Complete newbie at strategies/Bollinger bands

[margalis]

Hello,

First of all, I am really new with NT7... There must be simple thing I am missing so I created an account to ask you guys.

I am in my last year of bachelors and decided to write my thesis on EMH and technical analysis (challenging but really interesting topic to me). I chose Bollinger bands as my indicator and ran a super simple back-test. Results were interesting. This is my cumulative profit graph (for eur/usd 1980-2010 period):

https://gyazo.com/d35b9b2032a018b99863470a8ff6b26e

I thought by reversing (longing when I was shorting and shorting when I was longing) my strategy I should get exactly opposite results as I am using stop loss/take profit functions as well. However I got this:

https://gyazo.com/ddbf231f6e0be9297405a175d06013cb

So basically I backtest opposite strategies and I get same result... I found this confusing and tried checking my chart for when I enter/exit positions. I observed a really interesting thing for one of the months.

Normal Bollinger:

https://gyazo.com/fe7d9cc75118651ddaa84794cd1d6e4d

Reverse Bollinger:

https://gyazo.com/10d7620842f863ed13ca3553cfb1cc7b

Basically I enter positions at same prices with stop loss/take profit of 30 pips and these positions do not offset each other... Therefore in both strategies i get around 30% winning trades and 70% losing trades doing exactly opposite things...

I hope my problem here is clear and I would appreciate some help... If someone would be willing to guide me through my work and help me out with further questions I would really appreciate if you PMed me and we would exchange skypes..

Thank you all in advance

[sledge]

This sounds more like something I would read about on elitetrader.com.

my strategy I should get exactly opposite results as I am using stop loss/take profit functions as well.

Unfortunately, it will not.

[koganam]

Hello,

First of all, I am really new with NT7... There must be simple thing I am missing so I created an account to ask you guys.

I am in my last year of bachelors and decided to write my thesis on EMH and technical analysis (challenging but really interesting topic to me). I chose Bollinger bands as my indicator and ran a super simple back-test. Results were interesting. This is my cumulative profit graph (for eur/usd 1980-2010 period):

https://gyazo.com/d35b9b2032a018b99863470a8ff6b26e

I thought by reversing (longing when I was shorting and shorting when I was longing) my strategy I should get exactly opposite results as I am using stop loss/take profit functions as well. However I got this:

https://gyazo.com/ddbf231f6e0be9297405a175d06013cb

So basically I backtest opposite strategies and I get same result... I found this confusing and tried checking my chart for when I enter/exit positions. I observed a really interesting thing for one of the months.

Normal Bollinger:

https://gyazo.com/fe7d9cc75118651ddaa84794cd1d6e4d

Reverse Bollinger:

https://gyazo.com/10d7620842f863ed13ca3553cfb1cc7b

Basically I enter positions at same prices with stop loss/take profit of 30 pips and these positions do not offset each other... Therefore in both strategies i get around 30% winning trades and 70% losing trades doing exactly opposite things...

I hope my problem here is clear and I would appreciate some help... If someone would be willing to guide me through my work and help me out with further questions I would really appreciate if you PMed me and we would exchange skypes..

Thank you all in advance

That is because you have not taken into account that a market (even the manipulated one that we have had for the last 7 years) is probabilistic, not deterministic.

You have to account for the probabilities of your data points being hit, and hence the expectancy of the trading system. You have not done so. The result that you got is only to be expected in any non-deterministic system: the output of reversing inputs does not necessarily (in fact, rarely), produces the opposite result.

Short term movements in the market are a random walk, and it is provable. In fact, I presented the empirical statistical proof on a webinar that I held for NT last year.

cf: http://ninjatrader.com/support/forum...ad.php?t=43860

[margalis]

That is because you have not taken into account that a market (even the manipulated one that we have had for the last 7 years) is probabilistic, not deterministic.

You have to account for the probabilities of your data points being hit, and hence the expectancy of the trading system. You have not done so. The result that you got is only to be expected in any non-deterministic system: the output of reversing inputs does not necessarily (in fact, rarely), produces the opposite result.

Short term movements in the market are a random walk, and it is provable. In fact, I presented the empirical statistical proof on a webinar that I held for NT. last year.

cf: http://ninjatrader.com/support/forum...ad.php?t=43860

I PMed you and would appreciate if you replied, I have a few questions if you dont mind helping...

[koganam]

I PMed you and would appreciate if you replied, I have a few questions if you dont mind helping...

Just looked at your PM. Will add you to my Skype.