could breakout 10 mins tf and start falling again.

But when its sell signal, market falls rapidly so instead of waiting for 12 mins confirmation, we should sell immediately when 10 mins says so. When market moves up it doesnt move that much fast, thats why I said to wait for 12 mins confirmation for buy signal.

I have test this for couple of trades, if am comfortable with it I will follow it. Or else, I will stick to 10 mins time frame.

Thankfully, you did not try any other parameters for Supertrend, otherwise you would even switch to that as that setting "at the moment" gave better results !!! Whats the logic behind moving to 10 min TF? Sorry for being sarcastic , but it would be really good to trade for a pre-decided period of time without deviating from the Rules. Its fine if your rule says to trade long in 12 min and to trade 10 min for Short trades, but switching TF just because the current market movement favours 10 min chart should not be done. Change the rules only after you trade for a Set period and then trade with the changed rules ! Only this discipline can lead us to Long Term gains.
Please dont take any offence to this post !

 

[rkkarnani]

Did not notice that Vjay has posted similar views. Thanks Vjay, this is also what I mean to convey.

 

[shekharinvest]

This thread on Supertrend attracted me because I have been watching this indicator for a pretty long time and had started working on it since some time back.

I am a no big fan of automated backtest, but do look back on charts and see what had happened over a period of time. I believe in what meets the eye.

What I have observed that in Supertrend Stop loss are pretty wide, even if we have a good run over a period of time it does not makes much money. In comparison to losses it gives during minor whips.

For example I am posting an hourly chart of most recent happening on supertrend. (Please ignore PARSAR & RSI still working on them)

CHART:

I have selected the longest successful trade which is preceded and followed by minor whips.

You will notice that Such a nice run gave only 82 points on nifty
where as a small whip preceding the trend gave a loss of 111 pts
and the whip which followed the trend again gave a loss of 68 pts.
Now we are in NET Loss of 97 points.

If we assume 50% success rate and
take trade 1 and trade 2 we are in a loss 29 pts
take trade 2 and trade 3 there is a very small profit 14 pts

Any method which makes less money on successful trades than it looses on a failed trade is not going to work It needs to be used in conjuction with other methods for filtering purposes.

Whereas, I have found it to be little useful if we use 5min TF for positional trades. The higher up we move on time frame SL becomes wide.

This is in no means to belittle Cubt's efforts, just to put in my views if it helps.

 

[Cubt]

Thats right, stop loss are always wider irrespective of whatever system we use. I have back tested 5, 10, 12, 15, 30 minutes, 1 hr, 4 hr. I have uploaded the results as well, finally ended up with 10 mins time frame.

Trading in higher time frame with super trend will not result in higher profits in long run.