SH's 315 Strategy - how to make it more comfortable and profitable!!

[NTrader42]

Dear Ntrader,
Where you posted these 45EMa things??????????Please give me the link...

Same thread few pages back

 

[vikrit]

Okay!!! Then its fine.. will compare and comment in case I find any differences!! There was some difference in EMA vales as I started checking form June 2007 hence I thought it might be Spot Nifty!!
Thanks.

there can be diff in ema values for futures in every start of month. it depends on how calculation have done.

if calculation are on ohlc values for all those years then calculate, it will give continuous ema values, having no gap at the end of month.

if we are doing calculation for every series(t) differently, and keeping the data for current month only, then in start of every month there is a difference in 3ema as it's calculation not having last series(t-1) ema & ohlc.

another possibility of diff is for calculating ema we need last ema. so if software/excel at start take sma for first few. in that backtest intial data was from 14 may, 2007.

basically, i dont have another data to compare. plz share your data for ema's with ohlc if possible. so, i will be able to compare & will backtest other suggestion's on that data.

thanks.

 

[mangup]

Dear Ntrader42 & vikrit,

Thanks for the analysis done with various improvements. Both of you have done a good job and added diff colours to the 315 Strategy.

This is what i was saying. We need to keep on backtesting the suggested improvements over the basic 315 Strategy.

Vikirt - You have published what i have in my mind. But it is slightly adding to the confusion. To make it more clear can you mention the rules that you have followed while performing the backtesting?

 

[blackberry]

I am getting some difference in ema values in AMibroker and online Icharts..Today zee entertainment ema values difference is 0.2, very little, in icharts.Amibroker showing difference of 0.5.This difference could create confusion if some one follows this 315 strategy.

This could happen in Nifty futures charts also...(I dont have nifty f charts in Amibroker)

 

[NTrader42]

I am getting some difference in ema values in AMibroker and online Icharts..Today zee entertainment ema values difference is 0.2, very little, in icharts.Amibroker showing difference of 0.5.This difference could create confusion if some one follows this 315 strategy.

This could happen in Nifty futures charts also...(I dont have nifty f charts in Amibroker)

Yes confusion will happen, to avoid that consistently follow only 1 of the source for your trading decisions.

Thanks

 

[mangup]

hmm same marriage season problems here also

As i also trade commodities, that effectively leaves only sat/sun to do this kind of work. But I feel we already made good progress, just getting things together and running the tests (walk-through) is needed, and a few weeks should not make a difference for a system that trades only about 200 trades in 11 years


on the test attached unintentionally 4 times leverage was used (margin 25%), which was an mistake so consider everything like gains, drawdowns as 1/4th of that in the image.


Thanks

Dear Ntrader42,

I have one doubt. Ur analysis shows:-
1. Net Profit % -> 5562.96%
2. Annual Returns -> 42.52%
3. Max System Drawdown % -> -84.59%

Sorry i don't know much bactesting results. Can you pl explain me the meaning of these results as the value are something beyond imagination?

mangup

 

[NTrader42]

Dear Ntrader42,

I have one doubt. Ur analysis shows:-
1. Net Profit % -> 5562.96%
2. Annual Returns -> 42.52%
3. Max System Drawdown % -> -84.59%

Sorry i don't know much bactesting reuslts. Can you pl explain me the meaning of these results.

mangup

Lets see what does this Annual Returns -> 42.52%, translate to . . .

1 lot of nifty future ==> 50000 margin & M2M money.

50% of 50000==> 25,000/ 50 (lot size) gives 500 points


These results are with about 10-11 years of data and using an margin account, so the compounding shows some crazy results.

The basic stats to take home is the win/loss ratio was about 35%, the system can stay in red for few months, makes about 500-600 points per year without considering exit & re-entry etc. . .

using the 3ema exit / re-entry and 15 ema adds will double these gains


Thanks

 

[mangup]

Hello

The above concept is very easy to test using amibroker and many different combination or twists can also be easily tested.

For e.g. if instead of 2 times up/down we consider 3 times, the profit increases to 2700 and number of trades reduce to 240.

This concept of using the direction of the slope of MA line to decide the long/short works very well with higher MA

For e.g. if we choose EMA45 we get just 102 trades with gain of 7291 points and win/loss of 40/62

List of Trades for ema45

Dear Ntrader42,

I have not understood the logic of using the direction of the slope of MA line to decide the long/short works very well with higher MA. Can u pl explain a little bit?

Just basic rules, long (buy today's close) when ema45 > ema45 of previous day, and e45 of previous day > e45 of day before that.

even ema30 is giving comparable results. Using add on ema15 touch and part booking profit with ema3 come to mind, as a first step for improvement.

Thanks

Does it mean -
Buy --> Before today's close, If today's 45 EMA > Prev Day's 45 EMA
Sell --> Before today's close, If today's 45 EMA < Prev Day's 45 EMA

or i am missing something.

 

[NTrader42]

Dear Ntrader42,

I have not understood the logic of using the direction of the slope of MA line to decide the long/short works very well with higher MA. Can u pl explain a little bit?



Does it mean -
Buy --> Before today's close, If today's 45 EMA > Prev Day's 45 EMA
Sell --> Before today's close, If today's 45 EMA < Prev Day's 45 EMA

or i am missing something.

Long (buy today's close) when
1. ema45 > ema45 of previous day, and
2. ema45 of previous day > ema45 of day before that.

 

[timepass]

If EMA 45 (today) > EMA 45 (yesterday) > EMA 45 (day before yesterday), then buy..

If EMA 45 (today) < EMA 45 (yesterday) < EMA 45 (day before yesterday), then sell..

hmm.. dekhna padega.

Edit : But, maybe this will work better with Nifty spot, because 45days ago today's future prices may have been commanding serious premium... just a thought.